From 3d3522604a5557e80fdcab61c54bd04eaf15b525 Mon Sep 17 00:00:00 2001 From: Joachim Eastwood Date: Tue, 19 Feb 2013 22:44:57 +0100 Subject: [PATCH 001/198] spi/atmel: fix speed_hz check in atmel_spi_transfer() atmel_spi_transfer() would check speed_hz and fail if the speed was changed in the transfer. After commit "spi: make sure all transfer has proper speed set" this would happen on all transfers. Change speed_hz check to only fail if a lower speed than max is requested. Signed-off-by: Joachim Eastwood Acked-by: Nicolas Ferre Signed-off-by: Grant Likely --- drivers/spi/spi-atmel.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/spi/spi-atmel.c b/drivers/spi/spi-atmel.c index 656d137db253..80f5867c088b 100644 --- a/drivers/spi/spi-atmel.c +++ b/drivers/spi/spi-atmel.c @@ -847,8 +847,8 @@ static int atmel_spi_transfer(struct spi_device *spi, struct spi_message *msg) } /* FIXME implement these protocol options!! */ - if (xfer->speed_hz) { - dev_dbg(&spi->dev, "no protocol options yet\n"); + if (xfer->speed_hz < spi->max_speed_hz) { + dev_dbg(&spi->dev, "can't change speed in transfer\n"); return -ENOPROTOOPT; } From 446411e18b2cb17d153e45f634a3c9a79ada3ac2 Mon Sep 17 00:00:00 2001 From: Andreas Larsson Date: Wed, 13 Feb 2013 14:20:25 +0100 Subject: [PATCH 002/198] spi: Initialize cs_gpio and cs_gpios with -ENOENT The return value from of_get_named_gpio is -ENOENT when the given index matches a hole in the "cs-gpios" property phandle list. However, the default value of cs_gpio in struct spi_device and entries of cs_gpios in struct spi_master is -EINVAL, which is documented to indicate that a GPIO line should not be used for the given spi_device. This sets the default value of cs_gpio in struct spi_device and entries of cs_gpios in struct spi_master to -ENOENT. Thus, -ENOENT is the only value used to indicate that no GPIO line should be used. Signed-off-by: Andreas Larsson Signed-off-by: Grant Likely --- drivers/spi/spi.c | 4 ++-- include/linux/spi/spi.h | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/spi/spi.c b/drivers/spi/spi.c index 004b10f184d4..bcca66c5107f 100644 --- a/drivers/spi/spi.c +++ b/drivers/spi/spi.c @@ -334,7 +334,7 @@ struct spi_device *spi_alloc_device(struct spi_master *master) spi->dev.parent = &master->dev; spi->dev.bus = &spi_bus_type; spi->dev.release = spidev_release; - spi->cs_gpio = -EINVAL; + spi->cs_gpio = -ENOENT; device_initialize(&spi->dev); return spi; } @@ -1079,7 +1079,7 @@ static int of_spi_register_master(struct spi_master *master) return -ENOMEM; for (i = 0; i < master->num_chipselect; i++) - cs[i] = -EINVAL; + cs[i] = -ENOENT; for (i = 0; i < nb; i++) cs[i] = of_get_named_gpio(np, "cs-gpios", i); diff --git a/include/linux/spi/spi.h b/include/linux/spi/spi.h index 38c2b925923d..c395f32b53b3 100644 --- a/include/linux/spi/spi.h +++ b/include/linux/spi/spi.h @@ -57,7 +57,7 @@ extern struct bus_type spi_bus_type; * @modalias: Name of the driver to use with this device, or an alias * for that name. This appears in the sysfs "modalias" attribute * for driver coldplugging, and in uevents used for hotplugging - * @cs_gpio: gpio number of the chipselect line (optional, -EINVAL when + * @cs_gpio: gpio number of the chipselect line (optional, -ENOENT when * when not using a GPIO line) * * A @spi_device is used to interchange data between an SPI slave @@ -261,7 +261,7 @@ static inline void spi_unregister_driver(struct spi_driver *sdrv) * queue so the subsystem notifies the driver that it may relax the * hardware by issuing this call * @cs_gpios: Array of GPIOs to use as chip select lines; one per CS - * number. Any individual value may be -EINVAL for CS lines that + * number. Any individual value may be -ENOENT for CS lines that * are not GPIOs (driven by the SPI controller itself). * * Each SPI master controller can communicate with one or more @spi_device From 8ec5d84ef51cc64ed02bb9bf0e43a652178252c1 Mon Sep 17 00:00:00 2001 From: Andreas Larsson Date: Wed, 13 Feb 2013 14:23:24 +0100 Subject: [PATCH 003/198] spi: Return error from of_spi_register_master on bad "cs-gpios" property This makes sure that an error is returned on an incorrectly formed "cs-gpios" property, but reports success when the "cs-gpios" property is well formed or missing. When holes in the cs-gpios property phandle list is used to indicate that some other form of chipselect is to be used it is important that failure to read a broken "cs-gpios" property does not silently fail leading to the spi controller to use an unintended chipselect. Signed-off-by: Andreas Larsson Signed-off-by: Grant Likely --- drivers/spi/spi.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/spi/spi.c b/drivers/spi/spi.c index bcca66c5107f..3738e7cbff33 100644 --- a/drivers/spi/spi.c +++ b/drivers/spi/spi.c @@ -1067,8 +1067,11 @@ static int of_spi_register_master(struct spi_master *master) nb = of_gpio_named_count(np, "cs-gpios"); master->num_chipselect = max(nb, (int)master->num_chipselect); - if (nb < 1) + /* Return error only for an incorrectly formed cs-gpios property */ + if (nb == 0 || nb == -ENOENT) return 0; + else if (nb < 0) + return nb; cs = devm_kzalloc(&master->dev, sizeof(int) * master->num_chipselect, From 0d2d0cc5216db678070efc911ef47f4cb53b4aad Mon Sep 17 00:00:00 2001 From: "Manjunathappa, Prakash" Date: Mon, 25 Feb 2013 16:14:07 +0530 Subject: [PATCH 004/198] spi/davinci: fix module build error Fix below module build error: CC [M] drivers/spi/spi-davinci.o drivers/spi/spi-davinci.c:774:1: error: '__mod_of_device_table' aliased to undefined symbol 'davini_spi_of_match' drivers/spi/spi-davinci.c:774:1: error: '__mod_of_device_table' aliased to undefined symbol 'davini_spi_of_match' make[2]: *** [drivers/spi/spi-davinci.o] Error 1 make[1]: *** [drivers/spi] Error 2 make[1]: *** Waiting for unfinished jobs.... make: *** [drivers] Error 2 Signed-off-by: Manjunathappa, Prakash Acked-by: Sekhar Nori Signed-off-by: Grant Likely --- drivers/spi/spi-davinci.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/spi/spi-davinci.c b/drivers/spi/spi-davinci.c index 8234d2259722..6287c8315d0d 100644 --- a/drivers/spi/spi-davinci.c +++ b/drivers/spi/spi-davinci.c @@ -784,7 +784,7 @@ static const struct of_device_id davinci_spi_of_match[] = { }, { }, }; -MODULE_DEVICE_TABLE(of, davini_spi_of_match); +MODULE_DEVICE_TABLE(of, davinci_spi_of_match); /** * spi_davinci_get_pdata - Get platform data from DTS binding From 49717cb40410fe4b563968680ff7c513967504c6 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Thu, 11 Apr 2013 08:07:11 -0700 Subject: [PATCH 005/198] kthread: Document ways of reducing OS jitter due to per-CPU kthreads The Linux kernel uses a number of per-CPU kthreads, any of which might contribute to OS jitter at any time. The usual approach to normal kthreads, namely to bind them to a "housekeeping" CPU, does not work with these kthreads because they cannot operate correctly if moved to some other CPU. This commit therefore lists ways of controlling OS jitter from the Linux kernel's per-CPU kthreads. It also lists some ways of diagnosing excessive jitter. Signed-off-by: Paul E. McKenney Cc: Frederic Weisbecker Cc: Steven Rostedt Cc: Borislav Petkov Cc: Arjan van de Ven Cc: Kevin Hilman Cc: Christoph Lameter Cc: Thomas Gleixner Cc: Olivier Baetz Cc: Pradeep Satyanarayana Reviewed-by: Randy Dunlap Reviewed-by: Borislav Petkov --- Documentation/kernel-per-CPU-kthreads.txt | 202 ++++++++++++++++++++++ 1 file changed, 202 insertions(+) create mode 100644 Documentation/kernel-per-CPU-kthreads.txt diff --git a/Documentation/kernel-per-CPU-kthreads.txt b/Documentation/kernel-per-CPU-kthreads.txt new file mode 100644 index 000000000000..cbf7ae412da4 --- /dev/null +++ b/Documentation/kernel-per-CPU-kthreads.txt @@ -0,0 +1,202 @@ +REDUCING OS JITTER DUE TO PER-CPU KTHREADS + +This document lists per-CPU kthreads in the Linux kernel and presents +options to control their OS jitter. Note that non-per-CPU kthreads are +not listed here. To reduce OS jitter from non-per-CPU kthreads, bind +them to a "housekeeping" CPU dedicated to such work. + + +REFERENCES + +o Documentation/IRQ-affinity.txt: Binding interrupts to sets of CPUs. + +o Documentation/cgroups: Using cgroups to bind tasks to sets of CPUs. + +o man taskset: Using the taskset command to bind tasks to sets + of CPUs. + +o man sched_setaffinity: Using the sched_setaffinity() system + call to bind tasks to sets of CPUs. + +o /sys/devices/system/cpu/cpuN/online: Control CPU N's hotplug state, + writing "0" to offline and "1" to online. + +o In order to locate kernel-generated OS jitter on CPU N: + + cd /sys/kernel/debug/tracing + echo 1 > max_graph_depth # Increase the "1" for more detail + echo function_graph > current_tracer + # run workload + cat per_cpu/cpuN/trace + + +KTHREADS + +Name: ehca_comp/%u +Purpose: Periodically process Infiniband-related work. +To reduce its OS jitter, do any of the following: +1. Don't use eHCA Infiniband hardware, instead choosing hardware + that does not require per-CPU kthreads. This will prevent these + kthreads from being created in the first place. (This will + work for most people, as this hardware, though important, is + relatively old and is produced in relatively low unit volumes.) +2. Do all eHCA-Infiniband-related work on other CPUs, including + interrupts. +3. Rework the eHCA driver so that its per-CPU kthreads are + provisioned only on selected CPUs. + + +Name: irq/%d-%s +Purpose: Handle threaded interrupts. +To reduce its OS jitter, do the following: +1. Use irq affinity to force the irq threads to execute on + some other CPU. + +Name: kcmtpd_ctr_%d +Purpose: Handle Bluetooth work. +To reduce its OS jitter, do one of the following: +1. Don't use Bluetooth, in which case these kthreads won't be + created in the first place. +2. Use irq affinity to force Bluetooth-related interrupts to + occur on some other CPU and furthermore initiate all + Bluetooth activity on some other CPU. + +Name: ksoftirqd/%u +Purpose: Execute softirq handlers when threaded or when under heavy load. +To reduce its OS jitter, each softirq vector must be handled +separately as follows: +TIMER_SOFTIRQ: Do all of the following: +1. To the extent possible, keep the CPU out of the kernel when it + is non-idle, for example, by avoiding system calls and by forcing + both kernel threads and interrupts to execute elsewhere. +2. Build with CONFIG_HOTPLUG_CPU=y. After boot completes, force + the CPU offline, then bring it back online. This forces + recurring timers to migrate elsewhere. If you are concerned + with multiple CPUs, force them all offline before bringing the + first one back online. Once you have onlined the CPUs in question, + do not offline any other CPUs, because doing so could force the + timer back onto one of the CPUs in question. +NET_TX_SOFTIRQ and NET_RX_SOFTIRQ: Do all of the following: +1. Force networking interrupts onto other CPUs. +2. Initiate any network I/O on other CPUs. +3. Once your application has started, prevent CPU-hotplug operations + from being initiated from tasks that might run on the CPU to + be de-jittered. (It is OK to force this CPU offline and then + bring it back online before you start your application.) +BLOCK_SOFTIRQ: Do all of the following: +1. Force block-device interrupts onto some other CPU. +2. Initiate any block I/O on other CPUs. +3. Once your application has started, prevent CPU-hotplug operations + from being initiated from tasks that might run on the CPU to + be de-jittered. (It is OK to force this CPU offline and then + bring it back online before you start your application.) +BLOCK_IOPOLL_SOFTIRQ: Do all of the following: +1. Force block-device interrupts onto some other CPU. +2. Initiate any block I/O and block-I/O polling on other CPUs. +3. Once your application has started, prevent CPU-hotplug operations + from being initiated from tasks that might run on the CPU to + be de-jittered. (It is OK to force this CPU offline and then + bring it back online before you start your application.) +TASKLET_SOFTIRQ: Do one or more of the following: +1. Avoid use of drivers that use tasklets. (Such drivers will contain + calls to things like tasklet_schedule().) +2. Convert all drivers that you must use from tasklets to workqueues. +3. Force interrupts for drivers using tasklets onto other CPUs, + and also do I/O involving these drivers on other CPUs. +SCHED_SOFTIRQ: Do all of the following: +1. Avoid sending scheduler IPIs to the CPU to be de-jittered, + for example, ensure that at most one runnable kthread is present + on that CPU. If a thread that expects to run on the de-jittered + CPU awakens, the scheduler will send an IPI that can result in + a subsequent SCHED_SOFTIRQ. +2. Build with CONFIG_RCU_NOCB_CPU=y, CONFIG_RCU_NOCB_CPU_ALL=y, + CONFIG_NO_HZ_FULL=y, and, in addition, ensure that the CPU + to be de-jittered is marked as an adaptive-ticks CPU using the + "nohz_full=" boot parameter. This reduces the number of + scheduler-clock interrupts that the de-jittered CPU receives, + minimizing its chances of being selected to do the load balancing + work that runs in SCHED_SOFTIRQ context. +3. To the extent possible, keep the CPU out of the kernel when it + is non-idle, for example, by avoiding system calls and by + forcing both kernel threads and interrupts to execute elsewhere. + This further reduces the number of scheduler-clock interrupts + received by the de-jittered CPU. +HRTIMER_SOFTIRQ: Do all of the following: +1. To the extent possible, keep the CPU out of the kernel when it + is non-idle. For example, avoid system calls and force both + kernel threads and interrupts to execute elsewhere. +2. Build with CONFIG_HOTPLUG_CPU=y. Once boot completes, force the + CPU offline, then bring it back online. This forces recurring + timers to migrate elsewhere. If you are concerned with multiple + CPUs, force them all offline before bringing the first one + back online. Once you have onlined the CPUs in question, do not + offline any other CPUs, because doing so could force the timer + back onto one of the CPUs in question. +RCU_SOFTIRQ: Do at least one of the following: +1. Offload callbacks and keep the CPU in either dyntick-idle or + adaptive-ticks state by doing all of the following: + a. Build with CONFIG_RCU_NOCB_CPU=y, CONFIG_RCU_NOCB_CPU_ALL=y, + CONFIG_NO_HZ_FULL=y, and, in addition ensure that the CPU + to be de-jittered is marked as an adaptive-ticks CPU using + the "nohz_full=" boot parameter. Bind the rcuo kthreads + to housekeeping CPUs, which can tolerate OS jitter. + b. To the extent possible, keep the CPU out of the kernel + when it is non-idle, for example, by avoiding system + calls and by forcing both kernel threads and interrupts + to execute elsewhere. +2. Enable RCU to do its processing remotely via dyntick-idle by + doing all of the following: + a. Build with CONFIG_NO_HZ=y and CONFIG_RCU_FAST_NO_HZ=y. + b. Ensure that the CPU goes idle frequently, allowing other + CPUs to detect that it has passed through an RCU quiescent + state. If the kernel is built with CONFIG_NO_HZ_FULL=y, + userspace execution also allows other CPUs to detect that + the CPU in question has passed through a quiescent state. + c. To the extent possible, keep the CPU out of the kernel + when it is non-idle, for example, by avoiding system + calls and by forcing both kernel threads and interrupts + to execute elsewhere. + +Name: rcuc/%u +Purpose: Execute RCU callbacks in CONFIG_RCU_BOOST=y kernels. +To reduce its OS jitter, do at least one of the following: +1. Build the kernel with CONFIG_PREEMPT=n. This prevents these + kthreads from being created in the first place, and also obviates + the need for RCU priority boosting. This approach is feasible + for workloads that do not require high degrees of responsiveness. +2. Build the kernel with CONFIG_RCU_BOOST=n. This prevents these + kthreads from being created in the first place. This approach + is feasible only if your workload never requires RCU priority + boosting, for example, if you ensure frequent idle time on all + CPUs that might execute within the kernel. +3. Build with CONFIG_RCU_NOCB_CPU=y and CONFIG_RCU_NOCB_CPU_ALL=y, + which offloads all RCU callbacks to kthreads that can be moved + off of CPUs susceptible to OS jitter. This approach prevents the + rcuc/%u kthreads from having any work to do, so that they are + never awakened. +4. Ensure that the CPU never enters the kernel, and, in particular, + avoid initiating any CPU hotplug operations on this CPU. This is + another way of preventing any callbacks from being queued on the + CPU, again preventing the rcuc/%u kthreads from having any work + to do. + +Name: rcuob/%d, rcuop/%d, and rcuos/%d +Purpose: Offload RCU callbacks from the corresponding CPU. +To reduce its OS jitter, do at least one of the following: +1. Use affinity, cgroups, or other mechanism to force these kthreads + to execute on some other CPU. +2. Build with CONFIG_RCU_NOCB_CPUS=n, which will prevent these + kthreads from being created in the first place. However, please + note that this will not eliminate OS jitter, but will instead + shift it to RCU_SOFTIRQ. + +Name: watchdog/%u +Purpose: Detect software lockups on each CPU. +To reduce its OS jitter, do at least one of the following: +1. Build with CONFIG_LOCKUP_DETECTOR=n, which will prevent these + kthreads from being created in the first place. +2. Echo a zero to /proc/sys/kernel/watchdog to disable the + watchdog timer. +3. Echo a large number of /proc/sys/kernel/watchdog_thresh in + order to reduce the frequency of OS jitter due to the watchdog + timer down to a level that is acceptable for your workload. From 4f42f4ef0dba0cf837b628ac33231a7acb8fcc9d Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Fri, 26 Apr 2013 16:22:46 +0300 Subject: [PATCH 006/198] drm/i915: Always normalize return timeout for wait_timeout_ioctl MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit As we recompute the remaining timeout after waiting, there is a potential for that timeout to be less than zero and so need sanitizing. The timeout is always returned to userspace and validated, so we should always perform the sanitation. v2 [vsyrjala]: Only normalize the timespec if it's invalid v3: Add a comment to clarify the situation and remove the now useless WARN_ON() (ickle) Cc: Ben Widawsky Signed-off-by: Chris Wilson Signed-off-by: Ville Syrjälä Tested-by: Ben Widawsky Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_gem.c | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 6be940effefd..6165535d15f0 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -1045,6 +1045,8 @@ static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno, if (timeout) { struct timespec sleep_time = timespec_sub(now, before); *timeout = timespec_sub(*timeout, sleep_time); + if (!timespec_valid(timeout)) /* i.e. negative time remains */ + set_normalized_timespec(timeout, 0, 0); } switch (end) { @@ -1053,8 +1055,6 @@ static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno, case -ERESTARTSYS: /* Signal */ return (int)end; case 0: /* Timeout */ - if (timeout) - set_normalized_timespec(timeout, 0, 0); return -ETIME; default: /* Completed */ WARN_ON(end < 0); /* We're not aware of other errors */ @@ -2377,10 +2377,8 @@ i915_gem_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file) mutex_unlock(&dev->struct_mutex); ret = __wait_seqno(ring, seqno, reset_counter, true, timeout); - if (timeout) { - WARN_ON(!timespec_valid(timeout)); + if (timeout) args->timeout_ns = timespec_to_ns(timeout); - } return ret; out: From 3a359f0b21ab218c1bf7a6a1b638b6fd143d0b99 Mon Sep 17 00:00:00 2001 From: Daniel Vetter Date: Sat, 20 Apr 2013 12:08:11 +0200 Subject: [PATCH 007/198] drm/mm: fix dump table BUG In commit 9e8944ab564f2e3dde90a518cd32048c58918608 Author: Chris Wilson Date: Thu Nov 15 11:32:17 2012 +0000 drm: Introduce an iterator over holes in the drm_mm range manager helpers and iterators for hole handling have been introduced with some debug BUG_ONs sprinkled over. Unfortunately this broke the mm dumper which unconditionally tried to compute the size of the very first hole. While at it unify the code a bit with the hole dumping in the loop. v2: Extract a hole dump helper. Reported-by: Christopher Harvey Cc: Christopher Harvey Cc: Dave Airlie Cc: Chris Wilson Cc: stable@vger.kernel.org Acked-by: Dave Airlie Signed-off-by: Daniel Vetter --- drivers/gpu/drm/drm_mm.c | 34 ++++++++++++++++++---------------- 1 file changed, 18 insertions(+), 16 deletions(-) diff --git a/drivers/gpu/drm/drm_mm.c b/drivers/gpu/drm/drm_mm.c index db1e2d6f90d7..07cf99cc8862 100644 --- a/drivers/gpu/drm/drm_mm.c +++ b/drivers/gpu/drm/drm_mm.c @@ -755,33 +755,35 @@ void drm_mm_debug_table(struct drm_mm *mm, const char *prefix) EXPORT_SYMBOL(drm_mm_debug_table); #if defined(CONFIG_DEBUG_FS) +static unsigned long drm_mm_dump_hole(struct seq_file *m, struct drm_mm_node *entry) +{ + unsigned long hole_start, hole_end, hole_size; + + if (entry->hole_follows) { + hole_start = drm_mm_hole_node_start(entry); + hole_end = drm_mm_hole_node_end(entry); + hole_size = hole_end - hole_start; + seq_printf(m, "0x%08lx-0x%08lx: 0x%08lx: free\n", + hole_start, hole_end, hole_size); + return hole_size; + } + + return 0; +} + int drm_mm_dump_table(struct seq_file *m, struct drm_mm *mm) { struct drm_mm_node *entry; unsigned long total_used = 0, total_free = 0, total = 0; - unsigned long hole_start, hole_end, hole_size; - hole_start = drm_mm_hole_node_start(&mm->head_node); - hole_end = drm_mm_hole_node_end(&mm->head_node); - hole_size = hole_end - hole_start; - if (hole_size) - seq_printf(m, "0x%08lx-0x%08lx: 0x%08lx: free\n", - hole_start, hole_end, hole_size); - total_free += hole_size; + total_free += drm_mm_dump_hole(m, &mm->head_node); drm_mm_for_each_node(entry, mm) { seq_printf(m, "0x%08lx-0x%08lx: 0x%08lx: used\n", entry->start, entry->start + entry->size, entry->size); total_used += entry->size; - if (entry->hole_follows) { - hole_start = drm_mm_hole_node_start(entry); - hole_end = drm_mm_hole_node_end(entry); - hole_size = hole_end - hole_start; - seq_printf(m, "0x%08lx-0x%08lx: 0x%08lx: free\n", - hole_start, hole_end, hole_size); - total_free += hole_size; - } + total_free += drm_mm_dump_hole(m, entry); } total = total_free + total_used; From 51cea1f469ad473c8d8b7d4a227640b8c02bf167 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 21 Mar 2013 13:10:44 +0200 Subject: [PATCH 008/198] drm/i915: Fix pipe enabled mask for pipe C in WM calculations MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fix the incorrect enabled pipes mask for pipe C in the WM calculations. Additionally, in an effort to make the code easier to understand, populate the mask with 1 << PIPE_[ABC] instead of raw numbers. v2: Use 1 << PIPE_[ABC] (ickle/danvet) v3: Pass PIPE_[ABC] to g4x_compute_wm0() (ickle) Signed-off-by: Ville Syrjälä Reviewed-by: Chris Wilson Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 44 ++++++++++++++++----------------- 1 file changed, 22 insertions(+), 22 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index de3b0dc5658b..aa01128ff192 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -1301,17 +1301,17 @@ static void valleyview_update_wm(struct drm_device *dev) vlv_update_drain_latency(dev); - if (g4x_compute_wm0(dev, 0, + if (g4x_compute_wm0(dev, PIPE_A, &valleyview_wm_info, latency_ns, &valleyview_cursor_wm_info, latency_ns, &planea_wm, &cursora_wm)) - enabled |= 1; + enabled |= 1 << PIPE_A; - if (g4x_compute_wm0(dev, 1, + if (g4x_compute_wm0(dev, PIPE_B, &valleyview_wm_info, latency_ns, &valleyview_cursor_wm_info, latency_ns, &planeb_wm, &cursorb_wm)) - enabled |= 2; + enabled |= 1 << PIPE_B; if (single_plane_enabled(enabled) && g4x_compute_srwm(dev, ffs(enabled) - 1, @@ -1357,17 +1357,17 @@ static void g4x_update_wm(struct drm_device *dev) int plane_sr, cursor_sr; unsigned int enabled = 0; - if (g4x_compute_wm0(dev, 0, + if (g4x_compute_wm0(dev, PIPE_A, &g4x_wm_info, latency_ns, &g4x_cursor_wm_info, latency_ns, &planea_wm, &cursora_wm)) - enabled |= 1; + enabled |= 1 << PIPE_A; - if (g4x_compute_wm0(dev, 1, + if (g4x_compute_wm0(dev, PIPE_B, &g4x_wm_info, latency_ns, &g4x_cursor_wm_info, latency_ns, &planeb_wm, &cursorb_wm)) - enabled |= 2; + enabled |= 1 << PIPE_B; if (single_plane_enabled(enabled) && g4x_compute_srwm(dev, ffs(enabled) - 1, @@ -1716,7 +1716,7 @@ static void ironlake_update_wm(struct drm_device *dev) unsigned int enabled; enabled = 0; - if (g4x_compute_wm0(dev, 0, + if (g4x_compute_wm0(dev, PIPE_A, &ironlake_display_wm_info, ILK_LP0_PLANE_LATENCY, &ironlake_cursor_wm_info, @@ -1727,10 +1727,10 @@ static void ironlake_update_wm(struct drm_device *dev) DRM_DEBUG_KMS("FIFO watermarks For pipe A -" " plane %d, " "cursor: %d\n", plane_wm, cursor_wm); - enabled |= 1; + enabled |= 1 << PIPE_A; } - if (g4x_compute_wm0(dev, 1, + if (g4x_compute_wm0(dev, PIPE_B, &ironlake_display_wm_info, ILK_LP0_PLANE_LATENCY, &ironlake_cursor_wm_info, @@ -1741,7 +1741,7 @@ static void ironlake_update_wm(struct drm_device *dev) DRM_DEBUG_KMS("FIFO watermarks For pipe B -" " plane %d, cursor: %d\n", plane_wm, cursor_wm); - enabled |= 2; + enabled |= 1 << PIPE_B; } /* @@ -1801,7 +1801,7 @@ static void sandybridge_update_wm(struct drm_device *dev) unsigned int enabled; enabled = 0; - if (g4x_compute_wm0(dev, 0, + if (g4x_compute_wm0(dev, PIPE_A, &sandybridge_display_wm_info, latency, &sandybridge_cursor_wm_info, latency, &plane_wm, &cursor_wm)) { @@ -1812,10 +1812,10 @@ static void sandybridge_update_wm(struct drm_device *dev) DRM_DEBUG_KMS("FIFO watermarks For pipe A -" " plane %d, " "cursor: %d\n", plane_wm, cursor_wm); - enabled |= 1; + enabled |= 1 << PIPE_A; } - if (g4x_compute_wm0(dev, 1, + if (g4x_compute_wm0(dev, PIPE_B, &sandybridge_display_wm_info, latency, &sandybridge_cursor_wm_info, latency, &plane_wm, &cursor_wm)) { @@ -1826,7 +1826,7 @@ static void sandybridge_update_wm(struct drm_device *dev) DRM_DEBUG_KMS("FIFO watermarks For pipe B -" " plane %d, cursor: %d\n", plane_wm, cursor_wm); - enabled |= 2; + enabled |= 1 << PIPE_B; } /* @@ -1904,7 +1904,7 @@ static void ivybridge_update_wm(struct drm_device *dev) unsigned int enabled; enabled = 0; - if (g4x_compute_wm0(dev, 0, + if (g4x_compute_wm0(dev, PIPE_A, &sandybridge_display_wm_info, latency, &sandybridge_cursor_wm_info, latency, &plane_wm, &cursor_wm)) { @@ -1915,10 +1915,10 @@ static void ivybridge_update_wm(struct drm_device *dev) DRM_DEBUG_KMS("FIFO watermarks For pipe A -" " plane %d, " "cursor: %d\n", plane_wm, cursor_wm); - enabled |= 1; + enabled |= 1 << PIPE_A; } - if (g4x_compute_wm0(dev, 1, + if (g4x_compute_wm0(dev, PIPE_B, &sandybridge_display_wm_info, latency, &sandybridge_cursor_wm_info, latency, &plane_wm, &cursor_wm)) { @@ -1929,10 +1929,10 @@ static void ivybridge_update_wm(struct drm_device *dev) DRM_DEBUG_KMS("FIFO watermarks For pipe B -" " plane %d, cursor: %d\n", plane_wm, cursor_wm); - enabled |= 2; + enabled |= 1 << PIPE_B; } - if (g4x_compute_wm0(dev, 2, + if (g4x_compute_wm0(dev, PIPE_C, &sandybridge_display_wm_info, latency, &sandybridge_cursor_wm_info, latency, &plane_wm, &cursor_wm)) { @@ -1943,7 +1943,7 @@ static void ivybridge_update_wm(struct drm_device *dev) DRM_DEBUG_KMS("FIFO watermarks For pipe C -" " plane %d, cursor: %d\n", plane_wm, cursor_wm); - enabled |= 3; + enabled |= 1 << PIPE_C; } /* From f557c98b168a2860bfc7dedf4b4e3bafb59dc267 Mon Sep 17 00:00:00 2001 From: Richard Genoud Date: Thu, 2 May 2013 19:25:11 +0800 Subject: [PATCH 009/198] spi/spi-atmel: BUG: fix doesn' support 16 bits transfers using PIO Fix using PIO transfer mode only support 8 bits transfer, doesn't support 16 bits. Signed-off-by: Richard Genoud [wenyou.yang@atmel.com: submit the patch] Signed-off-by: Wenyou Yang Signed-off-by: Mark Brown --- drivers/spi/spi-atmel.c | 47 +++++++++++++++++++++++++++++++++-------- 1 file changed, 38 insertions(+), 9 deletions(-) diff --git a/drivers/spi/spi-atmel.c b/drivers/spi/spi-atmel.c index 787bd2c22bca..d8cb7da65efe 100644 --- a/drivers/spi/spi-atmel.c +++ b/drivers/spi/spi-atmel.c @@ -526,13 +526,17 @@ static void atmel_spi_next_xfer_pio(struct spi_master *master, } if (xfer->tx_buf) - spi_writel(as, TDR, *(u8 *)(xfer->tx_buf)); + if (xfer->bits_per_word > 8) + spi_writel(as, TDR, *(u16 *)(xfer->tx_buf)); + else + spi_writel(as, TDR, *(u8 *)(xfer->tx_buf)); else spi_writel(as, TDR, 0); dev_dbg(master->dev.parent, - " start pio xfer %p: len %u tx %p rx %p\n", - xfer, xfer->len, xfer->tx_buf, xfer->rx_buf); + " start pio xfer %p: len %u tx %p rx %p bitpw %d\n", + xfer, xfer->len, xfer->tx_buf, xfer->rx_buf, + xfer->bits_per_word); /* Enable relevant interrupts */ spi_writel(as, IER, SPI_BIT(RDRF) | SPI_BIT(OVRES)); @@ -950,21 +954,39 @@ atmel_spi_pump_pio_data(struct atmel_spi *as, struct spi_transfer *xfer) { u8 *txp; u8 *rxp; + u16 *txp16; + u16 *rxp16; unsigned long xfer_pos = xfer->len - as->current_remaining_bytes; if (xfer->rx_buf) { - rxp = ((u8 *)xfer->rx_buf) + xfer_pos; - *rxp = spi_readl(as, RDR); + if (xfer->bits_per_word > 8) { + rxp16 = (u16 *)(((u8 *)xfer->rx_buf) + xfer_pos); + *rxp16 = spi_readl(as, RDR); + } else { + rxp = ((u8 *)xfer->rx_buf) + xfer_pos; + *rxp = spi_readl(as, RDR); + } } else { spi_readl(as, RDR); } - - as->current_remaining_bytes--; + if (xfer->bits_per_word > 8) { + as->current_remaining_bytes -= 2; + if (as->current_remaining_bytes < 0) + as->current_remaining_bytes = 0; + } else { + as->current_remaining_bytes--; + } if (as->current_remaining_bytes) { if (xfer->tx_buf) { - txp = ((u8 *)xfer->tx_buf) + xfer_pos + 1; - spi_writel(as, TDR, *txp); + if (xfer->bits_per_word > 8) { + txp16 = (u16 *)(((u8 *)xfer->tx_buf) + + xfer_pos + 2); + spi_writel(as, TDR, *txp16); + } else { + txp = ((u8 *)xfer->tx_buf) + xfer_pos + 1; + spi_writel(as, TDR, *txp); + } } else { spi_writel(as, TDR, 0); } @@ -1378,6 +1400,13 @@ static int atmel_spi_transfer(struct spi_device *spi, struct spi_message *msg) } } + if (xfer->bits_per_word > 8) { + if (xfer->len % 2) { + dev_dbg(&spi->dev, "buffer len should be 16 bits aligned\n"); + return -EINVAL; + } + } + /* FIXME implement these protocol options!! */ if (xfer->speed_hz) { dev_dbg(&spi->dev, "no protocol options yet\n"); From a9b054e8ab06504c2afa0e307ee78d3778993a1d Mon Sep 17 00:00:00 2001 From: Daniel Vetter Date: Thu, 2 May 2013 09:43:05 +0200 Subject: [PATCH 010/198] drm: don't check modeset locks in panic handler Since we know that locking is broken in that case and it's more important to not flood the dmesg with random gunk. Cc: Dave Airlie Cc: Borislav Petkov References: http://lkml.kernel.org/r/20130502000206.GH15623@pd.tnic Cc: stable@vger.kernel.org Reported-and-tested-by: Borislav Petkov Signed-off-by: Daniel Vetter --- drivers/gpu/drm/drm_crtc.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/gpu/drm/drm_crtc.c b/drivers/gpu/drm/drm_crtc.c index 792c3e3795ca..3be0802c6797 100644 --- a/drivers/gpu/drm/drm_crtc.c +++ b/drivers/gpu/drm/drm_crtc.c @@ -78,6 +78,10 @@ void drm_warn_on_modeset_not_all_locked(struct drm_device *dev) { struct drm_crtc *crtc; + /* Locking is currently fubar in the panic handler. */ + if (oops_in_progress) + return; + list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) WARN_ON(!mutex_is_locked(&crtc->mutex)); From e30b5dca15dea86aa697f9d58ff646294fe80d3d Mon Sep 17 00:00:00 2001 From: "Yan, Zheng" Date: Fri, 3 May 2013 02:15:52 -0400 Subject: [PATCH 011/198] ext4: fix fio regression We (Linux Kernel Performance project) found a regression introduced by commit: f7fec032aa ext4: track all extent status in extent status tree The commit causes about 20% performance decrease in fio random write test. Profiler shows that rb_next() uses a lot of CPU time. The call stack is: rb_next ext4_es_find_delayed_extent ext4_map_blocks _ext4_get_block ext4_get_block_write __blockdev_direct_IO ext4_direct_IO generic_file_direct_write __generic_file_aio_write ext4_file_write aio_rw_vect_retry aio_run_iocb do_io_submit sys_io_submit system_call_fastpath io_submit td_io_getevents io_u_queued_complete thread_main main __libc_start_main The cause is that ext4_es_find_delayed_extent() doesn't have an upper bound, it keeps searching until a delayed extent is found. When there are a lots of non-delayed entries in the extent state tree, ext4_es_find_delayed_extent() may uses a lot of CPU time. Reported-by: LKP project Signed-off-by: Yan, Zheng Signed-off-by: Zheng Liu Cc: "Theodore Ts'o" --- fs/ext4/extents.c | 9 +++++---- fs/ext4/extents_status.c | 17 ++++++++++++----- fs/ext4/extents_status.h | 3 ++- fs/ext4/file.c | 4 ++-- include/trace/events/ext4.h | 4 ++-- 5 files changed, 23 insertions(+), 14 deletions(-) diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 107936db244e..bc0f1910b9cf 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -3642,7 +3642,7 @@ int ext4_find_delalloc_range(struct inode *inode, { struct extent_status es; - ext4_es_find_delayed_extent(inode, lblk_start, &es); + ext4_es_find_delayed_extent_range(inode, lblk_start, lblk_end, &es); if (es.es_len == 0) return 0; /* there is no delay extent in this tree */ else if (es.es_lblk <= lblk_start && @@ -4608,9 +4608,10 @@ static int ext4_find_delayed_extent(struct inode *inode, struct extent_status es; ext4_lblk_t block, next_del; - ext4_es_find_delayed_extent(inode, newes->es_lblk, &es); - if (newes->es_pblk == 0) { + ext4_es_find_delayed_extent_range(inode, newes->es_lblk, + newes->es_lblk + newes->es_len - 1, &es); + /* * No extent in extent-tree contains block @newes->es_pblk, * then the block may stay in 1)a hole or 2)delayed-extent. @@ -4630,7 +4631,7 @@ static int ext4_find_delayed_extent(struct inode *inode, } block = newes->es_lblk + newes->es_len; - ext4_es_find_delayed_extent(inode, block, &es); + ext4_es_find_delayed_extent_range(inode, block, EXT_MAX_BLOCKS, &es); if (es.es_len == 0) next_del = EXT_MAX_BLOCKS; else diff --git a/fs/ext4/extents_status.c b/fs/ext4/extents_status.c index fe3337a85ede..e6941e622d31 100644 --- a/fs/ext4/extents_status.c +++ b/fs/ext4/extents_status.c @@ -232,14 +232,16 @@ static struct extent_status *__es_tree_search(struct rb_root *root, } /* - * ext4_es_find_delayed_extent: find the 1st delayed extent covering @es->lblk - * if it exists, otherwise, the next extent after @es->lblk. + * ext4_es_find_delayed_extent_range: find the 1st delayed extent covering + * @es->lblk if it exists, otherwise, the next extent after @es->lblk. * * @inode: the inode which owns delayed extents * @lblk: the offset where we start to search + * @end: the offset where we stop to search * @es: delayed extent that we found */ -void ext4_es_find_delayed_extent(struct inode *inode, ext4_lblk_t lblk, +void ext4_es_find_delayed_extent_range(struct inode *inode, + ext4_lblk_t lblk, ext4_lblk_t end, struct extent_status *es) { struct ext4_es_tree *tree = NULL; @@ -247,7 +249,8 @@ void ext4_es_find_delayed_extent(struct inode *inode, ext4_lblk_t lblk, struct rb_node *node; BUG_ON(es == NULL); - trace_ext4_es_find_delayed_extent_enter(inode, lblk); + BUG_ON(end < lblk); + trace_ext4_es_find_delayed_extent_range_enter(inode, lblk); read_lock(&EXT4_I(inode)->i_es_lock); tree = &EXT4_I(inode)->i_es_tree; @@ -270,6 +273,10 @@ out: if (es1 && !ext4_es_is_delayed(es1)) { while ((node = rb_next(&es1->rb_node)) != NULL) { es1 = rb_entry(node, struct extent_status, rb_node); + if (es1->es_lblk > end) { + es1 = NULL; + break; + } if (ext4_es_is_delayed(es1)) break; } @@ -285,7 +292,7 @@ out: read_unlock(&EXT4_I(inode)->i_es_lock); ext4_es_lru_add(inode); - trace_ext4_es_find_delayed_extent_exit(inode, es); + trace_ext4_es_find_delayed_extent_range_exit(inode, es); } static struct extent_status * diff --git a/fs/ext4/extents_status.h b/fs/ext4/extents_status.h index d8e2d4dc311e..f740eb03b707 100644 --- a/fs/ext4/extents_status.h +++ b/fs/ext4/extents_status.h @@ -62,7 +62,8 @@ extern int ext4_es_insert_extent(struct inode *inode, ext4_lblk_t lblk, unsigned long long status); extern int ext4_es_remove_extent(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t len); -extern void ext4_es_find_delayed_extent(struct inode *inode, ext4_lblk_t lblk, +extern void ext4_es_find_delayed_extent_range(struct inode *inode, + ext4_lblk_t lblk, ext4_lblk_t end, struct extent_status *es); extern int ext4_es_lookup_extent(struct inode *inode, ext4_lblk_t lblk, struct extent_status *es); diff --git a/fs/ext4/file.c b/fs/ext4/file.c index 64848b595b24..1bc4523541d6 100644 --- a/fs/ext4/file.c +++ b/fs/ext4/file.c @@ -464,7 +464,7 @@ static loff_t ext4_seek_data(struct file *file, loff_t offset, loff_t maxsize) * If there is a delay extent at this offset, * it will be as a data. */ - ext4_es_find_delayed_extent(inode, last, &es); + ext4_es_find_delayed_extent_range(inode, last, last, &es); if (es.es_len != 0 && in_range(last, es.es_lblk, es.es_len)) { if (last != start) dataoff = last << blkbits; @@ -547,7 +547,7 @@ static loff_t ext4_seek_hole(struct file *file, loff_t offset, loff_t maxsize) * If there is a delay extent at this offset, * we will skip this extent. */ - ext4_es_find_delayed_extent(inode, last, &es); + ext4_es_find_delayed_extent_range(inode, last, last, &es); if (es.es_len != 0 && in_range(last, es.es_lblk, es.es_len)) { last = es.es_lblk + es.es_len; holeoff = last << blkbits; diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h index d0e686402df8..8ee15b97cd38 100644 --- a/include/trace/events/ext4.h +++ b/include/trace/events/ext4.h @@ -2139,7 +2139,7 @@ TRACE_EVENT(ext4_es_remove_extent, __entry->lblk, __entry->len) ); -TRACE_EVENT(ext4_es_find_delayed_extent_enter, +TRACE_EVENT(ext4_es_find_delayed_extent_range_enter, TP_PROTO(struct inode *inode, ext4_lblk_t lblk), TP_ARGS(inode, lblk), @@ -2161,7 +2161,7 @@ TRACE_EVENT(ext4_es_find_delayed_extent_enter, (unsigned long) __entry->ino, __entry->lblk) ); -TRACE_EVENT(ext4_es_find_delayed_extent_exit, +TRACE_EVENT(ext4_es_find_delayed_extent_range_exit, TP_PROTO(struct inode *inode, struct extent_status *es), TP_ARGS(inode, es), From 4ef69d0394cba8caa9f75d3f2e53429bfb8b3045 Mon Sep 17 00:00:00 2001 From: Felix Fietkau Date: Sat, 27 Apr 2013 11:47:01 +0200 Subject: [PATCH 012/198] ath9k: fix key allocation error handling for powersave keys If no keycache slots are available, ath_key_config can return -ENOSPC. If the key index is not checked for errors, it can lead to logspam that looks like this: "ath: wiphy0: keyreset: keycache entry 228 out of range" This can cause follow-up errors if the invalid keycache index gets used for tx. Cc: stable@vger.kernel.org Signed-off-by: Felix Fietkau Signed-off-by: John W. Linville --- drivers/net/wireless/ath/ath9k/main.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/net/wireless/ath/ath9k/main.c b/drivers/net/wireless/ath/ath9k/main.c index 6963862a1872..c4fc3d14ea62 100644 --- a/drivers/net/wireless/ath/ath9k/main.c +++ b/drivers/net/wireless/ath/ath9k/main.c @@ -1332,6 +1332,7 @@ static int ath9k_sta_add(struct ieee80211_hw *hw, struct ath_common *common = ath9k_hw_common(sc->sc_ah); struct ath_node *an = (struct ath_node *) sta->drv_priv; struct ieee80211_key_conf ps_key = { }; + int key; ath_node_attach(sc, sta, vif); @@ -1339,7 +1340,9 @@ static int ath9k_sta_add(struct ieee80211_hw *hw, vif->type != NL80211_IFTYPE_AP_VLAN) return 0; - an->ps_key = ath_key_config(common, vif, sta, &ps_key); + key = ath_key_config(common, vif, sta, &ps_key); + if (key > 0) + an->ps_key = key; return 0; } @@ -1356,6 +1359,7 @@ static void ath9k_del_ps_key(struct ath_softc *sc, return; ath_key_delete(common, &ps_key); + an->ps_key = 0; } static int ath9k_sta_remove(struct ieee80211_hw *hw, From db178340433f90e09ada35c174dfb2c84c2c71f4 Mon Sep 17 00:00:00 2001 From: Stanislaw Gruszka Date: Thu, 2 May 2013 09:43:57 +0200 Subject: [PATCH 013/198] ath5k: do not reschedule tx_complete_work on stop This patch claim to fix "WARNING: at net/mac80211/util.c:599 ieee80211_can_queue_work.isra.7+0x30/0x40", which was reported at: https://bugzilla.redhat.com/show_bug.cgi?id=922295 We use ATH_STAT_STARTED flag to disallow to perform ath5k_tx_complete_poll_work() code, hence reschedule ah->tx_complete_work, when we stop device. This flag was defined in ath5k code, but it was not used. I didn't get feedback if the fix works, so patch is compile only tested. Signed-off-by: Stanislaw Gruszka Signed-off-by: John W. Linville --- drivers/net/wireless/ath/ath5k/base.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/net/wireless/ath/ath5k/base.c b/drivers/net/wireless/ath/ath5k/base.c index 9b20d9ee2719..7f702fe3ecc2 100644 --- a/drivers/net/wireless/ath/ath5k/base.c +++ b/drivers/net/wireless/ath/ath5k/base.c @@ -2369,6 +2369,9 @@ ath5k_tx_complete_poll_work(struct work_struct *work) int i; bool needreset = false; + if (!test_bit(ATH_STAT_STARTED, ah->status)) + return; + mutex_lock(&ah->lock); for (i = 0; i < ARRAY_SIZE(ah->txqs); i++) { @@ -2676,6 +2679,7 @@ done: mmiowb(); mutex_unlock(&ah->lock); + set_bit(ATH_STAT_STARTED, ah->status); ieee80211_queue_delayed_work(ah->hw, &ah->tx_complete_work, msecs_to_jiffies(ATH5K_TX_COMPLETE_POLL_INT)); @@ -2737,6 +2741,7 @@ void ath5k_stop(struct ieee80211_hw *hw) ath5k_stop_tasklets(ah); + clear_bit(ATH_STAT_STARTED, ah->status); cancel_delayed_work_sync(&ah->tx_complete_work); if (!ath5k_modparam_no_hw_rfkill_switch) From 657445fe8660100ad174600ebfa61536392b7624 Mon Sep 17 00:00:00 2001 From: Daniel Vetter Date: Sat, 4 May 2013 10:09:18 +0200 Subject: [PATCH 014/198] Revert "drm/i915: revert eDP bpp clamping code changes" This reverts commit 57c219633275c7e7413f8bc7be250dc092887458. It's an ugly hack for a Haswell SDV platform where the vbt doesn't seem to fully agree with the panel. Since it seems to cause issues on real eDP platform let's just kill this hack again. Reported-and-tested-by: Josh Boyer References: https://lkml.org/lkml/2013/5/3/467 Cc: Paulo Zanoni Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_dp.c | 18 ++++-------------- 1 file changed, 4 insertions(+), 14 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index fc3bd9b4037c..0ab9813a79b5 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -702,6 +702,9 @@ intel_dp_compute_config(struct intel_encoder *encoder, /* Walk through all bpp values. Luckily they're all nicely spaced with 2 * bpc in between. */ bpp = min_t(int, 8*3, pipe_config->pipe_bpp); + if (is_edp(intel_dp) && dev_priv->edp.bpp) + bpp = min_t(int, bpp, dev_priv->edp.bpp); + for (; bpp >= 6*3; bpp -= 2*3) { mode_rate = intel_dp_link_required(target_clock, bpp); @@ -739,6 +742,7 @@ found: intel_dp->link_bw = bws[clock]; intel_dp->lane_count = lane_count; adjusted_mode->clock = drm_dp_bw_code_to_link_rate(intel_dp->link_bw); + pipe_config->pipe_bpp = bpp; pipe_config->pixel_target_clock = target_clock; DRM_DEBUG_KMS("DP link bw %02x lane count %d clock %d bpp %d\n", @@ -751,20 +755,6 @@ found: target_clock, adjusted_mode->clock, &pipe_config->dp_m_n); - /* - * XXX: We have a strange regression where using the vbt edp bpp value - * for the link bw computation results in black screens, the panel only - * works when we do the computation at the usual 24bpp (but still - * requires us to use 18bpp). Until that's fully debugged, stay - * bug-for-bug compatible with the old code. - */ - if (is_edp(intel_dp) && dev_priv->edp.bpp) { - DRM_DEBUG_KMS("clamping display bpc (was %d) to eDP (%d)\n", - bpp, dev_priv->edp.bpp); - bpp = min_t(int, bpp, dev_priv->edp.bpp); - } - pipe_config->pipe_bpp = bpp; - return true; } From 3ab9c63705cb7b1b9f83ddce725d8bd9ef7c66a9 Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Fri, 3 May 2013 12:57:41 +0300 Subject: [PATCH 015/198] drm/i915: hsw: fix link training for eDP on port-A According to BSpec the link training sequence for eDP on HSW port-A should be as follows: 1. link training: clock recovery 2. link training: equalization 3. link training: set idle transmission mode 4. display pipe enable 5. link training: disable (set normal mode) Contrary to this at the moment we don't do step 3. and we do step 5. before step 4. Fix this by setting idle transmission mode for eDP at the end of intel_dp_complete_link_train and adding a new intel_dp_stop_link_training function to disable link training. With these changes we'll end up with the following functions corresponding to the above steps: intel_dp_start_link_train -> step 1. intel_dp_complete_link_train -> step 2., step 3. intel_dp_stop_link_train -> step 5. For port-A we'll call intel_dp_stop_link_train only after enabling the pipe, for everything else we'll call it right after intel_dp_complete_link_train to preserve the current behavior. Tested on HSW/HSW-ULT. In v2: - Due to a HW issue we must set idle transmission mode for port-A too before enabling the pipe. Thanks for Arthur Runyan for explaining this. - Update the patch subject to make it clear that it's an eDP fix, DP is not affected. v3: - rename intel_dp_link_train() to intel_dp_set_link_train(), use 'val' instead 'l' as var name. (Paulo) Signed-off-by: Imre Deak Reviewed-by: Paulo Zanoni Tested-by: Paulo Zanoni Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_ddi.c | 5 +++ drivers/gpu/drm/i915/intel_dp.c | 59 ++++++++++++++++++++++++-------- drivers/gpu/drm/i915/intel_drv.h | 1 + 3 files changed, 50 insertions(+), 15 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c index 26a0a570f92e..fb961bb81903 100644 --- a/drivers/gpu/drm/i915/intel_ddi.c +++ b/drivers/gpu/drm/i915/intel_ddi.c @@ -1265,6 +1265,8 @@ static void intel_ddi_pre_enable(struct intel_encoder *intel_encoder) intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_ON); intel_dp_start_link_train(intel_dp); intel_dp_complete_link_train(intel_dp); + if (port != PORT_A) + intel_dp_stop_link_train(intel_dp); } } @@ -1326,6 +1328,9 @@ static void intel_enable_ddi(struct intel_encoder *intel_encoder) } else if (type == INTEL_OUTPUT_EDP) { struct intel_dp *intel_dp = enc_to_intel_dp(encoder); + if (port == PORT_A) + intel_dp_stop_link_train(intel_dp); + ironlake_edp_backlight_on(intel_dp); } diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 0ab9813a79b5..93694da7f717 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -1379,6 +1379,7 @@ static void intel_enable_dp(struct intel_encoder *encoder) ironlake_edp_panel_on(intel_dp); ironlake_edp_panel_vdd_off(intel_dp, true); intel_dp_complete_link_train(intel_dp); + intel_dp_stop_link_train(intel_dp); ironlake_edp_backlight_on(intel_dp); } @@ -1701,10 +1702,9 @@ intel_dp_set_link_train(struct intel_dp *intel_dp, struct drm_i915_private *dev_priv = dev->dev_private; enum port port = intel_dig_port->port; int ret; - uint32_t temp; if (HAS_DDI(dev)) { - temp = I915_READ(DP_TP_CTL(port)); + uint32_t temp = I915_READ(DP_TP_CTL(port)); if (dp_train_pat & DP_LINK_SCRAMBLING_DISABLE) temp |= DP_TP_CTL_SCRAMBLE_DISABLE; @@ -1714,18 +1714,6 @@ intel_dp_set_link_train(struct intel_dp *intel_dp, temp &= ~DP_TP_CTL_LINK_TRAIN_MASK; switch (dp_train_pat & DP_TRAINING_PATTERN_MASK) { case DP_TRAINING_PATTERN_DISABLE: - - if (port != PORT_A) { - temp |= DP_TP_CTL_LINK_TRAIN_IDLE; - I915_WRITE(DP_TP_CTL(port), temp); - - if (wait_for((I915_READ(DP_TP_STATUS(port)) & - DP_TP_STATUS_IDLE_DONE), 1)) - DRM_ERROR("Timed out waiting for DP idle patterns\n"); - - temp &= ~DP_TP_CTL_LINK_TRAIN_MASK; - } - temp |= DP_TP_CTL_LINK_TRAIN_NORMAL; break; @@ -1801,6 +1789,37 @@ intel_dp_set_link_train(struct intel_dp *intel_dp, return true; } +static void intel_dp_set_idle_link_train(struct intel_dp *intel_dp) +{ + struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp); + struct drm_device *dev = intel_dig_port->base.base.dev; + struct drm_i915_private *dev_priv = dev->dev_private; + enum port port = intel_dig_port->port; + uint32_t val; + + if (!HAS_DDI(dev)) + return; + + val = I915_READ(DP_TP_CTL(port)); + val &= ~DP_TP_CTL_LINK_TRAIN_MASK; + val |= DP_TP_CTL_LINK_TRAIN_IDLE; + I915_WRITE(DP_TP_CTL(port), val); + + /* + * On PORT_A we can have only eDP in SST mode. There the only reason + * we need to set idle transmission mode is to work around a HW issue + * where we enable the pipe while not in idle link-training mode. + * In this case there is requirement to wait for a minimum number of + * idle patterns to be sent. + */ + if (port == PORT_A) + return; + + if (wait_for((I915_READ(DP_TP_STATUS(port)) & DP_TP_STATUS_IDLE_DONE), + 1)) + DRM_ERROR("Timed out waiting for DP idle patterns\n"); +} + /* Enable corresponding port and start training pattern 1 */ void intel_dp_start_link_train(struct intel_dp *intel_dp) @@ -1943,10 +1962,19 @@ intel_dp_complete_link_train(struct intel_dp *intel_dp) ++tries; } + intel_dp_set_idle_link_train(intel_dp); + + intel_dp->DP = DP; + if (channel_eq) DRM_DEBUG_KMS("Channel EQ done. DP Training successfull\n"); - intel_dp_set_link_train(intel_dp, DP, DP_TRAINING_PATTERN_DISABLE); +} + +void intel_dp_stop_link_train(struct intel_dp *intel_dp) +{ + intel_dp_set_link_train(intel_dp, intel_dp->DP, + DP_TRAINING_PATTERN_DISABLE); } static void @@ -2154,6 +2182,7 @@ intel_dp_check_link_status(struct intel_dp *intel_dp) drm_get_encoder_name(&intel_encoder->base)); intel_dp_start_link_train(intel_dp); intel_dp_complete_link_train(intel_dp); + intel_dp_stop_link_train(intel_dp); } } diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index b5b6d19e6dd3..624a9e6b8d71 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -499,6 +499,7 @@ extern void intel_dp_init_connector(struct intel_digital_port *intel_dig_port, extern void intel_dp_init_link_config(struct intel_dp *intel_dp); extern void intel_dp_start_link_train(struct intel_dp *intel_dp); extern void intel_dp_complete_link_train(struct intel_dp *intel_dp); +extern void intel_dp_stop_link_train(struct intel_dp *intel_dp); extern void intel_dp_sink_dpms(struct intel_dp *intel_dp, int mode); extern void intel_dp_encoder_destroy(struct drm_encoder *encoder); extern void intel_dp_check_link_status(struct intel_dp *intel_dp); From 87c46b6c3d9c4d7f22bc553945728f5a82e9248a Mon Sep 17 00:00:00 2001 From: Russell King Date: Sat, 4 May 2013 14:38:59 +0100 Subject: [PATCH 016/198] ARM: finally enable IRQ time accounting config We've had IRQ time accounting for the last six months, except for the Kconfig symbol. This somehow got missed out of the original patch. Add this now. Signed-off-by: Russell King --- arch/arm/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index 62079d434581..b2d235d015d3 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -38,6 +38,7 @@ config ARM select HAVE_GENERIC_HARDIRQS select HAVE_HW_BREAKPOINT if (PERF_EVENTS && (CPU_V6 || CPU_V6K || CPU_V7)) select HAVE_IDE if PCI || ISA || PCMCIA + select HAVE_IRQ_TIME_ACCOUNTING select HAVE_KERNEL_GZIP select HAVE_KERNEL_LZMA select HAVE_KERNEL_LZO From 485cf5dac2966b93ef2b6211cf2fd42d66c823e6 Mon Sep 17 00:00:00 2001 From: Kevin Hilman Date: Wed, 24 Apr 2013 17:19:13 -0700 Subject: [PATCH 017/198] idle: Fix hlt/nohlt command-line handling in new generic idle commit d1669912 (idle: Implement generic idle function) added a new generic idle along with support for hlt/nohlt command line options to override default idle loop behavior. However, the command-line processing is never compiled. The command-line handling is wrapped by CONFIG_GENERIC_IDLE_POLL_SETUP and arches that use this feature select it in their Kconfigs. However, no Kconfig definition was created for this option, so it is never enabled, and therefore command-line override of the idle-loop behavior is broken after migrating to the generic idle loop. To fix, add a Kconfig definition for GENERIC_IDLE_POLL_SETUP. Tested on ARM (OMAP4/Panda) which enables the command-line overrides by default. Signed-off-by: Kevin Hilman Reviewed-by: Srivatsa S. Bhat Cc: Linus Torvalds Cc: Rusty Russell Cc: Paul McKenney Cc: Peter Zijlstra Cc: Magnus Damm Cc: linux-arm-kernel@lists.infradead.org Cc: linaro-kernel@lists.linaro.org Link: http://lkml.kernel.org/r/1366849153-25564-1-git-send-email-khilman@linaro.org Signed-off-by: Thomas Gleixner --- arch/Kconfig | 3 +++ 1 file changed, 3 insertions(+) diff --git a/arch/Kconfig b/arch/Kconfig index 99f0e17df429..3e644037195b 100644 --- a/arch/Kconfig +++ b/arch/Kconfig @@ -213,6 +213,9 @@ config USE_GENERIC_SMP_HELPERS config GENERIC_SMP_IDLE_THREAD bool +config GENERIC_IDLE_POLL_SETUP + bool + # Select if arch init_task initializer is different to init/init_task.c config ARCH_INIT_TASK bool From fbd44a607a1a5019bc32c3615cead8c5ee8f89c9 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Fri, 3 May 2013 20:22:36 +0200 Subject: [PATCH 018/198] tick: Use zalloc_cpumask_var for allocating offstack cpumasks commit b352bc1cbc (tick: Convert broadcast cpu bitmaps to cpumask_var_t) broke CONFIG_CPUMASK_OFFSTACK in a very subtle way. Instead of allocating the cpumasks with zalloc_cpumask_var it uses alloc_cpumask_var, so we can get random data there, which of course confuses the logic completely and causes random failures. Reported-and-tested-by: Dave Jones Reported-and-tested-by: Yinghai Lu Link: http://lkml.kernel.org/r/alpine.LFD.2.02.1305032015060.2990@ionos Signed-off-by: Thomas Gleixner Signed-off-by: Ingo Molnar --- kernel/time/tick-broadcast.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/kernel/time/tick-broadcast.c b/kernel/time/tick-broadcast.c index 61d00a8cdf2f..d70cdc42c829 100644 --- a/kernel/time/tick-broadcast.c +++ b/kernel/time/tick-broadcast.c @@ -785,11 +785,11 @@ bool tick_broadcast_oneshot_available(void) void __init tick_broadcast_init(void) { - alloc_cpumask_var(&tick_broadcast_mask, GFP_NOWAIT); - alloc_cpumask_var(&tmpmask, GFP_NOWAIT); + zalloc_cpumask_var(&tick_broadcast_mask, GFP_NOWAIT); + zalloc_cpumask_var(&tmpmask, GFP_NOWAIT); #ifdef CONFIG_TICK_ONESHOT - alloc_cpumask_var(&tick_broadcast_oneshot_mask, GFP_NOWAIT); - alloc_cpumask_var(&tick_broadcast_pending_mask, GFP_NOWAIT); - alloc_cpumask_var(&tick_broadcast_force_mask, GFP_NOWAIT); + zalloc_cpumask_var(&tick_broadcast_oneshot_mask, GFP_NOWAIT); + zalloc_cpumask_var(&tick_broadcast_pending_mask, GFP_NOWAIT); + zalloc_cpumask_var(&tick_broadcast_force_mask, GFP_NOWAIT); #endif } From e6155736ad76b2070652745f9e54cdea3f0d8567 Mon Sep 17 00:00:00 2001 From: Lachlan McIlroy Date: Sun, 5 May 2013 23:10:00 -0400 Subject: [PATCH 019/198] ext4: limit group search loop for non-extent files In the case where we are allocating for a non-extent file, we must limit the groups we allocate from to those below 2^32 blocks, and ext4_mb_regular_allocator() attempts to do this initially by putting a cap on ngroups for the subsequent search loop. However, the initial target group comes in from the allocation context (ac), and it may already be beyond the artificially limited ngroups. In this case, the limit if (group == ngroups) group = 0; at the top of the loop is never true, and the loop will run away. Catch this case inside the loop and reset the search to start at group 0. [sandeen@redhat.com: add commit msg & comments] Signed-off-by: Lachlan McIlroy Signed-off-by: Eric Sandeen Signed-off-by: "Theodore Ts'o" Cc: stable@vger.kernel.org --- fs/ext4/mballoc.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index a11ea4d6164c..39d30f12d98d 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -2105,7 +2105,11 @@ repeat: group = ac->ac_g_ex.fe_group; for (i = 0; i < ngroups; group++, i++) { - if (group == ngroups) + /* + * Artificially restricted ngroups for non-extent + * files makes group > ngroups possible on first loop. + */ + if (group >= ngroups) group = 0; /* This now checks without needing the buddy page */ From 756b253167be36ea74980220d606662713431503 Mon Sep 17 00:00:00 2001 From: Sebastian Hesselbarth Date: Thu, 2 May 2013 19:56:12 +0100 Subject: [PATCH 020/198] ARM: 7711/1: dove: fix Dove cpu type from V7 to PJ4 The CPU used in Marvell Dove SoCs is a PJ4 Sheeva core. Using CONFIG_CPU_PJ4 instead of CONFIG_CPU_V7 will enable iWMMXt extensions on Dove. Signed-off-by: Sebastian Hesselbarth Acked-by: Jason Cooper Acked-by: Andrew Lunn Signed-off-by: Russell King --- arch/arm/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index b2d235d015d3..7e3e2bfc5d7a 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -492,7 +492,7 @@ config ARCH_IXP4XX config ARCH_DOVE bool "Marvell Dove" select ARCH_REQUIRE_GPIOLIB - select CPU_V7 + select CPU_PJ4 select GENERIC_CLOCKEVENTS select MIGHT_HAVE_PCI select PINCTRL From c01c5a506b59a8248b7666c779eb4ed689937874 Mon Sep 17 00:00:00 2001 From: Daniel Tang Date: Fri, 3 May 2013 08:13:55 +0100 Subject: [PATCH 021/198] ARM: 7712/1: Remove trailing whitespace in arch/arm/Makefile Clean up some trailing whitespace issues in arch/arm/Makefile. Signed-off-by: Daniel Tang Signed-off-by: Russell King --- arch/arm/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm/Makefile b/arch/arm/Makefile index 47374085befd..1ba358ba16b8 100644 --- a/arch/arm/Makefile +++ b/arch/arm/Makefile @@ -309,7 +309,7 @@ define archhelp echo ' Image - Uncompressed kernel image (arch/$(ARCH)/boot/Image)' echo '* xipImage - XIP kernel image, if configured (arch/$(ARCH)/boot/xipImage)' echo ' uImage - U-Boot wrapped zImage' - echo ' bootpImage - Combined zImage and initial RAM disk' + echo ' bootpImage - Combined zImage and initial RAM disk' echo ' (supply initrd image via make variable INITRD=)' echo '* dtbs - Build device tree blobs for enabled boards' echo ' install - Install uncompressed kernel' From 524eff183f51d080a83b348d0ea97c08b3607b9a Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Mon, 6 May 2013 18:27:17 +0200 Subject: [PATCH 022/198] perf: Fix EXIT event notification The perf_event_task_ctx() function needs to be called with preemption disabled, since it's checking for currently scheduled cpu against event cpu. We disable preemption for task related perf event context if there's one defined, leaving up to the chance which cpu it gets scheduled in. Signed-off-by: Jiri Olsa Acked-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Namhyung Kim Cc: Corey Ashford Cc: Frederic Weisbecker Cc: Paul Mackerras Cc: Stephane Eranian Cc: Borislav Petkov Link: http://lkml.kernel.org/r/1367857638-27631-2-git-send-email-jolsa@redhat.com Signed-off-by: Ingo Molnar --- kernel/events/core.c | 26 ++++++++++++++------------ 1 file changed, 14 insertions(+), 12 deletions(-) diff --git a/kernel/events/core.c b/kernel/events/core.c index 6b41c1899a8b..38b68a05c3c6 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -4474,7 +4474,7 @@ static void perf_event_task_ctx(struct perf_event_context *ctx, static void perf_event_task_event(struct perf_task_event *task_event) { struct perf_cpu_context *cpuctx; - struct perf_event_context *ctx; + struct perf_event_context *ctx, *task_ctx = task_event->task_ctx; struct pmu *pmu; int ctxn; @@ -4485,20 +4485,22 @@ static void perf_event_task_event(struct perf_task_event *task_event) goto next; perf_event_task_ctx(&cpuctx->ctx, task_event); - ctx = task_event->task_ctx; - if (!ctx) { - ctxn = pmu->task_ctx_nr; - if (ctxn < 0) - goto next; - ctx = rcu_dereference(current->perf_event_ctxp[ctxn]); - if (ctx) - perf_event_task_ctx(ctx, task_event); - } + if (task_ctx) + goto next; + ctxn = pmu->task_ctx_nr; + if (ctxn < 0) + goto next; + ctx = rcu_dereference(current->perf_event_ctxp[ctxn]); + if (ctx) + perf_event_task_ctx(ctx, task_event); next: put_cpu_ptr(pmu->pmu_cpu_context); } - if (task_event->task_ctx) - perf_event_task_ctx(task_event->task_ctx, task_event); + if (task_ctx) { + preempt_disable(); + perf_event_task_ctx(task_ctx, task_event); + preempt_enable(); + } rcu_read_unlock(); } From 52d857a8784a09576215c71cebf368d61c12a754 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Mon, 6 May 2013 18:27:18 +0200 Subject: [PATCH 023/198] perf: Factor out auxiliary events notification Add perf_event_aux() function to send out all types of auxiliary events - mmap, task, comm events. For each type there's match and output functions defined and used as callbacks during perf_event_aux processing. This way we can centralize the pmu/context iterating and event matching logic. Also since lot of the code was duplicated, this patch reduces the .text size about 2kB on my setup: snipped output from 'objdump -x kernel/events/core.o' before: Idx Name Size 0 .text 0000d313 after: Idx Name Size 0 .text 0000cad3 Signed-off-by: Jiri Olsa Acked-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Namhyung Kim Cc: Corey Ashford Cc: Frederic Weisbecker Cc: Paul Mackerras Cc: Peter Zijlstra Cc: Stephane Eranian Cc: Borislav Petkov Link: http://lkml.kernel.org/r/1367857638-27631-3-git-send-email-jolsa@redhat.com Signed-off-by: Ingo Molnar --- kernel/events/core.c | 242 ++++++++++++++++--------------------------- 1 file changed, 89 insertions(+), 153 deletions(-) diff --git a/kernel/events/core.c b/kernel/events/core.c index 38b68a05c3c6..9dc297faf7c0 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -4394,6 +4394,64 @@ perf_event_read_event(struct perf_event *event, perf_output_end(&handle); } +typedef int (perf_event_aux_match_cb)(struct perf_event *event, void *data); +typedef void (perf_event_aux_output_cb)(struct perf_event *event, void *data); + +static void +perf_event_aux_ctx(struct perf_event_context *ctx, + perf_event_aux_match_cb match, + perf_event_aux_output_cb output, + void *data) +{ + struct perf_event *event; + + list_for_each_entry_rcu(event, &ctx->event_list, event_entry) { + if (event->state < PERF_EVENT_STATE_INACTIVE) + continue; + if (!event_filter_match(event)) + continue; + if (match(event, data)) + output(event, data); + } +} + +static void +perf_event_aux(perf_event_aux_match_cb match, + perf_event_aux_output_cb output, + void *data, + struct perf_event_context *task_ctx) +{ + struct perf_cpu_context *cpuctx; + struct perf_event_context *ctx; + struct pmu *pmu; + int ctxn; + + rcu_read_lock(); + list_for_each_entry_rcu(pmu, &pmus, entry) { + cpuctx = get_cpu_ptr(pmu->pmu_cpu_context); + if (cpuctx->unique_pmu != pmu) + goto next; + perf_event_aux_ctx(&cpuctx->ctx, match, output, data); + if (task_ctx) + goto next; + ctxn = pmu->task_ctx_nr; + if (ctxn < 0) + goto next; + ctx = rcu_dereference(current->perf_event_ctxp[ctxn]); + if (ctx) + perf_event_aux_ctx(ctx, match, output, data); +next: + put_cpu_ptr(pmu->pmu_cpu_context); + } + + if (task_ctx) { + preempt_disable(); + perf_event_aux_ctx(task_ctx, match, output, data); + preempt_enable(); + } + rcu_read_unlock(); +} + /* * task tracking -- fork/exit * @@ -4416,8 +4474,9 @@ struct perf_task_event { }; static void perf_event_task_output(struct perf_event *event, - struct perf_task_event *task_event) + void *data) { + struct perf_task_event *task_event = data; struct perf_output_handle handle; struct perf_sample_data sample; struct task_struct *task = task_event->task; @@ -4445,64 +4504,11 @@ out: task_event->event_id.header.size = size; } -static int perf_event_task_match(struct perf_event *event) +static int perf_event_task_match(struct perf_event *event, + void *data __maybe_unused) { - if (event->state < PERF_EVENT_STATE_INACTIVE) - return 0; - - if (!event_filter_match(event)) - return 0; - - if (event->attr.comm || event->attr.mmap || - event->attr.mmap_data || event->attr.task) - return 1; - - return 0; -} - -static void perf_event_task_ctx(struct perf_event_context *ctx, - struct perf_task_event *task_event) -{ - struct perf_event *event; - - list_for_each_entry_rcu(event, &ctx->event_list, event_entry) { - if (perf_event_task_match(event)) - perf_event_task_output(event, task_event); - } -} - -static void perf_event_task_event(struct perf_task_event *task_event) -{ - struct perf_cpu_context *cpuctx; - struct perf_event_context *ctx, *task_ctx = task_event->task_ctx; - struct pmu *pmu; - int ctxn; - - rcu_read_lock(); - list_for_each_entry_rcu(pmu, &pmus, entry) { - cpuctx = get_cpu_ptr(pmu->pmu_cpu_context); - if (cpuctx->unique_pmu != pmu) - goto next; - perf_event_task_ctx(&cpuctx->ctx, task_event); - - if (task_ctx) - goto next; - ctxn = pmu->task_ctx_nr; - if (ctxn < 0) - goto next; - ctx = rcu_dereference(current->perf_event_ctxp[ctxn]); - if (ctx) - perf_event_task_ctx(ctx, task_event); -next: - put_cpu_ptr(pmu->pmu_cpu_context); - } - if (task_ctx) { - preempt_disable(); - perf_event_task_ctx(task_ctx, task_event); - preempt_enable(); - } - - rcu_read_unlock(); + return event->attr.comm || event->attr.mmap || + event->attr.mmap_data || event->attr.task; } static void perf_event_task(struct task_struct *task, @@ -4533,7 +4539,10 @@ static void perf_event_task(struct task_struct *task, }, }; - perf_event_task_event(&task_event); + perf_event_aux(perf_event_task_match, + perf_event_task_output, + &task_event, + task_ctx); } void perf_event_fork(struct task_struct *task) @@ -4559,8 +4568,9 @@ struct perf_comm_event { }; static void perf_event_comm_output(struct perf_event *event, - struct perf_comm_event *comm_event) + void *data) { + struct perf_comm_event *comm_event = data; struct perf_output_handle handle; struct perf_sample_data sample; int size = comm_event->event_id.header.size; @@ -4587,39 +4597,16 @@ out: comm_event->event_id.header.size = size; } -static int perf_event_comm_match(struct perf_event *event) +static int perf_event_comm_match(struct perf_event *event, + void *data __maybe_unused) { - if (event->state < PERF_EVENT_STATE_INACTIVE) - return 0; - - if (!event_filter_match(event)) - return 0; - - if (event->attr.comm) - return 1; - - return 0; -} - -static void perf_event_comm_ctx(struct perf_event_context *ctx, - struct perf_comm_event *comm_event) -{ - struct perf_event *event; - - list_for_each_entry_rcu(event, &ctx->event_list, event_entry) { - if (perf_event_comm_match(event)) - perf_event_comm_output(event, comm_event); - } + return event->attr.comm; } static void perf_event_comm_event(struct perf_comm_event *comm_event) { - struct perf_cpu_context *cpuctx; - struct perf_event_context *ctx; char comm[TASK_COMM_LEN]; unsigned int size; - struct pmu *pmu; - int ctxn; memset(comm, 0, sizeof(comm)); strlcpy(comm, comm_event->task->comm, sizeof(comm)); @@ -4629,24 +4616,11 @@ static void perf_event_comm_event(struct perf_comm_event *comm_event) comm_event->comm_size = size; comm_event->event_id.header.size = sizeof(comm_event->event_id) + size; - rcu_read_lock(); - list_for_each_entry_rcu(pmu, &pmus, entry) { - cpuctx = get_cpu_ptr(pmu->pmu_cpu_context); - if (cpuctx->unique_pmu != pmu) - goto next; - perf_event_comm_ctx(&cpuctx->ctx, comm_event); - ctxn = pmu->task_ctx_nr; - if (ctxn < 0) - goto next; - - ctx = rcu_dereference(current->perf_event_ctxp[ctxn]); - if (ctx) - perf_event_comm_ctx(ctx, comm_event); -next: - put_cpu_ptr(pmu->pmu_cpu_context); - } - rcu_read_unlock(); + perf_event_aux(perf_event_comm_match, + perf_event_comm_output, + comm_event, + NULL); } void perf_event_comm(struct task_struct *task) @@ -4708,8 +4682,9 @@ struct perf_mmap_event { }; static void perf_event_mmap_output(struct perf_event *event, - struct perf_mmap_event *mmap_event) + void *data) { + struct perf_mmap_event *mmap_event = data; struct perf_output_handle handle; struct perf_sample_data sample; int size = mmap_event->event_id.header.size; @@ -4736,46 +4711,24 @@ out: } static int perf_event_mmap_match(struct perf_event *event, - struct perf_mmap_event *mmap_event, - int executable) + void *data) { - if (event->state < PERF_EVENT_STATE_INACTIVE) - return 0; + struct perf_mmap_event *mmap_event = data; + struct vm_area_struct *vma = mmap_event->vma; + int executable = vma->vm_flags & VM_EXEC; - if (!event_filter_match(event)) - return 0; - - if ((!executable && event->attr.mmap_data) || - (executable && event->attr.mmap)) - return 1; - - return 0; -} - -static void perf_event_mmap_ctx(struct perf_event_context *ctx, - struct perf_mmap_event *mmap_event, - int executable) -{ - struct perf_event *event; - - list_for_each_entry_rcu(event, &ctx->event_list, event_entry) { - if (perf_event_mmap_match(event, mmap_event, executable)) - perf_event_mmap_output(event, mmap_event); - } + return (!executable && event->attr.mmap_data) || + (executable && event->attr.mmap); } static void perf_event_mmap_event(struct perf_mmap_event *mmap_event) { - struct perf_cpu_context *cpuctx; - struct perf_event_context *ctx; struct vm_area_struct *vma = mmap_event->vma; struct file *file = vma->vm_file; unsigned int size; char tmp[16]; char *buf = NULL; const char *name; - struct pmu *pmu; - int ctxn; memset(tmp, 0, sizeof(tmp)); @@ -4831,27 +4784,10 @@ got_name: mmap_event->event_id.header.size = sizeof(mmap_event->event_id) + size; - rcu_read_lock(); - list_for_each_entry_rcu(pmu, &pmus, entry) { - cpuctx = get_cpu_ptr(pmu->pmu_cpu_context); - if (cpuctx->unique_pmu != pmu) - goto next; - perf_event_mmap_ctx(&cpuctx->ctx, mmap_event, - vma->vm_flags & VM_EXEC); - - ctxn = pmu->task_ctx_nr; - if (ctxn < 0) - goto next; - - ctx = rcu_dereference(current->perf_event_ctxp[ctxn]); - if (ctx) { - perf_event_mmap_ctx(ctx, mmap_event, - vma->vm_flags & VM_EXEC); - } -next: - put_cpu_ptr(pmu->pmu_cpu_context); - } - rcu_read_unlock(); + perf_event_aux(perf_event_mmap_match, + perf_event_mmap_output, + mmap_event, + NULL); kfree(buf); } From 97a5b81fa4d3a11dcdf224befc577f2e0abadc0b Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Thu, 2 May 2013 16:21:15 +0200 Subject: [PATCH 024/198] x86: Fix idle consolidation fallout The core code expects the arch idle code to return with interrupts enabled. The conversion missed two x86 cases which fail to do that. Reported-and-tested-by: Markus Trippelsdorf Tested-by: Borislav Petkov Link: http://lkml.kernel.org/r/alpine.LFD.2.02.1305021557030.3972@ionos Signed-off-by: Thomas Gleixner --- arch/x86/kernel/process.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c index 607af0d4d5ef..4e7a37ff03ab 100644 --- a/arch/x86/kernel/process.c +++ b/arch/x86/kernel/process.c @@ -312,6 +312,8 @@ void arch_cpu_idle(void) { if (cpuidle_idle_call()) x86_idle(); + else + local_irq_enable(); } /* @@ -368,9 +370,6 @@ void amd_e400_remove_cpu(int cpu) */ static void amd_e400_idle(void) { - if (need_resched()) - return; - if (!amd_e400_c1e_detected) { u32 lo, hi; From c4ae25ecdf93132fb69b37d8121a86046615b53b Mon Sep 17 00:00:00 2001 From: Ben Widawsky Date: Wed, 1 May 2013 11:00:34 -0700 Subject: [PATCH 025/198] Revert "drm/i915: Calculate correct stolen size for GEN7+" This reverts commit 03752f5b7b77b95d83479885040950fba1250850. This revert requires a bit of explanation on how I understand things work. Internally the architects/designers decide how the stolen encoding works. We put it in a doc. BIOS writers take these docs and implement it. Driver writers read the doc too, and read the value left by the BIOS writers, and then we make magic. The failing here is that in the docs we had[1] contained two different definitions for this register for Gen7. (We have both a PCI register, and an MMIO, and each of these were different). At the time [2] of 03752f5, we asked the architects what the correct value should be; but that doesn't match the reality (BIOS) unfortunately. So on all machines I can get my hands on, this revert is the right thing to do. I've also worked with the product group to confirm that they agree this revert is what we should do. People using HW made my "people" who both write their own BIOS, and have access to our docs (Apple?). Investigations are still ongoing about whether we need to add a list of machines needing special handling, but this patch should be the right thing for pretty much everyone. [1] The docs are still wrong on this one. Now instead of two registers with two definitions, we have one register with BOTH definitions, progress? [2] The open source PRMs have the "wrong" definitions in chapter Volume 1 part6, section 1.1.12. This digging was inspired by Paulo. Cc: Paulo Zanoni Signed-off-by: Ben Widawsky Acked-by: Jesse Barnes [danvet: Augment the patch saying that it's still a bit unclear whether there are any machines out there with "wrong" firmware and whether we need to add a list to handle them specially.] Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_gem_gtt.c | 15 +-------------- drivers/gpu/drm/i915/i915_reg.h | 2 -- 2 files changed, 1 insertion(+), 16 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index dca614de71b6..bdb0d7717bc7 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -709,15 +709,6 @@ static inline size_t gen6_get_stolen_size(u16 snb_gmch_ctl) return snb_gmch_ctl << 25; /* 32 MB units */ } -static inline size_t gen7_get_stolen_size(u16 snb_gmch_ctl) -{ - static const int stolen_decoder[] = { - 0, 0, 0, 0, 0, 32, 48, 64, 128, 256, 96, 160, 224, 352}; - snb_gmch_ctl >>= IVB_GMCH_GMS_SHIFT; - snb_gmch_ctl &= IVB_GMCH_GMS_MASK; - return stolen_decoder[snb_gmch_ctl] << 20; -} - static int gen6_gmch_probe(struct drm_device *dev, size_t *gtt_total, size_t *stolen, @@ -747,11 +738,7 @@ static int gen6_gmch_probe(struct drm_device *dev, pci_read_config_word(dev->pdev, SNB_GMCH_CTRL, &snb_gmch_ctl); gtt_size = gen6_get_total_gtt_size(snb_gmch_ctl); - if (IS_GEN7(dev) && !IS_VALLEYVIEW(dev)) - *stolen = gen7_get_stolen_size(snb_gmch_ctl); - else - *stolen = gen6_get_stolen_size(snb_gmch_ctl); - + *stolen = gen6_get_stolen_size(snb_gmch_ctl); *gtt_total = (gtt_size / sizeof(gen6_gtt_pte_t)) << PAGE_SHIFT; /* For Modern GENs the PTEs and register space are split in the BAR */ diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index 83f9c26e1adb..2d6b62e42daf 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -46,8 +46,6 @@ #define SNB_GMCH_GGMS_MASK 0x3 #define SNB_GMCH_GMS_SHIFT 3 /* Graphics Mode Select */ #define SNB_GMCH_GMS_MASK 0x1f -#define IVB_GMCH_GMS_SHIFT 4 -#define IVB_GMCH_GMS_MASK 0xf /* PCI config space */ From 1ffc5289bfcf7f4c4e4213240bb4be68c48ce603 Mon Sep 17 00:00:00 2001 From: Jani Nikula Date: Tue, 7 May 2013 18:54:05 +0300 Subject: [PATCH 026/198] drm/i915: clear the stolen fb before resuming Similar to commit 88afe715dd5469bc24ca7a19ac62dd3c241cab48 Author: Chris Wilson Date: Sun Dec 16 12:15:41 2012 +0000 drm/i915: Clear the stolen fb before enabling but on the resume path. Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=57191 Reported-and-tested-by: Nikolay Amiantov Signed-off-by: Jani Nikula Reviewed-by: Chris Wilson Cc: stable@vger.kernel.org (3.9 only) Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_fb.c | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_fb.c b/drivers/gpu/drm/i915/intel_fb.c index 0e19e575a1b4..6b7c3ca2c035 100644 --- a/drivers/gpu/drm/i915/intel_fb.c +++ b/drivers/gpu/drm/i915/intel_fb.c @@ -262,10 +262,22 @@ void intel_fbdev_fini(struct drm_device *dev) void intel_fbdev_set_suspend(struct drm_device *dev, int state) { drm_i915_private_t *dev_priv = dev->dev_private; - if (!dev_priv->fbdev) + struct intel_fbdev *ifbdev = dev_priv->fbdev; + struct fb_info *info; + + if (!ifbdev) return; - fb_set_suspend(dev_priv->fbdev->helper.fbdev, state); + info = ifbdev->helper.fbdev; + + /* On resume from hibernation: If the object is shmemfs backed, it has + * been restored from swap. If the object is stolen however, it will be + * full of whatever garbage was left in there. + */ + if (!state && ifbdev->ifb.obj->stolen) + memset_io(info->screen_base, 0, info->screen_size); + + fb_set_suspend(info, state); } MODULE_LICENSE("GPL and additional rights"); From 30650239adc9e4e9439256d6988e521518dccbb3 Mon Sep 17 00:00:00 2001 From: Alistair Popple Date: Mon, 29 Apr 2013 18:07:47 +0000 Subject: [PATCH 027/198] powerpc: Add an in memory udbg console This patch adds a new udbg early debug console which utilises statically defined input and output buffers stored within the kernel BSS. It is primarily designed to assist with bring up of new hardware which may not have a working console but which has a method of reading/writing kernel memory. This version incorporates comments made by Ben H (thanks!). Changes from v1: - Add memory barriers. - Ensure updating of read/write positions is atomic. Signed-off-by: Alistair Popple Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/Kconfig.debug | 23 +++++++ arch/powerpc/include/asm/udbg.h | 1 + arch/powerpc/kernel/udbg.c | 3 + arch/powerpc/sysdev/Makefile | 2 + arch/powerpc/sysdev/udbg_memcons.c | 105 +++++++++++++++++++++++++++++ 5 files changed, 134 insertions(+) create mode 100644 arch/powerpc/sysdev/udbg_memcons.c diff --git a/arch/powerpc/Kconfig.debug b/arch/powerpc/Kconfig.debug index 5416e28a7538..863d877e0b5f 100644 --- a/arch/powerpc/Kconfig.debug +++ b/arch/powerpc/Kconfig.debug @@ -262,8 +262,31 @@ config PPC_EARLY_DEBUG_OPAL_HVSI Select this to enable early debugging for the PowerNV platform using an "hvsi" console +config PPC_EARLY_DEBUG_MEMCONS + bool "In memory console" + help + Select this to enable early debugging using an in memory console. + This console provides input and output buffers stored within the + kernel BSS and should be safe to select on any system. A debugger + can then be used to read kernel output or send input to the console. endchoice +config PPC_MEMCONS_OUTPUT_SIZE + int "In memory console output buffer size" + depends on PPC_EARLY_DEBUG_MEMCONS + default 4096 + help + Selects the size of the output buffer (in bytes) of the in memory + console. + +config PPC_MEMCONS_INPUT_SIZE + int "In memory console input buffer size" + depends on PPC_EARLY_DEBUG_MEMCONS + default 128 + help + Selects the size of the input buffer (in bytes) of the in memory + console. + config PPC_EARLY_DEBUG_OPAL def_bool y depends on PPC_EARLY_DEBUG_OPAL_RAW || PPC_EARLY_DEBUG_OPAL_HVSI diff --git a/arch/powerpc/include/asm/udbg.h b/arch/powerpc/include/asm/udbg.h index 5a7510e9d09d..dc590919f8eb 100644 --- a/arch/powerpc/include/asm/udbg.h +++ b/arch/powerpc/include/asm/udbg.h @@ -52,6 +52,7 @@ extern void __init udbg_init_40x_realmode(void); extern void __init udbg_init_cpm(void); extern void __init udbg_init_usbgecko(void); extern void __init udbg_init_wsp(void); +extern void __init udbg_init_memcons(void); extern void __init udbg_init_ehv_bc(void); extern void __init udbg_init_ps3gelic(void); extern void __init udbg_init_debug_opal_raw(void); diff --git a/arch/powerpc/kernel/udbg.c b/arch/powerpc/kernel/udbg.c index 13b867093499..9d3fdcd66290 100644 --- a/arch/powerpc/kernel/udbg.c +++ b/arch/powerpc/kernel/udbg.c @@ -64,6 +64,9 @@ void __init udbg_early_init(void) udbg_init_usbgecko(); #elif defined(CONFIG_PPC_EARLY_DEBUG_WSP) udbg_init_wsp(); +#elif defined(CONFIG_PPC_EARLY_DEBUG_MEMCONS) + /* In memory console */ + udbg_init_memcons(); #elif defined(CONFIG_PPC_EARLY_DEBUG_EHV_BC) udbg_init_ehv_bc(); #elif defined(CONFIG_PPC_EARLY_DEBUG_PS3GELIC) diff --git a/arch/powerpc/sysdev/Makefile b/arch/powerpc/sysdev/Makefile index b0a518e97599..99464a7bdb3b 100644 --- a/arch/powerpc/sysdev/Makefile +++ b/arch/powerpc/sysdev/Makefile @@ -64,6 +64,8 @@ endif obj-$(CONFIG_PPC_SCOM) += scom.o +obj-$(CONFIG_PPC_EARLY_DEBUG_MEMCONS) += udbg_memcons.o + subdir-ccflags-$(CONFIG_PPC_WERROR) := -Werror obj-$(CONFIG_PPC_XICS) += xics/ diff --git a/arch/powerpc/sysdev/udbg_memcons.c b/arch/powerpc/sysdev/udbg_memcons.c new file mode 100644 index 000000000000..ce5a7b489e4b --- /dev/null +++ b/arch/powerpc/sysdev/udbg_memcons.c @@ -0,0 +1,105 @@ +/* + * A udbg backend which logs messages and reads input from in memory + * buffers. + * + * The console output can be read from memcons_output which is a + * circular buffer whose next write position is stored in memcons.output_pos. + * + * Input may be passed by writing into the memcons_input buffer when it is + * empty. The input buffer is empty when both input_pos == input_start and + * *input_start == '\0'. + * + * Copyright (C) 2003-2005 Anton Blanchard and Milton Miller, IBM Corp + * Copyright (C) 2013 Alistair Popple, IBM Corp + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ + +#include +#include +#include +#include +#include +#include + +struct memcons { + char *output_start; + char *output_pos; + char *output_end; + char *input_start; + char *input_pos; + char *input_end; +}; + +static char memcons_output[CONFIG_PPC_MEMCONS_OUTPUT_SIZE]; +static char memcons_input[CONFIG_PPC_MEMCONS_INPUT_SIZE]; + +struct memcons memcons = { + .output_start = memcons_output, + .output_pos = memcons_output, + .output_end = &memcons_output[CONFIG_PPC_MEMCONS_OUTPUT_SIZE], + .input_start = memcons_input, + .input_pos = memcons_input, + .input_end = &memcons_input[CONFIG_PPC_MEMCONS_INPUT_SIZE], +}; + +void memcons_putc(char c) +{ + char *new_output_pos; + + *memcons.output_pos = c; + wmb(); + new_output_pos = memcons.output_pos + 1; + if (new_output_pos >= memcons.output_end) + new_output_pos = memcons.output_start; + + memcons.output_pos = new_output_pos; +} + +int memcons_getc_poll(void) +{ + char c; + char *new_input_pos; + + if (*memcons.input_pos) { + c = *memcons.input_pos; + + new_input_pos = memcons.input_pos + 1; + if (new_input_pos >= memcons.input_end) + new_input_pos = memcons.input_start; + else if (*new_input_pos == '\0') + new_input_pos = memcons.input_start; + + *memcons.input_pos = '\0'; + wmb(); + memcons.input_pos = new_input_pos; + return c; + } + + return -1; +} + +int memcons_getc(void) +{ + int c; + + while (1) { + c = memcons_getc_poll(); + if (c == -1) + cpu_relax(); + else + break; + } + + return c; +} + +void udbg_init_memcons(void) +{ + udbg_putc = memcons_putc; + udbg_getc = memcons_getc; + udbg_getc_poll = memcons_getc_poll; +} From f558a845c3a043d032bb247cdbe50db9202476a1 Mon Sep 17 00:00:00 2001 From: Dave Jones Date: Fri, 3 May 2013 16:40:09 -0400 Subject: [PATCH 028/198] Add missing module license tag to vring helpers. [ 624.286653] vringh: module license 'unspecified' taints kernel. Signed-off-by: Dave Jones Signed-off-by: Rusty Russell --- drivers/vhost/vringh.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/vhost/vringh.c b/drivers/vhost/vringh.c index bff0775e258c..5174ebac288d 100644 --- a/drivers/vhost/vringh.c +++ b/drivers/vhost/vringh.c @@ -3,6 +3,7 @@ * * Since these may be in userspace, we use (inline) accessors. */ +#include #include #include #include @@ -1005,3 +1006,5 @@ int vringh_need_notify_kern(struct vringh *vrh) return __vringh_need_notify(vrh, getu16_kern); } EXPORT_SYMBOL(vringh_need_notify_kern); + +MODULE_LICENSE("GPL"); From f616fe4feee4000c2995d2d1e1981513bf9ab0d4 Mon Sep 17 00:00:00 2001 From: Rusty Russell Date: Wed, 8 May 2013 10:06:55 +0930 Subject: [PATCH 029/198] lguest: clear cached last cpu when guest_set_pgd() called. commit v3.9-rc1-53-g6d0cda9 "lguest: cache last cpu we ran on." missed one case, which causes a triple fault. The guest calls guest_set_pgd() on the top page, and we carefully remap the Switcher text page. But we didn't reset last_host_cpu, so map_switcher_in_guest() thinks the guest's regs and IDT/GDT etc are already mapped. Reported-by: Paul Bolle Tested-by: Paul Bolle Signed-off-by: Rusty Russell --- drivers/lguest/page_tables.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/lguest/page_tables.c b/drivers/lguest/page_tables.c index 699187ab3800..5b9ac32801c7 100644 --- a/drivers/lguest/page_tables.c +++ b/drivers/lguest/page_tables.c @@ -1002,6 +1002,7 @@ void guest_set_pgd(struct lguest *lg, unsigned long gpgdir, u32 idx) kill_guest(&lg->cpus[0], "Cannot populate switcher mapping"); } + lg->pgdirs[pgdir].last_host_cpu = -1; } } From 1de1455f33709a8afd8d41d26d09739a1148105b Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Wed, 8 May 2013 14:14:26 +1000 Subject: [PATCH 030/198] powerpc/powernv: Properly drop characters if console is closed If the firmware returns an error such as "closed" (or hardware error), we should drop characters. Currently we only do that when a firmware compatible with OPAL v2 APIs is detected, in the code that calls opal_console_write_buffer_space(), which didn't exist with OPAL v1 (or didn't work). However, when enabling early debug consoles, the flag indicating that v2 is supported isn't set yet, causing us, in case of errors or closed console, to spin forever. Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/platforms/powernv/opal.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/arch/powerpc/platforms/powernv/opal.c b/arch/powerpc/platforms/powernv/opal.c index ade4463226c6..12d9846aa87e 100644 --- a/arch/powerpc/platforms/powernv/opal.c +++ b/arch/powerpc/platforms/powernv/opal.c @@ -144,6 +144,13 @@ int opal_put_chars(uint32_t vtermno, const char *data, int total_len) rc == OPAL_BUSY_EVENT || rc == OPAL_SUCCESS)) { len = total_len; rc = opal_console_write(vtermno, &len, data); + + /* Closed or other error drop */ + if (rc != OPAL_SUCCESS && rc != OPAL_BUSY && + rc != OPAL_BUSY_EVENT) { + written = total_len; + break; + } if (rc == OPAL_SUCCESS) { total_len -= len; data += len; From 49ece554288caf1a8ea9e546ab1ff5bc4b175456 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 08:37:00 -0500 Subject: [PATCH 031/198] rbd: fix leak of format 2 snapshot context When rbd_dev_v2_refresh() is called, the rbd device already has a snapshot context associated with it. But that never gets freed, the pointer just gets overwritten. Fix this by dropping the rbd device's reference to the snapshot context before overwriting the pointer. Because ceph_put_snap_context() already handles for a null pointer we don't need to check for that (for the probe case, where no context has yet been assigned). This resolves: http://tracker.ceph.com/issues/4912 Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index c2ca1818f335..426374321d75 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -4004,6 +4004,7 @@ static int rbd_dev_v2_snap_context(struct rbd_device *rbd_dev) for (i = 0; i < snap_count; i++) snapc->snaps[i] = ceph_decode_64(&p); + ceph_put_snap_context(rbd_dev->header.snapc); rbd_dev->header.snapc = snapc; dout(" snap context seq = %llu, snap_count = %u\n", From e627db085e0dab7744b68f3c927be6ed6df2f7f9 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 07:40:30 -0500 Subject: [PATCH 032/198] rbd: revalidate only for mapping size changes This commit: d98df63e rbd: revalidate_disk upon rbd resize instituted a call to revalidate_disk() to notify interested parties that a mapped image has changed size. This works well, as long as the the rbd device doesn't map a snapshot. A snapshot will never change size. However, the base image the snapshot is associated with can, and it can do so while the snapshot is mapped. The problem is that the test for the size is looking at the size of the base image, not the size of the mapped snapshot. This patch corrects that. Update the warning message shown in the event of error, and move it into the callers. This resolves: http://tracker.ceph.com/issues/4911 Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 16 +++++++++------- 1 file changed, 9 insertions(+), 7 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 426374321d75..5d5e3f0b5fb4 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -2628,6 +2628,7 @@ out: static void rbd_watch_cb(u64 ver, u64 notify_id, u8 opcode, void *data) { struct rbd_device *rbd_dev = (struct rbd_device *)data; + int ret; if (!rbd_dev) return; @@ -2635,7 +2636,9 @@ static void rbd_watch_cb(u64 ver, u64 notify_id, u8 opcode, void *data) dout("%s: \"%s\" notify_id %llu opcode %u\n", __func__, rbd_dev->header_name, (unsigned long long)notify_id, (unsigned int)opcode); - (void)rbd_dev_refresh(rbd_dev); + ret = rbd_dev_refresh(rbd_dev); + if (ret) + rbd_warn(rbd_dev, ": header refresh error (%d)\n", ret); rbd_obj_notify_ack(rbd_dev, notify_id); } @@ -3182,11 +3185,11 @@ static void rbd_exists_validate(struct rbd_device *rbd_dev) static int rbd_dev_refresh(struct rbd_device *rbd_dev) { - u64 image_size; + u64 mapping_size; int ret; rbd_assert(rbd_image_format_valid(rbd_dev->image_format)); - image_size = rbd_dev->header.image_size; + mapping_size = rbd_dev->mapping.size; mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING); if (rbd_dev->image_format == 1) ret = rbd_dev_v1_refresh(rbd_dev); @@ -3197,10 +3200,7 @@ static int rbd_dev_refresh(struct rbd_device *rbd_dev) rbd_exists_validate(rbd_dev); mutex_unlock(&ctl_mutex); - if (ret) - rbd_warn(rbd_dev, "got notification but failed to " - " update snaps: %d\n", ret); - if (image_size != rbd_dev->header.image_size) + if (mapping_size != rbd_dev->mapping.size) revalidate_disk(rbd_dev->disk); return ret; @@ -3405,6 +3405,8 @@ static ssize_t rbd_image_refresh(struct device *dev, int ret; ret = rbd_dev_refresh(rbd_dev); + if (ret) + rbd_warn(rbd_dev, ": manual header refresh error (%d)\n", ret); return ret < 0 ? ret : size; } From 00a653e216a8427547774ab3f2cc92709c3e28c9 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 07:40:30 -0500 Subject: [PATCH 033/198] rbd: update capacity in rbd_dev_refresh() When a mapped image changes size, we change the capacity recorded for the Linux disk associated with it, in rbd_update_mapping_size(). That function is called in two places--the format 1 and format 2 refresh routines. There is no need to set the capacity while holding the header semaphore. Instead, do it in the common rbd_dev_refresh(), using the logic that's already there to initiate disk revalidation. Add handling in the request function, just in case a request that exceeds the capacity of the device comes in (perhaps one that was started before a refresh shrunk the device). Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 23 +++++++++++++++-------- 1 file changed, 15 insertions(+), 8 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 5d5e3f0b5fb4..9c094c67d778 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -2874,6 +2874,13 @@ static void rbd_request_fn(struct request_queue *q) goto end_request; /* Shouldn't happen */ } + result = -EIO; + if (offset + length > rbd_dev->mapping.size) { + rbd_warn(rbd_dev, "beyond EOD (%llu~%llu > %llu)\n", + offset, length, rbd_dev->mapping.size); + goto end_request; + } + result = -ENOMEM; img_request = rbd_img_request_create(rbd_dev, offset, length, write_request, false); @@ -3116,14 +3123,8 @@ static void rbd_update_mapping_size(struct rbd_device *rbd_dev) if (rbd_dev->spec->snap_id != CEPH_NOSNAP) return; - if (rbd_dev->mapping.size != rbd_dev->header.image_size) { - sector_t size; - + if (rbd_dev->mapping.size != rbd_dev->header.image_size) rbd_dev->mapping.size = rbd_dev->header.image_size; - size = (sector_t)rbd_dev->mapping.size / SECTOR_SIZE; - dout("setting size to %llu sectors", (unsigned long long)size); - set_capacity(rbd_dev->disk, size); - } } /* @@ -3200,8 +3201,14 @@ static int rbd_dev_refresh(struct rbd_device *rbd_dev) rbd_exists_validate(rbd_dev); mutex_unlock(&ctl_mutex); - if (mapping_size != rbd_dev->mapping.size) + if (mapping_size != rbd_dev->mapping.size) { + sector_t size; + + size = (sector_t)rbd_dev->mapping.size / SECTOR_SIZE; + dout("setting size to %llu sectors", (unsigned long long)size); + set_capacity(rbd_dev->disk, size); revalidate_disk(rbd_dev->disk); + } return ret; } From 29334ba49c3e3defd9a2697cd4a199c597c30dc9 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 07:40:30 -0500 Subject: [PATCH 034/198] rbd: kill rbd_update_mapping_size() Since rbd_update_mapping_size() is now a trivial wrapper, just open code it in its two callers. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 17 ++++++----------- 1 file changed, 6 insertions(+), 11 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 9c094c67d778..4c9869545073 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -3118,15 +3118,6 @@ static int rbd_read_header(struct rbd_device *rbd_dev, return ret; } -static void rbd_update_mapping_size(struct rbd_device *rbd_dev) -{ - if (rbd_dev->spec->snap_id != CEPH_NOSNAP) - return; - - if (rbd_dev->mapping.size != rbd_dev->header.image_size) - rbd_dev->mapping.size = rbd_dev->header.image_size; -} - /* * only read the first part of the ondisk header, without the snaps info */ @@ -3143,7 +3134,9 @@ static int rbd_dev_v1_refresh(struct rbd_device *rbd_dev) /* Update image size, and check for resize of mapped image */ rbd_dev->header.image_size = h.image_size; - rbd_update_mapping_size(rbd_dev); + if (rbd_dev->spec->snap_id == CEPH_NOSNAP) + if (rbd_dev->mapping.size != rbd_dev->header.image_size) + rbd_dev->mapping.size = rbd_dev->header.image_size; /* rbd_dev->header.object_prefix shouldn't change */ kfree(rbd_dev->header.snap_sizes); @@ -4074,7 +4067,9 @@ static int rbd_dev_v2_refresh(struct rbd_device *rbd_dev) ret = rbd_dev_v2_image_size(rbd_dev); if (ret) goto out; - rbd_update_mapping_size(rbd_dev); + if (rbd_dev->spec->snap_id == CEPH_NOSNAP) + if (rbd_dev->mapping.size != rbd_dev->header.image_size) + rbd_dev->mapping.size = rbd_dev->header.image_size; ret = rbd_dev_v2_snap_context(rbd_dev); dout("rbd_dev_v2_snap_context returned %d\n", ret); From c734b79655a91a24afcae73738a43a0db09a801a Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 07:40:30 -0500 Subject: [PATCH 035/198] rbd: don't print warning if not mapping a parent The presence of the LAYERING bit in an rbd image's feature mask does not guarantee the image actually has a parent image. Currently that bit is set only when a clone (i.e., image with a parent) is created, but it is (currently) not cleared if that clone gets flattened back into a "normal" image. A "parent_id" query will leave the parent_spec for the image being mapped a null pointer, but will not return an error. Currently, whenever an image with the LAYERED feature gets mapped, a warning about the use of layered images gets printed. But we don't want to do this for a flattened image, so print the warning only if we find there is a parent spec after the probe. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 4c9869545073..2d34aea772be 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -4567,13 +4567,14 @@ static int rbd_dev_v2_probe(struct rbd_device *rbd_dev) ret = rbd_dev_v2_parent_info(rbd_dev); if (ret) goto out_err; - /* - * Don't print a warning for parent images. We can - * tell this point because we won't know its pool - * name yet (just its pool id). + * Print a warning if this image has a parent. + * Don't print it if the image now being probed + * is itself a parent. We can tell at this point + * because we won't know its pool name yet (just its + * pool id). */ - if (rbd_dev->spec->pool_name) + if (rbd_dev->parent_spec && rbd_dev->spec->pool_name) rbd_warn(rbd_dev, "WARNING: kernel layering " "is EXPERIMENTAL!"); } From 8f4b7d9821715767ac28bbc2d401bbb5f3f9a448 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 07:40:30 -0500 Subject: [PATCH 036/198] rbd: don't look up snapshot id in rbd_dev_mapping_set() Currently rbd_dev_mapping_set() looks up the snapshot id for the snapshot whose name is found in the rbd device's spec structure. That function gets called by rbd_dev_device_setup(), which is called by rbd_add() *after* rbd_dev_image_probe(). If the image probe succeeds, the rbd device's spec will already have been updated to include names and ids for all fields. Therefore there's no need to look up the snapshot id in rbd_dev_mapping_set(). Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 11 +---------- 1 file changed, 1 insertion(+), 10 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 2d34aea772be..75a8cea1a2d0 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -936,20 +936,11 @@ static int rbd_snap_features(struct rbd_device *rbd_dev, u64 snap_id, static int rbd_dev_mapping_set(struct rbd_device *rbd_dev) { - const char *snap_name = rbd_dev->spec->snap_name; - u64 snap_id; + u64 snap_id = rbd_dev->spec->snap_id; u64 size = 0; u64 features = 0; int ret; - if (strcmp(snap_name, RBD_SNAP_HEAD_NAME)) { - snap_id = rbd_snap_id_by_name(rbd_dev, snap_name); - if (snap_id == CEPH_NOSNAP) - return -ENOENT; - } else { - snap_id = CEPH_NOSNAP; - } - ret = rbd_snap_size(rbd_dev, snap_id, &size); if (ret) return ret; From 6d80b130d516deef51666e210fde674c947b8b5c Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 07:40:30 -0500 Subject: [PATCH 037/198] rbd: kill rbd_dev_clear_mapping() This function is a duplicate of rbd_dev_mapping_clear(), and was added by mistake. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 9 +-------- 1 file changed, 1 insertion(+), 8 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 75a8cea1a2d0..a62e59a0aea7 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -966,13 +966,6 @@ static void rbd_dev_mapping_clear(struct rbd_device *rbd_dev) rbd_dev->mapping.read_only = true; } -static void rbd_dev_clear_mapping(struct rbd_device *rbd_dev) -{ - rbd_dev->mapping.size = 0; - rbd_dev->mapping.features = 0; - rbd_dev->mapping.read_only = true; -} - static const char *rbd_segment_name(struct rbd_device *rbd_dev, u64 offset) { char *name; @@ -4910,7 +4903,7 @@ static void rbd_dev_device_release(struct device *dev) rbd_free_disk(rbd_dev); clear_bit(RBD_DEV_FLAG_EXISTS, &rbd_dev->flags); - rbd_dev_clear_mapping(rbd_dev); + rbd_dev_mapping_clear(rbd_dev); unregister_blkdev(rbd_dev->major, rbd_dev->name); rbd_dev->major = 0; rbd_dev_id_put(rbd_dev); From 51344a38ba2033be18a4ec23e318845caeccdc04 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 07:40:30 -0500 Subject: [PATCH 038/198] rbd: always set read-only flag in rbd_add() Hold off setting the read-only flag in rbd_add() for an image being mapped until we have successfully probed the image. At that point we know whether it's a snapshot mapping or not, so we can set the read-only flag in that one place rather than doing so (for snapshots) in rbd_dev_mapping_set(). To do this, pass a flag to the image probe routine indicating whether we want a read-only mapping. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 28 ++++++++++++++-------------- 1 file changed, 14 insertions(+), 14 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index a62e59a0aea7..b5cac7931ffc 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -359,7 +359,7 @@ static ssize_t rbd_add(struct bus_type *bus, const char *buf, size_t count); static ssize_t rbd_remove(struct bus_type *bus, const char *buf, size_t count); -static int rbd_dev_image_probe(struct rbd_device *rbd_dev); +static int rbd_dev_image_probe(struct rbd_device *rbd_dev, bool read_only); static struct bus_attribute rbd_bus_attrs[] = { __ATTR(add, S_IWUSR, NULL, rbd_add), @@ -951,11 +951,6 @@ static int rbd_dev_mapping_set(struct rbd_device *rbd_dev) rbd_dev->mapping.size = size; rbd_dev->mapping.features = features; - /* If we are mapping a snapshot it must be marked read-only */ - - if (snap_id != CEPH_NOSNAP) - rbd_dev->mapping.read_only = true; - return 0; } @@ -963,7 +958,6 @@ static void rbd_dev_mapping_clear(struct rbd_device *rbd_dev) { rbd_dev->mapping.size = 0; rbd_dev->mapping.features = 0; - rbd_dev->mapping.read_only = true; } static const char *rbd_segment_name(struct rbd_device *rbd_dev, u64 offset) @@ -4620,7 +4614,7 @@ static int rbd_dev_probe_parent(struct rbd_device *rbd_dev) if (!parent) goto out_err; - ret = rbd_dev_image_probe(parent); + ret = rbd_dev_image_probe(parent, true); if (ret < 0) goto out_err; rbd_dev->parent = parent; @@ -4743,7 +4737,7 @@ static void rbd_dev_image_release(struct rbd_device *rbd_dev) * device. For format 2 images this includes determining the image * id. */ -static int rbd_dev_image_probe(struct rbd_device *rbd_dev) +static int rbd_dev_image_probe(struct rbd_device *rbd_dev, bool read_only) { int ret; int tmp; @@ -4778,6 +4772,12 @@ static int rbd_dev_image_probe(struct rbd_device *rbd_dev) if (ret) goto err_out_probe; + /* If we are mapping a snapshot it must be marked read-only */ + + if (rbd_dev->spec->snap_id != CEPH_NOSNAP) + read_only = true; + rbd_dev->mapping.read_only = read_only; + ret = rbd_dev_probe_parent(rbd_dev); if (!ret) return 0; @@ -4811,6 +4811,7 @@ static ssize_t rbd_add(struct bus_type *bus, struct rbd_spec *spec = NULL; struct rbd_client *rbdc; struct ceph_osd_client *osdc; + bool read_only; int rc = -ENOMEM; if (!try_module_get(THIS_MODULE)) @@ -4820,6 +4821,9 @@ static ssize_t rbd_add(struct bus_type *bus, rc = rbd_add_parse_args(buf, &ceph_opts, &rbd_opts, &spec); if (rc < 0) goto err_out_module; + read_only = rbd_opts->read_only; + kfree(rbd_opts); + rbd_opts = NULL; /* done with this */ rbdc = rbd_get_client(ceph_opts); if (IS_ERR(rbdc)) { @@ -4850,11 +4854,7 @@ static ssize_t rbd_add(struct bus_type *bus, rbdc = NULL; /* rbd_dev now owns this */ spec = NULL; /* rbd_dev now owns this */ - rbd_dev->mapping.read_only = rbd_opts->read_only; - kfree(rbd_opts); - rbd_opts = NULL; /* done with this */ - - rc = rbd_dev_image_probe(rbd_dev); + rc = rbd_dev_image_probe(rbd_dev, read_only); if (rc < 0) goto err_out_rbd_dev; From 73b82bf0bfbf58e6ff328d3726934370585f6e78 Mon Sep 17 00:00:00 2001 From: Thommy Jakobsson Date: Tue, 23 Apr 2013 21:45:11 +0200 Subject: [PATCH 039/198] B43: Handle DMA RX descriptor underrun Add handling of rx descriptor underflow. This fixes a fault that could happen on slow machines, where data is received faster than the CPU can handle. In such a case the device will use up all rx descriptors and refuse to send any more data before confirming that it is ok. This patch enables necessary interrupt to discover such a situation and will handle them by dropping everything in the ring buffer. Reviewed-by: Michael Buesch Signed-off-by: Thommy Jakobsson Cc: stable Signed-off-by: John W. Linville --- drivers/net/wireless/b43/dma.c | 19 +++++++++++++++ drivers/net/wireless/b43/dma.h | 4 ++- drivers/net/wireless/b43/main.c | 43 ++++++++++++++------------------- 3 files changed, 40 insertions(+), 26 deletions(-) diff --git a/drivers/net/wireless/b43/dma.c b/drivers/net/wireless/b43/dma.c index 122146943bf2..ee3d6403c795 100644 --- a/drivers/net/wireless/b43/dma.c +++ b/drivers/net/wireless/b43/dma.c @@ -1733,6 +1733,25 @@ drop_recycle_buffer: sync_descbuffer_for_device(ring, dmaaddr, ring->rx_buffersize); } +void b43_dma_handle_rx_overflow(struct b43_dmaring *ring) +{ + int current_slot, previous_slot; + + B43_WARN_ON(ring->tx); + + /* Device has filled all buffers, drop all packets and let TCP + * decrease speed. + * Decrement RX index by one will let the device to see all slots + * as free again + */ + /* + *TODO: How to increase rx_drop in mac80211? + */ + current_slot = ring->ops->get_current_rxslot(ring); + previous_slot = prev_slot(ring, current_slot); + ring->ops->set_current_rxslot(ring, previous_slot); +} + void b43_dma_rx(struct b43_dmaring *ring) { const struct b43_dma_ops *ops = ring->ops; diff --git a/drivers/net/wireless/b43/dma.h b/drivers/net/wireless/b43/dma.h index 9fdd1983079c..df8c8cdcbdb5 100644 --- a/drivers/net/wireless/b43/dma.h +++ b/drivers/net/wireless/b43/dma.h @@ -9,7 +9,7 @@ /* DMA-Interrupt reasons. */ #define B43_DMAIRQ_FATALMASK ((1 << 10) | (1 << 11) | (1 << 12) \ | (1 << 14) | (1 << 15)) -#define B43_DMAIRQ_NONFATALMASK (1 << 13) +#define B43_DMAIRQ_RDESC_UFLOW (1 << 13) #define B43_DMAIRQ_RX_DONE (1 << 16) /*** 32-bit DMA Engine. ***/ @@ -295,6 +295,8 @@ int b43_dma_tx(struct b43_wldev *dev, void b43_dma_handle_txstatus(struct b43_wldev *dev, const struct b43_txstatus *status); +void b43_dma_handle_rx_overflow(struct b43_dmaring *ring); + void b43_dma_rx(struct b43_dmaring *ring); void b43_dma_direct_fifo_rx(struct b43_wldev *dev, diff --git a/drivers/net/wireless/b43/main.c b/drivers/net/wireless/b43/main.c index d377f77d30b5..6dd07e2ec595 100644 --- a/drivers/net/wireless/b43/main.c +++ b/drivers/net/wireless/b43/main.c @@ -1902,30 +1902,18 @@ static void b43_do_interrupt_thread(struct b43_wldev *dev) } } - if (unlikely(merged_dma_reason & (B43_DMAIRQ_FATALMASK | - B43_DMAIRQ_NONFATALMASK))) { - if (merged_dma_reason & B43_DMAIRQ_FATALMASK) { - b43err(dev->wl, "Fatal DMA error: " - "0x%08X, 0x%08X, 0x%08X, " - "0x%08X, 0x%08X, 0x%08X\n", - dma_reason[0], dma_reason[1], - dma_reason[2], dma_reason[3], - dma_reason[4], dma_reason[5]); - b43err(dev->wl, "This device does not support DMA " + if (unlikely(merged_dma_reason & (B43_DMAIRQ_FATALMASK))) { + b43err(dev->wl, + "Fatal DMA error: 0x%08X, 0x%08X, 0x%08X, 0x%08X, 0x%08X, 0x%08X\n", + dma_reason[0], dma_reason[1], + dma_reason[2], dma_reason[3], + dma_reason[4], dma_reason[5]); + b43err(dev->wl, "This device does not support DMA " "on your system. It will now be switched to PIO.\n"); - /* Fall back to PIO transfers if we get fatal DMA errors! */ - dev->use_pio = true; - b43_controller_restart(dev, "DMA error"); - return; - } - if (merged_dma_reason & B43_DMAIRQ_NONFATALMASK) { - b43err(dev->wl, "DMA error: " - "0x%08X, 0x%08X, 0x%08X, " - "0x%08X, 0x%08X, 0x%08X\n", - dma_reason[0], dma_reason[1], - dma_reason[2], dma_reason[3], - dma_reason[4], dma_reason[5]); - } + /* Fall back to PIO transfers if we get fatal DMA errors! */ + dev->use_pio = true; + b43_controller_restart(dev, "DMA error"); + return; } if (unlikely(reason & B43_IRQ_UCODE_DEBUG)) @@ -1944,6 +1932,11 @@ static void b43_do_interrupt_thread(struct b43_wldev *dev) handle_irq_noise(dev); /* Check the DMA reason registers for received data. */ + if (dma_reason[0] & B43_DMAIRQ_RDESC_UFLOW) { + if (B43_DEBUG) + b43warn(dev->wl, "RX descriptor underrun\n"); + b43_dma_handle_rx_overflow(dev->dma.rx_ring); + } if (dma_reason[0] & B43_DMAIRQ_RX_DONE) { if (b43_using_pio_transfers(dev)) b43_pio_rx(dev->pio.rx_queue); @@ -2001,7 +1994,7 @@ static irqreturn_t b43_do_interrupt(struct b43_wldev *dev) return IRQ_NONE; dev->dma_reason[0] = b43_read32(dev, B43_MMIO_DMA0_REASON) - & 0x0001DC00; + & 0x0001FC00; dev->dma_reason[1] = b43_read32(dev, B43_MMIO_DMA1_REASON) & 0x0000DC00; dev->dma_reason[2] = b43_read32(dev, B43_MMIO_DMA2_REASON) @@ -3130,7 +3123,7 @@ static int b43_chip_init(struct b43_wldev *dev) b43_write32(dev, 0x018C, 0x02000000); } b43_write32(dev, B43_MMIO_GEN_IRQ_REASON, 0x00004000); - b43_write32(dev, B43_MMIO_DMA0_IRQ_MASK, 0x0001DC00); + b43_write32(dev, B43_MMIO_DMA0_IRQ_MASK, 0x0001FC00); b43_write32(dev, B43_MMIO_DMA1_IRQ_MASK, 0x0000DC00); b43_write32(dev, B43_MMIO_DMA2_IRQ_MASK, 0x0000DC00); b43_write32(dev, B43_MMIO_DMA3_IRQ_MASK, 0x0001DC00); From dd9c46408fdc07098333655ff27edf8cac8d9fcf Mon Sep 17 00:00:00 2001 From: Stanislaw Gruszka Date: Tue, 7 May 2013 18:07:06 +0200 Subject: [PATCH 040/198] iwl4965: workaround connection regression on passive channel Jake reported that since commit 1672c0e31917f49d31d30d79067103432bc20cc7 "mac80211: start auth/assoc timeout on frame status", he is unable to connect to his AP, which is configured to use passive channel. After switch to passive channel 4965 firmware drops any TX packet until it receives beacon. Before commit 1672c0e3 we waited on channel and retransmit packet after 200ms, that makes we receive beacon on the meantime and association process succeed. New mac80211 behaviour cause that any ASSOC frame fail immediately on iwl4965 and we can not associate. This patch restore old mac80211 behaviour for iwl4965, by removing IEEE80211_HW_REPORTS_TX_ACK_STATUS feature. This feature will be added again to iwl4965 driver, when different, more complex workaround for this firmware issue, will be added to the driver. Cc: stable@vger.kernel.org # 3.9 Bisected-by: Jake Edge Reported-and-tested-by: Jake Edge Signed-off-by: Stanislaw Gruszka Signed-off-by: John W. Linville --- drivers/net/wireless/iwlegacy/4965-mac.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/net/wireless/iwlegacy/4965-mac.c b/drivers/net/wireless/iwlegacy/4965-mac.c index 431ae6cc5f8f..289701e814c7 100644 --- a/drivers/net/wireless/iwlegacy/4965-mac.c +++ b/drivers/net/wireless/iwlegacy/4965-mac.c @@ -5741,8 +5741,7 @@ il4965_mac_setup_register(struct il_priv *il, u32 max_probe_length) hw->flags = IEEE80211_HW_SIGNAL_DBM | IEEE80211_HW_AMPDU_AGGREGATION | IEEE80211_HW_NEED_DTIM_BEFORE_ASSOC | IEEE80211_HW_SPECTRUM_MGMT | - IEEE80211_HW_REPORTS_TX_ACK_STATUS | IEEE80211_HW_SUPPORTS_PS | - IEEE80211_HW_SUPPORTS_DYNAMIC_PS; + IEEE80211_HW_SUPPORTS_PS | IEEE80211_HW_SUPPORTS_DYNAMIC_PS; if (il->cfg->sku & IL_SKU_N) hw->flags |= IEEE80211_HW_SUPPORTS_DYNAMIC_SMPS | From a67682804165c428d5fd52c1a70fcd2a7502c056 Mon Sep 17 00:00:00 2001 From: Sujith Manoharan Date: Mon, 6 May 2013 10:09:03 +0530 Subject: [PATCH 041/198] ath9k: Fix beacon reconfiguration When a reset or a channel-change happens, for managed mode, the HW beacon timers have to be programmed after the TSF has been synchronized. This is handled via the sync flags. Signed-off-by: Sujith Manoharan Signed-off-by: John W. Linville --- drivers/net/wireless/ath/ath9k/main.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/wireless/ath/ath9k/main.c b/drivers/net/wireless/ath/ath9k/main.c index c4fc3d14ea62..a18414b5948b 100644 --- a/drivers/net/wireless/ath/ath9k/main.c +++ b/drivers/net/wireless/ath/ath9k/main.c @@ -227,13 +227,13 @@ static bool ath_complete_reset(struct ath_softc *sc, bool start) if (!test_bit(SC_OP_BEACONS, &sc->sc_flags)) goto work; - ath9k_set_beacon(sc); - if (ah->opmode == NL80211_IFTYPE_STATION && test_bit(SC_OP_PRIM_STA_VIF, &sc->sc_flags)) { spin_lock_irqsave(&sc->sc_pm_lock, flags); sc->ps_flags |= PS_BEACON_SYNC | PS_WAIT_FOR_BEACON; spin_unlock_irqrestore(&sc->sc_pm_lock, flags); + } else { + ath9k_set_beacon(sc); } work: ath_restart_work(sc); From f0fb6af2918ea6b6f781f7d598d7cd72cea28915 Mon Sep 17 00:00:00 2001 From: Sujith Manoharan Date: Mon, 6 May 2013 10:09:33 +0530 Subject: [PATCH 042/198] ath9k: Update initvals for AR9565 * Register Modification for xLNA board. * TX gain table modification for zero calibration. * AUX chain (LNA2) sensitivity enhancement * Modify diversity bias default setting in INI. Signed-off-by: Sujith Manoharan Signed-off-by: John W. Linville --- .../wireless/ath/ath9k/ar9565_1p0_initvals.h | 138 +++++++++--------- 1 file changed, 70 insertions(+), 68 deletions(-) diff --git a/drivers/net/wireless/ath/ath9k/ar9565_1p0_initvals.h b/drivers/net/wireless/ath/ath9k/ar9565_1p0_initvals.h index 0c2ac0c6dc89..e85a8b076c22 100644 --- a/drivers/net/wireless/ath/ath9k/ar9565_1p0_initvals.h +++ b/drivers/net/wireless/ath/ath9k/ar9565_1p0_initvals.h @@ -233,9 +233,9 @@ static const u32 ar9565_1p0_baseband_core[][2] = { {0x00009d10, 0x01834061}, {0x00009d14, 0x00c00400}, {0x00009d18, 0x00000000}, - {0x00009e08, 0x0078230c}, - {0x00009e24, 0x990bb515}, - {0x00009e28, 0x126f0000}, + {0x00009e08, 0x0038230c}, + {0x00009e24, 0x9907b515}, + {0x00009e28, 0x126f0600}, {0x00009e30, 0x06336f77}, {0x00009e34, 0x6af6532f}, {0x00009e38, 0x0cc80c00}, @@ -337,7 +337,7 @@ static const u32 ar9565_1p0_baseband_core[][2] = { static const u32 ar9565_1p0_baseband_postamble[][5] = { /* Addr 5G_HT20 5G_HT40 2G_HT40 2G_HT20 */ - {0x00009810, 0xd00a8005, 0xd00a8005, 0xd00a8005, 0xd00a800d}, + {0x00009810, 0xd00a8005, 0xd00a8005, 0xd00a8005, 0xd00a8009}, {0x00009820, 0x206a022e, 0x206a022e, 0x206a012e, 0x206a01ae}, {0x00009824, 0x5ac640d0, 0x5ac640d0, 0x5ac640d0, 0x63c640da}, {0x00009828, 0x06903081, 0x06903081, 0x06903881, 0x09143c81}, @@ -345,9 +345,9 @@ static const u32 ar9565_1p0_baseband_postamble[][5] = { {0x00009830, 0x0000059c, 0x0000059c, 0x0000059c, 0x0000059c}, {0x00009c00, 0x000000c4, 0x000000c4, 0x000000c4, 0x000000c4}, {0x00009e00, 0x0372111a, 0x0372111a, 0x037216a0, 0x037216a0}, - {0x00009e04, 0x00802020, 0x00802020, 0x00802020, 0x00802020}, - {0x00009e0c, 0x6c4000e2, 0x6d4000e2, 0x6d4000e2, 0x6c4000d8}, - {0x00009e10, 0x7ec88d2e, 0x7ec88d2e, 0x7ec84d2e, 0x7ec86d2e}, + {0x00009e04, 0x00802020, 0x00802020, 0x00142020, 0x00142020}, + {0x00009e0c, 0x6c4000e2, 0x6d4000e2, 0x6d4000e2, 0x6c4000e2}, + {0x00009e10, 0x7ec88d2e, 0x7ec88d2e, 0x7ec84d2e, 0x7ec84d2e}, {0x00009e14, 0x37b95d5e, 0x37b9605e, 0x3379605e, 0x33795d5e}, {0x00009e18, 0x00000000, 0x00000000, 0x00000000, 0x00000000}, {0x00009e1c, 0x0001cf9c, 0x0001cf9c, 0x00021f9c, 0x00021f9c}, @@ -450,6 +450,8 @@ static const u32 ar9565_1p0_soc_postamble[][5] = { static const u32 ar9565_1p0_Common_rx_gain_table[][2] = { /* Addr allmodes */ + {0x00004050, 0x00300300}, + {0x0000406c, 0x00100000}, {0x0000a000, 0x00010000}, {0x0000a004, 0x00030002}, {0x0000a008, 0x00050004}, @@ -498,27 +500,27 @@ static const u32 ar9565_1p0_Common_rx_gain_table[][2] = { {0x0000a0b4, 0x00000000}, {0x0000a0b8, 0x00000000}, {0x0000a0bc, 0x00000000}, - {0x0000a0c0, 0x001f0000}, - {0x0000a0c4, 0x01000101}, - {0x0000a0c8, 0x011e011f}, - {0x0000a0cc, 0x011c011d}, - {0x0000a0d0, 0x02030204}, - {0x0000a0d4, 0x02010202}, - {0x0000a0d8, 0x021f0200}, - {0x0000a0dc, 0x0302021e}, - {0x0000a0e0, 0x03000301}, - {0x0000a0e4, 0x031e031f}, - {0x0000a0e8, 0x0402031d}, - {0x0000a0ec, 0x04000401}, - {0x0000a0f0, 0x041e041f}, - {0x0000a0f4, 0x0502041d}, - {0x0000a0f8, 0x05000501}, - {0x0000a0fc, 0x051e051f}, - {0x0000a100, 0x06010602}, - {0x0000a104, 0x061f0600}, - {0x0000a108, 0x061d061e}, - {0x0000a10c, 0x07020703}, - {0x0000a110, 0x07000701}, + {0x0000a0c0, 0x00bf00a0}, + {0x0000a0c4, 0x11a011a1}, + {0x0000a0c8, 0x11be11bf}, + {0x0000a0cc, 0x11bc11bd}, + {0x0000a0d0, 0x22632264}, + {0x0000a0d4, 0x22612262}, + {0x0000a0d8, 0x227f2260}, + {0x0000a0dc, 0x4322227e}, + {0x0000a0e0, 0x43204321}, + {0x0000a0e4, 0x433e433f}, + {0x0000a0e8, 0x4462433d}, + {0x0000a0ec, 0x44604461}, + {0x0000a0f0, 0x447e447f}, + {0x0000a0f4, 0x5582447d}, + {0x0000a0f8, 0x55805581}, + {0x0000a0fc, 0x559e559f}, + {0x0000a100, 0x66816682}, + {0x0000a104, 0x669f6680}, + {0x0000a108, 0x669d669e}, + {0x0000a10c, 0x77627763}, + {0x0000a110, 0x77607761}, {0x0000a114, 0x00000000}, {0x0000a118, 0x00000000}, {0x0000a11c, 0x00000000}, @@ -530,27 +532,27 @@ static const u32 ar9565_1p0_Common_rx_gain_table[][2] = { {0x0000a134, 0x00000000}, {0x0000a138, 0x00000000}, {0x0000a13c, 0x00000000}, - {0x0000a140, 0x001f0000}, - {0x0000a144, 0x01000101}, - {0x0000a148, 0x011e011f}, - {0x0000a14c, 0x011c011d}, - {0x0000a150, 0x02030204}, - {0x0000a154, 0x02010202}, - {0x0000a158, 0x021f0200}, - {0x0000a15c, 0x0302021e}, - {0x0000a160, 0x03000301}, - {0x0000a164, 0x031e031f}, - {0x0000a168, 0x0402031d}, - {0x0000a16c, 0x04000401}, - {0x0000a170, 0x041e041f}, - {0x0000a174, 0x0502041d}, - {0x0000a178, 0x05000501}, - {0x0000a17c, 0x051e051f}, - {0x0000a180, 0x06010602}, - {0x0000a184, 0x061f0600}, - {0x0000a188, 0x061d061e}, - {0x0000a18c, 0x07020703}, - {0x0000a190, 0x07000701}, + {0x0000a140, 0x00bf00a0}, + {0x0000a144, 0x11a011a1}, + {0x0000a148, 0x11be11bf}, + {0x0000a14c, 0x11bc11bd}, + {0x0000a150, 0x22632264}, + {0x0000a154, 0x22612262}, + {0x0000a158, 0x227f2260}, + {0x0000a15c, 0x4322227e}, + {0x0000a160, 0x43204321}, + {0x0000a164, 0x433e433f}, + {0x0000a168, 0x4462433d}, + {0x0000a16c, 0x44604461}, + {0x0000a170, 0x447e447f}, + {0x0000a174, 0x5582447d}, + {0x0000a178, 0x55805581}, + {0x0000a17c, 0x559e559f}, + {0x0000a180, 0x66816682}, + {0x0000a184, 0x669f6680}, + {0x0000a188, 0x669d669e}, + {0x0000a18c, 0x77e677e7}, + {0x0000a190, 0x77e477e5}, {0x0000a194, 0x00000000}, {0x0000a198, 0x00000000}, {0x0000a19c, 0x00000000}, @@ -770,7 +772,7 @@ static const u32 ar9565_1p0_Modes_lowest_ob_db_tx_gain_table[][5] = { static const u32 ar9565_1p0_pciephy_clkreq_disable_L1[][2] = { /* Addr allmodes */ - {0x00018c00, 0x18213ede}, + {0x00018c00, 0x18212ede}, {0x00018c04, 0x000801d8}, {0x00018c08, 0x0003780c}, }; @@ -889,8 +891,8 @@ static const u32 ar9565_1p0_common_wo_xlna_rx_gain_table[][2] = { {0x0000a180, 0x66816682}, {0x0000a184, 0x669f6680}, {0x0000a188, 0x669d669e}, - {0x0000a18c, 0x77627763}, - {0x0000a190, 0x77607761}, + {0x0000a18c, 0x77e677e7}, + {0x0000a190, 0x77e477e5}, {0x0000a194, 0x00000000}, {0x0000a198, 0x00000000}, {0x0000a19c, 0x00000000}, @@ -1114,7 +1116,7 @@ static const u32 ar9565_1p0_modes_high_ob_db_tx_gain_table[][5] = { {0x0000a2e0, 0xffecec00, 0xffecec00, 0xfd339c84, 0xfd339c84}, {0x0000a2e4, 0xfc0f0000, 0xfc0f0000, 0xfec3e000, 0xfec3e000}, {0x0000a2e8, 0xfc100000, 0xfc100000, 0xfffc0000, 0xfffc0000}, - {0x0000a410, 0x000050d9, 0x000050d9, 0x000050d9, 0x000050d9}, + {0x0000a410, 0x000050d9, 0x000050d9, 0x000050df, 0x000050df}, {0x0000a500, 0x00002220, 0x00002220, 0x00000000, 0x00000000}, {0x0000a504, 0x06002223, 0x06002223, 0x04000002, 0x04000002}, {0x0000a508, 0x0b022220, 0x0b022220, 0x08000004, 0x08000004}, @@ -1140,13 +1142,13 @@ static const u32 ar9565_1p0_modes_high_ob_db_tx_gain_table[][5] = { {0x0000a558, 0x69027f56, 0x69027f56, 0x53001ce5, 0x53001ce5}, {0x0000a55c, 0x6d029f56, 0x6d029f56, 0x57001ce9, 0x57001ce9}, {0x0000a560, 0x73049f56, 0x73049f56, 0x5b001ceb, 0x5b001ceb}, - {0x0000a564, 0x7804ff56, 0x7804ff56, 0x5d001eec, 0x5d001eec}, - {0x0000a568, 0x7804ff56, 0x7804ff56, 0x5d001eec, 0x5d001eec}, - {0x0000a56c, 0x7804ff56, 0x7804ff56, 0x5d001eec, 0x5d001eec}, - {0x0000a570, 0x7804ff56, 0x7804ff56, 0x5d001eec, 0x5d001eec}, - {0x0000a574, 0x7804ff56, 0x7804ff56, 0x5d001eec, 0x5d001eec}, - {0x0000a578, 0x7804ff56, 0x7804ff56, 0x5d001eec, 0x5d001eec}, - {0x0000a57c, 0x7804ff56, 0x7804ff56, 0x5d001eec, 0x5d001eec}, + {0x0000a564, 0x7804ff56, 0x7804ff56, 0x60001cf0, 0x60001cf0}, + {0x0000a568, 0x7804ff56, 0x7804ff56, 0x61001cf1, 0x61001cf1}, + {0x0000a56c, 0x7804ff56, 0x7804ff56, 0x62001cf2, 0x62001cf2}, + {0x0000a570, 0x7804ff56, 0x7804ff56, 0x63001cf3, 0x63001cf3}, + {0x0000a574, 0x7804ff56, 0x7804ff56, 0x64001cf4, 0x64001cf4}, + {0x0000a578, 0x7804ff56, 0x7804ff56, 0x66001ff6, 0x66001ff6}, + {0x0000a57c, 0x7804ff56, 0x7804ff56, 0x66001ff6, 0x66001ff6}, {0x0000a600, 0x00000000, 0x00000000, 0x00000000, 0x00000000}, {0x0000a604, 0x00000000, 0x00000000, 0x00000000, 0x00000000}, {0x0000a608, 0x00000000, 0x00000000, 0x00000000, 0x00000000}, @@ -1174,7 +1176,7 @@ static const u32 ar9565_1p0_modes_high_power_tx_gain_table[][5] = { {0x0000a2e0, 0xffecec00, 0xffecec00, 0xfd339c84, 0xfd339c84}, {0x0000a2e4, 0xfc0f0000, 0xfc0f0000, 0xfec3e000, 0xfec3e000}, {0x0000a2e8, 0xfc100000, 0xfc100000, 0xfffc0000, 0xfffc0000}, - {0x0000a410, 0x000050d9, 0x000050d9, 0x000050d9, 0x000050d9}, + {0x0000a410, 0x000050d9, 0x000050d9, 0x000050df, 0x000050df}, {0x0000a500, 0x00002220, 0x00002220, 0x00000000, 0x00000000}, {0x0000a504, 0x06002223, 0x06002223, 0x04000002, 0x04000002}, {0x0000a508, 0x0a022220, 0x0a022220, 0x08000004, 0x08000004}, @@ -1200,13 +1202,13 @@ static const u32 ar9565_1p0_modes_high_power_tx_gain_table[][5] = { {0x0000a558, 0x66027f56, 0x66027f56, 0x4c001ce5, 0x4c001ce5}, {0x0000a55c, 0x6a029f56, 0x6a029f56, 0x50001ce9, 0x50001ce9}, {0x0000a560, 0x70049f56, 0x70049f56, 0x54001ceb, 0x54001ceb}, - {0x0000a564, 0x7504ff56, 0x7504ff56, 0x56001eec, 0x56001eec}, - {0x0000a568, 0x7504ff56, 0x7504ff56, 0x56001eec, 0x56001eec}, - {0x0000a56c, 0x7504ff56, 0x7504ff56, 0x56001eec, 0x56001eec}, - {0x0000a570, 0x7504ff56, 0x7504ff56, 0x56001eec, 0x56001eec}, - {0x0000a574, 0x7504ff56, 0x7504ff56, 0x56001eec, 0x56001eec}, - {0x0000a578, 0x7504ff56, 0x7504ff56, 0x56001eec, 0x56001eec}, - {0x0000a57c, 0x7504ff56, 0x7504ff56, 0x56001eec, 0x56001eec}, + {0x0000a564, 0x7504ff56, 0x7504ff56, 0x59001cf0, 0x59001cf0}, + {0x0000a568, 0x7504ff56, 0x7504ff56, 0x5a001cf1, 0x5a001cf1}, + {0x0000a56c, 0x7504ff56, 0x7504ff56, 0x5b001cf2, 0x5b001cf2}, + {0x0000a570, 0x7504ff56, 0x7504ff56, 0x5c001cf3, 0x5c001cf3}, + {0x0000a574, 0x7504ff56, 0x7504ff56, 0x5d001cf4, 0x5d001cf4}, + {0x0000a578, 0x7504ff56, 0x7504ff56, 0x5f001ff6, 0x5f001ff6}, + {0x0000a57c, 0x7504ff56, 0x7504ff56, 0x5f001ff6, 0x5f001ff6}, {0x0000a600, 0x00000000, 0x00000000, 0x00000000, 0x00000000}, {0x0000a604, 0x00000000, 0x00000000, 0x00000000, 0x00000000}, {0x0000a608, 0x00000000, 0x00000000, 0x00000000, 0x00000000}, From 48795424acff7215d5eac0b52793a2c1eb3a6283 Mon Sep 17 00:00:00 2001 From: Bing Zhao Date: Mon, 6 May 2013 19:46:53 -0700 Subject: [PATCH 043/198] mwifiex: clear is_suspended flag when interrupt is received early When the XO-4 with 8787 wireless is woken up due to wake-on-WLAN mwifiex is often flooded with "not allowed while suspended" messages and the interface is unusable. [ 202.171609] int: sdio_ireg = 0x1 [ 202.180700] info: mwifiex_process_hs_config: auto cancelling host sleep since there is interrupt from the firmware [ 202.201880] event: wakeup device... [ 202.211452] event: hs_deactivated [ 202.514638] info: --- Rx: Data packet --- [ 202.514753] data: 4294957544 BSS(0-0): Data <= kernel [ 202.514825] PREP_CMD: device in suspended state [ 202.514839] data: dequeuing the packet ec7248c0 ec4869c0 [ 202.514886] mwifiex_write_data_sync: not allowed while suspended [ 202.514886] host_to_card, write iomem (1) failed: -1 [ 202.514917] mwifiex_write_data_sync: not allowed while suspended [ 202.514936] host_to_card, write iomem (2) failed: -1 [ 202.514949] mwifiex_write_data_sync: not allowed while suspended [ 202.514965] host_to_card, write iomem (3) failed: -1 [ 202.514976] mwifiex_write_data_async failed: 0xFFFFFFFF This can be readily reproduced when putting the XO-4 in a loop where it goes to sleep due to inactivity, but then wakes up due to an incoming ping. The error is hit within an hour or two. This issue happens when an interrupt comes in early while host sleep is still activated. Driver handles this case by auto cancelling host sleep. However is_suspended flag is still set which prevents any cmd or data from being sent to firmware. Fix it by clearing is_suspended flag in this path. Cc: Reported-by: Daniel Drake Tested-by: Daniel Drake Signed-off-by: Bing Zhao Signed-off-by: John W. Linville --- drivers/net/wireless/mwifiex/cmdevt.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/wireless/mwifiex/cmdevt.c b/drivers/net/wireless/mwifiex/cmdevt.c index 74db0d24a579..26755d9acb55 100644 --- a/drivers/net/wireless/mwifiex/cmdevt.c +++ b/drivers/net/wireless/mwifiex/cmdevt.c @@ -1191,6 +1191,7 @@ mwifiex_process_hs_config(struct mwifiex_adapter *adapter) adapter->if_ops.wakeup(adapter); adapter->hs_activated = false; adapter->is_hs_configured = false; + adapter->is_suspended = false; mwifiex_hs_activated_event(mwifiex_get_priv(adapter, MWIFIEX_BSS_ROLE_ANY), false); From f16fdc9d2dc1e5b270e9a08377587e831e0d36ac Mon Sep 17 00:00:00 2001 From: Amitkumar Karwar Date: Mon, 6 May 2013 19:46:54 -0700 Subject: [PATCH 044/198] mwifiex: fix memory leak issue when driver unload After unregister_netdevice() call the request is queued and reg_state is changed to NETREG_UNREGISTERING. As we check for NETREG_UNREGISTERED state, free_netdev() never gets executed causing memory leak. Initialize "dev->destructor" to free_netdev() to free device data after unregistration. Cc: Reported-by: Daniel Drake Tested-by: Daniel Drake Signed-off-by: Amitkumar Karwar Signed-off-by: Bing Zhao Signed-off-by: John W. Linville --- drivers/net/wireless/mwifiex/cfg80211.c | 3 --- drivers/net/wireless/mwifiex/main.c | 1 + 2 files changed, 1 insertion(+), 3 deletions(-) diff --git a/drivers/net/wireless/mwifiex/cfg80211.c b/drivers/net/wireless/mwifiex/cfg80211.c index a0cb0770d319..636145d1708b 100644 --- a/drivers/net/wireless/mwifiex/cfg80211.c +++ b/drivers/net/wireless/mwifiex/cfg80211.c @@ -2234,9 +2234,6 @@ int mwifiex_del_virtual_intf(struct wiphy *wiphy, struct wireless_dev *wdev) if (wdev->netdev->reg_state == NETREG_REGISTERED) unregister_netdevice(wdev->netdev); - if (wdev->netdev->reg_state == NETREG_UNREGISTERED) - free_netdev(wdev->netdev); - /* Clear the priv in adapter */ priv->netdev = NULL; diff --git a/drivers/net/wireless/mwifiex/main.c b/drivers/net/wireless/mwifiex/main.c index 121443a0f2a1..2eb88ea9acf7 100644 --- a/drivers/net/wireless/mwifiex/main.c +++ b/drivers/net/wireless/mwifiex/main.c @@ -655,6 +655,7 @@ void mwifiex_init_priv_params(struct mwifiex_private *priv, struct net_device *dev) { dev->netdev_ops = &mwifiex_netdev_ops; + dev->destructor = free_netdev; /* Initialize private structure */ priv->current_key_index = 0; priv->media_connected = false; From ccd384b10420ac81ba3fb9b0a7d18272c7173552 Mon Sep 17 00:00:00 2001 From: Daniel Drake Date: Wed, 8 May 2013 15:37:19 -0400 Subject: [PATCH 045/198] mwifiex: fix setting of multicast filter A small bug in this code was causing the ALLMULTI filter to be set when in fact we were just wanting to program a selective multicast list to the hardware. Fix that bug and remove a redundant if condition in the code that follows. This fixes wakeup behaviour when multicast WOL is enabled. Previously, all multicast packets would wake up the system. Now, only those that the host intended to receive trigger wakeups. Signed-off-by: Daniel Drake Cc: Acked-by: Bing Zhao Signed-off-by: John W. Linville --- drivers/net/wireless/mwifiex/sta_ioctl.c | 21 ++++++--------------- 1 file changed, 6 insertions(+), 15 deletions(-) diff --git a/drivers/net/wireless/mwifiex/sta_ioctl.c b/drivers/net/wireless/mwifiex/sta_ioctl.c index 311d0b26b81c..1a8a19dbd635 100644 --- a/drivers/net/wireless/mwifiex/sta_ioctl.c +++ b/drivers/net/wireless/mwifiex/sta_ioctl.c @@ -96,7 +96,7 @@ int mwifiex_request_set_multicast_list(struct mwifiex_private *priv, } else { /* Multicast */ priv->curr_pkt_filter &= ~HostCmd_ACT_MAC_PROMISCUOUS_ENABLE; - if (mcast_list->mode == MWIFIEX_MULTICAST_MODE) { + if (mcast_list->mode == MWIFIEX_ALL_MULTI_MODE) { dev_dbg(priv->adapter->dev, "info: Enabling All Multicast!\n"); priv->curr_pkt_filter |= @@ -108,20 +108,11 @@ int mwifiex_request_set_multicast_list(struct mwifiex_private *priv, dev_dbg(priv->adapter->dev, "info: Set multicast list=%d\n", mcast_list->num_multicast_addr); - /* Set multicast addresses to firmware */ - if (old_pkt_filter == priv->curr_pkt_filter) { - /* Send request to firmware */ - ret = mwifiex_send_cmd_async(priv, - HostCmd_CMD_MAC_MULTICAST_ADR, - HostCmd_ACT_GEN_SET, 0, - mcast_list); - } else { - /* Send request to firmware */ - ret = mwifiex_send_cmd_async(priv, - HostCmd_CMD_MAC_MULTICAST_ADR, - HostCmd_ACT_GEN_SET, 0, - mcast_list); - } + /* Send multicast addresses to firmware */ + ret = mwifiex_send_cmd_async(priv, + HostCmd_CMD_MAC_MULTICAST_ADR, + HostCmd_ACT_GEN_SET, 0, + mcast_list); } } } From f35a4dee14c31dc00807f3bcd59cc7a6959f63d7 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 09:51:29 -0500 Subject: [PATCH 046/198] rbd: set the mapping size and features later Defer setting the size and features fields of a mapped image until after the Linux disk structure is set up. Set the capacity of the disk after that. Rearrange the definition of rbd_image_header, separating the fields that are set only once from those that can be updated. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 27 ++++++++++++++------------- 1 file changed, 14 insertions(+), 13 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index b5cac7931ffc..a05b6e5dc362 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -100,21 +100,20 @@ * block device image metadata (in-memory version) */ struct rbd_image_header { - /* These four fields never change for a given rbd image */ + /* These six fields never change for a given rbd image */ char *object_prefix; - u64 features; __u8 obj_order; __u8 crypt_type; __u8 comp_type; + u64 stripe_unit; + u64 stripe_count; + u64 features; /* Might be changeable someday? */ /* The remaining fields need to be updated occasionally */ u64 image_size; struct ceph_snap_context *snapc; - char *snap_names; - u64 *snap_sizes; - - u64 stripe_unit; - u64 stripe_count; + char *snap_names; /* format 1 only */ + u64 *snap_sizes; /* format 1 only */ }; /* @@ -4637,10 +4636,6 @@ static int rbd_dev_device_setup(struct rbd_device *rbd_dev) { int ret; - ret = rbd_dev_mapping_set(rbd_dev); - if (ret) - return ret; - /* generate unique id: find highest unique id, add one */ rbd_dev_id_get(rbd_dev); @@ -4662,13 +4657,17 @@ static int rbd_dev_device_setup(struct rbd_device *rbd_dev) if (ret) goto err_out_blkdev; - ret = rbd_bus_add_dev(rbd_dev); + ret = rbd_dev_mapping_set(rbd_dev); if (ret) goto err_out_disk; + set_capacity(rbd_dev->disk, rbd_dev->mapping.size / SECTOR_SIZE); + + ret = rbd_bus_add_dev(rbd_dev); + if (ret) + goto err_out_mapping; /* Everything's ready. Announce the disk to the world. */ - set_capacity(rbd_dev->disk, rbd_dev->mapping.size / SECTOR_SIZE); set_bit(RBD_DEV_FLAG_EXISTS, &rbd_dev->flags); add_disk(rbd_dev->disk); @@ -4677,6 +4676,8 @@ static int rbd_dev_device_setup(struct rbd_device *rbd_dev) return ret; +err_out_mapping: + rbd_dev_mapping_clear(rbd_dev); err_out_disk: rbd_free_disk(rbd_dev); err_out_blkdev: From 46578dcdca951f3da70d3a5a9b5166b2a492a182 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 09:51:29 -0500 Subject: [PATCH 047/198] rbd: zero format 1 header structure earlier The passed-in header structure is zeroed in rbd_header_from_disk(). Instead, have the caller do it. Note that there are two callers, rbd_dev_v1_refresh() and rbd_dev_v1_probe(). The latter already has a zeroed header structure so zeroing it isn't necessary there. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index a05b6e5dc362..b0fbc38cc6f0 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -738,8 +738,6 @@ static int rbd_header_from_disk(struct rbd_image_header *header, size_t size; u32 i; - memset(header, 0, sizeof (*header)); - snap_count = le32_to_cpu(ondisk->snap_count); len = strnlen(ondisk->object_prefix, sizeof (ondisk->object_prefix)); @@ -3103,6 +3101,7 @@ static int rbd_dev_v1_refresh(struct rbd_device *rbd_dev) int ret; struct rbd_image_header h; + memset(&h, 0, sizeof (h)); ret = rbd_read_header(rbd_dev, &h); if (ret < 0) return ret; From bb23e37acb2ae9604130c4819fb8ae0f784a3a2b Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 09:51:29 -0500 Subject: [PATCH 048/198] rbd: refactor rbd_header_from_disk() This rearranges rbd_header_from_disk so that it: - allocates the snapshot context right away - keeps results in local variables, not changing the passed-in header until it's known we'll succeed - does initialization of set-once fields in a header only if they have not already been set The last point is moot at the moment, because rbd_read_header() (the only caller) always supplies a zero-filled header buffer. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 129 ++++++++++++++++++++++++++------------------ 1 file changed, 76 insertions(+), 53 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index b0fbc38cc6f0..a3b6bf5e9ae8 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -727,86 +727,109 @@ static bool rbd_dev_ondisk_valid(struct rbd_image_header_ondisk *ondisk) } /* - * Create a new header structure, translate header format from the on-disk - * header. + * Fill an rbd image header with information from the given format 1 + * on-disk header. */ static int rbd_header_from_disk(struct rbd_image_header *header, struct rbd_image_header_ondisk *ondisk) { + bool first_time = header->object_prefix == NULL; + struct ceph_snap_context *snapc; + char *object_prefix = NULL; + char *snap_names = NULL; + u64 *snap_sizes = NULL; u32 snap_count; - size_t len; size_t size; + int ret = -ENOMEM; u32 i; + /* Allocate this now to avoid having to handle failure below */ + + if (first_time) { + size_t len; + + len = strnlen(ondisk->object_prefix, + sizeof (ondisk->object_prefix)); + object_prefix = kmalloc(len + 1, GFP_KERNEL); + if (!object_prefix) + return -ENOMEM; + memcpy(object_prefix, ondisk->object_prefix, len); + object_prefix[len] = '\0'; + } + + /* Allocate the snapshot context and fill it in */ + snap_count = le32_to_cpu(ondisk->snap_count); - - len = strnlen(ondisk->object_prefix, sizeof (ondisk->object_prefix)); - header->object_prefix = kmalloc(len + 1, GFP_KERNEL); - if (!header->object_prefix) - return -ENOMEM; - memcpy(header->object_prefix, ondisk->object_prefix, len); - header->object_prefix[len] = '\0'; - + snapc = ceph_create_snap_context(snap_count, GFP_KERNEL); + if (!snapc) + goto out_err; + snapc->seq = le64_to_cpu(ondisk->snap_seq); if (snap_count) { + struct rbd_image_snap_ondisk *snaps; u64 snap_names_len = le64_to_cpu(ondisk->snap_names_len); - /* Save a copy of the snapshot names */ + /* We'll keep a copy of the snapshot names... */ - if (snap_names_len > (u64) SIZE_MAX) - return -EIO; - header->snap_names = kmalloc(snap_names_len, GFP_KERNEL); - if (!header->snap_names) + if (snap_names_len > (u64)SIZE_MAX) + goto out_2big; + snap_names = kmalloc(snap_names_len, GFP_KERNEL); + if (!snap_names) goto out_err; + + /* ...as well as the array of their sizes. */ + + size = snap_count * sizeof (*header->snap_sizes); + snap_sizes = kmalloc(size, GFP_KERNEL); + if (!snap_sizes) + goto out_err; + /* - * Note that rbd_dev_v1_header_read() guarantees - * the ondisk buffer we're working with has + * Copy the names, and fill in each snapshot's id + * and size. + * + * Note that rbd_dev_v1_header_read() guarantees the + * ondisk buffer we're working with has * snap_names_len bytes beyond the end of the * snapshot id array, this memcpy() is safe. */ - memcpy(header->snap_names, &ondisk->snaps[snap_count], - snap_names_len); - - /* Record each snapshot's size */ - - size = snap_count * sizeof (*header->snap_sizes); - header->snap_sizes = kmalloc(size, GFP_KERNEL); - if (!header->snap_sizes) - goto out_err; - for (i = 0; i < snap_count; i++) - header->snap_sizes[i] = - le64_to_cpu(ondisk->snaps[i].image_size); - } else { - header->snap_names = NULL; - header->snap_sizes = NULL; + memcpy(snap_names, &ondisk->snaps[snap_count], snap_names_len); + snaps = ondisk->snaps; + for (i = 0; i < snap_count; i++) { + snapc->snaps[i] = le64_to_cpu(snaps[i].id); + snap_sizes[i] = le64_to_cpu(snaps[i].image_size); + } } - header->features = 0; /* No features support in v1 images */ - header->obj_order = ondisk->options.order; - header->crypt_type = ondisk->options.crypt_type; - header->comp_type = ondisk->options.comp_type; + /* We won't fail any more, fill in the header */ - /* Allocate and fill in the snapshot context */ + if (first_time) { + header->object_prefix = object_prefix; + header->obj_order = ondisk->options.order; + header->crypt_type = ondisk->options.crypt_type; + header->comp_type = ondisk->options.comp_type; + /* The rest aren't used for format 1 images */ + header->stripe_unit = 0; + header->stripe_count = 0; + header->features = 0; + } + + /* The remaining fields always get updated (when we refresh) */ header->image_size = le64_to_cpu(ondisk->image_size); - - header->snapc = ceph_create_snap_context(snap_count, GFP_KERNEL); - if (!header->snapc) - goto out_err; - header->snapc->seq = le64_to_cpu(ondisk->snap_seq); - for (i = 0; i < snap_count; i++) - header->snapc->snaps[i] = le64_to_cpu(ondisk->snaps[i].id); + header->snapc = snapc; + header->snap_names = snap_names; + header->snap_sizes = snap_sizes; return 0; - +out_2big: + ret = -EIO; out_err: - kfree(header->snap_sizes); - header->snap_sizes = NULL; - kfree(header->snap_names); - header->snap_names = NULL; - kfree(header->object_prefix); - header->object_prefix = NULL; + kfree(snap_sizes); + kfree(snap_names); + ceph_put_snap_context(snapc); + kfree(object_prefix); - return -ENOMEM; + return ret; } static const char *_rbd_dev_v1_snap_name(struct rbd_device *rbd_dev, u32 which) From 662518b128c27def65e9af4bea2b56a1e04b3251 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 09:51:29 -0500 Subject: [PATCH 049/198] rbd: update in-core header directly Now that rbd_header_from_disk() only fills in one-time fields once, we can extend it slightly so it releases the other fields before replacing their values. This way there's no need to pass a temporary buffer and then copy all the results in. Just use the rbd device header structure in rbd_header_from_disk() so its values get updated directly. Note that this means we need to take the header semaphore at the point we update things. So pass the rbd_dev rather than the address of its header as its first argument to rbd_header_from_disk(), and have it return an error code. As a result, rbd_dev_v1_header_read() does all the work, rbd_read_header() becomes unnecessary, and rbd_dev_v1_refresh() becomes a very simple wrapper. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 98 +++++++++++++-------------------------------- 1 file changed, 27 insertions(+), 71 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index a3b6bf5e9ae8..e4586f2e04c2 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -730,9 +730,10 @@ static bool rbd_dev_ondisk_valid(struct rbd_image_header_ondisk *ondisk) * Fill an rbd image header with information from the given format 1 * on-disk header. */ -static int rbd_header_from_disk(struct rbd_image_header *header, +static int rbd_header_from_disk(struct rbd_device *rbd_dev, struct rbd_image_header_ondisk *ondisk) { + struct rbd_image_header *header = &rbd_dev->header; bool first_time = header->object_prefix == NULL; struct ceph_snap_context *snapc; char *object_prefix = NULL; @@ -802,6 +803,7 @@ static int rbd_header_from_disk(struct rbd_image_header *header, /* We won't fail any more, fill in the header */ + down_write(&rbd_dev->header_rwsem); if (first_time) { header->object_prefix = object_prefix; header->obj_order = ondisk->options.order; @@ -811,6 +813,10 @@ static int rbd_header_from_disk(struct rbd_image_header *header, header->stripe_unit = 0; header->stripe_count = 0; header->features = 0; + } else { + ceph_put_snap_context(header->snapc); + kfree(header->snap_names); + kfree(header->snap_sizes); } /* The remaining fields always get updated (when we refresh) */ @@ -820,6 +826,14 @@ static int rbd_header_from_disk(struct rbd_image_header *header, header->snap_names = snap_names; header->snap_sizes = snap_sizes; + /* Make sure mapping size is consistent with header info */ + + if (rbd_dev->spec->snap_id == CEPH_NOSNAP || first_time) + if (rbd_dev->mapping.size != header->image_size) + rbd_dev->mapping.size = header->image_size; + + up_write(&rbd_dev->header_rwsem); + return 0; out_2big: ret = -EIO; @@ -3032,17 +3046,11 @@ out: } /* - * Read the complete header for the given rbd device. - * - * Returns a pointer to a dynamically-allocated buffer containing - * the complete and validated header. Caller can pass the address - * of a variable that will be filled in with the version of the - * header object at the time it was read. - * - * Returns a pointer-coded errno if a failure occurs. + * Read the complete header for the given rbd device. On successful + * return, the rbd_dev->header field will contain up-to-date + * information about the image. */ -static struct rbd_image_header_ondisk * -rbd_dev_v1_header_read(struct rbd_device *rbd_dev) +static int rbd_dev_v1_header_read(struct rbd_device *rbd_dev) { struct rbd_image_header_ondisk *ondisk = NULL; u32 snap_count = 0; @@ -3067,22 +3075,22 @@ rbd_dev_v1_header_read(struct rbd_device *rbd_dev) size += names_size; ondisk = kmalloc(size, GFP_KERNEL); if (!ondisk) - return ERR_PTR(-ENOMEM); + return -ENOMEM; ret = rbd_obj_read_sync(rbd_dev, rbd_dev->header_name, 0, size, ondisk); if (ret < 0) - goto out_err; + goto out; if ((size_t)ret < size) { ret = -ENXIO; rbd_warn(rbd_dev, "short header read (want %zd got %d)", size, ret); - goto out_err; + goto out; } if (!rbd_dev_ondisk_valid(ondisk)) { ret = -ENXIO; rbd_warn(rbd_dev, "invalid header"); - goto out_err; + goto out; } names_size = le64_to_cpu(ondisk->snap_names_len); @@ -3090,27 +3098,8 @@ rbd_dev_v1_header_read(struct rbd_device *rbd_dev) snap_count = le32_to_cpu(ondisk->snap_count); } while (snap_count != want_count); - return ondisk; - -out_err: - kfree(ondisk); - - return ERR_PTR(ret); -} - -/* - * reload the ondisk the header - */ -static int rbd_read_header(struct rbd_device *rbd_dev, - struct rbd_image_header *header) -{ - struct rbd_image_header_ondisk *ondisk; - int ret; - - ondisk = rbd_dev_v1_header_read(rbd_dev); - if (IS_ERR(ondisk)) - return PTR_ERR(ondisk); - ret = rbd_header_from_disk(header, ondisk); + ret = rbd_header_from_disk(rbd_dev, ondisk); +out: kfree(ondisk); return ret; @@ -3121,40 +3110,7 @@ static int rbd_read_header(struct rbd_device *rbd_dev, */ static int rbd_dev_v1_refresh(struct rbd_device *rbd_dev) { - int ret; - struct rbd_image_header h; - - memset(&h, 0, sizeof (h)); - ret = rbd_read_header(rbd_dev, &h); - if (ret < 0) - return ret; - - down_write(&rbd_dev->header_rwsem); - - /* Update image size, and check for resize of mapped image */ - rbd_dev->header.image_size = h.image_size; - if (rbd_dev->spec->snap_id == CEPH_NOSNAP) - if (rbd_dev->mapping.size != rbd_dev->header.image_size) - rbd_dev->mapping.size = rbd_dev->header.image_size; - - /* rbd_dev->header.object_prefix shouldn't change */ - kfree(rbd_dev->header.snap_sizes); - kfree(rbd_dev->header.snap_names); - /* osd requests may still refer to snapc */ - ceph_put_snap_context(rbd_dev->header.snapc); - - rbd_dev->header.image_size = h.image_size; - rbd_dev->header.snapc = h.snapc; - rbd_dev->header.snap_names = h.snap_names; - rbd_dev->header.snap_sizes = h.snap_sizes; - /* Free the extra copy of the object prefix */ - if (strcmp(rbd_dev->header.object_prefix, h.object_prefix)) - rbd_warn(rbd_dev, "object prefix changed (ignoring)"); - kfree(h.object_prefix); - - up_write(&rbd_dev->header_rwsem); - - return ret; + return rbd_dev_v1_header_read(rbd_dev); } /* @@ -4517,7 +4473,7 @@ static int rbd_dev_v1_probe(struct rbd_device *rbd_dev) /* Populate rbd image metadata */ - ret = rbd_read_header(rbd_dev, &rbd_dev->header); + ret = rbd_dev_v1_header_read(rbd_dev); if (ret < 0) goto out_err; From 30d60ba2f258da24b91edb880338c7178f901de9 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 09:51:30 -0500 Subject: [PATCH 050/198] rbd: simplify rbd_dev_v1_probe() An rbd_dev structure's fields are all zero-filled for an initial probe, so there's no need to explicitly zero the parent_spec and parent_overlap fields in rbd_dev_v1_probe(). Removing these assignments makes rbd_dev_v1_probe() *almost* trivial. Move the dout() message that announces discovery of an image into rbd_dev_image_probe(), generalize to support images in either format and only show it if an image is fully discovered. This highlights that are some unnecessary cleanups in the error path for rbd_dev_v1_probe(), so they can be removed. Now rbd_dev_v1_probe() *is* a trivial wrapper function. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 37 +++++++------------------------------ 1 file changed, 7 insertions(+), 30 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index e4586f2e04c2..ae223819bbf0 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -4469,31 +4469,7 @@ static void rbd_dev_unprobe(struct rbd_device *rbd_dev) static int rbd_dev_v1_probe(struct rbd_device *rbd_dev) { - int ret; - - /* Populate rbd image metadata */ - - ret = rbd_dev_v1_header_read(rbd_dev); - if (ret < 0) - goto out_err; - - /* Version 1 images have no parent (no layering) */ - - rbd_dev->parent_spec = NULL; - rbd_dev->parent_overlap = 0; - - dout("discovered version 1 image, header name is %s\n", - rbd_dev->header_name); - - return 0; - -out_err: - kfree(rbd_dev->header_name); - rbd_dev->header_name = NULL; - kfree(rbd_dev->spec->image_id); - rbd_dev->spec->image_id = NULL; - - return ret; + return rbd_dev_v1_header_read(rbd_dev); } static int rbd_dev_v2_probe(struct rbd_device *rbd_dev) @@ -4553,9 +4529,6 @@ static int rbd_dev_v2_probe(struct rbd_device *rbd_dev) if (ret) goto out_err; - dout("discovered version 2 image, header name is %s\n", - rbd_dev->header_name); - return 0; out_err: rbd_dev->parent_overlap = 0; @@ -4758,9 +4731,13 @@ static int rbd_dev_image_probe(struct rbd_device *rbd_dev, bool read_only) rbd_dev->mapping.read_only = read_only; ret = rbd_dev_probe_parent(rbd_dev); - if (!ret) - return 0; + if (ret) + goto err_out_probe; + dout("discovered format %u image, header name is %s\n", + rbd_dev->image_format, rbd_dev->header_name); + + return 0; err_out_probe: rbd_dev_unprobe(rbd_dev); err_out_watch: From 99a41ebcee1a1ea0463b1b29d2e888de21a60c66 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 09:51:30 -0500 Subject: [PATCH 051/198] rbd: get rid of trivial v1 header wrappers Get rid of the trivial wrapper functions rbd_dev_v1_refresh() and rbd_dev_v1_probe(), substituting rbd_dev_v1_header_read() calls in their place. Rename rbd_dev_v1_header_read() to be rbd_dev_v1_header_info(), to be more generic (it will better reflect what happens with format 2 images). Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 21 ++++----------------- 1 file changed, 4 insertions(+), 17 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index ae223819bbf0..6748fe2d67d6 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -788,7 +788,7 @@ static int rbd_header_from_disk(struct rbd_device *rbd_dev, * Copy the names, and fill in each snapshot's id * and size. * - * Note that rbd_dev_v1_header_read() guarantees the + * Note that rbd_dev_v1_header_info() guarantees the * ondisk buffer we're working with has * snap_names_len bytes beyond the end of the * snapshot id array, this memcpy() is safe. @@ -3050,7 +3050,7 @@ out: * return, the rbd_dev->header field will contain up-to-date * information about the image. */ -static int rbd_dev_v1_header_read(struct rbd_device *rbd_dev) +static int rbd_dev_v1_header_info(struct rbd_device *rbd_dev) { struct rbd_image_header_ondisk *ondisk = NULL; u32 snap_count = 0; @@ -3105,14 +3105,6 @@ out: return ret; } -/* - * only read the first part of the ondisk header, without the snaps info - */ -static int rbd_dev_v1_refresh(struct rbd_device *rbd_dev) -{ - return rbd_dev_v1_header_read(rbd_dev); -} - /* * Clear the rbd device's EXISTS flag if the snapshot it's mapped to * has disappeared from the (just updated) snapshot context. @@ -3141,7 +3133,7 @@ static int rbd_dev_refresh(struct rbd_device *rbd_dev) mapping_size = rbd_dev->mapping.size; mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING); if (rbd_dev->image_format == 1) - ret = rbd_dev_v1_refresh(rbd_dev); + ret = rbd_dev_v1_header_info(rbd_dev); else ret = rbd_dev_v2_refresh(rbd_dev); @@ -4467,11 +4459,6 @@ static void rbd_dev_unprobe(struct rbd_device *rbd_dev) memset(header, 0, sizeof (*header)); } -static int rbd_dev_v1_probe(struct rbd_device *rbd_dev) -{ - return rbd_dev_v1_header_read(rbd_dev); -} - static int rbd_dev_v2_probe(struct rbd_device *rbd_dev) { int ret; @@ -4714,7 +4701,7 @@ static int rbd_dev_image_probe(struct rbd_device *rbd_dev, bool read_only) goto out_header_name; if (rbd_dev->image_format == 1) - ret = rbd_dev_v1_probe(rbd_dev); + ret = rbd_dev_v1_header_info(rbd_dev); else ret = rbd_dev_v2_probe(rbd_dev); if (ret) From 2df3fac75851dc4257b90dc72fdd3cf27ba177bc Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 09:51:30 -0500 Subject: [PATCH 052/198] rbd: define rbd_dev_v2_header_info() This rearranges rbd_dev_v2_refresh() so it works more like rbd_dev_v1_header_info(). While format 1 images need to read the whole header object to get any information, format 2 can collect almost all information selectively. So the one-time initialization will remain in a separate function--based on rbd_dev_v2_probe(). Rename rbd_dev_v2_refresh() to be rbd_dev_v2_header_info(), and have it call rbd_dev_v2_header_onetime() if it's being called for the first time for the given rbd device. Rename rbd_dev_v2_probe() to be rbd_dev_v2_header_onetime() and remove the image size and snapshot context calls it held in common with the refresh function. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 42 ++++++++++++++++++------------------------ 1 file changed, 18 insertions(+), 24 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 6748fe2d67d6..0d874a546949 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -425,7 +425,8 @@ static void rbd_img_parent_read(struct rbd_obj_request *obj_request); static void rbd_dev_remove_parent(struct rbd_device *rbd_dev); static int rbd_dev_refresh(struct rbd_device *rbd_dev); -static int rbd_dev_v2_refresh(struct rbd_device *rbd_dev); +static int rbd_dev_v2_header_onetime(struct rbd_device *rbd_dev); +static int rbd_dev_v2_header_info(struct rbd_device *rbd_dev); static const char *rbd_dev_v2_snap_name(struct rbd_device *rbd_dev, u64 snap_id); static int _rbd_dev_v2_snap_size(struct rbd_device *rbd_dev, u64 snap_id, @@ -3135,7 +3136,7 @@ static int rbd_dev_refresh(struct rbd_device *rbd_dev) if (rbd_dev->image_format == 1) ret = rbd_dev_v1_header_info(rbd_dev); else - ret = rbd_dev_v2_refresh(rbd_dev); + ret = rbd_dev_v2_header_info(rbd_dev); /* If it's a mapped snapshot, validate its EXISTS flag */ @@ -4005,12 +4006,19 @@ out: return snap_name; } -static int rbd_dev_v2_refresh(struct rbd_device *rbd_dev) +static int rbd_dev_v2_header_info(struct rbd_device *rbd_dev) { + bool first_time = rbd_dev->header.object_prefix == NULL; int ret; down_write(&rbd_dev->header_rwsem); + if (first_time) { + ret = rbd_dev_v2_header_onetime(rbd_dev); + if (ret) + goto out; + } + ret = rbd_dev_v2_image_size(rbd_dev); if (ret) goto out; @@ -4459,22 +4467,18 @@ static void rbd_dev_unprobe(struct rbd_device *rbd_dev) memset(header, 0, sizeof (*header)); } -static int rbd_dev_v2_probe(struct rbd_device *rbd_dev) +static int rbd_dev_v2_header_onetime(struct rbd_device *rbd_dev) { int ret; - ret = rbd_dev_v2_image_size(rbd_dev); - if (ret) - goto out_err; - - /* Get the object prefix (a.k.a. block_name) for the image */ - ret = rbd_dev_v2_object_prefix(rbd_dev); if (ret) goto out_err; - /* Get the and check features for the image */ - + /* + * Get the and check features for the image. Currently the + * features are assumed to never change. + */ ret = rbd_dev_v2_features(rbd_dev); if (ret) goto out_err; @@ -4504,17 +4508,7 @@ static int rbd_dev_v2_probe(struct rbd_device *rbd_dev) if (ret < 0) goto out_err; } - - /* crypto and compression type aren't (yet) supported for v2 images */ - - rbd_dev->header.crypt_type = 0; - rbd_dev->header.comp_type = 0; - - /* Get the snapshot context, plus the header version */ - - ret = rbd_dev_v2_snap_context(rbd_dev); - if (ret) - goto out_err; + /* No support for crypto and compression type format 2 images */ return 0; out_err: @@ -4703,7 +4697,7 @@ static int rbd_dev_image_probe(struct rbd_device *rbd_dev, bool read_only) if (rbd_dev->image_format == 1) ret = rbd_dev_v1_header_info(rbd_dev); else - ret = rbd_dev_v2_probe(rbd_dev); + ret = rbd_dev_v2_header_info(rbd_dev); if (ret) goto err_out_watch; From 91c6febb3817be576785ef06aeaaa8ed423e0a2a Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 17:40:32 -0500 Subject: [PATCH 053/198] rbd: fix an incorrect assertion condition In rbd_img_obj_parent_read_full_callback() there is an assertion intended to verify the size of the image request for a full parent read was the size of the original request's target object. But assertion was looking at the parent image order rather than the original one, and these values can differ. Fix that. This resolves: http://tracker.ceph.com/issues/4938 Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 0d874a546949..15ac2a54d4f3 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -2186,13 +2186,13 @@ rbd_img_obj_parent_read_full_callback(struct rbd_img_request *img_request) result = img_request->result; obj_size = img_request->length; xferred = img_request->xferred; + rbd_img_request_put(img_request); - rbd_dev = img_request->rbd_dev; + rbd_assert(orig_request->img_request); + rbd_dev = orig_request->img_request->rbd_dev; rbd_assert(rbd_dev); rbd_assert(obj_size == (u64)1 << rbd_dev->header.obj_order); - rbd_img_request_put(img_request); - if (result) goto out_err; From 5b2ab72d367d2682c1a237448fbc1595881a88fa Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 17:40:33 -0500 Subject: [PATCH 054/198] rbd: support reading parent page data Currently, rbd_img_parent_read() assumes the incoming object request contains bio data. But if a layered image is part of a multi-layer stack of images it will result in read requests of page data to parent images. Fortunately, it's not hard to add support for page data. This resolves: http://tracker.ceph.com/issues/4939 Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 15ac2a54d4f3..2a0e9b81be48 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -2574,7 +2574,7 @@ static void rbd_img_parent_read(struct rbd_obj_request *obj_request) rbd_assert(obj_request_img_data_test(obj_request)); rbd_assert(obj_request->img_request != NULL); rbd_assert(obj_request->result == (s32) -ENOENT); - rbd_assert(obj_request->type == OBJ_REQUEST_BIO); + rbd_assert(obj_request_type_valid(obj_request->type)); rbd_dev = obj_request->img_request->rbd_dev; rbd_assert(rbd_dev->parent != NULL); @@ -2590,8 +2590,12 @@ static void rbd_img_parent_read(struct rbd_obj_request *obj_request) rbd_obj_request_get(obj_request); img_request->obj_request = obj_request; - result = rbd_img_request_fill(img_request, OBJ_REQUEST_BIO, - obj_request->bio_list); + if (obj_request->type == OBJ_REQUEST_BIO) + result = rbd_img_request_fill(img_request, OBJ_REQUEST_BIO, + obj_request->bio_list); + else + result = rbd_img_request_fill(img_request, OBJ_REQUEST_PAGES, + obj_request->pages); if (result) goto out_err; From 7ce4eef7b5fad73b365b7e4b8892af3af72d4bd3 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 17:40:33 -0500 Subject: [PATCH 055/198] rbd: set mapping read-only flag in rbd_add() The rbd_dev->mapping field for a parent image is not meaningful. Since rbd_image_probe() is used both for images being mapped and their parents, it doesn't make sense to set that flag in that function. So move the setting of the mapping.read_only flag out of rbd_dev_image_probe() and into rbd_add() instead. This resolves: http://tracker.ceph.com/issues/4940 Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 23 +++++++++++------------ 1 file changed, 11 insertions(+), 12 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 2a0e9b81be48..dbfc44a9defd 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -358,7 +358,7 @@ static ssize_t rbd_add(struct bus_type *bus, const char *buf, size_t count); static ssize_t rbd_remove(struct bus_type *bus, const char *buf, size_t count); -static int rbd_dev_image_probe(struct rbd_device *rbd_dev, bool read_only); +static int rbd_dev_image_probe(struct rbd_device *rbd_dev); static struct bus_attribute rbd_bus_attrs[] = { __ATTR(add, S_IWUSR, NULL, rbd_add), @@ -4549,7 +4549,7 @@ static int rbd_dev_probe_parent(struct rbd_device *rbd_dev) if (!parent) goto out_err; - ret = rbd_dev_image_probe(parent, true); + ret = rbd_dev_image_probe(parent); if (ret < 0) goto out_err; rbd_dev->parent = parent; @@ -4671,10 +4671,9 @@ static void rbd_dev_image_release(struct rbd_device *rbd_dev) /* * Probe for the existence of the header object for the given rbd - * device. For format 2 images this includes determining the image - * id. + * device. */ -static int rbd_dev_image_probe(struct rbd_device *rbd_dev, bool read_only) +static int rbd_dev_image_probe(struct rbd_device *rbd_dev) { int ret; int tmp; @@ -4709,12 +4708,6 @@ static int rbd_dev_image_probe(struct rbd_device *rbd_dev, bool read_only) if (ret) goto err_out_probe; - /* If we are mapping a snapshot it must be marked read-only */ - - if (rbd_dev->spec->snap_id != CEPH_NOSNAP) - read_only = true; - rbd_dev->mapping.read_only = read_only; - ret = rbd_dev_probe_parent(rbd_dev); if (ret) goto err_out_probe; @@ -4795,10 +4788,16 @@ static ssize_t rbd_add(struct bus_type *bus, rbdc = NULL; /* rbd_dev now owns this */ spec = NULL; /* rbd_dev now owns this */ - rc = rbd_dev_image_probe(rbd_dev, read_only); + rc = rbd_dev_image_probe(rbd_dev); if (rc < 0) goto err_out_rbd_dev; + /* If we are mapping a snapshot it must be marked read-only */ + + if (rbd_dev->spec->snap_id != CEPH_NOSNAP) + read_only = true; + rbd_dev->mapping.read_only = read_only; + rc = rbd_dev_device_setup(rbd_dev); if (!rc) return count; From 1f3ef78861ac4b510175e177899b9b5ba4bbed91 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 17:40:33 -0500 Subject: [PATCH 056/198] rbd: only set up watch for mapped images Any changes to parent images are immaterial to any mapped clone. So there is no need to have a watch event registered on header objects except for the header object of an image that is mapped. In fact, a watch request is a write operation, and we may only have read access to a parent image. We can't set up the watch request until we know the name of the header object though. So pass a flag to rbd_dev_image_probe() to indicate whether this probe is for a mapping or for a parent image. Change the second parameter to rbd_dev_header_watch_sync() be Boolean while we're at it. This resolves: http://tracker.ceph.com/issues/4941 Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 43 ++++++++++++++++++++++++------------------- 1 file changed, 24 insertions(+), 19 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index dbfc44a9defd..e417704de8ca 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -358,7 +358,7 @@ static ssize_t rbd_add(struct bus_type *bus, const char *buf, size_t count); static ssize_t rbd_remove(struct bus_type *bus, const char *buf, size_t count); -static int rbd_dev_image_probe(struct rbd_device *rbd_dev); +static int rbd_dev_image_probe(struct rbd_device *rbd_dev, bool mapping); static struct bus_attribute rbd_bus_attrs[] = { __ATTR(add, S_IWUSR, NULL, rbd_add), @@ -2664,7 +2664,7 @@ static void rbd_watch_cb(u64 ver, u64 notify_id, u8 opcode, void *data) * Request sync osd watch/unwatch. The value of "start" determines * whether a watch request is being initiated or torn down. */ -static int rbd_dev_header_watch_sync(struct rbd_device *rbd_dev, int start) +static int rbd_dev_header_watch_sync(struct rbd_device *rbd_dev, bool start) { struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc; struct rbd_obj_request *obj_request; @@ -2698,7 +2698,7 @@ static int rbd_dev_header_watch_sync(struct rbd_device *rbd_dev, int start) rbd_dev->watch_request->osd_req); osd_req_op_watch_init(obj_request->osd_req, 0, CEPH_OSD_OP_WATCH, - rbd_dev->watch_event->cookie, 0, start); + rbd_dev->watch_event->cookie, 0, start ? 1 : 0); rbd_osd_req_format_write(obj_request); ret = rbd_obj_request_submit(osdc, obj_request); @@ -4549,7 +4549,7 @@ static int rbd_dev_probe_parent(struct rbd_device *rbd_dev) if (!parent) goto out_err; - ret = rbd_dev_image_probe(parent); + ret = rbd_dev_image_probe(parent, false); if (ret < 0) goto out_err; rbd_dev->parent = parent; @@ -4654,12 +4654,7 @@ static int rbd_dev_header_name(struct rbd_device *rbd_dev) static void rbd_dev_image_release(struct rbd_device *rbd_dev) { - int ret; - rbd_dev_unprobe(rbd_dev); - ret = rbd_dev_header_watch_sync(rbd_dev, 0); - if (ret) - rbd_warn(rbd_dev, "failed to cancel watch event (%d)\n", ret); kfree(rbd_dev->header_name); rbd_dev->header_name = NULL; rbd_dev->image_format = 0; @@ -4671,9 +4666,11 @@ static void rbd_dev_image_release(struct rbd_device *rbd_dev) /* * Probe for the existence of the header object for the given rbd - * device. + * device. If this image is the one being mapped (i.e., not a + * parent), initiate a watch on its header object before using that + * object to get detailed information about the rbd image. */ -static int rbd_dev_image_probe(struct rbd_device *rbd_dev) +static int rbd_dev_image_probe(struct rbd_device *rbd_dev, bool mapping) { int ret; int tmp; @@ -4693,9 +4690,11 @@ static int rbd_dev_image_probe(struct rbd_device *rbd_dev) if (ret) goto err_out_format; - ret = rbd_dev_header_watch_sync(rbd_dev, 1); - if (ret) - goto out_header_name; + if (mapping) { + ret = rbd_dev_header_watch_sync(rbd_dev, true); + if (ret) + goto out_header_name; + } if (rbd_dev->image_format == 1) ret = rbd_dev_v1_header_info(rbd_dev); @@ -4719,9 +4718,12 @@ static int rbd_dev_image_probe(struct rbd_device *rbd_dev) err_out_probe: rbd_dev_unprobe(rbd_dev); err_out_watch: - tmp = rbd_dev_header_watch_sync(rbd_dev, 0); - if (tmp) - rbd_warn(rbd_dev, "unable to tear down watch request\n"); + if (mapping) { + tmp = rbd_dev_header_watch_sync(rbd_dev, false); + if (tmp) + rbd_warn(rbd_dev, "unable to tear down " + "watch request (%d)\n", tmp); + } out_header_name: kfree(rbd_dev->header_name); rbd_dev->header_name = NULL; @@ -4788,7 +4790,7 @@ static ssize_t rbd_add(struct bus_type *bus, rbdc = NULL; /* rbd_dev now owns this */ spec = NULL; /* rbd_dev now owns this */ - rc = rbd_dev_image_probe(rbd_dev); + rc = rbd_dev_image_probe(rbd_dev, true); if (rc < 0) goto err_out_rbd_dev; @@ -4910,10 +4912,13 @@ static ssize_t rbd_remove(struct bus_type *bus, spin_unlock_irq(&rbd_dev->lock); if (ret < 0) goto done; - ret = count; rbd_bus_del_dev(rbd_dev); + ret = rbd_dev_header_watch_sync(rbd_dev, false); + if (ret) + rbd_warn(rbd_dev, "failed to cancel watch event (%d)\n", ret); rbd_dev_image_release(rbd_dev); module_put(THIS_MODULE); + ret = count; done: mutex_unlock(&ctl_mutex); From c48f3f86e248b1649ad22151dd45ef2610101ed3 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Wed, 8 May 2013 08:49:52 -0500 Subject: [PATCH 057/198] rbd: kill rbd_img_request_get() Get rid of rbd_img_request_get(), because it isn't used, and maybe won't ever be needed. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 10 ---------- 1 file changed, 10 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index e417704de8ca..51c45e793354 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -1357,13 +1357,6 @@ static void rbd_obj_request_put(struct rbd_obj_request *obj_request) kref_put(&obj_request->kref, rbd_obj_request_destroy); } -static void rbd_img_request_get(struct rbd_img_request *img_request) -{ - dout("%s: img %p (was %d)\n", __func__, img_request, - atomic_read(&img_request->kref.refcount)); - kref_get(&img_request->kref); -} - static void rbd_img_request_destroy(struct kref *kref); static void rbd_img_request_put(struct rbd_img_request *img_request) { @@ -1888,9 +1881,6 @@ static struct rbd_img_request *rbd_img_request_create( INIT_LIST_HEAD(&img_request->obj_requests); kref_init(&img_request->kref); - rbd_img_request_get(img_request); /* Avoid a warning */ - rbd_img_request_put(img_request); /* TEMPORARY */ - dout("%s: rbd_dev %p %s %llu/%llu -> img %p\n", __func__, rbd_dev, write_request ? "write" : "read", offset, length, img_request); From 074d72ff57f65de779e2f70d5906964c0ba1c123 Mon Sep 17 00:00:00 2001 From: Konrad Rzeszutek Wilk Date: Wed, 8 May 2013 12:13:03 -0400 Subject: [PATCH 058/198] x86/microcode: Add local mutex to fix physical CPU hot-add deadlock This can easily be triggered if a new CPU is added (via ACPI hotplug mechanism) and from user-space you do: echo 1 > /sys/devices/system/cpu/cpu3/online (or wait for UDEV to do it) on a newly appeared physical CPU. The deadlock is that the "store_online" in drivers/base/cpu.c takes the cpu_hotplug_driver_lock() lock, then calls "cpu_up". "cpu_up" eventually ends up calling "save_mc_for_early" which also takes the cpu_hotplug_driver_lock() lock. And here is that lockdep thinks of it: smpboot: Stack at about ffff880075c39f44 smpboot: CPU3: has booted. microcode: CPU3 sig=0x206a7, pf=0x2, revision=0x25 ============================================= [ INFO: possible recursive locking detected ] 3.9.0upstream-10129-g167af0e #1 Not tainted --------------------------------------------- sh/2487 is trying to acquire lock: (x86_cpu_hotplug_driver_mutex){+.+.+.}, at: [] cpu_hotplug_driver_lock+0x12/0x20 but task is already holding lock: (x86_cpu_hotplug_driver_mutex){+.+.+.}, at: [] cpu_hotplug_driver_lock+0x12/0x20 other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(x86_cpu_hotplug_driver_mutex); lock(x86_cpu_hotplug_driver_mutex); *** DEADLOCK *** May be due to missing lock nesting notation 6 locks held by sh/2487: #0: (sb_writers#5){.+.+.+}, at: [] vfs_write+0x17d/0x190 #1: (&buffer->mutex){+.+.+.}, at: [] sysfs_write_file+0x3f/0x160 #2: (s_active#20){.+.+.+}, at: [] sysfs_write_file+0xc8/0x160 #3: (x86_cpu_hotplug_driver_mutex){+.+.+.}, at: [] cpu_hotplug_driver_lock+0x12/0x20 #4: (cpu_add_remove_lock){+.+.+.}, at: [] cpu_maps_update_begin+0x12/0x20 #5: (cpu_hotplug.lock){+.+.+.}, at: [] cpu_hotplug_begin+0x27/0x60 Suggested-and-Acked-by: Borislav Petkov Signed-off-by: Konrad Rzeszutek Wilk Cc: fenghua.yu@intel.com Cc: xen-devel@lists.xensource.com Cc: stable@vger.kernel.org # for v3.9 Link: http://lkml.kernel.org/r/1368029583-23337-1-git-send-email-konrad.wilk@oracle.com Signed-off-by: Ingo Molnar --- arch/x86/kernel/microcode_intel_early.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/arch/x86/kernel/microcode_intel_early.c b/arch/x86/kernel/microcode_intel_early.c index d893e8ed8ac9..2e9e12871c2b 100644 --- a/arch/x86/kernel/microcode_intel_early.c +++ b/arch/x86/kernel/microcode_intel_early.c @@ -487,6 +487,7 @@ static inline void show_saved_mc(void) #endif #if defined(CONFIG_MICROCODE_INTEL_EARLY) && defined(CONFIG_HOTPLUG_CPU) +static DEFINE_MUTEX(x86_cpu_microcode_mutex); /* * Save this mc into mc_saved_data. So it will be loaded early when a CPU is * hot added or resumes. @@ -507,7 +508,7 @@ int save_mc_for_early(u8 *mc) * Hold hotplug lock so mc_saved_data is not accessed by a CPU in * hotplug. */ - cpu_hotplug_driver_lock(); + mutex_lock(&x86_cpu_microcode_mutex); mc_saved_count_init = mc_saved_data.mc_saved_count; mc_saved_count = mc_saved_data.mc_saved_count; @@ -544,7 +545,7 @@ int save_mc_for_early(u8 *mc) } out: - cpu_hotplug_driver_unlock(); + mutex_unlock(&x86_cpu_microcode_mutex); return ret; } From 293c9c1cef6be46ad72b528fa79c3ff034551fe0 Mon Sep 17 00:00:00 2001 From: Marek Lindner Date: Sat, 27 Apr 2013 16:22:28 +0800 Subject: [PATCH 059/198] batman-adv: check proto length before accessing proto string buffer batadv_param_set_ra() strips the trailing '\n' from the supplied string buffer without checking the length of the buffer first. This patches avoids random memory access and associated potential crashes. Reported-by: Sasha Levin Signed-off-by: Marek Lindner Signed-off-by: Antonio Quartulli --- net/batman-adv/main.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/batman-adv/main.c b/net/batman-adv/main.c index 3e30a0f1b908..9c620cd3b5f7 100644 --- a/net/batman-adv/main.c +++ b/net/batman-adv/main.c @@ -475,7 +475,7 @@ static int batadv_param_set_ra(const char *val, const struct kernel_param *kp) char *algo_name = (char *)val; size_t name_len = strlen(algo_name); - if (algo_name[name_len - 1] == '\n') + if (name_len > 0 && algo_name[name_len - 1] == '\n') algo_name[name_len - 1] = '\0'; bat_algo_ops = batadv_algo_get(algo_name); From 7da19971a9da9c5e63d259c9a8a566f071dc95d0 Mon Sep 17 00:00:00 2001 From: Marek Lindner Date: Tue, 7 May 2013 19:25:02 +0800 Subject: [PATCH 060/198] batman-adv: check return value of pskb_trim_rcsum() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Reported-by: Sven Eckelmann Signed-off-by: Marek Lindner Acked-by: Martin Hundebøll Signed-off-by: Antonio Quartulli --- net/batman-adv/network-coding.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/net/batman-adv/network-coding.c b/net/batman-adv/network-coding.c index f7c54305a918..e84629ece9b7 100644 --- a/net/batman-adv/network-coding.c +++ b/net/batman-adv/network-coding.c @@ -1514,6 +1514,7 @@ batadv_nc_skb_decode_packet(struct batadv_priv *bat_priv, struct sk_buff *skb, struct ethhdr *ethhdr, ethhdr_tmp; uint8_t *orig_dest, ttl, ttvn; unsigned int coding_len; + int err; /* Save headers temporarily */ memcpy(&coded_packet_tmp, skb->data, sizeof(coded_packet_tmp)); @@ -1568,8 +1569,11 @@ batadv_nc_skb_decode_packet(struct batadv_priv *bat_priv, struct sk_buff *skb, coding_len); /* Resize decoded skb if decoded with larger packet */ - if (nc_packet->skb->len > coding_len + h_size) - pskb_trim_rcsum(skb, coding_len + h_size); + if (nc_packet->skb->len > coding_len + h_size) { + err = pskb_trim_rcsum(skb, coding_len + h_size); + if (err) + return NULL; + } /* Create decoded unicast packet */ unicast_packet = (struct batadv_unicast_packet *)skb->data; From 88e48d7b3340ef07b108eb8a8b3813dd093cc7f7 Mon Sep 17 00:00:00 2001 From: Antonio Quartulli Date: Thu, 9 May 2013 09:35:45 +0200 Subject: [PATCH 061/198] batman-adv: make DAT drop ARP requests targeting local clients In the outgoing ARP request snooping routine in DAT, ARP Request sent by local clients which are supposed to be replied by other local clients can be silently dropped. The destination host will reply by itself through the LAN and therefore there is no need to involve DAT. Reported-by: Carlos Quijano Signed-off-by: Antonio Quartulli Tested-by: Carlos Quijano Signed-off-by: Marek Lindner --- net/batman-adv/distributed-arp-table.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) diff --git a/net/batman-adv/distributed-arp-table.c b/net/batman-adv/distributed-arp-table.c index 8e15d966d9b0..239992021b1d 100644 --- a/net/batman-adv/distributed-arp-table.c +++ b/net/batman-adv/distributed-arp-table.c @@ -837,6 +837,19 @@ bool batadv_dat_snoop_outgoing_arp_request(struct batadv_priv *bat_priv, dat_entry = batadv_dat_entry_hash_find(bat_priv, ip_dst); if (dat_entry) { + /* If the ARP request is destined for a local client the local + * client will answer itself. DAT would only generate a + * duplicate packet. + * + * Moreover, if the soft-interface is enslaved into a bridge, an + * additional DAT answer may trigger kernel warnings about + * a packet coming from the wrong port. + */ + if (batadv_is_my_client(bat_priv, dat_entry->mac_addr)) { + ret = true; + goto out; + } + skb_new = arp_create(ARPOP_REPLY, ETH_P_ARP, ip_src, bat_priv->soft_iface, ip_dst, hw_src, dat_entry->mac_addr, hw_src); From a4361860351e87876aebd9595906d928ce8572c6 Mon Sep 17 00:00:00 2001 From: Antonio Quartulli Date: Tue, 7 May 2013 01:06:18 +0200 Subject: [PATCH 062/198] batman-adv: reorder clean up routine in order to avoid race conditions nc_worker accesses the originator table during its periodic work, but since the originator table is freed before stopping the worker this leads to a global protection fault. Fix this by killing the worker (in nc_free) before freeing the originator table. Moreover tidy up the entire clean up routine by running all the subcomponents freeing procedures first and then killing the TT and the originator tables at the end. Signed-off-by: Antonio Quartulli Signed-off-by: Marek Lindner --- net/batman-adv/main.c | 18 +++++++++++++----- 1 file changed, 13 insertions(+), 5 deletions(-) diff --git a/net/batman-adv/main.c b/net/batman-adv/main.c index 9c620cd3b5f7..1240f07ad31d 100644 --- a/net/batman-adv/main.c +++ b/net/batman-adv/main.c @@ -163,14 +163,22 @@ void batadv_mesh_free(struct net_device *soft_iface) batadv_vis_quit(bat_priv); batadv_gw_node_purge(bat_priv); - batadv_originator_free(bat_priv); batadv_nc_free(bat_priv); - - batadv_tt_free(bat_priv); - + batadv_dat_free(bat_priv); batadv_bla_free(bat_priv); - batadv_dat_free(bat_priv); + /* Free the TT and the originator tables only after having terminated + * all the other depending components which may use these structures for + * their purposes. + */ + batadv_tt_free(bat_priv); + + /* Since the originator table clean up routine is accessing the TT + * tables as well, it has to be invoked after the TT tables have been + * freed and marked as empty. This ensures that no cleanup RCU callbacks + * accessing the TT data are scheduled for later execution. + */ + batadv_originator_free(bat_priv); free_percpu(bat_priv->bat_counters); From 97fbfbf40b678b95551276e9413517ba7858dba7 Mon Sep 17 00:00:00 2001 From: Damien Lespiau Date: Wed, 8 May 2013 17:03:30 +0100 Subject: [PATCH 063/198] drm: Add missing break in the command line mode parsing code As we parse the string given on the command line one char at a time, it seems that we do want a break at every case. Signed-off-by: Damien Lespiau Reviewed-by: Rodrigo Vivi Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_modes.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpu/drm/drm_modes.c b/drivers/gpu/drm/drm_modes.c index f264d08062f0..7a83836879bc 100644 --- a/drivers/gpu/drm/drm_modes.c +++ b/drivers/gpu/drm/drm_modes.c @@ -1143,6 +1143,7 @@ bool drm_mode_parse_command_line_for_connector(const char *mode_option, was_digit = false; } else goto done; + break; case '0' ... '9': was_digit = true; break; From ebbd97ad885c3c41f56a5f81c27b1ea8156a0f07 Mon Sep 17 00:00:00 2001 From: Damien Lespiau Date: Wed, 8 May 2013 17:03:32 +0100 Subject: [PATCH 064/198] drm: Don't prune modes loudly when a connector is disconnected drm_helper_probe_single_connector_modes() is responsible for pruning the previously detected modes on a disconnected connector. We don't really need to log, again, the full list of modes that used to be valid when connected. Signed-off-by: Damien Lespiau Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_crtc_helper.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/drm_crtc_helper.c b/drivers/gpu/drm/drm_crtc_helper.c index e974f9309b72..62f45ad56dc7 100644 --- a/drivers/gpu/drm/drm_crtc_helper.c +++ b/drivers/gpu/drm/drm_crtc_helper.c @@ -121,6 +121,7 @@ int drm_helper_probe_single_connector_modes(struct drm_connector *connector, connector->helper_private; int count = 0; int mode_flags = 0; + bool verbose_prune = true; DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n", connector->base.id, drm_get_connector_name(connector)); @@ -149,6 +150,7 @@ int drm_helper_probe_single_connector_modes(struct drm_connector *connector, DRM_DEBUG_KMS("[CONNECTOR:%d:%s] disconnected\n", connector->base.id, drm_get_connector_name(connector)); drm_mode_connector_update_edid_property(connector, NULL); + verbose_prune = false; goto prune; } @@ -182,7 +184,7 @@ int drm_helper_probe_single_connector_modes(struct drm_connector *connector, } prune: - drm_mode_prune_invalid(dev, &connector->modes, true); + drm_mode_prune_invalid(dev, &connector->modes, verbose_prune); if (list_empty(&connector->modes)) return 0; From d1fd3ddc469b576e36184f2bd02c0301d80118d3 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Tue, 7 May 2013 12:32:16 -0700 Subject: [PATCH 065/198] drm: refactor call to request_module This reduces the size of the stack frame when calling request_module(). Performing the sprintf before the call is not needed. Signed-off-by: Kees Cook Acked-by: Paul Menzel Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_encoder_slave.c | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-) diff --git a/drivers/gpu/drm/drm_encoder_slave.c b/drivers/gpu/drm/drm_encoder_slave.c index 48c52f7df4e6..0cfb60f54766 100644 --- a/drivers/gpu/drm/drm_encoder_slave.c +++ b/drivers/gpu/drm/drm_encoder_slave.c @@ -54,16 +54,12 @@ int drm_i2c_encoder_init(struct drm_device *dev, struct i2c_adapter *adap, const struct i2c_board_info *info) { - char modalias[sizeof(I2C_MODULE_PREFIX) - + I2C_NAME_SIZE]; struct module *module = NULL; struct i2c_client *client; struct drm_i2c_encoder_driver *encoder_drv; int err = 0; - snprintf(modalias, sizeof(modalias), - "%s%s", I2C_MODULE_PREFIX, info->type); - request_module(modalias); + request_module("%s%s", I2C_MODULE_PREFIX, info->type); client = i2c_new_device(adap, info); if (!client) { From 54afc1214133deddd49162b101971af5711f1b3a Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Wed, 8 May 2013 16:38:33 +0300 Subject: [PATCH 066/198] drm: Add kernel-doc for drm_fb_helper_funcs->initial_config MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: Ville Syrjälä Signed-off-by: Dave Airlie --- include/drm/drm_fb_helper.h | 1 + 1 file changed, 1 insertion(+) diff --git a/include/drm/drm_fb_helper.h b/include/drm/drm_fb_helper.h index 8230b46fdd73..61ebd51f6112 100644 --- a/include/drm/drm_fb_helper.h +++ b/include/drm/drm_fb_helper.h @@ -57,6 +57,7 @@ struct drm_fb_helper_surface_size { * @fb_probe: - Driver callback to allocate and initialize the fbdev info * structure. Futhermore it also needs to allocate the drm * framebuffer used to back the fbdev. + * @initial_config: - Setup an initial fbdev display configuration * * Driver callbacks used by the fbdev emulation helper library. */ From 7b97936f298af7991ca4717a4c23610d9ecd8927 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Wed, 8 May 2013 16:38:34 +0300 Subject: [PATCH 067/198] drm: Remove pointless '-' characters from drm_fb_helper documentation MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: Ville Syrjälä Signed-off-by: Dave Airlie --- include/drm/drm_fb_helper.h | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/include/drm/drm_fb_helper.h b/include/drm/drm_fb_helper.h index 61ebd51f6112..471f276ce8f7 100644 --- a/include/drm/drm_fb_helper.h +++ b/include/drm/drm_fb_helper.h @@ -50,14 +50,14 @@ struct drm_fb_helper_surface_size { /** * struct drm_fb_helper_funcs - driver callbacks for the fbdev emulation library - * @gamma_set: - Set the given gamma lut register on the given crtc. - * @gamma_get: - Read the given gamma lut register on the given crtc, used to - * save the current lut when force-restoring the fbdev for e.g. - * kdbg. - * @fb_probe: - Driver callback to allocate and initialize the fbdev info - * structure. Futhermore it also needs to allocate the drm - * framebuffer used to back the fbdev. - * @initial_config: - Setup an initial fbdev display configuration + * @gamma_set: Set the given gamma lut register on the given crtc. + * @gamma_get: Read the given gamma lut register on the given crtc, used to + * save the current lut when force-restoring the fbdev for e.g. + * kdbg. + * @fb_probe: Driver callback to allocate and initialize the fbdev info + * structure. Futhermore it also needs to allocate the drm + * framebuffer used to back the fbdev. + * @initial_config: Setup an initial fbdev display configuration * * Driver callbacks used by the fbdev emulation helper library. */ From b9434d0f162c351e02e0e8f66ca6b75a67595537 Mon Sep 17 00:00:00 2001 From: Chris Cummins Date: Thu, 9 May 2013 14:20:40 +0100 Subject: [PATCH 068/198] drm: Use names of ioctls in debug traces MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The intention here is to make the output of dmesg with full verbosity a bit easier for a human to parse. This commit transforms: [drm:drm_ioctl], pid=699, cmd=0x6458, nr=0x58, dev 0xe200, auth=1 [drm:drm_ioctl], pid=699, cmd=0xc010645b, nr=0x5b, dev 0xe200, auth=1 [drm:drm_ioctl], pid=699, cmd=0xc0106461, nr=0x61, dev 0xe200, auth=1 [drm:drm_ioctl], pid=699, cmd=0xc01c64ae, nr=0xae, dev 0xe200, auth=1 [drm:drm_mode_addfb], [FB:32] [drm:drm_ioctl], pid=699, cmd=0xc0106464, nr=0x64, dev 0xe200, auth=1 [drm:drm_vm_open_locked], 0x7fd9302fe000,0x00a00000 [drm:drm_ioctl], pid=699, cmd=0x400c645f, nr=0x5f, dev 0xe200, auth=1 [drm:drm_ioctl], pid=699, cmd=0xc00464af, nr=0xaf, dev 0xe200, auth=1 [drm:intel_crtc_set_config], [CRTC:3] [NOFB] into: [drm:drm_ioctl], pid=699, dev=0xe200, auth=1, I915_GEM_THROTTLE [drm:drm_ioctl], pid=699, dev=0xe200, auth=1, I915_GEM_CREATE [drm:drm_ioctl], pid=699, dev=0xe200, auth=1, I915_GEM_SET_TILING [drm:drm_ioctl], pid=699, dev=0xe200, auth=1, IOCTL_MODE_ADDFB [drm:drm_mode_addfb], [FB:32] [drm:drm_ioctl], pid=699, dev=0xe200, auth=1, I915_GEM_MMAP_GTT [drm:drm_vm_open_locked], 0x7fd9302fe000,0x00a00000 [drm:drm_ioctl], pid=699, dev=0xe200, auth=1, I915_GEM_SET_DOMAIN [drm:drm_ioctl], pid=699, dev=0xe200, auth=1, DRM_IOCTL_MODE_RMFB [drm:intel_crtc_set_config], [CRTC:3] [NOFB] v2: drm_ioctls is now a constant (Ville Syrjälä) Signed-off-by: Chris Cummins Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_drv.c | 20 +++++++++++++------- include/drm/drmP.h | 3 ++- 2 files changed, 15 insertions(+), 8 deletions(-) diff --git a/drivers/gpu/drm/drm_drv.c b/drivers/gpu/drm/drm_drv.c index 8d4f29075af5..9cc247f55502 100644 --- a/drivers/gpu/drm/drm_drv.c +++ b/drivers/gpu/drm/drm_drv.c @@ -57,7 +57,7 @@ static int drm_version(struct drm_device *dev, void *data, struct drm_file *file_priv); #define DRM_IOCTL_DEF(ioctl, _func, _flags) \ - [DRM_IOCTL_NR(ioctl)] = {.cmd = ioctl, .func = _func, .flags = _flags, .cmd_drv = 0} + [DRM_IOCTL_NR(ioctl)] = {.cmd = ioctl, .func = _func, .flags = _flags, .cmd_drv = 0, .name = #ioctl} /** Ioctl table */ static const struct drm_ioctl_desc drm_ioctls[] = { @@ -375,7 +375,7 @@ long drm_ioctl(struct file *filp, { struct drm_file *file_priv = filp->private_data; struct drm_device *dev; - const struct drm_ioctl_desc *ioctl; + const struct drm_ioctl_desc *ioctl = NULL; drm_ioctl_t *func; unsigned int nr = DRM_IOCTL_NR(cmd); int retcode = -EINVAL; @@ -392,11 +392,6 @@ long drm_ioctl(struct file *filp, atomic_inc(&dev->counts[_DRM_STAT_IOCTLS]); ++file_priv->ioctl_count; - DRM_DEBUG("pid=%d, cmd=0x%02x, nr=0x%02x, dev 0x%lx, auth=%d\n", - task_pid_nr(current), cmd, nr, - (long)old_encode_dev(file_priv->minor->device), - file_priv->authenticated); - if ((nr >= DRM_CORE_IOCTL_COUNT) && ((nr < DRM_COMMAND_BASE) || (nr >= DRM_COMMAND_END))) goto err_i1; @@ -417,6 +412,11 @@ long drm_ioctl(struct file *filp, } else goto err_i1; + DRM_DEBUG("pid=%d, dev=0x%lx, auth=%d, %s\n", + task_pid_nr(current), + (long)old_encode_dev(file_priv->minor->device), + file_priv->authenticated, ioctl->name); + /* Do not trust userspace, use our own definition */ func = ioctl->func; /* is there a local override? */ @@ -471,6 +471,12 @@ long drm_ioctl(struct file *filp, } err_i1: + if (!ioctl) + DRM_DEBUG("invalid iotcl: pid=%d, dev=0x%lx, auth=%d, cmd=0x%02x, nr=0x%02x\n", + task_pid_nr(current), + (long)old_encode_dev(file_priv->minor->device), + file_priv->authenticated, cmd, nr); + if (kdata != stack_kdata) kfree(kdata); atomic_dec(&dev->ioctl_count); diff --git a/include/drm/drmP.h b/include/drm/drmP.h index f1ce786736e4..b092b5bcf227 100644 --- a/include/drm/drmP.h +++ b/include/drm/drmP.h @@ -316,6 +316,7 @@ struct drm_ioctl_desc { int flags; drm_ioctl_t *func; unsigned int cmd_drv; + const char *name; }; /** @@ -324,7 +325,7 @@ struct drm_ioctl_desc { */ #define DRM_IOCTL_DEF_DRV(ioctl, _func, _flags) \ - [DRM_IOCTL_NR(DRM_##ioctl)] = {.cmd = DRM_##ioctl, .func = _func, .flags = _flags, .cmd_drv = DRM_IOCTL_##ioctl} + [DRM_IOCTL_NR(DRM_##ioctl)] = {.cmd = DRM_##ioctl, .func = _func, .flags = _flags, .cmd_drv = DRM_IOCTL_##ioctl, .name = #ioctl} struct drm_magic_entry { struct list_head head; From 73ed148aea9dc0508be7e30e7a447f55c1b2f378 Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Fri, 10 May 2013 16:59:18 +1000 Subject: [PATCH 069/198] powerpc/powernv: Improve kexec reliability We add a machine_shutdown hook that frees the OPAL interrupts (so they get masked at the source and don't fire while kexec'ing) and which triggers an IODA reset on all the PCIe host bridges which will have the effect of blocking all DMAs and subsequent PCIs interrupts. Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/include/asm/opal.h | 2 ++ arch/powerpc/platforms/powernv/opal.c | 17 +++++++++++++++++ arch/powerpc/platforms/powernv/pci-ioda.c | 9 +++++++++ arch/powerpc/platforms/powernv/pci.c | 12 ++++++++++++ arch/powerpc/platforms/powernv/pci.h | 2 ++ arch/powerpc/platforms/powernv/powernv.h | 2 ++ arch/powerpc/platforms/powernv/setup.c | 12 ++++++++++++ 7 files changed, 56 insertions(+) diff --git a/arch/powerpc/include/asm/opal.h b/arch/powerpc/include/asm/opal.h index b6c8b58b1d76..b2906adb89d3 100644 --- a/arch/powerpc/include/asm/opal.h +++ b/arch/powerpc/include/asm/opal.h @@ -563,6 +563,8 @@ extern void opal_nvram_init(void); extern int opal_machine_check(struct pt_regs *regs); +extern void opal_shutdown(void); + #endif /* __ASSEMBLY__ */ #endif /* __OPAL_H */ diff --git a/arch/powerpc/platforms/powernv/opal.c b/arch/powerpc/platforms/powernv/opal.c index 12d9846aa87e..27907cb18b8a 100644 --- a/arch/powerpc/platforms/powernv/opal.c +++ b/arch/powerpc/platforms/powernv/opal.c @@ -15,6 +15,7 @@ #include #include #include +#include #include #include @@ -28,6 +29,8 @@ struct opal { static struct device_node *opal_node; static DEFINE_SPINLOCK(opal_write_lock); extern u64 opal_mc_secondary_handler[]; +static unsigned int *opal_irqs; +static unsigned int opal_irq_count; int __init early_init_dt_scan_opal(unsigned long node, const char *uname, int depth, void *data) @@ -323,6 +326,8 @@ static int __init opal_init(void) irqs = of_get_property(opal_node, "opal-interrupts", &irqlen); pr_debug("opal: Found %d interrupts reserved for OPAL\n", irqs ? (irqlen / 4) : 0); + opal_irq_count = irqlen / 4; + opal_irqs = kzalloc(opal_irq_count * sizeof(unsigned int), GFP_KERNEL); for (i = 0; irqs && i < (irqlen / 4); i++, irqs++) { unsigned int hwirq = be32_to_cpup(irqs); unsigned int irq = irq_create_mapping(NULL, hwirq); @@ -334,7 +339,19 @@ static int __init opal_init(void) if (rc) pr_warning("opal: Error %d requesting irq %d" " (0x%x)\n", rc, irq, hwirq); + opal_irqs[i] = irq; } return 0; } subsys_initcall(opal_init); + +void opal_shutdown(void) +{ + unsigned int i; + + for (i = 0; i < opal_irq_count; i++) { + if (opal_irqs[i]) + free_irq(opal_irqs[i], 0); + opal_irqs[i] = 0; + } +} diff --git a/arch/powerpc/platforms/powernv/pci-ioda.c b/arch/powerpc/platforms/powernv/pci-ioda.c index 1da578b7c1bf..3937aaae5bc4 100644 --- a/arch/powerpc/platforms/powernv/pci-ioda.c +++ b/arch/powerpc/platforms/powernv/pci-ioda.c @@ -1048,6 +1048,12 @@ static u32 pnv_ioda_bdfn_to_pe(struct pnv_phb *phb, struct pci_bus *bus, return phb->ioda.pe_rmap[(bus->number << 8) | devfn]; } +static void pnv_pci_ioda_shutdown(struct pnv_phb *phb) +{ + opal_pci_reset(phb->opal_id, OPAL_PCI_IODA_TABLE_RESET, + OPAL_ASSERT_RESET); +} + void __init pnv_pci_init_ioda_phb(struct device_node *np, int ioda_type) { struct pci_controller *hose; @@ -1178,6 +1184,9 @@ void __init pnv_pci_init_ioda_phb(struct device_node *np, int ioda_type) /* Setup TCEs */ phb->dma_dev_setup = pnv_pci_ioda_dma_dev_setup; + /* Setup shutdown function for kexec */ + phb->shutdown = pnv_pci_ioda_shutdown; + /* Setup MSI support */ pnv_pci_init_ioda_msis(phb); diff --git a/arch/powerpc/platforms/powernv/pci.c b/arch/powerpc/platforms/powernv/pci.c index 55dfca844ddf..163bd7422f1c 100644 --- a/arch/powerpc/platforms/powernv/pci.c +++ b/arch/powerpc/platforms/powernv/pci.c @@ -450,6 +450,18 @@ static void pnv_pci_dma_dev_setup(struct pci_dev *pdev) pnv_pci_dma_fallback_setup(hose, pdev); } +void pnv_pci_shutdown(void) +{ + struct pci_controller *hose; + + list_for_each_entry(hose, &hose_list, list_node) { + struct pnv_phb *phb = hose->private_data; + + if (phb && phb->shutdown) + phb->shutdown(phb); + } +} + /* Fixup wrong class code in p7ioc and p8 root complex */ static void pnv_p7ioc_rc_quirk(struct pci_dev *dev) { diff --git a/arch/powerpc/platforms/powernv/pci.h b/arch/powerpc/platforms/powernv/pci.h index 48dc4bb856a1..25d76c4df50b 100644 --- a/arch/powerpc/platforms/powernv/pci.h +++ b/arch/powerpc/platforms/powernv/pci.h @@ -86,6 +86,7 @@ struct pnv_phb { void (*dma_dev_setup)(struct pnv_phb *phb, struct pci_dev *pdev); void (*fixup_phb)(struct pci_controller *hose); u32 (*bdfn_to_pe)(struct pnv_phb *phb, struct pci_bus *bus, u32 devfn); + void (*shutdown)(struct pnv_phb *phb); union { struct { @@ -158,4 +159,5 @@ extern void pnv_pci_init_ioda_hub(struct device_node *np); extern void pnv_pci_init_ioda2_phb(struct device_node *np); extern void pnv_pci_ioda_tce_invalidate(struct iommu_table *tbl, u64 *startp, u64 *endp); + #endif /* __POWERNV_PCI_H */ diff --git a/arch/powerpc/platforms/powernv/powernv.h b/arch/powerpc/platforms/powernv/powernv.h index 8a9df7f9667e..a1c6f83fc391 100644 --- a/arch/powerpc/platforms/powernv/powernv.h +++ b/arch/powerpc/platforms/powernv/powernv.h @@ -9,8 +9,10 @@ static inline void pnv_smp_init(void) { } #ifdef CONFIG_PCI extern void pnv_pci_init(void); +extern void pnv_pci_shutdown(void); #else static inline void pnv_pci_init(void) { } +static inline void pnv_pci_shutdown(void) { } #endif #endif /* _POWERNV_H */ diff --git a/arch/powerpc/platforms/powernv/setup.c b/arch/powerpc/platforms/powernv/setup.c index db1ad1c8f68f..c20381caaa3e 100644 --- a/arch/powerpc/platforms/powernv/setup.c +++ b/arch/powerpc/platforms/powernv/setup.c @@ -126,6 +126,17 @@ static void pnv_progress(char *s, unsigned short hex) { } +static void pnv_shutdown(void) +{ + /* Let the PCI code clear up IODA tables */ + pnv_pci_shutdown(); + + /* And unregister all OPAL interrupts so they don't fire + * up while we kexec + */ + opal_shutdown(); +} + #ifdef CONFIG_KEXEC static void pnv_kexec_cpu_down(int crash_shutdown, int secondary) { @@ -187,6 +198,7 @@ define_machine(powernv) { .init_IRQ = pnv_init_IRQ, .show_cpuinfo = pnv_show_cpuinfo, .progress = pnv_progress, + .machine_shutdown = pnv_shutdown, .power_save = power7_idle, .calibrate_decr = generic_calibrate_decr, #ifdef CONFIG_KEXEC From 4a3b8d0b833707811011ad14dd12dabf31fa690a Mon Sep 17 00:00:00 2001 From: Scott Wood Date: Thu, 9 May 2013 22:09:41 -0500 Subject: [PATCH 070/198] powerpc: hard_irq_disable(): Call trace_hardirqs_off after disabling lockdep.c has this: /* * So we're supposed to get called after you mask local IRQs, * but for some reason the hardware doesn't quite think you did * a proper job. */ if (DEBUG_LOCKS_WARN_ON(!irqs_disabled())) return; Since irqs_disabled() is based on soft_enabled(), that (not just the hard EE bit) needs to be 0 before we call trace_hardirqs_off. Signed-off-by: Scott Wood --- arch/powerpc/include/asm/hw_irq.h | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/arch/powerpc/include/asm/hw_irq.h b/arch/powerpc/include/asm/hw_irq.h index d615b28dda82..ba713f166fa5 100644 --- a/arch/powerpc/include/asm/hw_irq.h +++ b/arch/powerpc/include/asm/hw_irq.h @@ -96,11 +96,12 @@ static inline bool arch_irqs_disabled(void) #endif #define hard_irq_disable() do { \ + u8 _was_enabled = get_paca()->soft_enabled; \ __hard_irq_disable(); \ - if (local_paca->soft_enabled) \ - trace_hardirqs_off(); \ get_paca()->soft_enabled = 0; \ get_paca()->irq_happened |= PACA_IRQ_HARD_DIS; \ + if (_was_enabled) \ + trace_hardirqs_off(); \ } while(0) static inline bool lazy_irq_pending(void) From cf8b166d5c1c89aad6c436a954fa40fd18a75bfb Mon Sep 17 00:00:00 2001 From: Zhang Yanfei Date: Thu, 9 May 2013 23:57:42 +0800 Subject: [PATCH 071/198] x86/mm: Add missing comments for initial kernel direct mapping Two sets of comments were lost during patch-series shuffling: - comments for init_range_memory_mapping() - comments in init_mem_mapping that is helpful for reminding people that the pagetable is setup top-down The comments were written by Yinghai in his patch in: https://lkml.org/lkml/2012/11/28/620 This patch reintroduces them. Originally-From: Yinghai Lu Signed-off-by: Zhang Yanfei Cc: Yasuaki Ishimatsu Cc: Konrad Rzeszutek Wilk Cc: Andrew Morton Link: http://lkml.kernel.org/r/518BC776.7010506@gmail.com [ Tidied it all up a bit. ] Signed-off-by: Ingo Molnar --- arch/x86/mm/init.c | 19 ++++++++++++++++++- 1 file changed, 18 insertions(+), 1 deletion(-) diff --git a/arch/x86/mm/init.c b/arch/x86/mm/init.c index fdc5dca14fb3..eaac1743def7 100644 --- a/arch/x86/mm/init.c +++ b/arch/x86/mm/init.c @@ -359,7 +359,17 @@ unsigned long __init_refok init_memory_mapping(unsigned long start, } /* - * would have hole in the middle or ends, and only ram parts will be mapped. + * We need to iterate through the E820 memory map and create direct mappings + * for only E820_RAM and E820_KERN_RESERVED regions. We cannot simply + * create direct mappings for all pfns from [0 to max_low_pfn) and + * [4GB to max_pfn) because of possible memory holes in high addresses + * that cannot be marked as UC by fixed/variable range MTRRs. + * Depending on the alignment of E820 ranges, this may possibly result + * in using smaller size (i.e. 4K instead of 2M or 1G) page tables. + * + * init_mem_mapping() calls init_range_memory_mapping() with big range. + * That range would have hole in the middle or ends, and only ram parts + * will be mapped in init_range_memory_mapping(). */ static unsigned long __init init_range_memory_mapping( unsigned long r_start, @@ -419,6 +429,13 @@ void __init init_mem_mapping(void) max_pfn_mapped = 0; /* will get exact value next */ min_pfn_mapped = real_end >> PAGE_SHIFT; last_start = start = real_end; + + /* + * We start from the top (end of memory) and go to the bottom. + * The memblock_find_in_range() gets us a block of RAM from the + * end of RAM in [min_pfn_mapped, max_pfn_mapped) used as new pages + * for page table. + */ while (last_start > ISA_END_ADDRESS) { if (last_start > step_size) { start = round_down(last_start - 1, step_size); From db90f91f6f4a870583d5c11cda187f20e4d835ae Mon Sep 17 00:00:00 2001 From: Lee Jones Date: Fri, 3 May 2013 12:52:12 +0100 Subject: [PATCH 072/198] ARM: 7714/1: mmc: mmci: Ensure return value of regulator_enable() is checked MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This patch suppresses the warning below: drivers/mmc/host/mmci.c: In function ‘mmci_set_ios’: drivers/mmc/host/mmci.c:1165:20: warning: ignoring return value of ‘regulator_enable’, declared with attribute warn_unused_result [-Wunused-result] Cc: Chris Ball Acked-by: Srinidhi Kasagar Signed-off-by: Lee Jones Signed-off-by: Russell King --- drivers/mmc/host/mmci.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c index 375c109607ff..f4f3038c1df0 100644 --- a/drivers/mmc/host/mmci.c +++ b/drivers/mmc/host/mmci.c @@ -1130,6 +1130,7 @@ static void mmci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios) struct variant_data *variant = host->variant; u32 pwr = 0; unsigned long flags; + int ret; pm_runtime_get_sync(mmc_dev(mmc)); @@ -1161,8 +1162,12 @@ static void mmci_set_ios(struct mmc_host *mmc, struct mmc_ios *ios) break; case MMC_POWER_ON: if (!IS_ERR(mmc->supply.vqmmc) && - !regulator_is_enabled(mmc->supply.vqmmc)) - regulator_enable(mmc->supply.vqmmc); + !regulator_is_enabled(mmc->supply.vqmmc)) { + ret = regulator_enable(mmc->supply.vqmmc); + if (ret < 0) + dev_err(mmc_dev(mmc), + "failed to enable vqmmc regulator\n"); + } pwr |= MCI_PWR_ON; break; From 9e01573b5cf8b3188c1f33493d2e73a30e8d25fc Mon Sep 17 00:00:00 2001 From: Nicolas Pitre Date: Thu, 9 May 2013 19:19:20 +0100 Subject: [PATCH 073/198] ARM: 7715/1: MCPM: adapt to GIC changes after upstream merge Since commit c0114709ed85 ("irqchip: gic: Perform the gic_secondary_init() call via CPU notifier") it is no longer required nor possible to call gic_secondary_init() from platform code. Signed-off-by: Nicolas Pitre Signed-off-by: Russell King --- arch/arm/common/mcpm_platsmp.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/arch/arm/common/mcpm_platsmp.c b/arch/arm/common/mcpm_platsmp.c index 52b88d81b7bb..3caed0db6986 100644 --- a/arch/arm/common/mcpm_platsmp.c +++ b/arch/arm/common/mcpm_platsmp.c @@ -15,8 +15,6 @@ #include #include -#include - #include #include #include @@ -49,7 +47,6 @@ static int __cpuinit mcpm_boot_secondary(unsigned int cpu, struct task_struct *i static void __cpuinit mcpm_secondary_init(unsigned int cpu) { mcpm_cpu_powered_up(); - gic_secondary_init(0); } #ifdef CONFIG_HOTPLUG_CPU From d3251859168b0b12841e1b90d6d768ab478dc23d Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 10 May 2013 11:10:17 -0700 Subject: [PATCH 074/198] workqueue: workqueue_congested() shouldn't translate WORK_CPU_UNBOUND into node number df2d5ae499 ("workqueue: map an unbound workqueues to multiple per-node pool_workqueues") made unbound workqueues to map to multiple per-node pool_workqueues and accordingly updated workqueue_contested() so that, for unbound workqueues, it maps the specified @cpu to the NUMA node number to obtain the matching pool_workqueue to query the congested state. Before this change, workqueue_congested() ignored @cpu for unbound workqueues as there was only one pool_workqueue and some users (fscache) called it with WORK_CPU_UNBOUND. After the commit, this causes the following oops as WORK_CPU_UNBOUND gets translated to garbage by cpu_to_node(). BUG: unable to handle kernel paging request at ffff8803598d98b8 IP: [] unbound_pwq_by_node+0xa1/0xfa PGD 2421067 PUD 0 Oops: 0000 [#1] SMP CPU: 1 PID: 2689 Comm: cat Tainted: GF 3.9.0-fsdevel+ #4 task: ffff88003d801040 ti: ffff880025806000 task.ti: ffff880025806000 RIP: 0010:[] [] unbound_pwq_by_node+0xa1/0xfa RSP: 0018:ffff880025807ad8 EFLAGS: 00010202 RAX: 0000000000000001 RBX: ffff8800388a2400 RCX: 0000000000000003 RDX: ffff880025807fd8 RSI: ffffffff81a31420 RDI: ffff88003d8016e0 RBP: ffff880025807ae8 R08: ffff88003d801730 R09: ffffffffa00b4898 R10: ffffffff81044217 R11: ffff88003d801040 R12: 0000000064206e97 R13: ffff880036059d98 R14: ffff880038cc8080 R15: ffff880038cc82d0 FS: 00007f21afd9c740(0000) GS:ffff88003d100000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: ffff8803598d98b8 CR3: 000000003df49000 CR4: 00000000000007e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Stack: ffff8800388a2400 0000000000000002 ffff880025807b18 ffffffff810442ce ffffffff81044217 ffff880000000002 ffff8800371b4080 ffff88003d112ec0 ffff880025807b38 ffffffffa00810b0 ffff880036059d88 ffff880036059be8 Call Trace: [] workqueue_congested+0xb7/0x12c [] fscache_enqueue_object+0xb2/0xe8 [fscache] [] __fscache_acquire_cookie+0x3b9/0x56c [fscache] [] nfs_fscache_set_inode_cookie+0xee/0x132 [nfs] [] do_open+0x9/0xd [nfs] [] do_dentry_open+0x175/0x24b [] finish_open+0x41/0x51 Fix it by using smp_processor_id() if @cpu is WORK_CPU_UNBOUND. Signed-off-by: Tejun Heo Reported-by: David Howells Tested-and-Acked-by: David Howells --- kernel/workqueue.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 4aa9f5bc6b2d..1ae602809efb 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -4311,6 +4311,12 @@ bool current_is_workqueue_rescuer(void) * no synchronization around this function and the test result is * unreliable and only useful as advisory hints or for debugging. * + * If @cpu is WORK_CPU_UNBOUND, the test is performed on the local CPU. + * Note that both per-cpu and unbound workqueues may be associated with + * multiple pool_workqueues which have separate congested states. A + * workqueue being congested on one CPU doesn't mean the workqueue is also + * contested on other CPUs / NUMA nodes. + * * RETURNS: * %true if congested, %false otherwise. */ @@ -4321,6 +4327,9 @@ bool workqueue_congested(int cpu, struct workqueue_struct *wq) rcu_read_lock_sched(); + if (cpu == WORK_CPU_UNBOUND) + cpu = smp_processor_id(); + if (!(wq->flags & WQ_UNBOUND)) pwq = per_cpu_ptr(wq->cpu_pwqs, cpu); else From 416821d3d68164909b2cbcf398e4ba0797f5f8a2 Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Fri, 10 May 2013 21:24:01 +0000 Subject: [PATCH 075/198] parisc: implement irq stacks - part 2 (v2) This patch fixes few build issues which were introduced with the last irq stack patch, e.g. the combination of stack overflow check and irq stack. Furthermore we now do proper locking and change the irq bh handler to use the irq stack as well. In /proc/interrupts one now can monitor how huge the irq stack has grown and how often it was preferred over the kernel stack. IRQ stacks are now enabled by default just to make sure that we not overflow the kernel stack by accident. Signed-off-by: Helge Deller --- arch/parisc/Kconfig | 2 +- arch/parisc/include/asm/hardirq.h | 9 +++ arch/parisc/include/asm/processor.h | 3 + arch/parisc/kernel/irq.c | 101 ++++++++++++++++++++++++---- arch/parisc/mm/init.c | 4 +- 5 files changed, 102 insertions(+), 17 deletions(-) diff --git a/arch/parisc/Kconfig b/arch/parisc/Kconfig index cad060f288cf..6507dabdd5dd 100644 --- a/arch/parisc/Kconfig +++ b/arch/parisc/Kconfig @@ -245,7 +245,7 @@ config SMP config IRQSTACKS bool "Use separate kernel stacks when processing interrupts" - default n + default y help If you say Y here the kernel will use separate kernel stacks for handling hard and soft interrupts. This can help avoid diff --git a/arch/parisc/include/asm/hardirq.h b/arch/parisc/include/asm/hardirq.h index 12373c4dabab..c19f7138ba48 100644 --- a/arch/parisc/include/asm/hardirq.h +++ b/arch/parisc/include/asm/hardirq.h @@ -11,10 +11,18 @@ #include #include +#ifdef CONFIG_IRQSTACKS +#define __ARCH_HAS_DO_SOFTIRQ +#endif + typedef struct { unsigned int __softirq_pending; #ifdef CONFIG_DEBUG_STACKOVERFLOW unsigned int kernel_stack_usage; +#ifdef CONFIG_IRQSTACKS + unsigned int irq_stack_usage; + unsigned int irq_stack_counter; +#endif #endif #ifdef CONFIG_SMP unsigned int irq_resched_count; @@ -28,6 +36,7 @@ DECLARE_PER_CPU_SHARED_ALIGNED(irq_cpustat_t, irq_stat); #define __ARCH_IRQ_STAT #define __IRQ_STAT(cpu, member) (irq_stat[cpu].member) #define inc_irq_stat(member) this_cpu_inc(irq_stat.member) +#define __inc_irq_stat(member) __this_cpu_inc(irq_stat.member) #define local_softirq_pending() this_cpu_read(irq_stat.__softirq_pending) #define __ARCH_SET_SOFTIRQ_PENDING diff --git a/arch/parisc/include/asm/processor.h b/arch/parisc/include/asm/processor.h index 064015547d1e..cfbc43929cf6 100644 --- a/arch/parisc/include/asm/processor.h +++ b/arch/parisc/include/asm/processor.h @@ -63,10 +63,13 @@ */ #ifdef __KERNEL__ +#include + #define IRQ_STACK_SIZE (4096 << 2) /* 16k irq stack size */ union irq_stack_union { unsigned long stack[IRQ_STACK_SIZE/sizeof(unsigned long)]; + raw_spinlock_t lock; }; DECLARE_PER_CPU(union irq_stack_union, irq_stack_union); diff --git a/arch/parisc/kernel/irq.c b/arch/parisc/kernel/irq.c index e255db0bb761..55237a70e197 100644 --- a/arch/parisc/kernel/irq.c +++ b/arch/parisc/kernel/irq.c @@ -166,22 +166,32 @@ int arch_show_interrupts(struct seq_file *p, int prec) seq_printf(p, "%*s: ", prec, "STK"); for_each_online_cpu(j) seq_printf(p, "%10u ", irq_stats(j)->kernel_stack_usage); - seq_printf(p, " Kernel stack usage\n"); + seq_puts(p, " Kernel stack usage\n"); +# ifdef CONFIG_IRQSTACKS + seq_printf(p, "%*s: ", prec, "IST"); + for_each_online_cpu(j) + seq_printf(p, "%10u ", irq_stats(j)->irq_stack_usage); + seq_puts(p, " Interrupt stack usage\n"); + seq_printf(p, "%*s: ", prec, "ISC"); + for_each_online_cpu(j) + seq_printf(p, "%10u ", irq_stats(j)->irq_stack_counter); + seq_puts(p, " Interrupt stack usage counter\n"); +# endif #endif #ifdef CONFIG_SMP seq_printf(p, "%*s: ", prec, "RES"); for_each_online_cpu(j) seq_printf(p, "%10u ", irq_stats(j)->irq_resched_count); - seq_printf(p, " Rescheduling interrupts\n"); + seq_puts(p, " Rescheduling interrupts\n"); seq_printf(p, "%*s: ", prec, "CAL"); for_each_online_cpu(j) seq_printf(p, "%10u ", irq_stats(j)->irq_call_count); - seq_printf(p, " Function call interrupts\n"); + seq_puts(p, " Function call interrupts\n"); #endif seq_printf(p, "%*s: ", prec, "TLB"); for_each_online_cpu(j) seq_printf(p, "%10u ", irq_stats(j)->irq_tlb_count); - seq_printf(p, " TLB shootdowns\n"); + seq_puts(p, " TLB shootdowns\n"); return 0; } @@ -378,6 +388,7 @@ static inline void stack_overflow_check(struct pt_regs *regs) unsigned long sp = regs->gr[30]; unsigned long stack_usage; unsigned int *last_usage; + int cpu = smp_processor_id(); /* if sr7 != 0, we interrupted a userspace process which we do not want * to check for stack overflow. We will only check the kernel stack. */ @@ -386,7 +397,31 @@ static inline void stack_overflow_check(struct pt_regs *regs) /* calculate kernel stack usage */ stack_usage = sp - stack_start; - last_usage = &per_cpu(irq_stat.kernel_stack_usage, smp_processor_id()); +#ifdef CONFIG_IRQSTACKS + if (likely(stack_usage <= THREAD_SIZE)) + goto check_kernel_stack; /* found kernel stack */ + + /* check irq stack usage */ + stack_start = (unsigned long) &per_cpu(irq_stack_union, cpu).stack; + stack_usage = sp - stack_start; + + last_usage = &per_cpu(irq_stat.irq_stack_usage, cpu); + if (unlikely(stack_usage > *last_usage)) + *last_usage = stack_usage; + + if (likely(stack_usage < (IRQ_STACK_SIZE - STACK_MARGIN))) + return; + + pr_emerg("stackcheck: %s will most likely overflow irq stack " + "(sp:%lx, stk bottom-top:%lx-%lx)\n", + current->comm, sp, stack_start, stack_start + IRQ_STACK_SIZE); + goto panic_check; + +check_kernel_stack: +#endif + + /* check kernel stack usage */ + last_usage = &per_cpu(irq_stat.kernel_stack_usage, cpu); if (unlikely(stack_usage > *last_usage)) *last_usage = stack_usage; @@ -398,31 +433,69 @@ static inline void stack_overflow_check(struct pt_regs *regs) "(sp:%lx, stk bottom-top:%lx-%lx)\n", current->comm, sp, stack_start, stack_start + THREAD_SIZE); +#ifdef CONFIG_IRQSTACKS +panic_check: +#endif if (sysctl_panic_on_stackoverflow) panic("low stack detected by irq handler - check messages\n"); #endif } #ifdef CONFIG_IRQSTACKS -DEFINE_PER_CPU(union irq_stack_union, irq_stack_union); +DEFINE_PER_CPU(union irq_stack_union, irq_stack_union) = { + .lock = __RAW_SPIN_LOCK_UNLOCKED((irq_stack_union).lock) + }; static void execute_on_irq_stack(void *func, unsigned long param1) { - unsigned long *irq_stack_start; + union irq_stack_union *union_ptr; unsigned long irq_stack; - int cpu = smp_processor_id(); + raw_spinlock_t *irq_stack_in_use; - irq_stack_start = &per_cpu(irq_stack_union, cpu).stack[0]; - irq_stack = (unsigned long) irq_stack_start; - irq_stack = ALIGN(irq_stack, 16); /* align for stack frame usage */ + union_ptr = &per_cpu(irq_stack_union, smp_processor_id()); + irq_stack = (unsigned long) &union_ptr->stack; + irq_stack = ALIGN(irq_stack + sizeof(irq_stack_union.lock), + 64); /* align for stack frame usage */ - BUG_ON(*irq_stack_start); /* report bug if we were called recursive. */ - *irq_stack_start = 1; + /* We may be called recursive. If we are already using the irq stack, + * just continue to use it. Use spinlocks to serialize + * the irq stack usage. + */ + irq_stack_in_use = &union_ptr->lock; + if (!raw_spin_trylock(irq_stack_in_use)) { + void (*direct_call)(unsigned long p1) = func; + + /* We are using the IRQ stack already. + * Do direct call on current stack. */ + direct_call(param1); + return; + } /* This is where we switch to the IRQ stack. */ call_on_stack(param1, func, irq_stack); - *irq_stack_start = 0; + __inc_irq_stat(irq_stack_counter); + + /* free up irq stack usage. */ + do_raw_spin_unlock(irq_stack_in_use); +} + +asmlinkage void do_softirq(void) +{ + __u32 pending; + unsigned long flags; + + if (in_interrupt()) + return; + + local_irq_save(flags); + + pending = local_softirq_pending(); + + if (pending) + execute_on_irq_stack(__do_softirq, 0); + + local_irq_restore(flags); } #endif /* CONFIG_IRQSTACKS */ diff --git a/arch/parisc/mm/init.c b/arch/parisc/mm/init.c index ce939ac8622b..1c965642068b 100644 --- a/arch/parisc/mm/init.c +++ b/arch/parisc/mm/init.c @@ -1069,7 +1069,7 @@ void flush_tlb_all(void) { int do_recycle; - inc_irq_stat(irq_tlb_count); + __inc_irq_stat(irq_tlb_count); do_recycle = 0; spin_lock(&sid_lock); if (dirty_space_ids > RECYCLE_THRESHOLD) { @@ -1090,7 +1090,7 @@ void flush_tlb_all(void) #else void flush_tlb_all(void) { - inc_irq_stat(irq_tlb_count); + __inc_irq_stat(irq_tlb_count); spin_lock(&sid_lock); flush_tlb_all_local(NULL); recycle_sids(); From f0a18819e261afc5fdbd8c5c6f9943123c5461ba Mon Sep 17 00:00:00 2001 From: John David Anglin Date: Fri, 10 May 2013 23:21:38 +0000 Subject: [PATCH 076/198] parisc: fix SMP races when updating PTE and TLB entries in entry.S Currently, race conditions exist in the handling of TLB interruptions in entry.S. In particular, dirty bit updates can be lost if an accessed interruption occurs just after the dirty bit interruption on a different cpu. Lost dirty bit updates result in user pages not being flushed and general system instability. This change adds lock and unlock macros to synchronize all PTE and TLB updates done in entry.S. As a result, userspace stability is significantly improved. Signed-off-by: John David Anglin Signed-off-by: Helge Deller --- arch/parisc/kernel/entry.S | 155 ++++++++++++++++++++----------------- 1 file changed, 83 insertions(+), 72 deletions(-) diff --git a/arch/parisc/kernel/entry.S b/arch/parisc/kernel/entry.S index 4bb96ad9b0b1..ae27cb6ce19a 100644 --- a/arch/parisc/kernel/entry.S +++ b/arch/parisc/kernel/entry.S @@ -452,9 +452,41 @@ L2_ptep \pgd,\pte,\index,\va,\fault .endm + /* Acquire pa_dbit_lock lock. */ + .macro dbit_lock spc,tmp,tmp1 +#ifdef CONFIG_SMP + cmpib,COND(=),n 0,\spc,2f + load32 PA(pa_dbit_lock),\tmp +1: LDCW 0(\tmp),\tmp1 + cmpib,COND(=) 0,\tmp1,1b + nop +2: +#endif + .endm + + /* Release pa_dbit_lock lock without reloading lock address. */ + .macro dbit_unlock0 spc,tmp +#ifdef CONFIG_SMP + or,COND(=) %r0,\spc,%r0 + stw \spc,0(\tmp) +#endif + .endm + + /* Release pa_dbit_lock lock. */ + .macro dbit_unlock1 spc,tmp +#ifdef CONFIG_SMP + load32 PA(pa_dbit_lock),\tmp + dbit_unlock0 \spc,\tmp +#endif + .endm + /* Set the _PAGE_ACCESSED bit of the PTE. Be clever and * don't needlessly dirty the cache line if it was already set */ - .macro update_ptep ptep,pte,tmp,tmp1 + .macro update_ptep spc,ptep,pte,tmp,tmp1 +#ifdef CONFIG_SMP + or,COND(=) %r0,\spc,%r0 + LDREG 0(\ptep),\pte +#endif ldi _PAGE_ACCESSED,\tmp1 or \tmp1,\pte,\tmp and,COND(<>) \tmp1,\pte,%r0 @@ -463,7 +495,11 @@ /* Set the dirty bit (and accessed bit). No need to be * clever, this is only used from the dirty fault */ - .macro update_dirty ptep,pte,tmp + .macro update_dirty spc,ptep,pte,tmp +#ifdef CONFIG_SMP + or,COND(=) %r0,\spc,%r0 + LDREG 0(\ptep),\pte +#endif ldi _PAGE_ACCESSED|_PAGE_DIRTY,\tmp or \tmp,\pte,\pte STREG \pte,0(\ptep) @@ -1111,11 +1147,13 @@ dtlb_miss_20w: L3_ptep ptp,pte,t0,va,dtlb_check_alias_20w - update_ptep ptp,pte,t0,t1 + dbit_lock spc,t0,t1 + update_ptep spc,ptp,pte,t0,t1 make_insert_tlb spc,pte,prot idtlbt pte,prot + dbit_unlock1 spc,t0 rfir nop @@ -1135,11 +1173,13 @@ nadtlb_miss_20w: L3_ptep ptp,pte,t0,va,nadtlb_check_alias_20w - update_ptep ptp,pte,t0,t1 + dbit_lock spc,t0,t1 + update_ptep spc,ptp,pte,t0,t1 make_insert_tlb spc,pte,prot idtlbt pte,prot + dbit_unlock1 spc,t0 rfir nop @@ -1161,7 +1201,8 @@ dtlb_miss_11: L2_ptep ptp,pte,t0,va,dtlb_check_alias_11 - update_ptep ptp,pte,t0,t1 + dbit_lock spc,t0,t1 + update_ptep spc,ptp,pte,t0,t1 make_insert_tlb_11 spc,pte,prot @@ -1172,6 +1213,7 @@ dtlb_miss_11: idtlbp prot,(%sr1,va) mtsp t0, %sr1 /* Restore sr1 */ + dbit_unlock1 spc,t0 rfir nop @@ -1192,7 +1234,8 @@ nadtlb_miss_11: L2_ptep ptp,pte,t0,va,nadtlb_check_alias_11 - update_ptep ptp,pte,t0,t1 + dbit_lock spc,t0,t1 + update_ptep spc,ptp,pte,t0,t1 make_insert_tlb_11 spc,pte,prot @@ -1204,6 +1247,7 @@ nadtlb_miss_11: idtlbp prot,(%sr1,va) mtsp t0, %sr1 /* Restore sr1 */ + dbit_unlock1 spc,t0 rfir nop @@ -1224,13 +1268,15 @@ dtlb_miss_20: L2_ptep ptp,pte,t0,va,dtlb_check_alias_20 - update_ptep ptp,pte,t0,t1 + dbit_lock spc,t0,t1 + update_ptep spc,ptp,pte,t0,t1 make_insert_tlb spc,pte,prot f_extend pte,t0 idtlbt pte,prot + dbit_unlock1 spc,t0 rfir nop @@ -1250,13 +1296,15 @@ nadtlb_miss_20: L2_ptep ptp,pte,t0,va,nadtlb_check_alias_20 - update_ptep ptp,pte,t0,t1 + dbit_lock spc,t0,t1 + update_ptep spc,ptp,pte,t0,t1 make_insert_tlb spc,pte,prot f_extend pte,t0 idtlbt pte,prot + dbit_unlock1 spc,t0 rfir nop @@ -1357,11 +1405,13 @@ itlb_miss_20w: L3_ptep ptp,pte,t0,va,itlb_fault - update_ptep ptp,pte,t0,t1 + dbit_lock spc,t0,t1 + update_ptep spc,ptp,pte,t0,t1 make_insert_tlb spc,pte,prot iitlbt pte,prot + dbit_unlock1 spc,t0 rfir nop @@ -1379,11 +1429,13 @@ naitlb_miss_20w: L3_ptep ptp,pte,t0,va,naitlb_check_alias_20w - update_ptep ptp,pte,t0,t1 + dbit_lock spc,t0,t1 + update_ptep spc,ptp,pte,t0,t1 make_insert_tlb spc,pte,prot iitlbt pte,prot + dbit_unlock1 spc,t0 rfir nop @@ -1405,7 +1457,8 @@ itlb_miss_11: L2_ptep ptp,pte,t0,va,itlb_fault - update_ptep ptp,pte,t0,t1 + dbit_lock spc,t0,t1 + update_ptep spc,ptp,pte,t0,t1 make_insert_tlb_11 spc,pte,prot @@ -1416,6 +1469,7 @@ itlb_miss_11: iitlbp prot,(%sr1,va) mtsp t0, %sr1 /* Restore sr1 */ + dbit_unlock1 spc,t0 rfir nop @@ -1427,7 +1481,8 @@ naitlb_miss_11: L2_ptep ptp,pte,t0,va,naitlb_check_alias_11 - update_ptep ptp,pte,t0,t1 + dbit_lock spc,t0,t1 + update_ptep spc,ptp,pte,t0,t1 make_insert_tlb_11 spc,pte,prot @@ -1438,6 +1493,7 @@ naitlb_miss_11: iitlbp prot,(%sr1,va) mtsp t0, %sr1 /* Restore sr1 */ + dbit_unlock1 spc,t0 rfir nop @@ -1459,13 +1515,15 @@ itlb_miss_20: L2_ptep ptp,pte,t0,va,itlb_fault - update_ptep ptp,pte,t0,t1 + dbit_lock spc,t0,t1 + update_ptep spc,ptp,pte,t0,t1 make_insert_tlb spc,pte,prot f_extend pte,t0 iitlbt pte,prot + dbit_unlock1 spc,t0 rfir nop @@ -1477,13 +1535,15 @@ naitlb_miss_20: L2_ptep ptp,pte,t0,va,naitlb_check_alias_20 - update_ptep ptp,pte,t0,t1 + dbit_lock spc,t0,t1 + update_ptep spc,ptp,pte,t0,t1 make_insert_tlb spc,pte,prot f_extend pte,t0 iitlbt pte,prot + dbit_unlock1 spc,t0 rfir nop @@ -1507,29 +1567,13 @@ dbit_trap_20w: L3_ptep ptp,pte,t0,va,dbit_fault -#ifdef CONFIG_SMP - cmpib,COND(=),n 0,spc,dbit_nolock_20w - load32 PA(pa_dbit_lock),t0 - -dbit_spin_20w: - LDCW 0(t0),t1 - cmpib,COND(=) 0,t1,dbit_spin_20w - nop - -dbit_nolock_20w: -#endif - update_dirty ptp,pte,t1 + dbit_lock spc,t0,t1 + update_dirty spc,ptp,pte,t1 make_insert_tlb spc,pte,prot idtlbt pte,prot -#ifdef CONFIG_SMP - cmpib,COND(=),n 0,spc,dbit_nounlock_20w - ldi 1,t1 - stw t1,0(t0) - -dbit_nounlock_20w: -#endif + dbit_unlock0 spc,t0 rfir nop @@ -1543,18 +1587,8 @@ dbit_trap_11: L2_ptep ptp,pte,t0,va,dbit_fault -#ifdef CONFIG_SMP - cmpib,COND(=),n 0,spc,dbit_nolock_11 - load32 PA(pa_dbit_lock),t0 - -dbit_spin_11: - LDCW 0(t0),t1 - cmpib,= 0,t1,dbit_spin_11 - nop - -dbit_nolock_11: -#endif - update_dirty ptp,pte,t1 + dbit_lock spc,t0,t1 + update_dirty spc,ptp,pte,t1 make_insert_tlb_11 spc,pte,prot @@ -1565,13 +1599,7 @@ dbit_nolock_11: idtlbp prot,(%sr1,va) mtsp t1, %sr1 /* Restore sr1 */ -#ifdef CONFIG_SMP - cmpib,COND(=),n 0,spc,dbit_nounlock_11 - ldi 1,t1 - stw t1,0(t0) - -dbit_nounlock_11: -#endif + dbit_unlock0 spc,t0 rfir nop @@ -1583,32 +1611,15 @@ dbit_trap_20: L2_ptep ptp,pte,t0,va,dbit_fault -#ifdef CONFIG_SMP - cmpib,COND(=),n 0,spc,dbit_nolock_20 - load32 PA(pa_dbit_lock),t0 - -dbit_spin_20: - LDCW 0(t0),t1 - cmpib,= 0,t1,dbit_spin_20 - nop - -dbit_nolock_20: -#endif - update_dirty ptp,pte,t1 + dbit_lock spc,t0,t1 + update_dirty spc,ptp,pte,t1 make_insert_tlb spc,pte,prot f_extend pte,t1 idtlbt pte,prot - -#ifdef CONFIG_SMP - cmpib,COND(=),n 0,spc,dbit_nounlock_20 - ldi 1,t1 - stw t1,0(t0) - -dbit_nounlock_20: -#endif + dbit_unlock0 spc,t0 rfir nop From cffe52f38eb6f633d63a3313661d9ce8daaae9dd Mon Sep 17 00:00:00 2001 From: Manish Chopra Date: Thu, 9 May 2013 09:25:09 +0000 Subject: [PATCH 077/198] qlcnic: Fix setting MAC address o Delete previous unicast MAC which is already programmed in adapter before setting new unicast MAC Signed-off-by: Manish Chopra Signed-off-by: Shahed Shaikh Signed-off-by: David S. Miller --- .../net/ethernet/qlogic/qlcnic/qlcnic_main.c | 21 +++++++++++++++++++ 1 file changed, 21 insertions(+) diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c index 264d5a4f8153..04037b5d4d8f 100644 --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c @@ -308,6 +308,23 @@ int qlcnic_read_mac_addr(struct qlcnic_adapter *adapter) return 0; } +static void qlcnic_delete_adapter_mac(struct qlcnic_adapter *adapter) +{ + struct qlcnic_mac_list_s *cur; + struct list_head *head; + + list_for_each(head, &adapter->mac_list) { + cur = list_entry(head, struct qlcnic_mac_list_s, list); + if (!memcmp(adapter->mac_addr, cur->mac_addr, ETH_ALEN)) { + qlcnic_sre_macaddr_change(adapter, cur->mac_addr, + 0, QLCNIC_MAC_DEL); + list_del(&cur->list); + kfree(cur); + return; + } + } +} + static int qlcnic_set_mac(struct net_device *netdev, void *p) { struct qlcnic_adapter *adapter = netdev_priv(netdev); @@ -322,11 +339,15 @@ static int qlcnic_set_mac(struct net_device *netdev, void *p) if (!is_valid_ether_addr(addr->sa_data)) return -EINVAL; + if (!memcmp(adapter->mac_addr, addr->sa_data, ETH_ALEN)) + return 0; + if (test_bit(__QLCNIC_DEV_UP, &adapter->state)) { netif_device_detach(netdev); qlcnic_napi_disable(adapter); } + qlcnic_delete_adapter_mac(adapter); memcpy(adapter->mac_addr, addr->sa_data, netdev->addr_len); memcpy(netdev->dev_addr, addr->sa_data, netdev->addr_len); qlcnic_set_multi(adapter->netdev); From 8c0464108a79c7698d8af26a7b299c7f6651e02e Mon Sep 17 00:00:00 2001 From: Shahed Shaikh Date: Thu, 9 May 2013 09:25:10 +0000 Subject: [PATCH 078/198] qlcnic: Fix ethtool strings o Add missing information in ethtool statistics information array. o Fix the typo in the statistics information string. Signed-off-by: Shahed Shaikh Signed-off-by: David S. Miller --- drivers/net/ethernet/qlogic/qlcnic/qlcnic_ethtool.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_ethtool.c b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_ethtool.c index 08efb4635007..6005b06dbf64 100644 --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_ethtool.c +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_ethtool.c @@ -131,12 +131,13 @@ static const char qlcnic_83xx_rx_stats_strings[][ETH_GSTRING_LEN] = { "ctx_lro_pkt_cnt", "ctx_ip_csum_error", "ctx_rx_pkts_wo_ctx", - "ctx_rx_pkts_dropped_wo_sts", + "ctx_rx_pkts_drop_wo_sds_on_card", + "ctx_rx_pkts_drop_wo_sds_on_host", "ctx_rx_osized_pkts", "ctx_rx_pkts_dropped_wo_rds", "ctx_rx_unexpected_mcast_pkts", "ctx_invalid_mac_address", - "ctx_rx_rds_ring_prim_attemoted", + "ctx_rx_rds_ring_prim_attempted", "ctx_rx_rds_ring_prim_success", "ctx_num_lro_flows_added", "ctx_num_lro_flows_removed", From 9fd13331ab8f108ade0e8077b69ff4663b63d43f Mon Sep 17 00:00:00 2001 From: Himanshu Madhani Date: Thu, 9 May 2013 09:25:11 +0000 Subject: [PATCH 079/198] qlcnic: Fix missing bracket in module parameter. Signed-off-by: Himanshu Madhani Signed-off-by: Shahed Shaikh Signed-off-by: David S. Miller --- drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c index 04037b5d4d8f..d97446a3bf05 100644 --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c @@ -37,24 +37,24 @@ MODULE_PARM_DESC(qlcnic_mac_learn, "Mac Filter (0=learning is disabled, 1=Driver learning is enabled, 2=FDB learning is enabled)"); int qlcnic_use_msi = 1; -MODULE_PARM_DESC(use_msi, "MSI interrupt (0=disabled, 1=enabled"); +MODULE_PARM_DESC(use_msi, "MSI interrupt (0=disabled, 1=enabled)"); module_param_named(use_msi, qlcnic_use_msi, int, 0444); int qlcnic_use_msi_x = 1; -MODULE_PARM_DESC(use_msi_x, "MSI-X interrupt (0=disabled, 1=enabled"); +MODULE_PARM_DESC(use_msi_x, "MSI-X interrupt (0=disabled, 1=enabled)"); module_param_named(use_msi_x, qlcnic_use_msi_x, int, 0444); int qlcnic_auto_fw_reset = 1; -MODULE_PARM_DESC(auto_fw_reset, "Auto firmware reset (0=disabled, 1=enabled"); +MODULE_PARM_DESC(auto_fw_reset, "Auto firmware reset (0=disabled, 1=enabled)"); module_param_named(auto_fw_reset, qlcnic_auto_fw_reset, int, 0644); int qlcnic_load_fw_file; -MODULE_PARM_DESC(load_fw_file, "Load firmware from (0=flash, 1=file"); +MODULE_PARM_DESC(load_fw_file, "Load firmware from (0=flash, 1=file)"); module_param_named(load_fw_file, qlcnic_load_fw_file, int, 0444); int qlcnic_config_npars; module_param(qlcnic_config_npars, int, 0444); -MODULE_PARM_DESC(qlcnic_config_npars, "Configure NPARs (0=disabled, 1=enabled"); +MODULE_PARM_DESC(qlcnic_config_npars, "Configure NPARs (0=disabled, 1=enabled)"); static int qlcnic_probe(struct pci_dev *pdev, const struct pci_device_id *ent); static void qlcnic_remove(struct pci_dev *pdev); From b938662d88264c1a92611ca1b82fdff5a4e87121 Mon Sep 17 00:00:00 2001 From: Himanshu Madhani Date: Thu, 9 May 2013 09:25:12 +0000 Subject: [PATCH 080/198] qlcnic: Fix ethtool supported port status for 83xx o Fix display for interface while using 'ethtool ' for 83xx adapter Signed-off-by: Himanshu Madhani Signed-off-by: Shahed Shaikh Signed-off-by: David S. Miller --- drivers/net/ethernet/qlogic/qlcnic/qlcnic.h | 2 + .../ethernet/qlogic/qlcnic/qlcnic_83xx_hw.c | 68 ++++++++++++++++++- .../ethernet/qlogic/qlcnic/qlcnic_83xx_hw.h | 2 +- .../ethernet/qlogic/qlcnic/qlcnic_ethtool.c | 49 +++++++------ 4 files changed, 94 insertions(+), 27 deletions(-) diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic.h b/drivers/net/ethernet/qlogic/qlcnic/qlcnic.h index 90c253b145ef..019c5f78732e 100644 --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic.h +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic.h @@ -429,6 +429,7 @@ struct qlcnic_hardware_context { u16 port_type; u16 board_type; + u16 supported_type; u16 link_speed; u16 link_duplex; @@ -1514,6 +1515,7 @@ void qlcnic_create_diag_entries(struct qlcnic_adapter *adapter); void qlcnic_remove_diag_entries(struct qlcnic_adapter *adapter); void qlcnic_82xx_add_sysfs(struct qlcnic_adapter *adapter); void qlcnic_82xx_remove_sysfs(struct qlcnic_adapter *adapter); +int qlcnic_82xx_get_settings(struct qlcnic_adapter *, struct ethtool_cmd *); int qlcnicvf_config_bridged_mode(struct qlcnic_adapter *, u32); int qlcnicvf_config_led(struct qlcnic_adapter *, u32, u32); diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.c b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.c index ea790a93ee7c..642fca0b15d2 100644 --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.c +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.c @@ -2830,6 +2830,23 @@ int qlcnic_83xx_test_link(struct qlcnic_adapter *adapter) break; } config = cmd.rsp.arg[3]; + if (QLC_83XX_SFP_PRESENT(config)) { + switch (ahw->module_type) { + case LINKEVENT_MODULE_OPTICAL_UNKNOWN: + case LINKEVENT_MODULE_OPTICAL_SRLR: + case LINKEVENT_MODULE_OPTICAL_LRM: + case LINKEVENT_MODULE_OPTICAL_SFP_1G: + ahw->supported_type = PORT_FIBRE; + break; + case LINKEVENT_MODULE_TWINAX_UNSUPPORTED_CABLE: + case LINKEVENT_MODULE_TWINAX_UNSUPPORTED_CABLELEN: + case LINKEVENT_MODULE_TWINAX: + ahw->supported_type = PORT_TP; + break; + default: + ahw->supported_type = PORT_OTHER; + } + } if (config & 1) err = 1; } @@ -2838,7 +2855,8 @@ out: return config; } -int qlcnic_83xx_get_settings(struct qlcnic_adapter *adapter) +int qlcnic_83xx_get_settings(struct qlcnic_adapter *adapter, + struct ethtool_cmd *ecmd) { u32 config = 0; int status = 0; @@ -2851,6 +2869,54 @@ int qlcnic_83xx_get_settings(struct qlcnic_adapter *adapter) ahw->module_type = QLC_83XX_SFP_MODULE_TYPE(config); /* hard code until there is a way to get it from flash */ ahw->board_type = QLCNIC_BRDTYPE_83XX_10G; + + if (netif_running(adapter->netdev) && ahw->has_link_events) { + ethtool_cmd_speed_set(ecmd, ahw->link_speed); + ecmd->duplex = ahw->link_duplex; + ecmd->autoneg = ahw->link_autoneg; + } else { + ethtool_cmd_speed_set(ecmd, SPEED_UNKNOWN); + ecmd->duplex = DUPLEX_UNKNOWN; + ecmd->autoneg = AUTONEG_DISABLE; + } + + if (ahw->port_type == QLCNIC_XGBE) { + ecmd->supported = SUPPORTED_1000baseT_Full; + ecmd->advertising = ADVERTISED_1000baseT_Full; + } else { + ecmd->supported = (SUPPORTED_10baseT_Half | + SUPPORTED_10baseT_Full | + SUPPORTED_100baseT_Half | + SUPPORTED_100baseT_Full | + SUPPORTED_1000baseT_Half | + SUPPORTED_1000baseT_Full); + ecmd->advertising = (ADVERTISED_100baseT_Half | + ADVERTISED_100baseT_Full | + ADVERTISED_1000baseT_Half | + ADVERTISED_1000baseT_Full); + } + + switch (ahw->supported_type) { + case PORT_FIBRE: + ecmd->supported |= SUPPORTED_FIBRE; + ecmd->advertising |= ADVERTISED_FIBRE; + ecmd->port = PORT_FIBRE; + ecmd->transceiver = XCVR_EXTERNAL; + break; + case PORT_TP: + ecmd->supported |= SUPPORTED_TP; + ecmd->advertising |= ADVERTISED_TP; + ecmd->port = PORT_TP; + ecmd->transceiver = XCVR_INTERNAL; + break; + default: + ecmd->supported |= SUPPORTED_FIBRE; + ecmd->advertising |= ADVERTISED_FIBRE; + ecmd->port = PORT_OTHER; + ecmd->transceiver = XCVR_EXTERNAL; + break; + } + ecmd->phy_address = ahw->physical_port; return status; } diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.h b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.h index 1f1d85e6f2af..faf5553681e1 100644 --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.h +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.h @@ -603,7 +603,7 @@ int qlcnic_83xx_get_vnic_pf_info(struct qlcnic_adapter *, struct qlcnic_info *); void qlcnic_83xx_get_minidump_template(struct qlcnic_adapter *); void qlcnic_83xx_get_stats(struct qlcnic_adapter *adapter, u64 *data); -int qlcnic_83xx_get_settings(struct qlcnic_adapter *); +int qlcnic_83xx_get_settings(struct qlcnic_adapter *, struct ethtool_cmd *); int qlcnic_83xx_set_settings(struct qlcnic_adapter *, struct ethtool_cmd *); void qlcnic_83xx_get_pauseparam(struct qlcnic_adapter *, struct ethtool_pauseparam *); diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_ethtool.c b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_ethtool.c index 6005b06dbf64..f67652de5a63 100644 --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_ethtool.c +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_ethtool.c @@ -252,6 +252,18 @@ static int qlcnic_get_settings(struct net_device *dev, struct ethtool_cmd *ecmd) { struct qlcnic_adapter *adapter = netdev_priv(dev); + + if (qlcnic_82xx_check(adapter)) + return qlcnic_82xx_get_settings(adapter, ecmd); + else if (qlcnic_83xx_check(adapter)) + return qlcnic_83xx_get_settings(adapter, ecmd); + + return -EIO; +} + +int qlcnic_82xx_get_settings(struct qlcnic_adapter *adapter, + struct ethtool_cmd *ecmd) +{ struct qlcnic_hardware_context *ahw = adapter->ahw; u32 speed, reg; int check_sfp_module = 0; @@ -277,10 +289,7 @@ qlcnic_get_settings(struct net_device *dev, struct ethtool_cmd *ecmd) } else if (adapter->ahw->port_type == QLCNIC_XGBE) { u32 val = 0; - if (qlcnic_83xx_check(adapter)) - qlcnic_83xx_get_settings(adapter); - else - val = QLCRD32(adapter, QLCNIC_PORT_MODE_ADDR); + val = QLCRD32(adapter, QLCNIC_PORT_MODE_ADDR); if (val == QLCNIC_PORT_MODE_802_3_AP) { ecmd->supported = SUPPORTED_1000baseT_Full; @@ -290,16 +299,13 @@ qlcnic_get_settings(struct net_device *dev, struct ethtool_cmd *ecmd) ecmd->advertising = ADVERTISED_10000baseT_Full; } - if (netif_running(dev) && adapter->ahw->has_link_events) { - if (qlcnic_82xx_check(adapter)) { - reg = QLCRD32(adapter, - P3P_LINK_SPEED_REG(pcifn)); - speed = P3P_LINK_SPEED_VAL(pcifn, reg); - ahw->link_speed = speed * P3P_LINK_SPEED_MHZ; - } - ethtool_cmd_speed_set(ecmd, adapter->ahw->link_speed); - ecmd->autoneg = adapter->ahw->link_autoneg; - ecmd->duplex = adapter->ahw->link_duplex; + if (netif_running(adapter->netdev) && ahw->has_link_events) { + reg = QLCRD32(adapter, P3P_LINK_SPEED_REG(pcifn)); + speed = P3P_LINK_SPEED_VAL(pcifn, reg); + ahw->link_speed = speed * P3P_LINK_SPEED_MHZ; + ethtool_cmd_speed_set(ecmd, ahw->link_speed); + ecmd->autoneg = ahw->link_autoneg; + ecmd->duplex = ahw->link_duplex; goto skip; } @@ -341,8 +347,8 @@ skip: case QLCNIC_BRDTYPE_P3P_10G_SFP_QT: ecmd->advertising |= ADVERTISED_TP; ecmd->supported |= SUPPORTED_TP; - check_sfp_module = netif_running(dev) && - adapter->ahw->has_link_events; + check_sfp_module = netif_running(adapter->netdev) && + ahw->has_link_events; case QLCNIC_BRDTYPE_P3P_10G_XFP: ecmd->supported |= SUPPORTED_FIBRE; ecmd->advertising |= ADVERTISED_FIBRE; @@ -356,8 +362,8 @@ skip: ecmd->advertising |= (ADVERTISED_FIBRE | ADVERTISED_TP); ecmd->port = PORT_FIBRE; - check_sfp_module = netif_running(dev) && - adapter->ahw->has_link_events; + check_sfp_module = netif_running(adapter->netdev) && + ahw->has_link_events; } else { ecmd->autoneg = AUTONEG_ENABLE; ecmd->supported |= (SUPPORTED_TP | SUPPORTED_Autoneg); @@ -366,13 +372,6 @@ skip: ecmd->port = PORT_TP; } break; - case QLCNIC_BRDTYPE_83XX_10G: - ecmd->autoneg = AUTONEG_DISABLE; - ecmd->supported |= (SUPPORTED_FIBRE | SUPPORTED_TP); - ecmd->advertising |= (ADVERTISED_FIBRE | ADVERTISED_TP); - ecmd->port = PORT_FIBRE; - check_sfp_module = netif_running(dev) && ahw->has_link_events; - break; default: dev_err(&adapter->pdev->dev, "Unsupported board model %d\n", adapter->ahw->board_type); From 536faa61825c8e92f520264228388d69783370d3 Mon Sep 17 00:00:00 2001 From: Sony Chacko Date: Thu, 9 May 2013 09:25:13 +0000 Subject: [PATCH 081/198] qlcnic: Fix reset recovery after transmit timeout o When transmit timeout happens, recovery attempt should start with adapter soft reset. If soft reset fails to resume traffic, firmware dump will be collected and driver will perform a hard reset of the adapter. Reset recovery on 83xx was failing after a hard reset. This patch fixes that issue. Signed-off-by: Sony Chacko Signed-off-by: Shahed Shaikh Signed-off-by: David S. Miller --- .../ethernet/qlogic/qlcnic/qlcnic_83xx_init.c | 22 +++++++++++-------- .../net/ethernet/qlogic/qlcnic/qlcnic_main.c | 15 ++++++++----- 2 files changed, 23 insertions(+), 14 deletions(-) diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_init.c b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_init.c index ab1d8d99cbd5..c67d1eb35e8f 100644 --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_init.c +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_init.c @@ -435,10 +435,6 @@ static void qlcnic_83xx_idc_attach_driver(struct qlcnic_adapter *adapter) } done: netif_device_attach(netdev); - if (netif_running(netdev)) { - netif_carrier_on(netdev); - netif_wake_queue(netdev); - } } static int qlcnic_83xx_idc_enter_failed_state(struct qlcnic_adapter *adapter, @@ -642,15 +638,21 @@ static int qlcnic_83xx_idc_reattach_driver(struct qlcnic_adapter *adapter) static void qlcnic_83xx_idc_update_idc_params(struct qlcnic_adapter *adapter) { + struct qlcnic_hardware_context *ahw = adapter->ahw; + qlcnic_83xx_idc_update_drv_presence_reg(adapter, 1, 1); - clear_bit(__QLCNIC_RESETTING, &adapter->state); set_bit(QLC_83XX_MBX_READY, &adapter->ahw->idc.status); qlcnic_83xx_idc_update_audit_reg(adapter, 0, 1); set_bit(QLC_83XX_MODULE_LOADED, &adapter->ahw->idc.status); - adapter->ahw->idc.quiesce_req = 0; - adapter->ahw->idc.delay = QLC_83XX_IDC_FW_POLL_DELAY; - adapter->ahw->idc.err_code = 0; - adapter->ahw->idc.collect_dump = 0; + + ahw->idc.quiesce_req = 0; + ahw->idc.delay = QLC_83XX_IDC_FW_POLL_DELAY; + ahw->idc.err_code = 0; + ahw->idc.collect_dump = 0; + ahw->reset_context = 0; + adapter->tx_timeo_cnt = 0; + + clear_bit(__QLCNIC_RESETTING, &adapter->state); } /** @@ -851,6 +853,7 @@ static int qlcnic_83xx_idc_ready_state(struct qlcnic_adapter *adapter) /* Check for soft reset request */ if (ahw->reset_context && !(val & QLC_83XX_IDC_DISABLE_FW_RESET_RECOVERY)) { + adapter->ahw->reset_context = 0; qlcnic_83xx_idc_tx_soft_reset(adapter); return ret; } @@ -914,6 +917,7 @@ static int qlcnic_83xx_idc_need_quiesce_state(struct qlcnic_adapter *adapter) static int qlcnic_83xx_idc_failed_state(struct qlcnic_adapter *adapter) { dev_err(&adapter->pdev->dev, "%s: please restart!!\n", __func__); + clear_bit(__QLCNIC_RESETTING, &adapter->state); adapter->ahw->idc.err_code = -EIO; return 0; diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c index d97446a3bf05..8fb836d4129f 100644 --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c @@ -2502,12 +2502,17 @@ static void qlcnic_tx_timeout(struct net_device *netdev) if (test_bit(__QLCNIC_RESETTING, &adapter->state)) return; - dev_err(&netdev->dev, "transmit timeout, resetting.\n"); - - if (++adapter->tx_timeo_cnt >= QLCNIC_MAX_TX_TIMEOUTS) - adapter->need_fw_reset = 1; - else + if (++adapter->tx_timeo_cnt >= QLCNIC_MAX_TX_TIMEOUTS) { + netdev_info(netdev, "Tx timeout, reset the adapter.\n"); + if (qlcnic_82xx_check(adapter)) + adapter->need_fw_reset = 1; + else if (qlcnic_83xx_check(adapter)) + qlcnic_83xx_idc_request_reset(adapter, + QLCNIC_FORCE_FW_DUMP_KEY); + } else { + netdev_info(netdev, "Tx timeout, reset adapter context.\n"); adapter->ahw->reset_context = 1; + } } static struct net_device_stats *qlcnic_get_stats(struct net_device *netdev) From 13a82b44fd761fdb6aeceb2ce9930310569efa3b Mon Sep 17 00:00:00 2001 From: Manish Chopra Date: Thu, 9 May 2013 09:25:14 +0000 Subject: [PATCH 082/198] qlcnic: Fix bug in diagnostics test reset recovery path o In order to perform reset recovery during diagnostics tests, current device status information need to be preserved. This patch makes the required changes in diagnostics routines Signed-off-by: Manish Chopra Signed-off-by: Shahed Shaikh Signed-off-by: David S. Miller --- .../net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.c | 14 +++++++++++--- 1 file changed, 11 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.c b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.c index 642fca0b15d2..7b5cb610e7b6 100644 --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.c +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.c @@ -1276,11 +1276,13 @@ out: return err; } -static int qlcnic_83xx_diag_alloc_res(struct net_device *netdev, int test) +static int qlcnic_83xx_diag_alloc_res(struct net_device *netdev, int test, + int num_sds_ring) { struct qlcnic_adapter *adapter = netdev_priv(netdev); struct qlcnic_host_sds_ring *sds_ring; struct qlcnic_host_rds_ring *rds_ring; + u16 adapter_state = adapter->is_up; u8 ring; int ret; @@ -1304,6 +1306,10 @@ static int qlcnic_83xx_diag_alloc_res(struct net_device *netdev, int test) ret = qlcnic_fw_create_ctx(adapter); if (ret) { qlcnic_detach(adapter); + if (adapter_state == QLCNIC_ADAPTER_UP_MAGIC) { + adapter->max_sds_rings = num_sds_ring; + qlcnic_attach(adapter); + } netif_device_attach(netdev); return ret; } @@ -1596,7 +1602,8 @@ int qlcnic_83xx_loopback_test(struct net_device *netdev, u8 mode) if (test_and_set_bit(__QLCNIC_RESETTING, &adapter->state)) return -EBUSY; - ret = qlcnic_83xx_diag_alloc_res(netdev, QLCNIC_LOOPBACK_TEST); + ret = qlcnic_83xx_diag_alloc_res(netdev, QLCNIC_LOOPBACK_TEST, + max_sds_rings); if (ret) goto fail_diag_alloc; @@ -3112,7 +3119,8 @@ int qlcnic_83xx_interrupt_test(struct net_device *netdev) if (test_and_set_bit(__QLCNIC_RESETTING, &adapter->state)) return -EIO; - ret = qlcnic_83xx_diag_alloc_res(netdev, QLCNIC_INTERRUPT_TEST); + ret = qlcnic_83xx_diag_alloc_res(netdev, QLCNIC_INTERRUPT_TEST, + max_sds_rings); if (ret) goto fail_diag_irq; From 9106e5db8336d5048ac79f72056572c461a2566b Mon Sep 17 00:00:00 2001 From: Rajesh Borundia Date: Thu, 9 May 2013 09:25:15 +0000 Subject: [PATCH 083/198] qlcnic: Fix mailbox response handling. o Fix mailbox response poll time to maximum 5 seconds which includes mailbox response time as well as time required for processing AEN if any. o Driver need to read firmware control mailbox register instead of host control mailbox register. Signed-off-by: Jitendra Kalsaria Signed-off-by: Rajesh Borundia Signed-off-by: Shahed Shaikh Signed-off-by: David S. Miller --- drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.c | 13 +++++-------- drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.h | 2 +- drivers/net/ethernet/qlogic/qlcnic/qlcnic_hw.h | 2 +- .../ethernet/qlogic/qlcnic/qlcnic_sriov_common.c | 8 +++----- 4 files changed, 10 insertions(+), 15 deletions(-) diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.c b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.c index 7b5cb610e7b6..b4ff1e35a11d 100644 --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.c +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.c @@ -696,15 +696,14 @@ u32 qlcnic_83xx_mac_rcode(struct qlcnic_adapter *adapter) return 1; } -u32 qlcnic_83xx_mbx_poll(struct qlcnic_adapter *adapter) +u32 qlcnic_83xx_mbx_poll(struct qlcnic_adapter *adapter, u32 *wait_time) { u32 data; - unsigned long wait_time = 0; struct qlcnic_hardware_context *ahw = adapter->ahw; /* wait for mailbox completion */ do { data = QLCRDX(ahw, QLCNIC_FW_MBX_CTRL); - if (++wait_time > QLCNIC_MBX_TIMEOUT) { + if (++(*wait_time) > QLCNIC_MBX_TIMEOUT) { data = QLCNIC_RCODE_TIMEOUT; break; } @@ -720,8 +719,8 @@ int qlcnic_83xx_mbx_op(struct qlcnic_adapter *adapter, u16 opcode; u8 mbx_err_code; unsigned long flags; - u32 rsp, mbx_val, fw_data, rsp_num, mbx_cmd; struct qlcnic_hardware_context *ahw = adapter->ahw; + u32 rsp, mbx_val, fw_data, rsp_num, mbx_cmd, wait_time = 0; opcode = LSW(cmd->req.arg[0]); if (!test_bit(QLC_83XX_MBX_READY, &adapter->ahw->idc.status)) { @@ -754,15 +753,13 @@ int qlcnic_83xx_mbx_op(struct qlcnic_adapter *adapter, /* Signal FW about the impending command */ QLCWRX(ahw, QLCNIC_HOST_MBX_CTRL, QLCNIC_SET_OWNER); poll: - rsp = qlcnic_83xx_mbx_poll(adapter); + rsp = qlcnic_83xx_mbx_poll(adapter, &wait_time); if (rsp != QLCNIC_RCODE_TIMEOUT) { /* Get the FW response data */ fw_data = readl(QLCNIC_MBX_FW(ahw, 0)); if (fw_data & QLCNIC_MBX_ASYNC_EVENT) { __qlcnic_83xx_process_aen(adapter); - mbx_val = QLCRDX(ahw, QLCNIC_HOST_MBX_CTRL); - if (mbx_val) - goto poll; + goto poll; } mbx_err_code = QLCNIC_MBX_STATUS(fw_data); rsp_num = QLCNIC_MBX_NUM_REGS(fw_data); diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.h b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.h index faf5553681e1..f5db67fc9f55 100644 --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.h +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_83xx_hw.h @@ -620,7 +620,7 @@ int qlcnic_83xx_flash_test(struct qlcnic_adapter *); int qlcnic_83xx_enable_flash_write(struct qlcnic_adapter *); int qlcnic_83xx_disable_flash_write(struct qlcnic_adapter *); u32 qlcnic_83xx_mac_rcode(struct qlcnic_adapter *); -u32 qlcnic_83xx_mbx_poll(struct qlcnic_adapter *); +u32 qlcnic_83xx_mbx_poll(struct qlcnic_adapter *, u32 *); void qlcnic_83xx_enable_mbx_poll(struct qlcnic_adapter *); void qlcnic_83xx_disable_mbx_poll(struct qlcnic_adapter *); #endif diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_hw.h b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_hw.h index 95b1b5732838..b6818f4356b9 100644 --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_hw.h +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_hw.h @@ -134,7 +134,7 @@ struct qlcnic_mailbox_metadata { #define QLCNIC_SET_OWNER 1 #define QLCNIC_CLR_OWNER 0 -#define QLCNIC_MBX_TIMEOUT 10000 +#define QLCNIC_MBX_TIMEOUT 5000 #define QLCNIC_MBX_RSP_OK 1 #define QLCNIC_MBX_PORT_RSP_OK 0x1a diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_sriov_common.c b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_sriov_common.c index 44d547d78b84..3869c3864deb 100644 --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_sriov_common.c +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_sriov_common.c @@ -280,9 +280,9 @@ void qlcnic_sriov_cleanup(struct qlcnic_adapter *adapter) static int qlcnic_sriov_post_bc_msg(struct qlcnic_adapter *adapter, u32 *hdr, u32 *pay, u8 pci_func, u8 size) { + u32 rsp, mbx_val, fw_data, rsp_num, mbx_cmd, val, wait_time = 0; struct qlcnic_hardware_context *ahw = adapter->ahw; unsigned long flags; - u32 rsp, mbx_val, fw_data, rsp_num, mbx_cmd, val; u16 opcode; u8 mbx_err_code; int i, j; @@ -330,15 +330,13 @@ static int qlcnic_sriov_post_bc_msg(struct qlcnic_adapter *adapter, u32 *hdr, * assume something is wrong. */ poll: - rsp = qlcnic_83xx_mbx_poll(adapter); + rsp = qlcnic_83xx_mbx_poll(adapter, &wait_time); if (rsp != QLCNIC_RCODE_TIMEOUT) { /* Get the FW response data */ fw_data = readl(QLCNIC_MBX_FW(ahw, 0)); if (fw_data & QLCNIC_MBX_ASYNC_EVENT) { __qlcnic_83xx_process_aen(adapter); - mbx_val = QLCRDX(ahw, QLCNIC_HOST_MBX_CTRL); - if (mbx_val) - goto poll; + goto poll; } mbx_err_code = QLCNIC_MBX_STATUS(fw_data); rsp_num = QLCNIC_MBX_NUM_REGS(fw_data); From cbccb18f4f937132af64a95ae04186a6c2c05236 Mon Sep 17 00:00:00 2001 From: Rajesh Borundia Date: Thu, 9 May 2013 09:25:16 +0000 Subject: [PATCH 084/198] qlcnic: Fix validation of link event command. o VF driver that has enabled asynchronous link events may not set BIT_8 in the request, if it does not require link state in the response. Signed-off-by: Pratik Pujar Signed-off-by: Rajesh Borundia Signed-off-by: Shahed Shaikh Signed-off-by: David S. Miller --- drivers/net/ethernet/qlogic/qlcnic/qlcnic_sriov_pf.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_sriov_pf.c b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_sriov_pf.c index c81be2da119b..1a66ccded235 100644 --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_sriov_pf.c +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_sriov_pf.c @@ -1133,9 +1133,6 @@ static int qlcnic_sriov_validate_linkevent(struct qlcnic_vf_info *vf, if ((cmd->req.arg[1] >> 16) != vf->rx_ctx_id) return -EINVAL; - if (!(cmd->req.arg[1] & BIT_8)) - return -EINVAL; - return 0; } From a549984b8c95acbecefd1fdd4bfdbea4d29b0588 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 11 May 2013 19:07:42 -0400 Subject: [PATCH 085/198] ext4: revert "ext4: use io_end for multiple bios" This reverts commit 4eec708d263f0ee10861d69251708a225b64cac7. Multiple users have reported crashes which is apparently caused by this commit. Thanks to Dmitry Monakhov for bisecting it. Signed-off-by: "Theodore Ts'o" Cc: Dmitry Monakhov Cc: Jan Kara --- fs/ext4/ext4.h | 8 +-- fs/ext4/inode.c | 87 +++++++++++++++------------------ fs/ext4/page-io.c | 121 +++++++++++++++++----------------------------- 3 files changed, 86 insertions(+), 130 deletions(-) diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 0aabb344b02e..5aae3d12d400 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -209,7 +209,6 @@ typedef struct ext4_io_end { ssize_t size; /* size of the extent */ struct kiocb *iocb; /* iocb struct for AIO */ int result; /* error value for AIO */ - atomic_t count; /* reference counter */ } ext4_io_end_t; struct ext4_io_submit { @@ -2651,14 +2650,11 @@ extern int ext4_move_extents(struct file *o_filp, struct file *d_filp, /* page-io.c */ extern int __init ext4_init_pageio(void); +extern void ext4_add_complete_io(ext4_io_end_t *io_end); extern void ext4_exit_pageio(void); extern void ext4_ioend_shutdown(struct inode *); +extern void ext4_free_io_end(ext4_io_end_t *io); extern ext4_io_end_t *ext4_init_io_end(struct inode *inode, gfp_t flags); -extern ext4_io_end_t *ext4_get_io_end(ext4_io_end_t *io_end); -extern int ext4_put_io_end(ext4_io_end_t *io_end); -extern void ext4_put_io_end_defer(ext4_io_end_t *io_end); -extern void ext4_io_submit_init(struct ext4_io_submit *io, - struct writeback_control *wbc); extern void ext4_end_io_work(struct work_struct *work); extern void ext4_io_submit(struct ext4_io_submit *io); extern int ext4_bio_write_page(struct ext4_io_submit *io, diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 793d44b84d7f..d66656992358 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -1487,10 +1487,7 @@ static int mpage_da_submit_io(struct mpage_da_data *mpd, struct ext4_io_submit io_submit; BUG_ON(mpd->next_page <= mpd->first_page); - ext4_io_submit_init(&io_submit, mpd->wbc); - io_submit.io_end = ext4_init_io_end(inode, GFP_NOFS); - if (!io_submit.io_end) - return -ENOMEM; + memset(&io_submit, 0, sizeof(io_submit)); /* * We need to start from the first_page to the next_page - 1 * to make sure we also write the mapped dirty buffer_heads. @@ -1578,8 +1575,6 @@ static int mpage_da_submit_io(struct mpage_da_data *mpd, pagevec_release(&pvec); } ext4_io_submit(&io_submit); - /* Drop io_end reference we got from init */ - ext4_put_io_end_defer(io_submit.io_end); return ret; } @@ -2238,16 +2233,9 @@ static int ext4_writepage(struct page *page, */ return __ext4_journalled_writepage(page, len); - ext4_io_submit_init(&io_submit, wbc); - io_submit.io_end = ext4_init_io_end(inode, GFP_NOFS); - if (!io_submit.io_end) { - redirty_page_for_writepage(wbc, page); - return -ENOMEM; - } + memset(&io_submit, 0, sizeof(io_submit)); ret = ext4_bio_write_page(&io_submit, page, len, wbc); ext4_io_submit(&io_submit); - /* Drop io_end reference we got from init */ - ext4_put_io_end_defer(io_submit.io_end); return ret; } @@ -3078,13 +3066,9 @@ static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset, struct inode *inode = file_inode(iocb->ki_filp); ext4_io_end_t *io_end = iocb->private; - /* if not async direct IO just return */ - if (!io_end) { - inode_dio_done(inode); - if (is_async) - aio_complete(iocb, ret, 0); - return; - } + /* if not async direct IO or dio with 0 bytes write, just return */ + if (!io_end || !size) + goto out; ext_debug("ext4_end_io_dio(): io_end 0x%p " "for inode %lu, iocb 0x%p, offset %llu, size %zd\n", @@ -3092,13 +3076,25 @@ static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset, size); iocb->private = NULL; + + /* if not aio dio with unwritten extents, just free io and return */ + if (!(io_end->flag & EXT4_IO_END_UNWRITTEN)) { + ext4_free_io_end(io_end); +out: + inode_dio_done(inode); + if (is_async) + aio_complete(iocb, ret, 0); + return; + } + io_end->offset = offset; io_end->size = size; if (is_async) { io_end->iocb = iocb; io_end->result = ret; } - ext4_put_io_end_defer(io_end); + + ext4_add_complete_io(io_end); } /* @@ -3132,7 +3128,6 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb, get_block_t *get_block_func = NULL; int dio_flags = 0; loff_t final_size = offset + count; - ext4_io_end_t *io_end = NULL; /* Use the old path for reads and writes beyond i_size. */ if (rw != WRITE || final_size > inode->i_size) @@ -3171,16 +3166,13 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb, iocb->private = NULL; ext4_inode_aio_set(inode, NULL); if (!is_sync_kiocb(iocb)) { - io_end = ext4_init_io_end(inode, GFP_NOFS); + ext4_io_end_t *io_end = ext4_init_io_end(inode, GFP_NOFS); if (!io_end) { ret = -ENOMEM; goto retake_lock; } io_end->flag |= EXT4_IO_END_DIRECT; - /* - * Grab reference for DIO. Will be dropped in ext4_end_io_dio() - */ - iocb->private = ext4_get_io_end(io_end); + iocb->private = io_end; /* * we save the io structure for current async direct * IO, so that later ext4_map_blocks() could flag the @@ -3204,27 +3196,26 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb, NULL, dio_flags); - /* - * Put our reference to io_end. This can free the io_end structure e.g. - * in sync IO case or in case of error. It can even perform extent - * conversion if all bios we submitted finished before we got here. - * Note that in that case iocb->private can be already set to NULL - * here. - */ - if (io_end) { + if (iocb->private) ext4_inode_aio_set(inode, NULL); - ext4_put_io_end(io_end); - /* - * In case of error or no write ext4_end_io_dio() was not - * called so we have to put iocb's reference. - */ - if (ret <= 0 && ret != -EIOCBQUEUED) { - WARN_ON(iocb->private != io_end); - ext4_put_io_end(io_end); - iocb->private = NULL; - } - } - if (ret > 0 && !overwrite && ext4_test_inode_state(inode, + /* + * The io_end structure takes a reference to the inode, that + * structure needs to be destroyed and the reference to the + * inode need to be dropped, when IO is complete, even with 0 + * byte write, or failed. + * + * In the successful AIO DIO case, the io_end structure will + * be destroyed and the reference to the inode will be dropped + * after the end_io call back function is called. + * + * In the case there is 0 byte write, or error case, since VFS + * direct IO won't invoke the end_io call back function, we + * need to free the end_io structure here. + */ + if (ret != -EIOCBQUEUED && ret <= 0 && iocb->private) { + ext4_free_io_end(iocb->private); + iocb->private = NULL; + } else if (ret > 0 && !overwrite && ext4_test_inode_state(inode, EXT4_STATE_DIO_UNWRITTEN)) { int err; /* diff --git a/fs/ext4/page-io.c b/fs/ext4/page-io.c index 5929cd0baa20..6626aba57ebb 100644 --- a/fs/ext4/page-io.c +++ b/fs/ext4/page-io.c @@ -61,28 +61,15 @@ void ext4_ioend_shutdown(struct inode *inode) cancel_work_sync(&EXT4_I(inode)->i_unwritten_work); } -static void ext4_release_io_end(ext4_io_end_t *io_end) +void ext4_free_io_end(ext4_io_end_t *io) { - BUG_ON(!list_empty(&io_end->list)); - BUG_ON(io_end->flag & EXT4_IO_END_UNWRITTEN); + BUG_ON(!io); + BUG_ON(!list_empty(&io->list)); + BUG_ON(io->flag & EXT4_IO_END_UNWRITTEN); - if (atomic_dec_and_test(&EXT4_I(io_end->inode)->i_ioend_count)) - wake_up_all(ext4_ioend_wq(io_end->inode)); - if (io_end->flag & EXT4_IO_END_DIRECT) - inode_dio_done(io_end->inode); - if (io_end->iocb) - aio_complete(io_end->iocb, io_end->result, 0); - kmem_cache_free(io_end_cachep, io_end); -} - -static void ext4_clear_io_unwritten_flag(ext4_io_end_t *io_end) -{ - struct inode *inode = io_end->inode; - - io_end->flag &= ~EXT4_IO_END_UNWRITTEN; - /* Wake up anyone waiting on unwritten extent conversion */ - if (atomic_dec_and_test(&EXT4_I(inode)->i_unwritten)) - wake_up_all(ext4_ioend_wq(inode)); + if (atomic_dec_and_test(&EXT4_I(io->inode)->i_ioend_count)) + wake_up_all(ext4_ioend_wq(io->inode)); + kmem_cache_free(io_end_cachep, io); } /* check a range of space and convert unwritten extents to written. */ @@ -105,8 +92,13 @@ static int ext4_end_io(ext4_io_end_t *io) "(inode %lu, offset %llu, size %zd, error %d)", inode->i_ino, offset, size, ret); } - ext4_clear_io_unwritten_flag(io); - ext4_release_io_end(io); + /* Wake up anyone waiting on unwritten extent conversion */ + if (atomic_dec_and_test(&EXT4_I(inode)->i_unwritten)) + wake_up_all(ext4_ioend_wq(inode)); + if (io->flag & EXT4_IO_END_DIRECT) + inode_dio_done(inode); + if (io->iocb) + aio_complete(io->iocb, io->result, 0); return ret; } @@ -137,7 +129,7 @@ static void dump_completed_IO(struct inode *inode) } /* Add the io_end to per-inode completed end_io list. */ -static void ext4_add_complete_io(ext4_io_end_t *io_end) +void ext4_add_complete_io(ext4_io_end_t *io_end) { struct ext4_inode_info *ei = EXT4_I(io_end->inode); struct workqueue_struct *wq; @@ -174,6 +166,8 @@ static int ext4_do_flush_completed_IO(struct inode *inode) err = ext4_end_io(io); if (unlikely(!ret && err)) ret = err; + io->flag &= ~EXT4_IO_END_UNWRITTEN; + ext4_free_io_end(io); } return ret; } @@ -205,43 +199,10 @@ ext4_io_end_t *ext4_init_io_end(struct inode *inode, gfp_t flags) atomic_inc(&EXT4_I(inode)->i_ioend_count); io->inode = inode; INIT_LIST_HEAD(&io->list); - atomic_set(&io->count, 1); } return io; } -void ext4_put_io_end_defer(ext4_io_end_t *io_end) -{ - if (atomic_dec_and_test(&io_end->count)) { - if (!(io_end->flag & EXT4_IO_END_UNWRITTEN) || !io_end->size) { - ext4_release_io_end(io_end); - return; - } - ext4_add_complete_io(io_end); - } -} - -int ext4_put_io_end(ext4_io_end_t *io_end) -{ - int err = 0; - - if (atomic_dec_and_test(&io_end->count)) { - if (io_end->flag & EXT4_IO_END_UNWRITTEN) { - err = ext4_convert_unwritten_extents(io_end->inode, - io_end->offset, io_end->size); - ext4_clear_io_unwritten_flag(io_end); - } - ext4_release_io_end(io_end); - } - return err; -} - -ext4_io_end_t *ext4_get_io_end(ext4_io_end_t *io_end) -{ - atomic_inc(&io_end->count); - return io_end; -} - /* * Print an buffer I/O error compatible with the fs/buffer.c. This * provides compatibility with dmesg scrapers that look for a specific @@ -324,7 +285,12 @@ static void ext4_end_bio(struct bio *bio, int error) bi_sector >> (inode->i_blkbits - 9)); } - ext4_put_io_end_defer(io_end); + if (!(io_end->flag & EXT4_IO_END_UNWRITTEN)) { + ext4_free_io_end(io_end); + return; + } + + ext4_add_complete_io(io_end); } void ext4_io_submit(struct ext4_io_submit *io) @@ -338,37 +304,40 @@ void ext4_io_submit(struct ext4_io_submit *io) bio_put(io->io_bio); } io->io_bio = NULL; -} - -void ext4_io_submit_init(struct ext4_io_submit *io, - struct writeback_control *wbc) -{ - io->io_op = (wbc->sync_mode == WB_SYNC_ALL ? WRITE_SYNC : WRITE); - io->io_bio = NULL; + io->io_op = 0; io->io_end = NULL; } -static int io_submit_init_bio(struct ext4_io_submit *io, - struct buffer_head *bh) +static int io_submit_init(struct ext4_io_submit *io, + struct inode *inode, + struct writeback_control *wbc, + struct buffer_head *bh) { + ext4_io_end_t *io_end; + struct page *page = bh->b_page; int nvecs = bio_get_nr_vecs(bh->b_bdev); struct bio *bio; + io_end = ext4_init_io_end(inode, GFP_NOFS); + if (!io_end) + return -ENOMEM; bio = bio_alloc(GFP_NOIO, min(nvecs, BIO_MAX_PAGES)); bio->bi_sector = bh->b_blocknr * (bh->b_size >> 9); bio->bi_bdev = bh->b_bdev; + bio->bi_private = io->io_end = io_end; bio->bi_end_io = ext4_end_bio; - bio->bi_private = ext4_get_io_end(io->io_end); - if (!io->io_end->size) - io->io_end->offset = (bh->b_page->index << PAGE_CACHE_SHIFT) - + bh_offset(bh); + + io_end->offset = (page->index << PAGE_CACHE_SHIFT) + bh_offset(bh); + io->io_bio = bio; + io->io_op = (wbc->sync_mode == WB_SYNC_ALL ? WRITE_SYNC : WRITE); io->io_next_block = bh->b_blocknr; return 0; } static int io_submit_add_bh(struct ext4_io_submit *io, struct inode *inode, + struct writeback_control *wbc, struct buffer_head *bh) { ext4_io_end_t *io_end; @@ -379,18 +348,18 @@ submit_and_retry: ext4_io_submit(io); } if (io->io_bio == NULL) { - ret = io_submit_init_bio(io, bh); + ret = io_submit_init(io, inode, wbc, bh); if (ret) return ret; } - ret = bio_add_page(io->io_bio, bh->b_page, bh->b_size, bh_offset(bh)); - if (ret != bh->b_size) - goto submit_and_retry; io_end = io->io_end; if (test_clear_buffer_uninit(bh)) ext4_set_io_unwritten_flag(inode, io_end); - io_end->size += bh->b_size; + io->io_end->size += bh->b_size; io->io_next_block++; + ret = bio_add_page(io->io_bio, bh->b_page, bh->b_size, bh_offset(bh)); + if (ret != bh->b_size) + goto submit_and_retry; return 0; } @@ -462,7 +431,7 @@ int ext4_bio_write_page(struct ext4_io_submit *io, do { if (!buffer_async_write(bh)) continue; - ret = io_submit_add_bh(io, inode, bh); + ret = io_submit_add_bh(io, inode, wbc, bh); if (ret) { /* * We only get here on ENOMEM. Not much else From 4849625569874d474ce351efd84b1f487ba08c94 Mon Sep 17 00:00:00 2001 From: Shawn Guo Date: Wed, 8 May 2013 21:08:22 +0000 Subject: [PATCH 086/198] net: fec: enable hardware checksum only on imx6q-fec Commit 4c09eed (net: fec: Enable imx6 enet checksum acceleration.) enables hardware checksum acceleration unconditionally for all fec variants. This is inappropriate, because some variants like imx5 have no such support on hardware. Consequently, fec is broken on these platforms. Fix it by enabling hardware checksum only on imx6q-fec type of controllers. Signed-off-by: Shawn Guo Reviewed-by: Jim Baxter Signed-off-by: David S. Miller --- drivers/net/ethernet/freescale/fec_main.c | 20 +++++++++++++------- 1 file changed, 13 insertions(+), 7 deletions(-) diff --git a/drivers/net/ethernet/freescale/fec_main.c b/drivers/net/ethernet/freescale/fec_main.c index aff0310a778b..ca9825ca88c9 100644 --- a/drivers/net/ethernet/freescale/fec_main.c +++ b/drivers/net/ethernet/freescale/fec_main.c @@ -87,6 +87,8 @@ #define FEC_QUIRK_HAS_GBIT (1 << 3) /* Controller has extend desc buffer */ #define FEC_QUIRK_HAS_BUFDESC_EX (1 << 4) +/* Controller has hardware checksum support */ +#define FEC_QUIRK_HAS_CSUM (1 << 5) static struct platform_device_id fec_devtype[] = { { @@ -105,7 +107,7 @@ static struct platform_device_id fec_devtype[] = { }, { .name = "imx6q-fec", .driver_data = FEC_QUIRK_ENET_MAC | FEC_QUIRK_HAS_GBIT | - FEC_QUIRK_HAS_BUFDESC_EX, + FEC_QUIRK_HAS_BUFDESC_EX | FEC_QUIRK_HAS_CSUM, }, { .name = "mvf-fec", .driver_data = FEC_QUIRK_ENET_MAC, @@ -1744,6 +1746,8 @@ static const struct net_device_ops fec_netdev_ops = { static int fec_enet_init(struct net_device *ndev) { struct fec_enet_private *fep = netdev_priv(ndev); + const struct platform_device_id *id_entry = + platform_get_device_id(fep->pdev); struct bufdesc *cbd_base; /* Allocate memory for buffer descriptors. */ @@ -1775,12 +1779,14 @@ static int fec_enet_init(struct net_device *ndev) writel(FEC_RX_DISABLED_IMASK, fep->hwp + FEC_IMASK); netif_napi_add(ndev, &fep->napi, fec_enet_rx_napi, FEC_NAPI_WEIGHT); - /* enable hw accelerator */ - ndev->features |= (NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM - | NETIF_F_RXCSUM); - ndev->hw_features |= (NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM - | NETIF_F_RXCSUM); - fep->csum_flags |= FLAG_RX_CSUM_ENABLED; + if (id_entry->driver_data & FEC_QUIRK_HAS_CSUM) { + /* enable hw accelerator */ + ndev->features |= (NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM + | NETIF_F_RXCSUM); + ndev->hw_features |= (NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM + | NETIF_F_RXCSUM); + fep->csum_flags |= FLAG_RX_CSUM_ENABLED; + } fec_restart(ndev, 0); From 4e8cf5b8a1da8dbe1ece5f084b99f5c902ea709b Mon Sep 17 00:00:00 2001 From: Or Gerlitz Date: Wed, 8 May 2013 22:22:34 +0000 Subject: [PATCH 087/198] net/mlx4_core: Add missing report on VST and spoof-checking dev caps Commits e6b6a23 "net/mlx4: Add VF MAC spoof checking support" and 3f7fb021 "net/mlx4: Add set VF default vlan ID and priority support" missed reporting in the device capabilities dump when these features are actually supported. Also two too noisy debug messages which produce message on every QP opened by a VF, were left in the code, fix that. Signed-off-by: Rony Efraim Signed-off-by: Or Gerlitz Signed-off-by: David S. Miller --- drivers/net/ethernet/mellanox/mlx4/fw.c | 4 +++- drivers/net/ethernet/mellanox/mlx4/resource_tracker.c | 9 --------- 2 files changed, 3 insertions(+), 10 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx4/fw.c b/drivers/net/ethernet/mellanox/mlx4/fw.c index b147bdd40768..58a8e535d698 100644 --- a/drivers/net/ethernet/mellanox/mlx4/fw.c +++ b/drivers/net/ethernet/mellanox/mlx4/fw.c @@ -131,7 +131,9 @@ static void dump_dev_cap_flags2(struct mlx4_dev *dev, u64 flags) [2] = "RSS XOR Hash Function support", [3] = "Device manage flow steering support", [4] = "Automatic MAC reassignment support", - [5] = "Time stamping support" + [5] = "Time stamping support", + [6] = "VST (control vlan insertion/stripping) support", + [7] = "FSM (MAC anti-spoofing) support" }; int i; diff --git a/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c b/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c index e12e0d2e0ee0..d4a9de666fbd 100644 --- a/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c +++ b/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c @@ -377,19 +377,10 @@ static int update_vport_qp_param(struct mlx4_dev *dev, qpc->pri_path.feup |= 1 << 3; /* set fvl bit */ qpc->pri_path.sched_queue &= 0xC7; qpc->pri_path.sched_queue |= (vp_oper->state.default_qos) << 3; - mlx4_dbg(dev, "qp %d port %d Q 0x%x set vlan to %d vidx %d feup %x fl %x\n", - be32_to_cpu(qpc->local_qpn) & 0xffffff, port, - (int)(qpc->pri_path.sched_queue), vp_oper->state.default_vlan, - vp_oper->vlan_idx, (int)(qpc->pri_path.feup), - (int)(qpc->pri_path.fl)); } if (vp_oper->state.spoofchk) { qpc->pri_path.feup |= 1 << 5; /* set fsm bit */; qpc->pri_path.grh_mylmc = (0x80 & qpc->pri_path.grh_mylmc) + vp_oper->mac_idx; - mlx4_dbg(dev, "spoof qp %d port %d feup 0x%x, myLmc 0x%x mindx %d\n", - be32_to_cpu(qpc->local_qpn) & 0xffffff, port, - (int)qpc->pri_path.feup, (int)qpc->pri_path.grh_mylmc, - vp_oper->mac_idx); } return 0; } From 7677fc965fba41d1386fa3b76a1f00303f02bb2d Mon Sep 17 00:00:00 2001 From: Rony Efraim Date: Wed, 8 May 2013 22:22:35 +0000 Subject: [PATCH 088/198] net/mlx4: Strengthen VLAN tags/priorities enforcement in VST mode Make sure that the following steps are taken: - drop packets sent by the VF with vlan tag - block packets with vlan tag which are steered to the VF - drop/block tagged packets when the policy is priority-tagged - make sure VLAN stripping for received packets is set - make sure force UP bit for the VF QP is set Use enum values for all the above instead of numerical bit offsets. Signed-off-by: Rony Efraim Signed-off-by: Or Gerlitz Signed-off-by: David S. Miller --- .../net/ethernet/mellanox/mlx4/en_resources.c | 2 +- .../ethernet/mellanox/mlx4/resource_tracker.c | 20 +++++++++++-- include/linux/mlx4/qp.h | 29 +++++++++++++++++-- 3 files changed, 45 insertions(+), 6 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx4/en_resources.c b/drivers/net/ethernet/mellanox/mlx4/en_resources.c index 91f2b2c43c12..d3f508697a3d 100644 --- a/drivers/net/ethernet/mellanox/mlx4/en_resources.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_resources.c @@ -60,7 +60,7 @@ void mlx4_en_fill_qp_context(struct mlx4_en_priv *priv, int size, int stride, context->pri_path.sched_queue = 0x83 | (priv->port - 1) << 6; if (user_prio >= 0) { context->pri_path.sched_queue |= user_prio << 3; - context->pri_path.feup = 1 << 6; + context->pri_path.feup = MLX4_FEUP_FORCE_ETH_UP; } context->pri_path.counter_index = 0xff; context->cqn_send = cpu_to_be32(cqn); diff --git a/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c b/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c index d4a9de666fbd..1157f028a90f 100644 --- a/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c +++ b/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c @@ -372,14 +372,28 @@ static int update_vport_qp_param(struct mlx4_dev *dev, if (MLX4_QP_ST_RC == qp_type) return -EINVAL; + /* force strip vlan by clear vsd */ + qpc->param3 &= ~cpu_to_be32(MLX4_STRIP_VLAN); + if (0 != vp_oper->state.default_vlan) { + qpc->pri_path.vlan_control = + MLX4_VLAN_CTRL_ETH_TX_BLOCK_TAGGED | + MLX4_VLAN_CTRL_ETH_RX_BLOCK_PRIO_TAGGED | + MLX4_VLAN_CTRL_ETH_RX_BLOCK_UNTAGGED; + } else { /* priority tagged */ + qpc->pri_path.vlan_control = + MLX4_VLAN_CTRL_ETH_TX_BLOCK_TAGGED | + MLX4_VLAN_CTRL_ETH_RX_BLOCK_TAGGED; + } + + qpc->pri_path.fvl_rx |= MLX4_FVL_RX_FORCE_ETH_VLAN; qpc->pri_path.vlan_index = vp_oper->vlan_idx; - qpc->pri_path.fl = (1 << 6) | (1 << 2); /* set cv bit and hide_cqe_vlan bit*/ - qpc->pri_path.feup |= 1 << 3; /* set fvl bit */ + qpc->pri_path.fl |= MLX4_FL_CV | MLX4_FL_ETH_HIDE_CQE_VLAN; + qpc->pri_path.feup |= MLX4_FEUP_FORCE_ETH_UP | MLX4_FVL_FORCE_ETH_VLAN; qpc->pri_path.sched_queue &= 0xC7; qpc->pri_path.sched_queue |= (vp_oper->state.default_qos) << 3; } if (vp_oper->state.spoofchk) { - qpc->pri_path.feup |= 1 << 5; /* set fsm bit */; + qpc->pri_path.feup |= MLX4_FSM_FORCE_ETH_SRC_MAC; qpc->pri_path.grh_mylmc = (0x80 & qpc->pri_path.grh_mylmc) + vp_oper->mac_idx; } return 0; diff --git a/include/linux/mlx4/qp.h b/include/linux/mlx4/qp.h index 67f46ad6920a..352eec9df1b8 100644 --- a/include/linux/mlx4/qp.h +++ b/include/linux/mlx4/qp.h @@ -126,7 +126,7 @@ struct mlx4_rss_context { struct mlx4_qp_path { u8 fl; - u8 reserved1[1]; + u8 vlan_control; u8 disable_pkey_check; u8 pkey_index; u8 counter_index; @@ -141,11 +141,32 @@ struct mlx4_qp_path { u8 sched_queue; u8 vlan_index; u8 feup; - u8 reserved3; + u8 fvl_rx; u8 reserved4[2]; u8 dmac[6]; }; +enum { /* fl */ + MLX4_FL_CV = 1 << 6, + MLX4_FL_ETH_HIDE_CQE_VLAN = 1 << 2 +}; +enum { /* vlan_control */ + MLX4_VLAN_CTRL_ETH_TX_BLOCK_TAGGED = 1 << 6, + MLX4_VLAN_CTRL_ETH_RX_BLOCK_TAGGED = 1 << 2, + MLX4_VLAN_CTRL_ETH_RX_BLOCK_PRIO_TAGGED = 1 << 1, /* 802.1p priority tag */ + MLX4_VLAN_CTRL_ETH_RX_BLOCK_UNTAGGED = 1 << 0 +}; + +enum { /* feup */ + MLX4_FEUP_FORCE_ETH_UP = 1 << 6, /* force Eth UP */ + MLX4_FSM_FORCE_ETH_SRC_MAC = 1 << 5, /* force Source MAC */ + MLX4_FVL_FORCE_ETH_VLAN = 1 << 3 /* force Eth vlan */ +}; + +enum { /* fvl_rx */ + MLX4_FVL_RX_FORCE_ETH_VLAN = 1 << 0 /* enforce Eth rx vlan */ +}; + struct mlx4_qp_context { __be32 flags; __be32 pd; @@ -185,6 +206,10 @@ struct mlx4_qp_context { u32 reserved5[10]; }; +enum { /* param3 */ + MLX4_STRIP_VLAN = 1 << 30 +}; + /* Which firmware version adds support for NEC (NoErrorCompletion) bit */ #define MLX4_FW_VER_WQE_CTRL_NEC mlx4_fw_ver(2, 2, 232) From 2fbd967973ae6ae1a989f5638da8bbed93cad2c5 Mon Sep 17 00:00:00 2001 From: Denis Efremov Date: Wed, 8 May 2013 23:19:42 +0000 Subject: [PATCH 089/198] ipv4: ip_output: remove inline marking of EXPORT_SYMBOL functions EXPORT_SYMBOL and inline directives are contradictory to each other. The patch fixes this inconsistency. Found by Linux Driver Verification project (linuxtesting.org). Signed-off-by: Denis Efremov Signed-off-by: David S. Miller --- net/ipv4/ip_output.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c index 147abf5275aa..4bcabf3ab4ca 100644 --- a/net/ipv4/ip_output.c +++ b/net/ipv4/ip_output.c @@ -84,7 +84,7 @@ int sysctl_ip_default_ttl __read_mostly = IPDEFTTL; EXPORT_SYMBOL(sysctl_ip_default_ttl); /* Generate a checksum for an outgoing IP datagram. */ -__inline__ void ip_send_check(struct iphdr *iph) +void ip_send_check(struct iphdr *iph) { iph->check = 0; iph->check = ip_fast_csum((unsigned char *)iph, iph->ihl); From 233c7df0821c4190e2d3f4be0f2ca0ab40a5ed8c Mon Sep 17 00:00:00 2001 From: Jiri Pirko Date: Thu, 9 May 2013 04:23:40 +0000 Subject: [PATCH 090/198] macvlan: fix passthru mode race between dev removal and rx path Currently, if macvlan in passthru mode is created and data are rxed and you remove this device, following panic happens: NULL pointer dereference at 0000000000000198 IP: [] macvlan_handle_frame+0x153/0x1f7 [macvlan] I'm using following script to trigger this: I run this script while "ping -f" is running on another machine to send packets to e1 rx. Reason of the panic is that list_first_entry() is blindly called in macvlan_handle_frame() even if the list was empty. vlan is set to incorrect pointer which leads to the crash. I'm fixing this by protecting port->vlans list by rcu and by preventing from getting incorrect pointer in case the list is empty. Introduced by: commit eb06acdc85585f2 "macvlan: Introduce 'passthru' mode to takeover the underlying device" Signed-off-by: Jiri Pirko Acked-by: Eric Dumazet Signed-off-by: David S. Miller --- drivers/net/macvlan.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/drivers/net/macvlan.c b/drivers/net/macvlan.c index d5a141c7c4e7..1c502bb0c916 100644 --- a/drivers/net/macvlan.c +++ b/drivers/net/macvlan.c @@ -229,7 +229,8 @@ static rx_handler_result_t macvlan_handle_frame(struct sk_buff **pskb) } if (port->passthru) - vlan = list_first_entry(&port->vlans, struct macvlan_dev, list); + vlan = list_first_or_null_rcu(&port->vlans, + struct macvlan_dev, list); else vlan = macvlan_hash_lookup(port, eth->h_dest); if (vlan == NULL) @@ -814,7 +815,7 @@ int macvlan_common_newlink(struct net *src_net, struct net_device *dev, if (err < 0) goto upper_dev_unlink; - list_add_tail(&vlan->list, &port->vlans); + list_add_tail_rcu(&vlan->list, &port->vlans); netif_stacked_transfer_operstate(lowerdev, dev); return 0; @@ -842,7 +843,7 @@ void macvlan_dellink(struct net_device *dev, struct list_head *head) { struct macvlan_dev *vlan = netdev_priv(dev); - list_del(&vlan->list); + list_del_rcu(&vlan->list); unregister_netdevice_queue(dev, head); netdev_upper_dev_unlink(vlan->lowerdev, dev); } From f77d602124d865c38705df7fa25c03de9c284ad2 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 9 May 2013 10:28:16 +0000 Subject: [PATCH 091/198] ipv6: do not clear pinet6 field We have seen multiple NULL dereferences in __inet6_lookup_established() After analysis, I found that inet6_sk() could be NULL while the check for sk_family == AF_INET6 was true. Bug was added in linux-2.6.29 when RCU lookups were introduced in UDP and TCP stacks. Once an IPv6 socket, using SLAB_DESTROY_BY_RCU is inserted in a hash table, we no longer can clear pinet6 field. This patch extends logic used in commit fcbdf09d9652c891 ("net: fix nulls list corruptions in sk_prot_alloc") TCP/UDP/UDPLite IPv6 protocols provide their own .clear_sk() method to make sure we do not clear pinet6 field. At socket clone phase, we do not really care, as cloning the parent (non NULL) pinet6 is not adding a fatal race. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/sock.h | 12 ++++++++++++ net/core/sock.c | 12 ------------ net/ipv6/tcp_ipv6.c | 12 ++++++++++++ net/ipv6/udp.c | 13 ++++++++++++- net/ipv6/udp_impl.h | 2 ++ net/ipv6/udplite.c | 2 +- 6 files changed, 39 insertions(+), 14 deletions(-) diff --git a/include/net/sock.h b/include/net/sock.h index 5c97b0fc5623..66772cf8c3c5 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -866,6 +866,18 @@ struct inet_hashinfo; struct raw_hashinfo; struct module; +/* + * caches using SLAB_DESTROY_BY_RCU should let .next pointer from nulls nodes + * un-modified. Special care is taken when initializing object to zero. + */ +static inline void sk_prot_clear_nulls(struct sock *sk, int size) +{ + if (offsetof(struct sock, sk_node.next) != 0) + memset(sk, 0, offsetof(struct sock, sk_node.next)); + memset(&sk->sk_node.pprev, 0, + size - offsetof(struct sock, sk_node.pprev)); +} + /* Networking protocol blocks we attach to sockets. * socket layer -> transport layer interface * transport -> network interface is defined by struct inet_proto diff --git a/net/core/sock.c b/net/core/sock.c index d4f4cea726e7..6ba327da79e1 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -1217,18 +1217,6 @@ static void sock_copy(struct sock *nsk, const struct sock *osk) #endif } -/* - * caches using SLAB_DESTROY_BY_RCU should let .next pointer from nulls nodes - * un-modified. Special care is taken when initializing object to zero. - */ -static inline void sk_prot_clear_nulls(struct sock *sk, int size) -{ - if (offsetof(struct sock, sk_node.next) != 0) - memset(sk, 0, offsetof(struct sock, sk_node.next)); - memset(&sk->sk_node.pprev, 0, - size - offsetof(struct sock, sk_node.pprev)); -} - void sk_prot_clear_portaddr_nulls(struct sock *sk, int size) { unsigned long nulls1, nulls2; diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 71167069b394..0a17ed9eaf39 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -1890,6 +1890,17 @@ void tcp6_proc_exit(struct net *net) } #endif +static void tcp_v6_clear_sk(struct sock *sk, int size) +{ + struct inet_sock *inet = inet_sk(sk); + + /* we do not want to clear pinet6 field, because of RCU lookups */ + sk_prot_clear_nulls(sk, offsetof(struct inet_sock, pinet6)); + + size -= offsetof(struct inet_sock, pinet6) + sizeof(inet->pinet6); + memset(&inet->pinet6 + 1, 0, size); +} + struct proto tcpv6_prot = { .name = "TCPv6", .owner = THIS_MODULE, @@ -1933,6 +1944,7 @@ struct proto tcpv6_prot = { #ifdef CONFIG_MEMCG_KMEM .proto_cgroup = tcp_proto_cgroup, #endif + .clear_sk = tcp_v6_clear_sk, }; static const struct inet6_protocol tcpv6_protocol = { diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c index d4defdd44937..42923b14dfa6 100644 --- a/net/ipv6/udp.c +++ b/net/ipv6/udp.c @@ -1432,6 +1432,17 @@ void udp6_proc_exit(struct net *net) { } #endif /* CONFIG_PROC_FS */ +void udp_v6_clear_sk(struct sock *sk, int size) +{ + struct inet_sock *inet = inet_sk(sk); + + /* we do not want to clear pinet6 field, because of RCU lookups */ + sk_prot_clear_portaddr_nulls(sk, offsetof(struct inet_sock, pinet6)); + + size -= offsetof(struct inet_sock, pinet6) + sizeof(inet->pinet6); + memset(&inet->pinet6 + 1, 0, size); +} + /* ------------------------------------------------------------------------ */ struct proto udpv6_prot = { @@ -1462,7 +1473,7 @@ struct proto udpv6_prot = { .compat_setsockopt = compat_udpv6_setsockopt, .compat_getsockopt = compat_udpv6_getsockopt, #endif - .clear_sk = sk_prot_clear_portaddr_nulls, + .clear_sk = udp_v6_clear_sk, }; static struct inet_protosw udpv6_protosw = { diff --git a/net/ipv6/udp_impl.h b/net/ipv6/udp_impl.h index d7571046bfc4..4691ed50a928 100644 --- a/net/ipv6/udp_impl.h +++ b/net/ipv6/udp_impl.h @@ -31,6 +31,8 @@ extern int udpv6_recvmsg(struct kiocb *iocb, struct sock *sk, extern int udpv6_queue_rcv_skb(struct sock * sk, struct sk_buff *skb); extern void udpv6_destroy_sock(struct sock *sk); +extern void udp_v6_clear_sk(struct sock *sk, int size); + #ifdef CONFIG_PROC_FS extern int udp6_seq_show(struct seq_file *seq, void *v); #endif diff --git a/net/ipv6/udplite.c b/net/ipv6/udplite.c index 1d08e21d9f69..dfcc4be46898 100644 --- a/net/ipv6/udplite.c +++ b/net/ipv6/udplite.c @@ -56,7 +56,7 @@ struct proto udplitev6_prot = { .compat_setsockopt = compat_udpv6_setsockopt, .compat_getsockopt = compat_udpv6_getsockopt, #endif - .clear_sk = sk_prot_clear_portaddr_nulls, + .clear_sk = udp_v6_clear_sk, }; static struct inet_protosw udplite6_protosw = { From 3b76b3c37b89302db188b743f2f7b31aa666bb00 Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Thu, 9 May 2013 11:04:45 +0000 Subject: [PATCH 092/198] net/ethernet: NET_CALXEDA_XGMAC should depend on HAS_DMA If NO_DMA=y: drivers/built-in.o: In function `xgmac_xmit': drivers/net/ethernet/calxeda/xgmac.c:1102: undefined reference to `dma_mapping_error' Signed-off-by: Geert Uytterhoeven Cc: Rob Herring Cc: David S. Miller Cc: netdev@vger.kernel.org Signed-off-by: David S. Miller --- drivers/net/ethernet/calxeda/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/calxeda/Kconfig b/drivers/net/ethernet/calxeda/Kconfig index aba435c3d4ae..184a063bed5f 100644 --- a/drivers/net/ethernet/calxeda/Kconfig +++ b/drivers/net/ethernet/calxeda/Kconfig @@ -1,6 +1,6 @@ config NET_CALXEDA_XGMAC tristate "Calxeda 1G/10G XGMAC Ethernet driver" - depends on HAS_IOMEM + depends on HAS_IOMEM && HAS_DMA select CRC32 help This is the driver for the XGMAC Ethernet IP block found on Calxeda From fd1eb9e660c7521884387eaf45c12e5225e4ccfb Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Thu, 9 May 2013 11:04:46 +0000 Subject: [PATCH 093/198] net/ethernet: STMMAC_ETH should depend on HAS_DMA If NO_DMA=y: drivers/built-in.o: In function `dma_free_tx_skbufs': drivers/net/ethernet/stmicro/stmmac/stmmac_main.c:1141: undefined reference to `dma_unmap_single' drivers/built-in.o: In function `dma_free_rx_skbufs': drivers/net/ethernet/stmicro/stmmac/stmmac_main.c:1120: undefined reference to `dma_unmap_single' drivers/built-in.o: In function `free_dma_desc_resources': drivers/net/ethernet/stmicro/stmmac/stmmac_main.c:1159: undefined reference to `dma_free_coherent' drivers/built-in.o: In function `stmmac_init_rx_buffers': drivers/net/ethernet/stmicro/stmmac/stmmac_main.c:980: undefined reference to `dma_map_single' drivers/built-in.o: In function `init_dma_desc_rings': drivers/net/ethernet/stmicro/stmmac/stmmac_main.c:1015: undefined reference to `dma_alloc_coherent' drivers/built-in.o: In function `stmmac_tx_clean': drivers/net/ethernet/stmicro/stmmac/stmmac_main.c:1250: undefined reference to `dma_unmap_single' drivers/built-in.o: In function `stmmac_rx': drivers/net/ethernet/stmicro/stmmac/stmmac_main.c:2044: undefined reference to `dma_unmap_single' drivers/net/ethernet/stmicro/stmmac/stmmac_main.c:2082: undefined reference to `dma_unmap_single' drivers/built-in.o: In function `stmmac_rx_refill': drivers/net/ethernet/stmicro/stmmac/stmmac_main.c:1967: undefined reference to `dma_map_single' drivers/built-in.o: In function `stmmac_xmit': drivers/net/ethernet/stmicro/stmmac/stmmac_main.c:1845: undefined reference to `dma_map_single' drivers/built-in.o: In function `skb_frag_dma_map': include/linux/skbuff.h:2184: undefined reference to `dma_map_page' drivers/built-in.o: In function `stmmac_jumbo_frm': drivers/net/ethernet/stmicro/stmmac/ring_mode.c:40: undefined reference to `dma_map_single' drivers/built-in.o: In function `stmmac_jumbo_frm': drivers/net/ethernet/stmicro/stmmac/chain_mode.c:48: undefined reference to `dma_map_single' drivers/net/ethernet/stmicro/stmmac/chain_mode.c:55: undefined reference to `dma_map_single' Signed-off-by: Geert Uytterhoeven Cc: Giuseppe Cavallaro Cc: David S. Miller Cc: netdev@vger.kernel.org Signed-off-by: David S. Miller --- drivers/net/ethernet/stmicro/stmmac/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/Kconfig b/drivers/net/ethernet/stmicro/stmmac/Kconfig index f695a50bac47..43c1f3223322 100644 --- a/drivers/net/ethernet/stmicro/stmmac/Kconfig +++ b/drivers/net/ethernet/stmicro/stmmac/Kconfig @@ -1,6 +1,6 @@ config STMMAC_ETH tristate "STMicroelectronics 10/100/1000 Ethernet driver" - depends on HAS_IOMEM + depends on HAS_IOMEM && HAS_DMA select NET_CORE select MII select PHYLIB From 4837ef42fff894f6ad2bc65431455bc5d9259431 Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Thu, 9 May 2013 11:04:47 +0000 Subject: [PATCH 094/198] net/wireless: ATH9K should depend on HAS_DMA If NO_DMA=y: drivers/built-in.o: In function `ath9k_beacon_generate': drivers/net/wireless/ath/ath9k/beacon.c:146: undefined reference to `dma_unmap_single' drivers/net/wireless/ath/ath9k/beacon.c:174: undefined reference to `dma_map_single' drivers/net/wireless/ath/ath9k/beacon.c:176: undefined reference to `dma_mapping_error' drivers/built-in.o: In function `ath9k_beacon_remove_slot': drivers/net/wireless/ath/ath9k/beacon.c:252: undefined reference to `dma_unmap_single' drivers/built-in.o: In function `ath_descdma_setup': drivers/net/wireless/ath/ath9k/init.c:382: undefined reference to `dmam_alloc_coherent' drivers/built-in.o: In function `ath_edma_get_buffers': drivers/net/wireless/ath/ath9k/recv.c:616: undefined reference to `dma_sync_single_for_cpu' drivers/built-in.o: In function `ath_get_next_rx_buf': drivers/net/wireless/ath/ath9k/recv.c:740: undefined reference to `dma_sync_single_for_cpu' drivers/built-in.o: In function `ath_rx_edma_cleanup': drivers/net/wireless/ath/ath9k/recv.c:176: undefined reference to `dma_unmap_single' drivers/built-in.o: In function `ath_rx_cleanup': drivers/net/wireless/ath/ath9k/recv.c:340: undefined reference to `dma_unmap_single' drivers/built-in.o: In function `ath_rx_edma_buf_link': drivers/net/wireless/ath/ath9k/recv.c:122: undefined reference to `dma_sync_single_for_cpu' drivers/built-in.o: In function `ath_rx_tasklet': drivers/net/wireless/ath/ath9k/recv.c:1275: undefined reference to `dma_map_single' drivers/net/wireless/ath/ath9k/recv.c:1277: undefined reference to `dma_mapping_error' drivers/net/wireless/ath/ath9k/recv.c:1283: undefined reference to `dma_unmap_single' drivers/built-in.o: In function `ath_rx_edma_init': drivers/net/wireless/ath/ath9k/recv.c:226: undefined reference to `dma_map_single' drivers/net/wireless/ath/ath9k/recv.c:229: undefined reference to `dma_mapping_error' drivers/built-in.o: In function `ath_rx_init': drivers/net/wireless/ath/ath9k/recv.c:303: undefined reference to `dma_map_single' drivers/net/wireless/ath/ath9k/recv.c:306: undefined reference to `dma_mapping_error' drivers/built-in.o: In function `ath_tx_complete_buf': drivers/net/wireless/ath/ath9k/xmit.c:2088: undefined reference to `dma_unmap_single' drivers/built-in.o: In function `ath_txstatus_setup': drivers/net/wireless/ath/ath9k/xmit.c:2344: undefined reference to `dmam_alloc_coherent' drivers/built-in.o: In function `ath_tx_set_retry': drivers/net/wireless/ath/ath9k/xmit.c:307: undefined reference to `dma_sync_single_for_cpu' drivers/built-in.o: In function `ath_tx_setup_buffer': drivers/net/wireless/ath/ath9k/xmit.c:1887: undefined reference to `dma_map_single' drivers/net/wireless/ath/ath9k/xmit.c:1889: undefined reference to `dma_mapping_error' Signed-off-by: Geert Uytterhoeven Cc: Luis R. Rodriguez Cc: John W. Linville Cc: linux-wireless@vger.kernel.org Cc: netdev@vger.kernel.org Signed-off-by: David S. Miller --- drivers/net/wireless/ath/ath9k/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/wireless/ath/ath9k/Kconfig b/drivers/net/wireless/ath/ath9k/Kconfig index 17507dc8a1e7..f3dc124c60c7 100644 --- a/drivers/net/wireless/ath/ath9k/Kconfig +++ b/drivers/net/wireless/ath/ath9k/Kconfig @@ -17,7 +17,7 @@ config ATH9K_BTCOEX_SUPPORT config ATH9K tristate "Atheros 802.11n wireless cards support" - depends on MAC80211 + depends on MAC80211 && HAS_DMA select ATH9K_HW select MAC80211_LEDS select LEDS_CLASS From d914ae80b93c8ae8b1c9c55cf431426eff3982dc Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Thu, 9 May 2013 11:04:49 +0000 Subject: [PATCH 095/198] net/ethernet: ARM_AT91_ETHER should depend on HAS_DMA If NO_DMA=y: drivers/built-in.o: In function `at91ether_start': drivers/net/ethernet/cadence/at91_ether.c:49: undefined reference to `dma_alloc_coherent' drivers/net/ethernet/cadence/at91_ether.c:60: undefined reference to `dma_free_coherent' drivers/built-in.o: In function `at91ether_interrupt': drivers/net/ethernet/cadence/at91_ether.c:250: undefined reference to `dma_unmap_single' drivers/built-in.o: In function `at91ether_start_xmit': drivers/net/ethernet/cadence/at91_ether.c:169: undefined reference to `dma_map_single' drivers/built-in.o: In function `at91ether_close': drivers/net/ethernet/cadence/at91_ether.c:145: undefined reference to `dma_free_coherent' Signed-off-by: Geert Uytterhoeven Cc: Nicolas Ferre Cc: David S. Miller Cc: netdev@vger.kernel.org Acked-by: Nicolas Ferre Signed-off-by: David S. Miller --- drivers/net/ethernet/cadence/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/cadence/Kconfig b/drivers/net/ethernet/cadence/Kconfig index 1194446f859a..0d9d6e64d26f 100644 --- a/drivers/net/ethernet/cadence/Kconfig +++ b/drivers/net/ethernet/cadence/Kconfig @@ -22,7 +22,7 @@ if NET_CADENCE config ARM_AT91_ETHER tristate "AT91RM9200 Ethernet support" - depends on GENERIC_HARDIRQS + depends on GENERIC_HARDIRQS && HAS_DMA select NET_CORE select MACB ---help--- From 822bc329695ac1612c87efe68bd8f7d5cbb47be9 Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Thu, 9 May 2013 11:04:50 +0000 Subject: [PATCH 096/198] net/ethernet: MACB should depend on HAS_DMA If NO_DMA=y: drivers/built-in.o: In function `macb_free_consistent': drivers/net/ethernet/cadence/macb.c:878: undefined reference to `dma_free_coherent' drivers/net/ethernet/cadence/macb.c:883: undefined reference to `dma_free_coherent' drivers/net/ethernet/cadence/macb.c:888: undefined reference to `dma_free_coherent' drivers/built-in.o: In function `macb_alloc_consistent': drivers/net/ethernet/cadence/macb.c:905: undefined reference to `dma_alloc_coherent' drivers/built-in.o: In function `macb_tx_interrupt': drivers/net/ethernet/cadence/macb.c:515: undefined reference to `dma_unmap_single' drivers/built-in.o: In function `macb_tx_error_task': drivers/net/ethernet/cadence/macb.c:457: undefined reference to `dma_unmap_single' drivers/built-in.o: In function `macb_start_xmit': drivers/net/ethernet/cadence/macb.c:838: undefined reference to `dma_map_single' Signed-off-by: Geert Uytterhoeven Cc: Nicolas Ferre Cc: David S. Miller Cc: netdev@vger.kernel.org Acked-by: Nicolas Ferre Signed-off-by: David S. Miller --- drivers/net/ethernet/cadence/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/ethernet/cadence/Kconfig b/drivers/net/ethernet/cadence/Kconfig index 0d9d6e64d26f..768285ec10f4 100644 --- a/drivers/net/ethernet/cadence/Kconfig +++ b/drivers/net/ethernet/cadence/Kconfig @@ -31,6 +31,7 @@ config ARM_AT91_ETHER config MACB tristate "Cadence MACB/GEM support" + depends on HAS_DMA select PHYLIB ---help--- The Cadence MACB ethernet interface is found on many Atmel AT32 and From 79e0c19e85cd1617efcf441988a206ffc96b56a1 Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Thu, 9 May 2013 11:04:52 +0000 Subject: [PATCH 097/198] caif: CAIF_VIRTIO should depend on HAS_DMA If NO_DMA=y: drivers/built-in.o: In function `cfv_destroy_genpool': drivers/net/caif/caif_virtio.c:364: undefined reference to `dma_free_coherent' drivers/built-in.o: In function `cfv_create_genpool': drivers/net/caif/caif_virtio.c:397: undefined reference to `dma_alloc_coherent' Signed-off-by: Geert Uytterhoeven Cc: Dmitry Tarnyagin Cc: David S. Miller Cc: netdev@vger.kernel.org Signed-off-by: David S. Miller --- drivers/net/caif/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/caif/Kconfig b/drivers/net/caif/Kconfig index 7ffc756131a2..547098086773 100644 --- a/drivers/net/caif/Kconfig +++ b/drivers/net/caif/Kconfig @@ -43,7 +43,7 @@ config CAIF_HSI config CAIF_VIRTIO tristate "CAIF virtio transport driver" - depends on CAIF + depends on CAIF && HAS_DMA select VHOST_RING select VIRTIO select GENERIC_ALLOCATOR From 4b264a1676e70dc656ba53a8cac690f2d4b65f4e Mon Sep 17 00:00:00 2001 From: Sergei Shtylyov Date: Thu, 9 May 2013 11:14:07 +0000 Subject: [PATCH 098/198] 3c59x: fix PCI resource management The driver wrongly claimed I/O ports at an address returned by pci_iomap() -- even if it was passed an MMIO address. Fix this by claiming/releasing all PCI resources in the PCI driver's probe()/remove() methods instead and get rid of 'must_free_region' flag weirdness (why would Cardbus claim anything for us?). Signed-off-by: Sergei Shtylyov Signed-off-by: David S. Miller --- drivers/net/ethernet/3com/3c59x.c | 25 +++++++++++++------------ 1 file changed, 13 insertions(+), 12 deletions(-) diff --git a/drivers/net/ethernet/3com/3c59x.c b/drivers/net/ethernet/3com/3c59x.c index de570a8f8967..072c6f14e8fc 100644 --- a/drivers/net/ethernet/3com/3c59x.c +++ b/drivers/net/ethernet/3com/3c59x.c @@ -632,7 +632,6 @@ struct vortex_private { pm_state_valid:1, /* pci_dev->saved_config_space has sane contents */ open:1, medialock:1, - must_free_region:1, /* Flag: if zero, Cardbus owns the I/O region */ large_frames:1, /* accept large frames */ handling_irq:1; /* private in_irq indicator */ /* {get|set}_wol operations are already serialized by rtnl. @@ -1012,6 +1011,12 @@ static int vortex_init_one(struct pci_dev *pdev, if (rc < 0) goto out; + rc = pci_request_regions(pdev, DRV_NAME); + if (rc < 0) { + pci_disable_device(pdev); + goto out; + } + unit = vortex_cards_found; if (global_use_mmio < 0 && (unit >= MAX_UNITS || use_mmio[unit] < 0)) { @@ -1027,6 +1032,7 @@ static int vortex_init_one(struct pci_dev *pdev, if (!ioaddr) /* If mapping fails, fall-back to BAR 0... */ ioaddr = pci_iomap(pdev, 0, 0); if (!ioaddr) { + pci_release_regions(pdev); pci_disable_device(pdev); rc = -ENOMEM; goto out; @@ -1036,6 +1042,7 @@ static int vortex_init_one(struct pci_dev *pdev, ent->driver_data, unit); if (rc < 0) { pci_iounmap(pdev, ioaddr); + pci_release_regions(pdev); pci_disable_device(pdev); goto out; } @@ -1178,11 +1185,6 @@ static int vortex_probe1(struct device *gendev, void __iomem *ioaddr, int irq, /* PCI-only startup logic */ if (pdev) { - /* EISA resources already marked, so only PCI needs to do this here */ - /* Ignore return value, because Cardbus drivers already allocate for us */ - if (request_region(dev->base_addr, vci->io_size, print_name) != NULL) - vp->must_free_region = 1; - /* enable bus-mastering if necessary */ if (vci->flags & PCI_USES_MASTER) pci_set_master(pdev); @@ -1220,7 +1222,7 @@ static int vortex_probe1(struct device *gendev, void __iomem *ioaddr, int irq, &vp->rx_ring_dma); retval = -ENOMEM; if (!vp->rx_ring) - goto free_region; + goto free_device; vp->tx_ring = (struct boom_tx_desc *)(vp->rx_ring + RX_RING_SIZE); vp->tx_ring_dma = vp->rx_ring_dma + sizeof(struct boom_rx_desc) * RX_RING_SIZE; @@ -1484,9 +1486,7 @@ free_ring: + sizeof(struct boom_tx_desc) * TX_RING_SIZE, vp->rx_ring, vp->rx_ring_dma); -free_region: - if (vp->must_free_region) - release_region(dev->base_addr, vci->io_size); +free_device: free_netdev(dev); pr_err(PFX "vortex_probe1 fails. Returns %d\n", retval); out: @@ -3254,8 +3254,9 @@ static void vortex_remove_one(struct pci_dev *pdev) + sizeof(struct boom_tx_desc) * TX_RING_SIZE, vp->rx_ring, vp->rx_ring_dma); - if (vp->must_free_region) - release_region(dev->base_addr, vp->io_size); + + pci_release_regions(pdev); + free_netdev(dev); } From 23fbb5a87c56e98a7a4cfac9d6f2ac70f135c4df Mon Sep 17 00:00:00 2001 From: Petri Gynther Date: Thu, 9 May 2013 16:50:00 +0000 Subject: [PATCH 099/198] emac: Fix EMAC soft reset on 460EX/GT Fix EMAC soft reset on 460EX/GT to select the right PHY clock source before and after the soft reset. EMAC with PHY should use the clock from PHY during soft reset. EMAC without PHY should use the internal clock during soft reset. PPC460EX/GT Embedded Processor Advanced User's Manual section 28.10.1 Mode Register 0 (EMACx_MR0) states: Note: The PHY must provide a TX Clk in order to perform a soft reset of the EMAC. If none is present, select the internal clock (SDR0_ETH_CFG[EMACx_PHY_CLK] = 1). After a soft reset, select the external clock. Without the fix, 460EX/GT-based boards with RGMII PHYs attached to EMACs experience EMAC interrupt storm and system watchdog reset when issuing "ifconfig eth0 down" + "ifconfig eth0 up" a few times. The system enters endless loop of serving emac_irq() with EMACx_ISR register stuck at value 0x10000000 (Rx parity error). With the fix, the above issue is no longer observed. Signed-off-by: Petri Gynther Signed-off-by: David S. Miller --- drivers/net/ethernet/ibm/emac/core.c | 36 +++++++++++++++++++++------- 1 file changed, 28 insertions(+), 8 deletions(-) diff --git a/drivers/net/ethernet/ibm/emac/core.c b/drivers/net/ethernet/ibm/emac/core.c index 4989481c19f0..d300a0c0eafc 100644 --- a/drivers/net/ethernet/ibm/emac/core.c +++ b/drivers/net/ethernet/ibm/emac/core.c @@ -359,10 +359,26 @@ static int emac_reset(struct emac_instance *dev) } #ifdef CONFIG_PPC_DCR_NATIVE - /* Enable internal clock source */ - if (emac_has_feature(dev, EMAC_FTR_460EX_PHY_CLK_FIX)) - dcri_clrset(SDR0, SDR0_ETH_CFG, - 0, SDR0_ETH_CFG_ECS << dev->cell_index); + /* + * PPC460EX/GT Embedded Processor Advanced User's Manual + * section 28.10.1 Mode Register 0 (EMACx_MR0) states: + * Note: The PHY must provide a TX Clk in order to perform a soft reset + * of the EMAC. If none is present, select the internal clock + * (SDR0_ETH_CFG[EMACx_PHY_CLK] = 1). + * After a soft reset, select the external clock. + */ + if (emac_has_feature(dev, EMAC_FTR_460EX_PHY_CLK_FIX)) { + if (dev->phy_address == 0xffffffff && + dev->phy_map == 0xffffffff) { + /* No PHY: select internal loop clock before reset */ + dcri_clrset(SDR0, SDR0_ETH_CFG, + 0, SDR0_ETH_CFG_ECS << dev->cell_index); + } else { + /* PHY present: select external clock before reset */ + dcri_clrset(SDR0, SDR0_ETH_CFG, + SDR0_ETH_CFG_ECS << dev->cell_index, 0); + } + } #endif out_be32(&p->mr0, EMAC_MR0_SRST); @@ -370,10 +386,14 @@ static int emac_reset(struct emac_instance *dev) --n; #ifdef CONFIG_PPC_DCR_NATIVE - /* Enable external clock source */ - if (emac_has_feature(dev, EMAC_FTR_460EX_PHY_CLK_FIX)) - dcri_clrset(SDR0, SDR0_ETH_CFG, - SDR0_ETH_CFG_ECS << dev->cell_index, 0); + if (emac_has_feature(dev, EMAC_FTR_460EX_PHY_CLK_FIX)) { + if (dev->phy_address == 0xffffffff && + dev->phy_map == 0xffffffff) { + /* No PHY: restore external clock source after reset */ + dcri_clrset(SDR0, SDR0_ETH_CFG, + SDR0_ETH_CFG_ECS << dev->cell_index, 0); + } + } #endif if (n) { From d34710e3e30fee26f4b0617b5458aa1c48236bb0 Mon Sep 17 00:00:00 2001 From: Amerigo Wang Date: Thu, 9 May 2013 19:50:51 +0000 Subject: [PATCH 100/198] virtio_net: use default napi weight by default Since commit 82dc3c63c692b1e1d5937 ("net: introduce NAPI_POLL_WEIGHT") we warn drivers when they use napi weight higher than NAPI_POLL_WEIGHT, but virtio_net still uses 128 by default. This patch makes its default value to NAPI_POLL_WEIGHT. Cc: "Michael S. Tsirkin" Cc: Eric Dumazet Cc: David S. Miller Signed-off-by: Cong Wang Signed-off-by: David S. Miller --- drivers/net/virtio_net.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index 3c23fdc27bf0..655bb25eed2b 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -28,7 +28,7 @@ #include #include -static int napi_weight = 128; +static int napi_weight = NAPI_POLL_WEIGHT; module_param(napi_weight, int, 0444); static bool csum = true, gso = true; From 5dbd5068430b8bd1c19387d46d6c1a88b261257f Mon Sep 17 00:00:00 2001 From: Amerigo Wang Date: Thu, 9 May 2013 21:56:37 +0000 Subject: [PATCH 101/198] ipv6,gre: do not leak info to user-space There is a hole in struct ip6_tnl_parm2, so we have to zero the struct on stack before copying it to user-space. Cc: David S. Miller Signed-off-by: Cong Wang Signed-off-by: David S. Miller --- net/ipv6/ip6_gre.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/net/ipv6/ip6_gre.c b/net/ipv6/ip6_gre.c index d3ddd8400354..ecd60733e5e2 100644 --- a/net/ipv6/ip6_gre.c +++ b/net/ipv6/ip6_gre.c @@ -1081,6 +1081,7 @@ static int ip6gre_tunnel_ioctl(struct net_device *dev, } if (t == NULL) t = netdev_priv(dev); + memset(&p, 0, sizeof(p)); ip6gre_tnl_parm_to_user(&p, &t->parms); if (copy_to_user(ifr->ifr_ifru.ifru_data, &p, sizeof(p))) err = -EFAULT; @@ -1128,6 +1129,7 @@ static int ip6gre_tunnel_ioctl(struct net_device *dev, if (t) { err = 0; + memset(&p, 0, sizeof(p)); ip6gre_tnl_parm_to_user(&p, &t->parms); if (copy_to_user(ifr->ifr_ifru.ifru_data, &p, sizeof(p))) err = -EFAULT; From 84c4a9dfbf430861e7588d95ae3ff61535dca351 Mon Sep 17 00:00:00 2001 From: Cong Wang Date: Thu, 9 May 2013 22:40:00 +0000 Subject: [PATCH 102/198] xfrm6: release dev before returning error We forget to call dev_put() on error path in xfrm6_fill_dst(), its caller doesn't handle this. Cc: Herbert Xu Cc: Steffen Klassert Cc: David S. Miller Signed-off-by: Cong Wang Signed-off-by: David S. Miller --- net/ipv6/xfrm6_policy.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/net/ipv6/xfrm6_policy.c b/net/ipv6/xfrm6_policy.c index 4ef7bdb65440..23ed03d786c8 100644 --- a/net/ipv6/xfrm6_policy.c +++ b/net/ipv6/xfrm6_policy.c @@ -103,8 +103,10 @@ static int xfrm6_fill_dst(struct xfrm_dst *xdst, struct net_device *dev, dev_hold(dev); xdst->u.rt6.rt6i_idev = in6_dev_get(dev); - if (!xdst->u.rt6.rt6i_idev) + if (!xdst->u.rt6.rt6i_idev) { + dev_put(dev); return -ENODEV; + } rt6_transfer_peer(&xdst->u.rt6, rt); From 4b0c0f294f60abcdd20994a8341a95c8ac5eeb96 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Fri, 3 May 2013 15:02:50 +0200 Subject: [PATCH 103/198] tick: Cleanup NOHZ per cpu data on cpu down Prarit reported a crash on CPU offline/online. The reason is that on CPU down the NOHZ related per cpu data of the dead cpu is not cleaned up. If at cpu online an interrupt happens before the per cpu tick device is registered the irq_enter() check potentially sees stale data and dereferences a NULL pointer. Cleanup the data after the cpu is dead. Reported-by: Prarit Bhargava Cc: stable@vger.kernel.org Cc: Mike Galbraith Link: http://lkml.kernel.org/r/alpine.LFD.2.02.1305031451561.2886@ionos Signed-off-by: Thomas Gleixner --- kernel/time/tick-sched.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c index 225f8bf19095..0eed1db2d792 100644 --- a/kernel/time/tick-sched.c +++ b/kernel/time/tick-sched.c @@ -904,7 +904,7 @@ void tick_cancel_sched_timer(int cpu) hrtimer_cancel(&ts->sched_timer); # endif - ts->nohz_mode = NOHZ_MODE_INACTIVE; + memset(ts, 0, sizeof(*ts)); } #endif From ed7951dc13aad4a14695ec8122e9f0e2ef25d39e Mon Sep 17 00:00:00 2001 From: "Lespiau, Damien" Date: Fri, 10 May 2013 12:36:42 +0000 Subject: [PATCH 104/198] drm: Make the HPD status updates debug logs more readable MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Instead of just printing "status updated from 1 to 2", make those enum numbers immediately readable. v2: Also patch output_poll_execute() (Daniel Vetter) v3: Use drm_get_connector_status_name (Ville Syrjälä) Signed-off-by: Damien Lespiau Reviewed-by: Jesse Barnes (for v1) Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_crtc.c | 1 + drivers/gpu/drm/drm_crtc_helper.c | 10 ++++++---- 2 files changed, 7 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/drm_crtc.c b/drivers/gpu/drm/drm_crtc.c index d7c449f0b110..e7e92429d10f 100644 --- a/drivers/gpu/drm/drm_crtc.c +++ b/drivers/gpu/drm/drm_crtc.c @@ -250,6 +250,7 @@ char *drm_get_connector_status_name(enum drm_connector_status status) else return "unknown"; } +EXPORT_SYMBOL(drm_get_connector_status_name); /** * drm_mode_object_get - allocate a new modeset identifier diff --git a/drivers/gpu/drm/drm_crtc_helper.c b/drivers/gpu/drm/drm_crtc_helper.c index 62f45ad56dc7..9085db634092 100644 --- a/drivers/gpu/drm/drm_crtc_helper.c +++ b/drivers/gpu/drm/drm_crtc_helper.c @@ -1007,10 +1007,11 @@ static void output_poll_execute(struct work_struct *work) continue; connector->status = connector->funcs->detect(connector, false); - DRM_DEBUG_KMS("[CONNECTOR:%d:%s] status updated from %d to %d\n", + DRM_DEBUG_KMS("[CONNECTOR:%d:%s] status updated from %s to %s\n", connector->base.id, drm_get_connector_name(connector), - old_status, connector->status); + drm_get_connector_status_name(old_status), + drm_get_connector_status_name(connector->status)); if (old_status != connector->status) changed = true; } @@ -1085,10 +1086,11 @@ void drm_helper_hpd_irq_event(struct drm_device *dev) old_status = connector->status; connector->status = connector->funcs->detect(connector, false); - DRM_DEBUG_KMS("[CONNECTOR:%d:%s] status updated from %d to %d\n", + DRM_DEBUG_KMS("[CONNECTOR:%d:%s] status updated from %s to %s\n", connector->base.id, drm_get_connector_name(connector), - old_status, connector->status); + drm_get_connector_status_name(old_status), + drm_get_connector_status_name(connector->status)); if (old_status != connector->status) changed = true; } From b2dfcae3cc6802e897556e09541080309cfdec60 Mon Sep 17 00:00:00 2001 From: "Lespiau, Damien" Date: Fri, 10 May 2013 12:36:44 +0000 Subject: [PATCH 105/198] drm: Only print a debug message when the polled connector has changed Suggested-by: Chris Wilson Signed-off-by: Damien Lespiau Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_crtc_helper.c | 19 +++++++++++++------ 1 file changed, 13 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/drm_crtc_helper.c b/drivers/gpu/drm/drm_crtc_helper.c index 9085db634092..ed1334e27c33 100644 --- a/drivers/gpu/drm/drm_crtc_helper.c +++ b/drivers/gpu/drm/drm_crtc_helper.c @@ -1007,13 +1007,20 @@ static void output_poll_execute(struct work_struct *work) continue; connector->status = connector->funcs->detect(connector, false); - DRM_DEBUG_KMS("[CONNECTOR:%d:%s] status updated from %s to %s\n", - connector->base.id, - drm_get_connector_name(connector), - drm_get_connector_status_name(old_status), - drm_get_connector_status_name(connector->status)); - if (old_status != connector->status) + if (old_status != connector->status) { + const char *old, *new; + + old = drm_get_connector_status_name(old_status); + new = drm_get_connector_status_name(connector->status); + + DRM_DEBUG_KMS("[CONNECTOR:%d:%s] " + "status updated from %s to %s\n", + connector->base.id, + drm_get_connector_name(connector), + old, new); + changed = true; + } } mutex_unlock(&dev->mode_config.mutex); From 9d8aa55ff6bd21f573581a29da32ad746f2edcfb Mon Sep 17 00:00:00 2001 From: Christopher Harvey Date: Fri, 12 Apr 2013 20:42:19 +0000 Subject: [PATCH 106/198] drm/mgag200: Don't change unrelated registers during modeset Registers in indices below 0x18 are totally unrelated to modesetting, so don't write 0's, or anything else into them on modeset. Most of these registers are hardware cursor related, so this existing code interferes with hardware cursor development. Signed-off-by: Christopher Harvey Tested-by: Julia Lemire Acked-by: Julia Lemire Acked-by: Mathieu Larouche Signed-off-by: Dave Airlie --- drivers/gpu/drm/mgag200/mgag200_mode.c | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-) diff --git a/drivers/gpu/drm/mgag200/mgag200_mode.c b/drivers/gpu/drm/mgag200/mgag200_mode.c index 7337013a4a0e..2affe7af6cbb 100644 --- a/drivers/gpu/drm/mgag200/mgag200_mode.c +++ b/drivers/gpu/drm/mgag200/mgag200_mode.c @@ -832,11 +832,7 @@ static int mga_crtc_mode_set(struct drm_crtc *crtc, for (i = 0; i < sizeof(dacvalue); i++) { - if ((i <= 0x03) || - (i == 0x07) || - (i == 0x0b) || - (i == 0x0f) || - ((i >= 0x13) && (i <= 0x17)) || + if ((i <= 0x17) || (i == 0x1b) || (i == 0x1c) || ((i >= 0x1f) && (i <= 0x29)) || From fb70a6690875315a3a1454e52fa339441ee7612b Mon Sep 17 00:00:00 2001 From: Christopher Harvey Date: Fri, 12 Apr 2013 22:24:05 +0000 Subject: [PATCH 107/198] drm/mgag200: Fix writes into MGA1064_PIX_CLK_CTL register The original line, WREG_DAC(MGA1064_PIX_CLK_CTL_CLK_DIS, tmp); wrote tmp into MGA1064_PIX_CLK_CTL_CLK_DIS, where MGA1064_PIX_CLK_CTL_CLK_DIS is an offset into MGA1064_PIX_CLK_CTL. Change the line to write properly into MGA1064_PIX_CLK_CTL. There were other chunks of code nearby that use the same pattern (but work correctly), so this patch updates them all to use this new (slightly more efficient) write pattern. The WREG_DAC macro was causing the DAC_INDEX register to be set to the same value twice. WREG8(DAC_DATA, foo) takes advantage of the fact that DAC_INDEX is already at the value we want. Signed-off-by: Christopher Harvey Acked-by: Julia Lemire Tested-by: Julia Lemire Acked-by: Mathieu Larouche Cc: stable@vger.kernel.org Signed-off-by: Dave Airlie --- drivers/gpu/drm/mgag200/mgag200_mode.c | 42 +++++++++++++------------- 1 file changed, 21 insertions(+), 21 deletions(-) diff --git a/drivers/gpu/drm/mgag200/mgag200_mode.c b/drivers/gpu/drm/mgag200/mgag200_mode.c index 2affe7af6cbb..d983868ebd35 100644 --- a/drivers/gpu/drm/mgag200/mgag200_mode.c +++ b/drivers/gpu/drm/mgag200/mgag200_mode.c @@ -189,12 +189,12 @@ static int mga_g200wb_set_plls(struct mga_device *mdev, long clock) WREG8(DAC_INDEX, MGA1064_PIX_CLK_CTL); tmp = RREG8(DAC_DATA); tmp |= MGA1064_PIX_CLK_CTL_CLK_DIS; - WREG_DAC(MGA1064_PIX_CLK_CTL_CLK_DIS, tmp); + WREG8(DAC_DATA, tmp); WREG8(DAC_INDEX, MGA1064_REMHEADCTL); tmp = RREG8(DAC_DATA); tmp |= MGA1064_REMHEADCTL_CLKDIS; - WREG_DAC(MGA1064_REMHEADCTL, tmp); + WREG8(DAC_DATA, tmp); /* select PLL Set C */ tmp = RREG8(MGAREG_MEM_MISC_READ); @@ -204,7 +204,7 @@ static int mga_g200wb_set_plls(struct mga_device *mdev, long clock) WREG8(DAC_INDEX, MGA1064_PIX_CLK_CTL); tmp = RREG8(DAC_DATA); tmp |= MGA1064_PIX_CLK_CTL_CLK_POW_DOWN | 0x80; - WREG_DAC(MGA1064_PIX_CLK_CTL, tmp); + WREG8(DAC_DATA, tmp); udelay(500); @@ -212,7 +212,7 @@ static int mga_g200wb_set_plls(struct mga_device *mdev, long clock) WREG8(DAC_INDEX, MGA1064_VREF_CTL); tmp = RREG8(DAC_DATA); tmp &= ~0x04; - WREG_DAC(MGA1064_VREF_CTL, tmp); + WREG8(DAC_DATA, tmp); udelay(50); @@ -236,13 +236,13 @@ static int mga_g200wb_set_plls(struct mga_device *mdev, long clock) tmp = RREG8(DAC_DATA); tmp &= ~MGA1064_PIX_CLK_CTL_SEL_MSK; tmp |= MGA1064_PIX_CLK_CTL_SEL_PLL; - WREG_DAC(MGA1064_PIX_CLK_CTL, tmp); + WREG8(DAC_DATA, tmp); WREG8(DAC_INDEX, MGA1064_REMHEADCTL); tmp = RREG8(DAC_DATA); tmp &= ~MGA1064_REMHEADCTL_CLKSL_MSK; tmp |= MGA1064_REMHEADCTL_CLKSL_PLL; - WREG_DAC(MGA1064_REMHEADCTL, tmp); + WREG8(DAC_DATA, tmp); /* reset dotclock rate bit */ WREG8(MGAREG_SEQ_INDEX, 1); @@ -253,7 +253,7 @@ static int mga_g200wb_set_plls(struct mga_device *mdev, long clock) WREG8(DAC_INDEX, MGA1064_PIX_CLK_CTL); tmp = RREG8(DAC_DATA); tmp &= ~MGA1064_PIX_CLK_CTL_CLK_DIS; - WREG_DAC(MGA1064_PIX_CLK_CTL, tmp); + WREG8(DAC_DATA, tmp); vcount = RREG8(MGAREG_VCOUNT); @@ -318,7 +318,7 @@ static int mga_g200ev_set_plls(struct mga_device *mdev, long clock) WREG8(DAC_INDEX, MGA1064_PIX_CLK_CTL); tmp = RREG8(DAC_DATA); tmp |= MGA1064_PIX_CLK_CTL_CLK_DIS; - WREG_DAC(MGA1064_PIX_CLK_CTL_CLK_DIS, tmp); + WREG8(DAC_DATA, tmp); tmp = RREG8(MGAREG_MEM_MISC_READ); tmp |= 0x3 << 2; @@ -326,12 +326,12 @@ static int mga_g200ev_set_plls(struct mga_device *mdev, long clock) WREG8(DAC_INDEX, MGA1064_PIX_PLL_STAT); tmp = RREG8(DAC_DATA); - WREG_DAC(MGA1064_PIX_PLL_STAT, tmp & ~0x40); + WREG8(DAC_DATA, tmp & ~0x40); WREG8(DAC_INDEX, MGA1064_PIX_CLK_CTL); tmp = RREG8(DAC_DATA); tmp |= MGA1064_PIX_CLK_CTL_CLK_POW_DOWN; - WREG_DAC(MGA1064_PIX_CLK_CTL, tmp); + WREG8(DAC_DATA, tmp); WREG_DAC(MGA1064_EV_PIX_PLLC_M, m); WREG_DAC(MGA1064_EV_PIX_PLLC_N, n); @@ -342,7 +342,7 @@ static int mga_g200ev_set_plls(struct mga_device *mdev, long clock) WREG8(DAC_INDEX, MGA1064_PIX_CLK_CTL); tmp = RREG8(DAC_DATA); tmp &= ~MGA1064_PIX_CLK_CTL_CLK_POW_DOWN; - WREG_DAC(MGA1064_PIX_CLK_CTL, tmp); + WREG8(DAC_DATA, tmp); udelay(500); @@ -350,11 +350,11 @@ static int mga_g200ev_set_plls(struct mga_device *mdev, long clock) tmp = RREG8(DAC_DATA); tmp &= ~MGA1064_PIX_CLK_CTL_SEL_MSK; tmp |= MGA1064_PIX_CLK_CTL_SEL_PLL; - WREG_DAC(MGA1064_PIX_CLK_CTL, tmp); + WREG8(DAC_DATA, tmp); WREG8(DAC_INDEX, MGA1064_PIX_PLL_STAT); tmp = RREG8(DAC_DATA); - WREG_DAC(MGA1064_PIX_PLL_STAT, tmp | 0x40); + WREG8(DAC_DATA, tmp | 0x40); tmp = RREG8(MGAREG_MEM_MISC_READ); tmp |= (0x3 << 2); @@ -363,7 +363,7 @@ static int mga_g200ev_set_plls(struct mga_device *mdev, long clock) WREG8(DAC_INDEX, MGA1064_PIX_CLK_CTL); tmp = RREG8(DAC_DATA); tmp &= ~MGA1064_PIX_CLK_CTL_CLK_DIS; - WREG_DAC(MGA1064_PIX_CLK_CTL, tmp); + WREG8(DAC_DATA, tmp); return 0; } @@ -416,7 +416,7 @@ static int mga_g200eh_set_plls(struct mga_device *mdev, long clock) WREG8(DAC_INDEX, MGA1064_PIX_CLK_CTL); tmp = RREG8(DAC_DATA); tmp |= MGA1064_PIX_CLK_CTL_CLK_DIS; - WREG_DAC(MGA1064_PIX_CLK_CTL_CLK_DIS, tmp); + WREG8(DAC_DATA, tmp); tmp = RREG8(MGAREG_MEM_MISC_READ); tmp |= 0x3 << 2; @@ -425,7 +425,7 @@ static int mga_g200eh_set_plls(struct mga_device *mdev, long clock) WREG8(DAC_INDEX, MGA1064_PIX_CLK_CTL); tmp = RREG8(DAC_DATA); tmp |= MGA1064_PIX_CLK_CTL_CLK_POW_DOWN; - WREG_DAC(MGA1064_PIX_CLK_CTL, tmp); + WREG8(DAC_DATA, tmp); udelay(500); @@ -439,13 +439,13 @@ static int mga_g200eh_set_plls(struct mga_device *mdev, long clock) tmp = RREG8(DAC_DATA); tmp &= ~MGA1064_PIX_CLK_CTL_SEL_MSK; tmp |= MGA1064_PIX_CLK_CTL_SEL_PLL; - WREG_DAC(MGA1064_PIX_CLK_CTL, tmp); + WREG8(DAC_DATA, tmp); WREG8(DAC_INDEX, MGA1064_PIX_CLK_CTL); tmp = RREG8(DAC_DATA); tmp &= ~MGA1064_PIX_CLK_CTL_CLK_DIS; tmp &= ~MGA1064_PIX_CLK_CTL_CLK_POW_DOWN; - WREG_DAC(MGA1064_PIX_CLK_CTL, tmp); + WREG8(DAC_DATA, tmp); vcount = RREG8(MGAREG_VCOUNT); @@ -515,12 +515,12 @@ static int mga_g200er_set_plls(struct mga_device *mdev, long clock) WREG8(DAC_INDEX, MGA1064_PIX_CLK_CTL); tmp = RREG8(DAC_DATA); tmp |= MGA1064_PIX_CLK_CTL_CLK_DIS; - WREG_DAC(MGA1064_PIX_CLK_CTL_CLK_DIS, tmp); + WREG8(DAC_DATA, tmp); WREG8(DAC_INDEX, MGA1064_REMHEADCTL); tmp = RREG8(DAC_DATA); tmp |= MGA1064_REMHEADCTL_CLKDIS; - WREG_DAC(MGA1064_REMHEADCTL, tmp); + WREG8(DAC_DATA, tmp); tmp = RREG8(MGAREG_MEM_MISC_READ); tmp |= (0x3<<2) | 0xc0; @@ -530,7 +530,7 @@ static int mga_g200er_set_plls(struct mga_device *mdev, long clock) tmp = RREG8(DAC_DATA); tmp &= ~MGA1064_PIX_CLK_CTL_CLK_DIS; tmp |= MGA1064_PIX_CLK_CTL_CLK_POW_DOWN; - WREG_DAC(MGA1064_PIX_CLK_CTL, tmp); + WREG8(DAC_DATA, tmp); udelay(500); From 3cdc0e8d6142aaf3eb84a53eab6de1160da290a3 Mon Sep 17 00:00:00 2001 From: Christopher Harvey Date: Mon, 6 May 2013 15:56:17 +0000 Subject: [PATCH 108/198] drm/mgag200: Convert counter delays to jiffies Signed-off-by: Christopher Harvey Acked-by: Julia Lemire Tested-by: Julia Lemire Acked-by: Mathieu Larouche Signed-off-by: Dave Airlie --- drivers/gpu/drm/mgag200/mgag200_mode.c | 15 ++++++--------- 1 file changed, 6 insertions(+), 9 deletions(-) diff --git a/drivers/gpu/drm/mgag200/mgag200_mode.c b/drivers/gpu/drm/mgag200/mgag200_mode.c index d983868ebd35..0326989a6aec 100644 --- a/drivers/gpu/drm/mgag200/mgag200_mode.c +++ b/drivers/gpu/drm/mgag200/mgag200_mode.c @@ -46,29 +46,26 @@ static void mga_crtc_load_lut(struct drm_crtc *crtc) static inline void mga_wait_vsync(struct mga_device *mdev) { - unsigned int count = 0; + unsigned long timeout = jiffies + HZ/10; unsigned int status = 0; do { status = RREG32(MGAREG_Status); - count++; - } while ((status & 0x08) && (count < 250000)); - count = 0; + } while ((status & 0x08) && time_before(jiffies, timeout)); + timeout = jiffies + HZ/10; status = 0; do { status = RREG32(MGAREG_Status); - count++; - } while (!(status & 0x08) && (count < 250000)); + } while (!(status & 0x08) && time_before(jiffies, timeout)); } static inline void mga_wait_busy(struct mga_device *mdev) { - unsigned int count = 0; + unsigned long timeout = jiffies + HZ; unsigned int status = 0; do { status = RREG8(MGAREG_Status + 2); - count++; - } while ((status & 0x01) && (count < 500000)); + } while ((status & 0x01) && time_before(jiffies, timeout)); } /* From 9f1d036648c1c5ed81b0e98d7a06d55df972701e Mon Sep 17 00:00:00 2001 From: Christopher Harvey Date: Wed, 8 May 2013 19:10:38 +0000 Subject: [PATCH 109/198] drm/mgag200: Fix framebuffer base address programming Higher bits of the base address of framebuffers weren't being programmed properly. This caused framebuffers that didn't happen to be allocated at a low enough address to not be displayed properly. Signed-off-by: Christopher Harvey Signed-off-by: Mathieu Larouche Acked-by: Julia Lemire Tested-by: Julia Lemire Cc: stable@vger.kernel.org Signed-off-by: Dave Airlie --- drivers/gpu/drm/mgag200/mgag200_mode.c | 27 +++++++++++++++++++++++--- 1 file changed, 24 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/mgag200/mgag200_mode.c b/drivers/gpu/drm/mgag200/mgag200_mode.c index 0326989a6aec..e8e20c653b1d 100644 --- a/drivers/gpu/drm/mgag200/mgag200_mode.c +++ b/drivers/gpu/drm/mgag200/mgag200_mode.c @@ -654,12 +654,26 @@ static void mga_g200wb_commit(struct drm_crtc *crtc) WREG_DAC(MGA1064_GEN_IO_DATA, tmp); } - +/* + This is how the framebuffer base address is stored in g200 cards: + * Assume @offset is the gpu_addr variable of the framebuffer object + * Then addr is the number of _pixels_ (not bytes) from the start of + VRAM to the first pixel we want to display. (divided by 2 for 32bit + framebuffers) + * addr is stored in the CRTCEXT0, CRTCC and CRTCD registers + addr<20> -> CRTCEXT0<6> + addr<19-16> -> CRTCEXT0<3-0> + addr<15-8> -> CRTCC<7-0> + addr<7-0> -> CRTCD<7-0> + CRTCEXT0 has to be programmed last to trigger an update and make the + new addr variable take effect. + */ void mga_set_start_address(struct drm_crtc *crtc, unsigned offset) { struct mga_device *mdev = crtc->dev->dev_private; u32 addr; int count; + u8 crtcext0; while (RREG8(0x1fda) & 0x08); while (!(RREG8(0x1fda) & 0x08)); @@ -667,10 +681,17 @@ void mga_set_start_address(struct drm_crtc *crtc, unsigned offset) count = RREG8(MGAREG_VCOUNT) + 2; while (RREG8(MGAREG_VCOUNT) < count); - addr = offset >> 2; + WREG8(MGAREG_CRTCEXT_INDEX, 0); + crtcext0 = RREG8(MGAREG_CRTCEXT_DATA); + crtcext0 &= 0xB0; + addr = offset / 8; + /* Can't store addresses any higher than that... + but we also don't have more than 16MB of memory, so it should be fine. */ + WARN_ON(addr > 0x1fffff); + crtcext0 |= (!!(addr & (1<<20)))<<6; WREG_CRT(0x0d, (u8)(addr & 0xff)); WREG_CRT(0x0c, (u8)(addr >> 8) & 0xff); - WREG_CRT(0xaf, (u8)(addr >> 16) & 0xf); + WREG_ECRT(0x0, ((u8)(addr >> 16) & 0xf) | crtcext0); } From e2555fde4159467fb579e6ae3c0a8fc33015d0f5 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Mon, 13 May 2013 09:45:01 -0400 Subject: [PATCH 110/198] jbd,jbd2: fix oops in jbd2_journal_put_journal_head() Commit ae4647fb (jbd2: reduce journal_head size) introduced a regression where we occasionally hit panic in jbd2_journal_put_journal_head() because of wrong b_jcount. The bug is caused by gcc making 64-bit access to 32-bit bitfield and thus clobbering b_jcount. At least for now, those 8 bytes saved in struct journal_head are not worth the trouble with gcc bitfield handling so revert that part of the patch. Reported-by: EUNBONG SONG Reported-by: Tony Luck Signed-off-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- include/linux/journal-head.h | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/include/linux/journal-head.h b/include/linux/journal-head.h index 13a3da25ff07..98cd41bb39c8 100644 --- a/include/linux/journal-head.h +++ b/include/linux/journal-head.h @@ -30,15 +30,19 @@ struct journal_head { /* * Journalling list for this buffer [jbd_lock_bh_state()] + * NOTE: We *cannot* combine this with b_modified into a bitfield + * as gcc would then (which the C standard allows but which is + * very unuseful) make 64-bit accesses to the bitfield and clobber + * b_jcount if its update races with bitfield modification. */ - unsigned b_jlist:4; + unsigned b_jlist; /* * This flag signals the buffer has been modified by * the currently running transaction * [jbd_lock_bh_state()] */ - unsigned b_modified:1; + unsigned b_modified; /* * Copy of the buffer data frozen for writing to the log. From 1aa3d8d9935dfb459d64f06ccd828b56147fd971 Mon Sep 17 00:00:00 2001 From: Stefano Stabellini Date: Wed, 8 May 2013 11:59:01 +0000 Subject: [PATCH 111/198] xen/arm: initialize pm functions later If we are running in dom0, we have to wait for the arch specific code to complete the initialization in order for us to successfully reset the power_off and pm_restart functions. Signed-off-by: Stefano Stabellini --- arch/arm/xen/enlighten.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/arch/arm/xen/enlighten.c b/arch/arm/xen/enlighten.c index d30042e39974..83d13b0d5f74 100644 --- a/arch/arm/xen/enlighten.c +++ b/arch/arm/xen/enlighten.c @@ -272,12 +272,18 @@ static int __init xen_guest_init(void) if (!xen_initial_domain()) xenbus_probe(NULL); + return 0; +} +core_initcall(xen_guest_init); + +static int __init xen_pm_init(void) +{ pm_power_off = xen_power_off; arm_pm_restart = xen_restart; return 0; } -core_initcall(xen_guest_init); +subsys_initcall(xen_pm_init); static irqreturn_t xen_arm_callback(int irq, void *arg) { From d7266d7894b111dee9b081c05944533bc8780eca Mon Sep 17 00:00:00 2001 From: Stefano Stabellini Date: Wed, 8 May 2013 13:02:38 +0000 Subject: [PATCH 112/198] xen/arm: do not handle VCPUOP_register_vcpu_info failures We expect VCPUOP_register_vcpu_info to succeed, do not try to handle failures. Signed-off-by: Stefano Stabellini Acked-by: Ian Campbell --- arch/arm/xen/enlighten.c | 10 +++------- 1 file changed, 3 insertions(+), 7 deletions(-) diff --git a/arch/arm/xen/enlighten.c b/arch/arm/xen/enlighten.c index 83d13b0d5f74..3dfa2b5c787d 100644 --- a/arch/arm/xen/enlighten.c +++ b/arch/arm/xen/enlighten.c @@ -165,13 +165,9 @@ static int __init xen_secondary_init(unsigned int cpu) info.offset = offset_in_page(vcpup); err = HYPERVISOR_vcpu_op(VCPUOP_register_vcpu_info, cpu, &info); - if (err) { - pr_debug("register_vcpu_info failed: err=%d\n", err); - } else { - /* This cpu is using the registered vcpu info, even if - later ones fail to. */ - per_cpu(xen_vcpu, cpu) = vcpup; - } + BUG_ON(err); + per_cpu(xen_vcpu, cpu) = vcpup; + return 0; } From 3cc8e40e8ff8e232a9dd672da81beabd09f87366 Mon Sep 17 00:00:00 2001 From: Stefano Stabellini Date: Wed, 8 May 2013 11:59:01 +0000 Subject: [PATCH 113/198] xen/arm: rename xen_secondary_init and run it on every online cpu Rename xen_secondary_init to xen_percpu_init. Run xen_percpu_init on the each online cpu, reuse the current on_each_cpu call. Merge xen_percpu_enable_events into xen_percpu_init. Signed-off-by: Stefano Stabellini --- arch/arm/xen/enlighten.c | 15 ++++----------- 1 file changed, 4 insertions(+), 11 deletions(-) diff --git a/arch/arm/xen/enlighten.c b/arch/arm/xen/enlighten.c index 3dfa2b5c787d..13609e01f4b7 100644 --- a/arch/arm/xen/enlighten.c +++ b/arch/arm/xen/enlighten.c @@ -152,11 +152,12 @@ int xen_unmap_domain_mfn_range(struct vm_area_struct *vma, } EXPORT_SYMBOL_GPL(xen_unmap_domain_mfn_range); -static int __init xen_secondary_init(unsigned int cpu) +static void __init xen_percpu_init(void *unused) { struct vcpu_register_vcpu_info info; struct vcpu_info *vcpup; int err; + int cpu = get_cpu(); pr_info("Xen: initializing cpu%d\n", cpu); vcpup = per_cpu_ptr(xen_vcpu_info, cpu); @@ -168,7 +169,7 @@ static int __init xen_secondary_init(unsigned int cpu) BUG_ON(err); per_cpu(xen_vcpu, cpu) = vcpup; - return 0; + enable_percpu_irq(xen_events_irq, 0); } static void xen_restart(char str, const char *cmd) @@ -204,7 +205,6 @@ static int __init xen_guest_init(void) const char *version = NULL; const char *xen_prefix = "xen,xen-"; struct resource res; - int i; node = of_find_compatible_node(NULL, NULL, "xen,xen"); if (!node) { @@ -261,8 +261,6 @@ static int __init xen_guest_init(void) sizeof(struct vcpu_info)); if (xen_vcpu_info == NULL) return -ENOMEM; - for_each_online_cpu(i) - xen_secondary_init(i); gnttab_init(); if (!xen_initial_domain()) @@ -287,11 +285,6 @@ static irqreturn_t xen_arm_callback(int irq, void *arg) return IRQ_HANDLED; } -static __init void xen_percpu_enable_events(void *unused) -{ - enable_percpu_irq(xen_events_irq, 0); -} - static int __init xen_init_events(void) { if (!xen_domain() || xen_events_irq < 0) @@ -305,7 +298,7 @@ static int __init xen_init_events(void) return -EINVAL; } - on_each_cpu(xen_percpu_enable_events, NULL, 0); + on_each_cpu(xen_percpu_init, NULL, 0); return 0; } From c10ebbf55b40503a46fb8b29824c9ca1ca089826 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Thu, 9 May 2013 14:56:32 -0500 Subject: [PATCH 114/198] libceph: init sent and completed when starting The rbd code has a need to be able to restart an osd request that has already been started and completed once before. This currently wouldn't work right because the osd client code assumes an osd request will be started exactly once Certain fields in a request are never cleared and this leads to trouble if you try to reuse it. Specifically, the r_sent, r_got_reply, and r_completed fields are never cleared. The r_sent field records the osd incarnation at the time the request was sent to that osd. If that's non-zero, the message won't get re-mapped to a target osd properly, and won't be put on the unsafe requests list the first time it's sent as it should. The r_got_reply field is used in handle_reply() to ensure the reply to a request is processed only once. And the r_completed field is used for lingering requests to avoid calling the callback function every time the osd client re-sends the request on behalf of its initiator. Each osd request passes through ceph_osdc_start_request() when responsibility for the request is handed over to the osd client for completion. We can safely zero these three fields there each time a request gets started. One last related change--clear the r_linger flag when a request is no longer registered as a linger request. This resolves: http://tracker.ceph.com/issues/5026 Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- net/ceph/osd_client.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/net/ceph/osd_client.c b/net/ceph/osd_client.c index a3395fdfbd4f..d5953b87918c 100644 --- a/net/ceph/osd_client.c +++ b/net/ceph/osd_client.c @@ -1204,6 +1204,7 @@ void ceph_osdc_unregister_linger_request(struct ceph_osd_client *osdc, mutex_lock(&osdc->request_mutex); if (req->r_linger) { __unregister_linger_request(osdc, req); + req->r_linger = 0; ceph_osdc_put_request(req); } mutex_unlock(&osdc->request_mutex); @@ -2120,7 +2121,9 @@ int ceph_osdc_start_request(struct ceph_osd_client *osdc, down_read(&osdc->map_sem); mutex_lock(&osdc->request_mutex); __register_request(osdc, req); - WARN_ON(req->r_sent); + req->r_sent = 0; + req->r_got_reply = 0; + req->r_completed = 0; rc = __map_request(osdc, req, 0); if (rc < 0) { if (nofail) { From ebda6408f2227a774343c3cc2861384942143ff3 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Fri, 10 May 2013 16:29:22 -0500 Subject: [PATCH 115/198] rbd: fix parent request size assumption The code that reads object data from the parent for a copyup on write request currently assumes that the size of that request is the size of a "full" object from the original target image. That is not necessarily the case. The parent overlap could reduce the request size below that. To fix that assumption we need to record the number of pages in the copyup_pages array, for both an image request and an object request. Rename a local variable in rbd_img_obj_parent_read_full_callback() to reflect we're recording the length of the parent read request, not the size of the target object. This resolves: http://tracker.ceph.com/issues/5038 Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 31 ++++++++++++++++++++----------- 1 file changed, 20 insertions(+), 11 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 51c45e793354..597b9bbe2fc7 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -224,6 +224,7 @@ struct rbd_obj_request { }; }; struct page **copyup_pages; + u32 copyup_page_count; struct ceph_osd_request *osd_req; @@ -256,6 +257,7 @@ struct rbd_img_request { struct rbd_obj_request *obj_request; /* obj req initiator */ }; struct page **copyup_pages; + u32 copyup_page_count; spinlock_t completion_lock;/* protects next_completion */ u32 next_completion; rbd_img_callback_t callback; @@ -2119,7 +2121,7 @@ rbd_img_obj_copyup_callback(struct rbd_obj_request *obj_request) { struct rbd_img_request *img_request; struct rbd_device *rbd_dev; - u64 length; + struct page **pages; u32 page_count; rbd_assert(obj_request->type == OBJ_REQUEST_BIO); @@ -2129,12 +2131,14 @@ rbd_img_obj_copyup_callback(struct rbd_obj_request *obj_request) rbd_dev = img_request->rbd_dev; rbd_assert(rbd_dev); - length = (u64)1 << rbd_dev->header.obj_order; - page_count = (u32)calc_pages_for(0, length); - rbd_assert(obj_request->copyup_pages); - ceph_release_page_vector(obj_request->copyup_pages, page_count); + pages = obj_request->copyup_pages; + rbd_assert(pages != NULL); obj_request->copyup_pages = NULL; + page_count = obj_request->copyup_page_count; + rbd_assert(page_count); + obj_request->copyup_page_count = 0; + ceph_release_page_vector(pages, page_count); /* * We want the transfer count to reflect the size of the @@ -2158,9 +2162,9 @@ rbd_img_obj_parent_read_full_callback(struct rbd_img_request *img_request) struct ceph_osd_client *osdc; struct rbd_device *rbd_dev; struct page **pages; + u32 page_count; int result; - u64 obj_size; - u64 xferred; + u64 parent_length; rbd_assert(img_request_child_test(img_request)); @@ -2169,19 +2173,21 @@ rbd_img_obj_parent_read_full_callback(struct rbd_img_request *img_request) pages = img_request->copyup_pages; rbd_assert(pages != NULL); img_request->copyup_pages = NULL; + page_count = img_request->copyup_page_count; + rbd_assert(page_count); + img_request->copyup_page_count = 0; orig_request = img_request->obj_request; rbd_assert(orig_request != NULL); rbd_assert(orig_request->type == OBJ_REQUEST_BIO); result = img_request->result; - obj_size = img_request->length; - xferred = img_request->xferred; + parent_length = img_request->length; + rbd_assert(parent_length == img_request->xferred); rbd_img_request_put(img_request); rbd_assert(orig_request->img_request); rbd_dev = orig_request->img_request->rbd_dev; rbd_assert(rbd_dev); - rbd_assert(obj_size == (u64)1 << rbd_dev->header.obj_order); if (result) goto out_err; @@ -2195,11 +2201,12 @@ rbd_img_obj_parent_read_full_callback(struct rbd_img_request *img_request) goto out_err; orig_request->osd_req = osd_req; orig_request->copyup_pages = pages; + orig_request->copyup_page_count = page_count; /* Initialize the copyup op */ osd_req_op_cls_init(osd_req, 0, CEPH_OSD_OP_CALL, "rbd", "copyup"); - osd_req_op_cls_request_data_pages(osd_req, 0, pages, obj_size, 0, + osd_req_op_cls_request_data_pages(osd_req, 0, pages, parent_length, 0, false, false); /* Then the original write request op */ @@ -2312,6 +2319,7 @@ static int rbd_img_obj_parent_read_full(struct rbd_obj_request *obj_request) if (result) goto out_err; parent_request->copyup_pages = pages; + parent_request->copyup_page_count = page_count; parent_request->callback = rbd_img_obj_parent_read_full_callback; result = rbd_img_request_submit(parent_request); @@ -2319,6 +2327,7 @@ static int rbd_img_obj_parent_read_full(struct rbd_obj_request *obj_request) return 0; parent_request->copyup_pages = NULL; + parent_request->copyup_page_count = 0; parent_request->obj_request = NULL; rbd_obj_request_put(obj_request); out_err: From b91f09f17b2a302f07022e2f766969e2536d71b3 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Fri, 10 May 2013 16:29:22 -0500 Subject: [PATCH 116/198] rbd: support reading parent page data for writes Currently, rbd_img_obj_parent_read_full() assumes the incoming object request contains bio data. But if a layered image is part of a multi-layer stack of images it will result in read requests of page data to parent images. This is handling the same kind of issue as was resolved by this commit: 5b2ab72d rbd: support reading parent page data This resolves: http://tracker.ceph.com/issues/5027 Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 20 ++++++++++++++------ 1 file changed, 14 insertions(+), 6 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 597b9bbe2fc7..5161e80a38ef 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -2165,6 +2165,8 @@ rbd_img_obj_parent_read_full_callback(struct rbd_img_request *img_request) u32 page_count; int result; u64 parent_length; + u64 offset; + u64 length; rbd_assert(img_request_child_test(img_request)); @@ -2179,7 +2181,7 @@ rbd_img_obj_parent_read_full_callback(struct rbd_img_request *img_request) orig_request = img_request->obj_request; rbd_assert(orig_request != NULL); - rbd_assert(orig_request->type == OBJ_REQUEST_BIO); + rbd_assert(obj_request_type_valid(orig_request->type)); result = img_request->result; parent_length = img_request->length; rbd_assert(parent_length == img_request->xferred); @@ -2211,11 +2213,17 @@ rbd_img_obj_parent_read_full_callback(struct rbd_img_request *img_request) /* Then the original write request op */ + offset = orig_request->offset; + length = orig_request->length; osd_req_op_extent_init(osd_req, 1, CEPH_OSD_OP_WRITE, - orig_request->offset, - orig_request->length, 0, 0); - osd_req_op_extent_osd_data_bio(osd_req, 1, orig_request->bio_list, - orig_request->length); + offset, length, 0, 0); + if (orig_request->type == OBJ_REQUEST_BIO) + osd_req_op_extent_osd_data_bio(osd_req, 1, + orig_request->bio_list, length); + else + osd_req_op_extent_osd_data_pages(osd_req, 1, + orig_request->pages, length, + offset & ~PAGE_MASK, false, false); rbd_osd_req_format_write(orig_request); @@ -2261,7 +2269,7 @@ static int rbd_img_obj_parent_read_full(struct rbd_obj_request *obj_request) int result; rbd_assert(obj_request_img_data_test(obj_request)); - rbd_assert(obj_request->type == OBJ_REQUEST_BIO); + rbd_assert(obj_request_type_valid(obj_request->type)); img_request = obj_request->img_request; rbd_assert(img_request != NULL); From 70cf49cfc7a4d1eb4aeea6cd128b88230be9d0b1 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 17:40:33 -0500 Subject: [PATCH 117/198] rbd: ignore zero-overlap parent An rbd clone image that has an overlap with its parent of 0 is effectively not a layered image at all. Detect this case and treat such an image as non-layered. Issue a warning to be sure the user knows what's going on. This resolves: http://tracker.ceph.com/issues/5028 Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 5161e80a38ef..b67ecda1e7ef 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -3665,9 +3665,13 @@ static int rbd_dev_v2_parent_info(struct rbd_device *rbd_dev) ceph_decode_64_safe(&p, end, parent_spec->snap_id, out_err); ceph_decode_64_safe(&p, end, overlap, out_err); - rbd_dev->parent_overlap = overlap; - rbd_dev->parent_spec = parent_spec; - parent_spec = NULL; /* rbd_dev now owns this */ + if (overlap) { + rbd_dev->parent_spec = parent_spec; + parent_spec = NULL; /* rbd_dev now owns this */ + rbd_dev->parent_overlap = overlap; + } else { + rbd_warn(rbd_dev, "ignoring parent of clone with overlap 0\n"); + } out: ret = 0; out_err: From ef38079401c03fab7e0eb72085635c841b9b8ee7 Mon Sep 17 00:00:00 2001 From: Thadeu Lima de Souza Cascardo Date: Sat, 11 May 2013 09:15:37 +0000 Subject: [PATCH 118/198] qlge: fix dma map leak when the last chunk is not allocated qlge allocates chunks from a page that it maps and unmaps that page when the last chunk is released. When the driver is unloaded or the card is removed, all chunks are released and the page is unmapped for the last chunk. However, when the last chunk of a page is not allocated and the device is removed, that page is not unmapped. In fact, its last reference is not put and there's also a page leak. This bug prevents a device from being properly hotplugged. When the DMA API debug option is enabled, the following messages show the pending DMA allocation after we remove the driver. This patch fixes the bug by unmapping and putting the page from the ring if its last chunk has not been allocated. pci 0005:98:00.0: DMA-API: device driver has pending DMA allocations while released from device [count=1] One of leaked entries details: [device address=0x0000000060a80000] [size=65536 bytes] [mapped with DMA_FROM_DEVICE] [mapped as page] ------------[ cut here ]------------ WARNING: at lib/dma-debug.c:746 Modules linked in: qlge(-) rpadlpar_io rpaphp pci_hotplug fuse [last unloaded: qlge] NIP: c0000000003fc3ec LR: c0000000003fc3e8 CTR: c00000000054de60 REGS: c0000003ee9c74e0 TRAP: 0700 Tainted: G O (3.7.2) MSR: 8000000000029032 CR: 28002424 XER: 00000001 SOFTE: 1 CFAR: c0000000007a39c8 TASK = c0000003ee8d5c90[8406] 'rmmod' THREAD: c0000003ee9c4000 CPU: 31 GPR00: c0000000003fc3e8 c0000003ee9c7760 c000000000c789f8 00000000000000ee GPR04: 0000000000000000 00000000000000ef 0000000000004000 0000000000010000 GPR08: 00000000000000be c000000000b22088 c000000000c4c218 00000000007c0000 GPR12: 0000000028002422 c00000000ff26c80 0000000000000000 000001001b0f1b40 GPR16: 00000000100cb9d8 0000000010093088 c000000000cdf910 0000000000000001 GPR20: 0000000000000000 c000000000dbfc00 0000000000000000 c000000000dbfb80 GPR24: c0000003fafc9d80 0000000000000001 000000000001ff80 c0000003f38f7888 GPR28: c000000000ddfc00 0000000000000400 c000000000bd7790 c000000000ddfb80 NIP [c0000000003fc3ec] .dma_debug_device_change+0x22c/0x2b0 LR [c0000000003fc3e8] .dma_debug_device_change+0x228/0x2b0 Call Trace: [c0000003ee9c7760] [c0000000003fc3e8] .dma_debug_device_change+0x228/0x2b0 (unreliable) [c0000003ee9c7840] [c00000000079a098] .notifier_call_chain+0x78/0xf0 [c0000003ee9c78e0] [c0000000000acc20] .__blocking_notifier_call_chain+0x70/0xb0 [c0000003ee9c7990] [c0000000004a9580] .__device_release_driver+0x100/0x140 [c0000003ee9c7a20] [c0000000004a9708] .driver_detach+0x148/0x150 [c0000003ee9c7ac0] [c0000000004a8144] .bus_remove_driver+0xc4/0x150 [c0000003ee9c7b60] [c0000000004aa58c] .driver_unregister+0x8c/0xe0 [c0000003ee9c7bf0] [c0000000004090b4] .pci_unregister_driver+0x34/0xf0 [c0000003ee9c7ca0] [d000000002231194] .qlge_exit+0x1c/0x34 [qlge] [c0000003ee9c7d20] [c0000000000e36d8] .SyS_delete_module+0x1e8/0x290 [c0000003ee9c7e30] [c0000000000098d4] syscall_exit+0x0/0x94 Instruction dump: 7f26cb78 e818003a e87e81a0 e8f80028 e9180030 796b1f24 78001f24 7d6a5a14 7d2a002a e94b0020 483a7595 60000000 <0fe00000> 2fb80000 40de0048 80120050 ---[ end trace 4294f9abdb01031d ]--- Mapped at: [] .ql_update_lbq+0x384/0x580 [qlge] [] .ql_clean_inbound_rx_ring+0x300/0xc60 [qlge] [] .ql_napi_poll_msix+0x39c/0x5a0 [qlge] [] .net_rx_action+0x170/0x300 [] .__do_softirq+0x170/0x300 Signed-off-by: Thadeu Lima de Souza Cascardo Acked-by: Jitendra Kalsaria Signed-off-by: David S. Miller --- drivers/net/ethernet/qlogic/qlge/qlge_main.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/net/ethernet/qlogic/qlge/qlge_main.c b/drivers/net/ethernet/qlogic/qlge/qlge_main.c index 87463bc701a6..50235d201592 100644 --- a/drivers/net/ethernet/qlogic/qlge/qlge_main.c +++ b/drivers/net/ethernet/qlogic/qlge/qlge_main.c @@ -1106,6 +1106,7 @@ static int ql_get_next_chunk(struct ql_adapter *qdev, struct rx_ring *rx_ring, if (pci_dma_mapping_error(qdev->pdev, map)) { __free_pages(rx_ring->pg_chunk.page, qdev->lbq_buf_order); + rx_ring->pg_chunk.page = NULL; netif_err(qdev, drv, qdev->ndev, "PCI mapping failed.\n"); return -ENOMEM; @@ -2777,6 +2778,12 @@ static void ql_free_lbq_buffers(struct ql_adapter *qdev, struct rx_ring *rx_ring curr_idx = 0; } + if (rx_ring->pg_chunk.page) { + pci_unmap_page(qdev->pdev, rx_ring->pg_chunk.map, + ql_lbq_block_size(qdev), PCI_DMA_FROMDEVICE); + put_page(rx_ring->pg_chunk.page); + rx_ring->pg_chunk.page = NULL; + } } static void ql_free_sbq_buffers(struct ql_adapter *qdev, struct rx_ring *rx_ring) From ba21fc696dd28ea7a5880345faf0168619a478d2 Mon Sep 17 00:00:00 2001 From: Wei Yongjun Date: Mon, 13 May 2013 04:26:06 +0000 Subject: [PATCH 119/198] bna: add missing iounmap() on error in bnad_init() Add the missing iounmap() before return from bnad_init() in the error handling case. Introduced by commit 01b54b1451853593739816a392485c4e2bee7dda (bna: tx rx cleanup fix). Signed-off-by: Wei Yongjun Signed-off-by: David S. Miller --- drivers/net/ethernet/brocade/bna/bnad.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/brocade/bna/bnad.c b/drivers/net/ethernet/brocade/bna/bnad.c index ce4a030d3d0c..07f7ef05c3f2 100644 --- a/drivers/net/ethernet/brocade/bna/bnad.c +++ b/drivers/net/ethernet/brocade/bna/bnad.c @@ -3236,9 +3236,10 @@ bnad_init(struct bnad *bnad, sprintf(bnad->wq_name, "%s_wq_%d", BNAD_NAME, bnad->id); bnad->work_q = create_singlethread_workqueue(bnad->wq_name); - - if (!bnad->work_q) + if (!bnad->work_q) { + iounmap(bnad->bar0); return -ENOMEM; + } return 0; } From 642a25375f4c863607d2170f4471aec8becf7788 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 17:40:33 -0500 Subject: [PATCH 120/198] rbd: get parent info on refresh Get parent info for format 2 images on every refresh (rather than just during the initial probe). This will be needed to detect the disappearance of the parent image in the event a mapped image becomes unlayered (i.e., flattened). Avoid leaking the previous parent spec on the second and subsequent times this information is requested by dropping the previous one (if any) before updating it. (Also, extract the pool id into a local variable before assigning it into the parent spec.) Switch to using a non-zero parent overlap value rather than the existence of a parent (a non-null parent_spec pointer) to determine whether to mark a request layered. It will soon be possible for a layered image to become unlayered while a request is in flight. This means that the layered flag for an image request indicates that there was a non-zero parent overlap at the time the image request was created. The parent overlap can change thereafter, which may lead to special handling at request submission or completion time. This and the next several patches are related to: http://tracker.ceph.com/issues/3763 NOTE: If an error occurs while refreshing the parent info (i.e., requesting it after initial probe), the old parent info will persist. This is not really correct, and is a scenario that needs to be addressed. For now we'll assert that the failure mode is unlikely, but the issue has been documented in tracker issue 5040. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 67 +++++++++++++++++++++++++-------------------- 1 file changed, 37 insertions(+), 30 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index b67ecda1e7ef..fcef63c2c30b 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -1873,7 +1873,7 @@ static struct rbd_img_request *rbd_img_request_create( } if (child_request) img_request_child_set(img_request); - if (rbd_dev->parent_spec) + if (rbd_dev->parent_overlap) img_request_layered_set(img_request); spin_lock_init(&img_request->completion_lock); img_request->next_completion = 0; @@ -3613,6 +3613,7 @@ static int rbd_dev_v2_parent_info(struct rbd_device *rbd_dev) __le64 snapid; void *p; void *end; + u64 pool_id; char *image_id; u64 overlap; int ret; @@ -3643,18 +3644,19 @@ static int rbd_dev_v2_parent_info(struct rbd_device *rbd_dev) p = reply_buf; end = reply_buf + ret; ret = -ERANGE; - ceph_decode_64_safe(&p, end, parent_spec->pool_id, out_err); - if (parent_spec->pool_id == CEPH_NOPOOL) + ceph_decode_64_safe(&p, end, pool_id, out_err); + if (pool_id == CEPH_NOPOOL) goto out; /* No parent? No problem. */ /* The ceph file layout needs to fit pool id in 32 bits */ ret = -EIO; - if (parent_spec->pool_id > (u64)U32_MAX) { + if (pool_id > (u64)U32_MAX) { rbd_warn(NULL, "parent pool id too large (%llu > %u)\n", - (unsigned long long)parent_spec->pool_id, U32_MAX); + (unsigned long long)pool_id, U32_MAX); goto out_err; } + parent_spec->pool_id = pool_id; image_id = ceph_extract_encoded_string(&p, end, NULL, GFP_KERNEL); if (IS_ERR(image_id)) { @@ -3666,6 +3668,7 @@ static int rbd_dev_v2_parent_info(struct rbd_device *rbd_dev) ceph_decode_64_safe(&p, end, overlap, out_err); if (overlap) { + rbd_spec_put(rbd_dev->parent_spec); rbd_dev->parent_spec = parent_spec; parent_spec = NULL; /* rbd_dev now owns this */ rbd_dev->parent_overlap = overlap; @@ -4034,17 +4037,43 @@ static int rbd_dev_v2_header_info(struct rbd_device *rbd_dev) goto out; } + /* + * If the image supports layering, get the parent info. We + * need to probe the first time regardless. Thereafter we + * only need to if there's a parent, to see if it has + * disappeared due to the mapped image getting flattened. + */ + if (rbd_dev->header.features & RBD_FEATURE_LAYERING && + (first_time || rbd_dev->parent_spec)) { + bool warn; + + ret = rbd_dev_v2_parent_info(rbd_dev); + if (ret) + goto out; + + /* + * Print a warning if this is the initial probe and + * the image has a parent. Don't print it if the + * image now being probed is itself a parent. We + * can tell at this point because we won't know its + * pool name yet (just its pool id). + */ + warn = rbd_dev->parent_spec && rbd_dev->spec->pool_name; + if (first_time && warn) + rbd_warn(rbd_dev, "WARNING: kernel layering " + "is EXPERIMENTAL!"); + } + ret = rbd_dev_v2_image_size(rbd_dev); if (ret) goto out; + if (rbd_dev->spec->snap_id == CEPH_NOSNAP) if (rbd_dev->mapping.size != rbd_dev->header.image_size) rbd_dev->mapping.size = rbd_dev->header.image_size; ret = rbd_dev_v2_snap_context(rbd_dev); dout("rbd_dev_v2_snap_context returned %d\n", ret); - if (ret) - goto out; out: up_write(&rbd_dev->header_rwsem); @@ -4498,24 +4527,6 @@ static int rbd_dev_v2_header_onetime(struct rbd_device *rbd_dev) if (ret) goto out_err; - /* If the image supports layering, get the parent info */ - - if (rbd_dev->header.features & RBD_FEATURE_LAYERING) { - ret = rbd_dev_v2_parent_info(rbd_dev); - if (ret) - goto out_err; - /* - * Print a warning if this image has a parent. - * Don't print it if the image now being probed - * is itself a parent. We can tell at this point - * because we won't know its pool name yet (just its - * pool id). - */ - if (rbd_dev->parent_spec && rbd_dev->spec->pool_name) - rbd_warn(rbd_dev, "WARNING: kernel layering " - "is EXPERIMENTAL!"); - } - /* If the image supports fancy striping, get its parameters */ if (rbd_dev->header.features & RBD_FEATURE_STRIPINGV2) { @@ -4527,11 +4538,7 @@ static int rbd_dev_v2_header_onetime(struct rbd_device *rbd_dev) return 0; out_err: - rbd_dev->parent_overlap = 0; - rbd_spec_put(rbd_dev->parent_spec); - rbd_dev->parent_spec = NULL; - kfree(rbd_dev->header_name); - rbd_dev->header_name = NULL; + rbd_dev->header.features = 0; kfree(rbd_dev->header.object_prefix); rbd_dev->header.object_prefix = NULL; From 8785b1d487f0a31afd2c802499786d3b355eccea Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Thu, 9 May 2013 10:08:49 -0500 Subject: [PATCH 121/198] rbd: don't release write request until necessary Previously when a layered write was going to involve a copyup request, the original osd request was released before submitting the parent full-object read. The osd request for the copyup would then be allocated in rbd_img_obj_parent_read_full_callback(). Shortly we will be handling the event of mapped layered images getting flattened, and when that occurs we need to resubmit the original request. We therefore don't want to release the osd request until we really konw we're going to replace it--in the callback function. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 19 +++++++------------ 1 file changed, 7 insertions(+), 12 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index fcef63c2c30b..d861c71b4005 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -2194,13 +2194,17 @@ rbd_img_obj_parent_read_full_callback(struct rbd_img_request *img_request) if (result) goto out_err; - /* Allocate the new copyup osd request for the original request */ - + /* + * The original osd request is of no use to use any more. + * We need a new one that can hold the two ops in a copyup + * request. Allocate the new copyup osd request for the + * original request, and release the old one. + */ result = -ENOMEM; - rbd_assert(!orig_request->osd_req); osd_req = rbd_osd_req_create_copyup(orig_request); if (!osd_req) goto out_err; + rbd_osd_req_destroy(orig_request->osd_req); orig_request->osd_req = osd_req; orig_request->copyup_pages = pages; orig_request->copyup_page_count = page_count; @@ -2276,15 +2280,6 @@ static int rbd_img_obj_parent_read_full(struct rbd_obj_request *obj_request) rbd_dev = img_request->rbd_dev; rbd_assert(rbd_dev->parent != NULL); - /* - * First things first. The original osd request is of no - * use to use any more, we'll need a new one that can hold - * the two ops in a copyup request. We'll get that later, - * but for now we can release the old one. - */ - rbd_osd_req_destroy(obj_request->osd_req); - obj_request->osd_req = NULL; - /* * Determine the byte range covered by the object in the * child image to which the original request was to be sent. From fb65d2284c117cfc28d30217d25a14a8e7a75a94 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Wed, 8 May 2013 22:50:04 -0500 Subject: [PATCH 122/198] rbd: define rbd_dev_unparent() Define rbd_dev_unparent() to encapsulate cleaning up parent data structures from a layered rbd image. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 18 +++++++++++++----- 1 file changed, 13 insertions(+), 5 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index d861c71b4005..9c2b20a88be2 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -1837,6 +1837,17 @@ static void rbd_obj_request_destroy(struct kref *kref) kmem_cache_free(rbd_obj_request_cache, obj_request); } +/* It's OK to call this for a device with no parent */ + +static void rbd_spec_put(struct rbd_spec *spec); +static void rbd_dev_unparent(struct rbd_device *rbd_dev) +{ + rbd_dev_remove_parent(rbd_dev); + rbd_spec_put(rbd_dev->parent_spec); + rbd_dev->parent_spec = NULL; + rbd_dev->parent_overlap = 0; +} + /* * Caller is responsible for filling in the list of object requests * that comprises the image request, and the Linux request pointer @@ -4491,10 +4502,7 @@ static void rbd_dev_unprobe(struct rbd_device *rbd_dev) { struct rbd_image_header *header; - rbd_dev_remove_parent(rbd_dev); - rbd_spec_put(rbd_dev->parent_spec); - rbd_dev->parent_spec = NULL; - rbd_dev->parent_overlap = 0; + rbd_dev_unparent(rbd_dev); /* Free dynamic fields from the header, then zero it out */ @@ -4570,7 +4578,7 @@ static int rbd_dev_probe_parent(struct rbd_device *rbd_dev) return 0; out_err: if (parent) { - rbd_spec_put(rbd_dev->parent_spec); + rbd_dev_unparent(rbd_dev); kfree(rbd_dev->header_name); rbd_dev_destroy(parent); } else { From e93f3152357ca75284284bef8eeea7d45fe1bab1 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Wed, 8 May 2013 22:50:04 -0500 Subject: [PATCH 123/198] rbd: define parent image request routines Define rbd_parent_request_create() and rbd_parent_request_destroy() to handle the creation of parent image requests submitted for layered image objects. For simplicity, let rbd_img_request_put() handle dropping the reference to any image request (parent or not), and call whichever destructor is appropriate on the last put. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 78 ++++++++++++++++++++++++++++++++------------- 1 file changed, 55 insertions(+), 23 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 9c2b20a88be2..1ffdfbfbf3c4 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -1359,13 +1359,18 @@ static void rbd_obj_request_put(struct rbd_obj_request *obj_request) kref_put(&obj_request->kref, rbd_obj_request_destroy); } +static bool img_request_child_test(struct rbd_img_request *img_request); +static void rbd_parent_request_destroy(struct kref *kref); static void rbd_img_request_destroy(struct kref *kref); static void rbd_img_request_put(struct rbd_img_request *img_request) { rbd_assert(img_request != NULL); dout("%s: img %p (was %d)\n", __func__, img_request, atomic_read(&img_request->kref.refcount)); - kref_put(&img_request->kref, rbd_img_request_destroy); + if (img_request_child_test(img_request)) + kref_put(&img_request->kref, rbd_parent_request_destroy); + else + kref_put(&img_request->kref, rbd_img_request_destroy); } static inline void rbd_img_obj_request_add(struct rbd_img_request *img_request, @@ -1482,6 +1487,12 @@ static void img_request_child_set(struct rbd_img_request *img_request) smp_mb(); } +static void img_request_child_clear(struct rbd_img_request *img_request) +{ + clear_bit(IMG_REQ_CHILD, &img_request->flags); + smp_mb(); +} + static bool img_request_child_test(struct rbd_img_request *img_request) { smp_mb(); @@ -1856,8 +1867,7 @@ static void rbd_dev_unparent(struct rbd_device *rbd_dev) static struct rbd_img_request *rbd_img_request_create( struct rbd_device *rbd_dev, u64 offset, u64 length, - bool write_request, - bool child_request) + bool write_request) { struct rbd_img_request *img_request; @@ -1882,8 +1892,6 @@ static struct rbd_img_request *rbd_img_request_create( } else { img_request->snap_id = rbd_dev->spec->snap_id; } - if (child_request) - img_request_child_set(img_request); if (rbd_dev->parent_overlap) img_request_layered_set(img_request); spin_lock_init(&img_request->completion_lock); @@ -1918,12 +1926,46 @@ static void rbd_img_request_destroy(struct kref *kref) if (img_request_write_test(img_request)) ceph_put_snap_context(img_request->snapc); - if (img_request_child_test(img_request)) - rbd_obj_request_put(img_request->obj_request); - kmem_cache_free(rbd_img_request_cache, img_request); } +static struct rbd_img_request *rbd_parent_request_create( + struct rbd_obj_request *obj_request, + u64 img_offset, u64 length) +{ + struct rbd_img_request *parent_request; + struct rbd_device *rbd_dev; + + rbd_assert(obj_request->img_request); + rbd_dev = obj_request->img_request->rbd_dev; + + parent_request = rbd_img_request_create(rbd_dev->parent, + img_offset, length, false); + if (!parent_request) + return NULL; + + img_request_child_set(parent_request); + rbd_obj_request_get(obj_request); + parent_request->obj_request = obj_request; + + return parent_request; +} + +static void rbd_parent_request_destroy(struct kref *kref) +{ + struct rbd_img_request *parent_request; + struct rbd_obj_request *orig_request; + + parent_request = container_of(kref, struct rbd_img_request, kref); + orig_request = parent_request->obj_request; + + parent_request->obj_request = NULL; + rbd_obj_request_put(orig_request); + img_request_child_clear(parent_request); + + rbd_img_request_destroy(kref); +} + static bool rbd_img_obj_end_request(struct rbd_obj_request *obj_request) { struct rbd_img_request *img_request; @@ -2321,13 +2363,10 @@ static int rbd_img_obj_parent_read_full(struct rbd_obj_request *obj_request) } result = -ENOMEM; - parent_request = rbd_img_request_create(rbd_dev->parent, - img_offset, length, - false, true); + parent_request = rbd_parent_request_create(obj_request, + img_offset, length); if (!parent_request) goto out_err; - rbd_obj_request_get(obj_request); - parent_request->obj_request = obj_request; result = rbd_img_request_fill(parent_request, OBJ_REQUEST_PAGES, pages); if (result) @@ -2580,7 +2619,6 @@ out: static void rbd_img_parent_read(struct rbd_obj_request *obj_request) { - struct rbd_device *rbd_dev; struct rbd_img_request *img_request; int result; @@ -2589,20 +2627,14 @@ static void rbd_img_parent_read(struct rbd_obj_request *obj_request) rbd_assert(obj_request->result == (s32) -ENOENT); rbd_assert(obj_request_type_valid(obj_request->type)); - rbd_dev = obj_request->img_request->rbd_dev; - rbd_assert(rbd_dev->parent != NULL); /* rbd_read_finish(obj_request, obj_request->length); */ - img_request = rbd_img_request_create(rbd_dev->parent, + img_request = rbd_parent_request_create(obj_request, obj_request->img_offset, - obj_request->length, - false, true); + obj_request->length); result = -ENOMEM; if (!img_request) goto out_err; - rbd_obj_request_get(obj_request); - img_request->obj_request = obj_request; - if (obj_request->type == OBJ_REQUEST_BIO) result = rbd_img_request_fill(img_request, OBJ_REQUEST_BIO, obj_request->bio_list); @@ -2913,7 +2945,7 @@ static void rbd_request_fn(struct request_queue *q) result = -ENOMEM; img_request = rbd_img_request_create(rbd_dev, offset, length, - write_request, false); + write_request); if (!img_request) goto end_request; From a2acd00e7964dbb1668a5956c9d0a4bdeb838c4a Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Wed, 8 May 2013 22:50:04 -0500 Subject: [PATCH 124/198] rbd: reference count parent requests Keep a reference count for uses of the parent information for an rbd device. An initial reference is set in rbd_img_request_create() if the target image has a parent (with non-zero overlap). Each image request for an image with a non-zero parent overlap gets another reference when it's created, and that reference is dropped when the request is destroyed. The initial reference is dropped when the image gets torn down. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 104 +++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 102 insertions(+), 2 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 1ffdfbfbf3c4..8b6091b6d5cb 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -55,6 +55,39 @@ #define SECTOR_SHIFT 9 #define SECTOR_SIZE (1ULL << SECTOR_SHIFT) +/* + * Increment the given counter and return its updated value. + * If the counter is already 0 it will not be incremented. + * If the counter is already at its maximum value returns + * -EINVAL without updating it. + */ +static int atomic_inc_return_safe(atomic_t *v) +{ + unsigned int counter; + + counter = (unsigned int)__atomic_add_unless(v, 1, 0); + if (counter <= (unsigned int)INT_MAX) + return (int)counter; + + atomic_dec(v); + + return -EINVAL; +} + +/* Decrement the counter. Return the resulting value, or -EINVAL */ +static int atomic_dec_return_safe(atomic_t *v) +{ + int counter; + + counter = atomic_dec_return(v); + if (counter >= 0) + return counter; + + atomic_inc(v); + + return -EINVAL; +} + #define RBD_DRV_NAME "rbd" #define RBD_DRV_NAME_LONG "rbd (rados block device)" @@ -312,6 +345,7 @@ struct rbd_device { struct rbd_spec *parent_spec; u64 parent_overlap; + atomic_t parent_ref; struct rbd_device *parent; /* protects updating the header */ @@ -361,6 +395,7 @@ static ssize_t rbd_add(struct bus_type *bus, const char *buf, static ssize_t rbd_remove(struct bus_type *bus, const char *buf, size_t count); static int rbd_dev_image_probe(struct rbd_device *rbd_dev, bool mapping); +static void rbd_spec_put(struct rbd_spec *spec); static struct bus_attribute rbd_bus_attrs[] = { __ATTR(add, S_IWUSR, NULL, rbd_add), @@ -1505,6 +1540,12 @@ static void img_request_layered_set(struct rbd_img_request *img_request) smp_mb(); } +static void img_request_layered_clear(struct rbd_img_request *img_request) +{ + clear_bit(IMG_REQ_LAYERED, &img_request->flags); + smp_mb(); +} + static bool img_request_layered_test(struct rbd_img_request *img_request) { smp_mb(); @@ -1859,6 +1900,58 @@ static void rbd_dev_unparent(struct rbd_device *rbd_dev) rbd_dev->parent_overlap = 0; } +/* + * Parent image reference counting is used to determine when an + * image's parent fields can be safely torn down--after there are no + * more in-flight requests to the parent image. When the last + * reference is dropped, cleaning them up is safe. + */ +static void rbd_dev_parent_put(struct rbd_device *rbd_dev) +{ + int counter; + + if (!rbd_dev->parent_spec) + return; + + counter = atomic_dec_return_safe(&rbd_dev->parent_ref); + if (counter > 0) + return; + + /* Last reference; clean up parent data structures */ + + if (!counter) + rbd_dev_unparent(rbd_dev); + else + rbd_warn(rbd_dev, "parent reference underflow\n"); +} + +/* + * If an image has a non-zero parent overlap, get a reference to its + * parent. + * + * Returns true if the rbd device has a parent with a non-zero + * overlap and a reference for it was successfully taken, or + * false otherwise. + */ +static bool rbd_dev_parent_get(struct rbd_device *rbd_dev) +{ + int counter; + + if (!rbd_dev->parent_spec) + return false; + + counter = atomic_inc_return_safe(&rbd_dev->parent_ref); + if (counter > 0 && rbd_dev->parent_overlap) + return true; + + /* Image was flattened, but parent is not yet torn down */ + + if (counter < 0) + rbd_warn(rbd_dev, "parent reference overflow\n"); + + return false; +} + /* * Caller is responsible for filling in the list of object requests * that comprises the image request, and the Linux request pointer @@ -1892,7 +1985,7 @@ static struct rbd_img_request *rbd_img_request_create( } else { img_request->snap_id = rbd_dev->spec->snap_id; } - if (rbd_dev->parent_overlap) + if (rbd_dev_parent_get(rbd_dev)) img_request_layered_set(img_request); spin_lock_init(&img_request->completion_lock); img_request->next_completion = 0; @@ -1923,6 +2016,11 @@ static void rbd_img_request_destroy(struct kref *kref) rbd_img_obj_request_del(img_request, obj_request); rbd_assert(img_request->obj_request_count == 0); + if (img_request_layered_test(img_request)) { + img_request_layered_clear(img_request); + rbd_dev_parent_put(img_request->rbd_dev); + } + if (img_request_write_test(img_request)) ceph_put_snap_context(img_request->snapc); @@ -3502,6 +3600,7 @@ static struct rbd_device *rbd_dev_create(struct rbd_client *rbdc, spin_lock_init(&rbd_dev->lock); rbd_dev->flags = 0; + atomic_set(&rbd_dev->parent_ref, 0); INIT_LIST_HEAD(&rbd_dev->node); init_rwsem(&rbd_dev->header_rwsem); @@ -4534,7 +4633,7 @@ static void rbd_dev_unprobe(struct rbd_device *rbd_dev) { struct rbd_image_header *header; - rbd_dev_unparent(rbd_dev); + rbd_dev_parent_put(rbd_dev); /* Free dynamic fields from the header, then zero it out */ @@ -4606,6 +4705,7 @@ static int rbd_dev_probe_parent(struct rbd_device *rbd_dev) if (ret < 0) goto out_err; rbd_dev->parent = parent; + atomic_set(&rbd_dev->parent_ref, 1); return 0; out_err: From 392a9dad7e777296fe79d97a6b3acd735ad2eb5f Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 17:40:33 -0500 Subject: [PATCH 125/198] rbd: detect when clone image is flattened A format 2 clone image can be the subject of a "flatten" operation, during which all of its data gets "copied up" from its parent image, leaving the image fully populated. Once this is complete, the clone's association with the parent is abolished. Since this can occur when a clone is mapped, we need to detect when it has occurred and handle it accordingly. We know an image has been flattened when we know it at one time had a parent, but we have learned (via a "get_parent" object class method call) it no longer has one. There might be in-flight requests at the point we learn an image has been flattened, so we can't simply clean up parent data structures right away. Instead, we'll drop the initial parent reference when the parent has disappeared (rather than when the image gets destroyed), which will allow the last in-flight reference to clean things up when it's complete. We leverage the fact that a zero parent overlap renders an image effectively unlayered. We set the overlap to 0 at the point we detect the clone image has flattened, which allows the unlayered behavior to take effect immediately, while keeping other parent structures in place until in-flight requests to complete. This and the next few patches resolve: http://tracker.ceph.com/issues/3763 Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 30 ++++++++++++++++++++++++++++-- 1 file changed, 28 insertions(+), 2 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 8b6091b6d5cb..9717e20f3477 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -1929,6 +1929,11 @@ static void rbd_dev_parent_put(struct rbd_device *rbd_dev) * If an image has a non-zero parent overlap, get a reference to its * parent. * + * We must get the reference before checking for the overlap to + * coordinate properly with zeroing the parent overlap in + * rbd_dev_v2_parent_info() when an image gets flattened. We + * drop it again if there is no overlap. + * * Returns true if the rbd device has a parent with a non-zero * overlap and a reference for it was successfully taken, or * false otherwise. @@ -3782,8 +3787,26 @@ static int rbd_dev_v2_parent_info(struct rbd_device *rbd_dev) end = reply_buf + ret; ret = -ERANGE; ceph_decode_64_safe(&p, end, pool_id, out_err); - if (pool_id == CEPH_NOPOOL) + if (pool_id == CEPH_NOPOOL) { + /* + * Either the parent never existed, or we have + * record of it but the image got flattened so it no + * longer has a parent. When the parent of a + * layered image disappears we immediately set the + * overlap to 0. The effect of this is that all new + * requests will be treated as if the image had no + * parent. + */ + if (rbd_dev->parent_overlap) { + rbd_dev->parent_overlap = 0; + smp_mb(); + rbd_dev_parent_put(rbd_dev); + pr_info("%s: clone image has been flattened\n", + rbd_dev->disk->disk_name); + } + goto out; /* No parent? No problem. */ + } /* The ceph file layout needs to fit pool id in 32 bits */ @@ -4633,7 +4656,10 @@ static void rbd_dev_unprobe(struct rbd_device *rbd_dev) { struct rbd_image_header *header; - rbd_dev_parent_put(rbd_dev); + /* Drop parent reference unless it's already been done (or none) */ + + if (rbd_dev->parent_overlap) + rbd_dev_parent_put(rbd_dev); /* Free dynamic fields from the header, then zero it out */ From 02c74fbad9d4a5149756eb35be7300736e4904e9 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 17:40:33 -0500 Subject: [PATCH 126/198] rbd: re-submit read request for flattened clone If a clone image gets flattened while a parent read request is underway, the original rbd object request needs to be resubmitted. The reason is that by the time we get the response to the parent read request, the data read from the parent may be out of date. In other words, we could see this sequence of events: rbd client parent image/osd ---------- ---------------- original object ENOENT; issue parent read respond to parent read child image flattened original image header refresh <--- original object written independently here parent read response received Add code to rbd_img_parent_read_callback() to detect when a clone's parent image has disappeared (as evidenced by its parent overlap becoming 0), and re-submit the original read request in that case. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 30 +++++++++++++++++++++++++----- 1 file changed, 25 insertions(+), 5 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 9717e20f3477..4edcb6d85f01 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -2682,14 +2682,36 @@ static void rbd_img_parent_read_callback(struct rbd_img_request *img_request) struct rbd_obj_request *obj_request; struct rbd_device *rbd_dev; u64 obj_end; + u64 img_xferred; + int img_result; rbd_assert(img_request_child_test(img_request)); + /* First get what we need from the image request and release it */ + obj_request = img_request->obj_request; + img_xferred = img_request->xferred; + img_result = img_request->result; + rbd_img_request_put(img_request); + + /* + * If the overlap has become 0 (most likely because the + * image has been flattened) we need to re-submit the + * original request. + */ rbd_assert(obj_request); rbd_assert(obj_request->img_request); + rbd_dev = obj_request->img_request->rbd_dev; + if (!rbd_dev->parent_overlap) { + struct ceph_osd_client *osdc; - obj_request->result = img_request->result; + osdc = &rbd_dev->rbd_client->client->osdc; + img_result = rbd_obj_request_submit(osdc, obj_request); + if (!img_result) + return; + } + + obj_request->result = img_result; if (obj_request->result) goto out; @@ -2702,7 +2724,6 @@ static void rbd_img_parent_read_callback(struct rbd_img_request *img_request) */ rbd_assert(obj_request->img_offset < U64_MAX - obj_request->length); obj_end = obj_request->img_offset + obj_request->length; - rbd_dev = obj_request->img_request->rbd_dev; if (obj_end > rbd_dev->parent_overlap) { u64 xferred = 0; @@ -2710,12 +2731,11 @@ static void rbd_img_parent_read_callback(struct rbd_img_request *img_request) xferred = rbd_dev->parent_overlap - obj_request->img_offset; - obj_request->xferred = min(img_request->xferred, xferred); + obj_request->xferred = min(img_xferred, xferred); } else { - obj_request->xferred = img_request->xferred; + obj_request->xferred = img_xferred; } out: - rbd_img_request_put(img_request); rbd_img_obj_request_read_callback(obj_request); rbd_obj_request_complete(obj_request); } From bbea1c1a31b1128d34b2b5b4f28276969cce14e9 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 17:40:33 -0500 Subject: [PATCH 127/198] rbd: re-submit write request for flattened clone Add code to rbd_img_parent_read_full_callback() to detect when a clone's parent image has disappeared, and re-submit the original write request in that case. (See the previous commit for more reasoning about why this is appropriate.) Rename some variables in rbd_img_obj_parent_read_full_callback() to match the convention used in the previous patch. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 29 ++++++++++++++++++++++------- 1 file changed, 22 insertions(+), 7 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 4edcb6d85f01..5c2731859e8a 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -2319,7 +2319,7 @@ rbd_img_obj_parent_read_full_callback(struct rbd_img_request *img_request) struct rbd_device *rbd_dev; struct page **pages; u32 page_count; - int result; + int img_result; u64 parent_length; u64 offset; u64 length; @@ -2338,7 +2338,7 @@ rbd_img_obj_parent_read_full_callback(struct rbd_img_request *img_request) orig_request = img_request->obj_request; rbd_assert(orig_request != NULL); rbd_assert(obj_request_type_valid(orig_request->type)); - result = img_request->result; + img_result = img_request->result; parent_length = img_request->length; rbd_assert(parent_length == img_request->xferred); rbd_img_request_put(img_request); @@ -2347,7 +2347,22 @@ rbd_img_obj_parent_read_full_callback(struct rbd_img_request *img_request) rbd_dev = orig_request->img_request->rbd_dev; rbd_assert(rbd_dev); - if (result) + /* + * If the overlap has become 0 (most likely because the + * image has been flattened) we need to free the pages + * and re-submit the original write request. + */ + if (!rbd_dev->parent_overlap) { + struct ceph_osd_client *osdc; + + ceph_release_page_vector(pages, page_count); + osdc = &rbd_dev->rbd_client->client->osdc; + img_result = rbd_obj_request_submit(osdc, orig_request); + if (!img_result) + return; + } + + if (img_result) goto out_err; /* @@ -2356,7 +2371,7 @@ rbd_img_obj_parent_read_full_callback(struct rbd_img_request *img_request) * request. Allocate the new copyup osd request for the * original request, and release the old one. */ - result = -ENOMEM; + img_result = -ENOMEM; osd_req = rbd_osd_req_create_copyup(orig_request); if (!osd_req) goto out_err; @@ -2391,13 +2406,13 @@ rbd_img_obj_parent_read_full_callback(struct rbd_img_request *img_request) orig_request->callback = rbd_img_obj_copyup_callback; osdc = &rbd_dev->rbd_client->client->osdc; - result = rbd_obj_request_submit(osdc, orig_request); - if (!result) + img_result = rbd_obj_request_submit(osdc, orig_request); + if (!img_result) return; out_err: /* Record the error code and complete the request */ - orig_request->result = result; + orig_request->result = img_result; orig_request->xferred = 0; obj_request_done_set(orig_request); rbd_obj_request_complete(orig_request); From 638f5abed3f7d8a7fc24087bd760fa3d99f68a39 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 6 May 2013 17:40:33 -0500 Subject: [PATCH 128/198] rbd: re-submit flattened write request (part 2) Add code to rbd_img_obj_exists_callback() to detect when a clone's parent image has disappeared, and re-submit the original write request in that case. Kill off some redundant assertions. This completes the resolution for: http://tracker.ceph.com/issues/3763 Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- drivers/block/rbd.c | 18 ++++++++++++++++-- 1 file changed, 16 insertions(+), 2 deletions(-) diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 5c2731859e8a..6b872f219774 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -2516,6 +2516,7 @@ out_err: static void rbd_img_obj_exists_callback(struct rbd_obj_request *obj_request) { struct rbd_obj_request *orig_request; + struct rbd_device *rbd_dev; int result; rbd_assert(!obj_request_img_data_test(obj_request)); @@ -2538,8 +2539,21 @@ static void rbd_img_obj_exists_callback(struct rbd_obj_request *obj_request) obj_request->xferred, obj_request->length); rbd_obj_request_put(obj_request); - rbd_assert(orig_request); - rbd_assert(orig_request->img_request); + /* + * If the overlap has become 0 (most likely because the + * image has been flattened) we need to free the pages + * and re-submit the original write request. + */ + rbd_dev = orig_request->img_request->rbd_dev; + if (!rbd_dev->parent_overlap) { + struct ceph_osd_client *osdc; + + rbd_obj_request_put(orig_request); + osdc = &rbd_dev->rbd_client->client->osdc; + result = rbd_obj_request_submit(osdc, orig_request); + if (!result) + return; + } /* * Our only purpose here is to determine whether the object From 6880b0150a7c25fd75c5ece80abc49ebf53c38c1 Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Sat, 11 May 2013 19:04:09 +0000 Subject: [PATCH 129/198] parisc: make default cross compiler search more robust (v3) People/distros vary how they prefix the toolchain name for 64bit builds. Rather than enforce one convention over another, add a for loop which does a search for all the general prefixes. For 64bit builds, we now search for (in order): hppa64-unknown-linux-gnu hppa64-linux-gnu hppa64-linux For 32bit builds, we look for: hppa-unknown-linux-gnu hppa-linux-gnu hppa-linux hppa2.0-unknown-linux-gnu hppa2.0-linux-gnu hppa2.0-linux hppa1.1-unknown-linux-gnu hppa1.1-linux-gnu hppa1.1-linux This patch was initiated by Mike Frysinger, with feedback from Jeroen Roovers, John David Anglin and Helge Deller. Signed-off-by: Mike Frysinger Signed-off-by: Jeroen Roovers Signed-off-by: John David Anglin Signed-off-by: Helge Deller --- arch/parisc/Makefile | 21 +++++++++------------ 1 file changed, 9 insertions(+), 12 deletions(-) diff --git a/arch/parisc/Makefile b/arch/parisc/Makefile index 2f967cc6649e..197690068f88 100644 --- a/arch/parisc/Makefile +++ b/arch/parisc/Makefile @@ -23,24 +23,21 @@ NM = sh $(srctree)/arch/parisc/nm CHECKFLAGS += -D__hppa__=1 LIBGCC = $(shell $(CC) $(KBUILD_CFLAGS) -print-libgcc-file-name) -MACHINE := $(shell uname -m) -NATIVE := $(if $(filter parisc%,$(MACHINE)),1,0) - ifdef CONFIG_64BIT UTS_MACHINE := parisc64 CHECKFLAGS += -D__LP64__=1 -m64 -WIDTH := 64 +CC_ARCHES = hppa64 else # 32-bit -WIDTH := +CC_ARCHES = hppa hppa2.0 hppa1.1 endif -# attempt to help out folks who are cross-compiling -ifeq ($(NATIVE),1) -CROSS_COMPILE := hppa$(WIDTH)-linux- -else - ifeq ($(CROSS_COMPILE),) - CROSS_COMPILE := hppa$(WIDTH)-linux-gnu- - endif +ifneq ($(SUBARCH),$(UTS_MACHINE)) + ifeq ($(CROSS_COMPILE),) + CC_SUFFIXES = linux linux-gnu unknown-linux-gnu + CROSS_COMPILE := $(call cc-cross-prefix, \ + $(foreach a,$(CC_ARCHES), \ + $(foreach s,$(CC_SUFFIXES),$(a)-$(s)-))) + endif endif OBJCOPY_FLAGS =-O binary -R .note -R .comment -S From 6eabb3301b1facee669d9938f7c5a0295c21d71d Mon Sep 17 00:00:00 2001 From: Jaccon Bastiaansen Date: Mon, 13 May 2013 17:28:27 +0100 Subject: [PATCH 130/198] ARM: 7720/1: ARM v6/v7 cmpxchg64 shouldn't clear upper 32 bits of the old/new value The implementation of cmpxchg64() for the ARM v6 and v7 architecture casts parameter 2 and 3 (the old and new 64bit values) to an unsigned long before calling the atomic_cmpxchg64() function. This clears the top 32 bits of the old and new values, resulting in the wrong values being compare-exchanged. Luckily, this only appears to be used for 64-bit sched_clock, which we don't (yet) have on ARM. This bug was introduced by commit 3e0f5a15f500 ("ARM: 7404/1: cmpxchg64: use atomic64 and local64 routines for cmpxchg64"). Cc: Acked-by: Will Deacon Signed-off-by: Jaccon Bastiaansen Signed-off-by: Russell King --- arch/arm/include/asm/cmpxchg.h | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/arch/arm/include/asm/cmpxchg.h b/arch/arm/include/asm/cmpxchg.h index 7eb18c1d8d6c..4f009c10540d 100644 --- a/arch/arm/include/asm/cmpxchg.h +++ b/arch/arm/include/asm/cmpxchg.h @@ -233,15 +233,15 @@ static inline unsigned long __cmpxchg_local(volatile void *ptr, ((__typeof__(*(ptr)))atomic64_cmpxchg(container_of((ptr), \ atomic64_t, \ counter), \ - (unsigned long)(o), \ - (unsigned long)(n))) + (unsigned long long)(o), \ + (unsigned long long)(n))) #define cmpxchg64_local(ptr, o, n) \ ((__typeof__(*(ptr)))local64_cmpxchg(container_of((ptr), \ local64_t, \ a), \ - (unsigned long)(o), \ - (unsigned long)(n))) + (unsigned long long)(o), \ + (unsigned long long)(n))) #endif /* __LINUX_ARM_ARCH__ >= 6 */ From 613e60a66181acc5c8b63268e939622162da2393 Mon Sep 17 00:00:00 2001 From: "Aneesh Kumar K.V" Date: Sat, 11 May 2013 22:33:19 +0000 Subject: [PATCH 131/198] powerpc/mm: Use the correct mask value when looking at pgtable address Our pgtable are 2*sizeof(pte_t)*PTRS_PER_PTE which is PTE_FRAG_SIZE. Instead of depending on frag size, mask with PMD_MASKED_BITS. Signed-off-by: Aneesh Kumar K.V Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/include/asm/pgalloc-64.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/powerpc/include/asm/pgalloc-64.h b/arch/powerpc/include/asm/pgalloc-64.h index 91acb12bac92..b66ae722a8e9 100644 --- a/arch/powerpc/include/asm/pgalloc-64.h +++ b/arch/powerpc/include/asm/pgalloc-64.h @@ -186,7 +186,7 @@ static inline void pmd_populate(struct mm_struct *mm, pmd_t *pmd, static inline pgtable_t pmd_pgtable(pmd_t pmd) { - return (pgtable_t)(pmd_val(pmd) & -sizeof(pte_t)*PTRS_PER_PTE); + return (pgtable_t)(pmd_val(pmd) & ~PMD_MASKED_BITS); } static inline pte_t *pte_alloc_one_kernel(struct mm_struct *mm, From 83d5e64b7efa7f39b10ff5e92792e807a720289c Mon Sep 17 00:00:00 2001 From: "Aneesh Kumar K.V" Date: Mon, 6 May 2013 10:51:00 +0000 Subject: [PATCH 132/198] powerpc: Fix build errors STRICT_MM_TYPECHECKS Signed-off-by: Aneesh Kumar K.V Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/include/asm/pte-hash64-64k.h | 2 +- arch/powerpc/kernel/pci-common.c | 5 ++--- arch/powerpc/mm/init_64.c | 3 ++- 3 files changed, 5 insertions(+), 5 deletions(-) diff --git a/arch/powerpc/include/asm/pte-hash64-64k.h b/arch/powerpc/include/asm/pte-hash64-64k.h index 3e13e23e4fdf..d836d945068d 100644 --- a/arch/powerpc/include/asm/pte-hash64-64k.h +++ b/arch/powerpc/include/asm/pte-hash64-64k.h @@ -47,7 +47,7 @@ * generic accessors and iterators here */ #define __real_pte(e,p) ((real_pte_t) { \ - (e), ((e) & _PAGE_COMBO) ? \ + (e), (pte_val(e) & _PAGE_COMBO) ? \ (pte_val(*((p) + PTRS_PER_PTE))) : 0 }) #define __rpte_to_hidx(r,index) ((pte_val((r).pte) & _PAGE_COMBO) ? \ (((r).hidx >> ((index)<<2)) & 0xf) : ((pte_val((r).pte) >> 12) & 0xf)) diff --git a/arch/powerpc/kernel/pci-common.c b/arch/powerpc/kernel/pci-common.c index f5c5c90799a7..6053f037ef0a 100644 --- a/arch/powerpc/kernel/pci-common.c +++ b/arch/powerpc/kernel/pci-common.c @@ -359,7 +359,6 @@ static pgprot_t __pci_mmap_set_pgprot(struct pci_dev *dev, struct resource *rp, enum pci_mmap_state mmap_state, int write_combine) { - unsigned long prot = pgprot_val(protection); /* Write combine is always 0 on non-memory space mappings. On * memory space, if the user didn't pass 1, we check for a @@ -376,9 +375,9 @@ static pgprot_t __pci_mmap_set_pgprot(struct pci_dev *dev, struct resource *rp, /* XXX would be nice to have a way to ask for write-through */ if (write_combine) - return pgprot_noncached_wc(prot); + return pgprot_noncached_wc(protection); else - return pgprot_noncached(prot); + return pgprot_noncached(protection); } /* diff --git a/arch/powerpc/mm/init_64.c b/arch/powerpc/mm/init_64.c index c2787bf779ca..a90b9c458990 100644 --- a/arch/powerpc/mm/init_64.c +++ b/arch/powerpc/mm/init_64.c @@ -215,7 +215,8 @@ static void __meminit vmemmap_create_mapping(unsigned long start, unsigned long phys) { int mapped = htab_bolt_mapping(start, start + page_size, phys, - PAGE_KERNEL, mmu_vmemmap_psize, + pgprot_val(PAGE_KERNEL), + mmu_vmemmap_psize, mmu_kernel_ssize); BUG_ON(mapped < 0); } From 79c66ce8f6448a3295a32efeac88c9debd7f7094 Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Sun, 12 May 2013 15:04:53 +0000 Subject: [PATCH 133/198] powerpc/kexec: Fix kexec when using VMX optimised memcpy commit b3f271e86e5a (powerpc: POWER7 optimised memcpy using VMX and enhanced prefetch) uses VMX when it is safe to do so (ie not in interrupt). It also looks at the task struct to decide if we have to save the current tasks' VMX state. kexec calls memcpy() at a point where the task struct may have been overwritten by the new kexec segments. If it has been overwritten then when memcpy -> enable_altivec looks up current->thread.regs->msr we get a cryptic oops or lockup. I also notice we aren't initialising thread_info->cpu, which means smp_processor_id is broken. Fix that too. Signed-off-by: Anton Blanchard Cc: # 3.6+ Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/kernel/machine_kexec_64.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/arch/powerpc/kernel/machine_kexec_64.c b/arch/powerpc/kernel/machine_kexec_64.c index 466a2908bb63..611acdf30096 100644 --- a/arch/powerpc/kernel/machine_kexec_64.c +++ b/arch/powerpc/kernel/machine_kexec_64.c @@ -17,6 +17,7 @@ #include #include #include +#include #include #include @@ -335,10 +336,13 @@ void default_machine_kexec(struct kimage *image) pr_debug("kexec: Starting switchover sequence.\n"); /* switch to a staticly allocated stack. Based on irq stack code. + * We setup preempt_count to avoid using VMX in memcpy. * XXX: the task struct will likely be invalid once we do the copy! */ kexec_stack.thread_info.task = current_thread_info()->task; kexec_stack.thread_info.flags = 0; + kexec_stack.thread_info.preempt_count = HARDIRQ_OFFSET; + kexec_stack.thread_info.cpu = current_thread_info()->cpu; /* We need a static PACA, too; copy this CPU's PACA over and switch to * it. Also poison per_cpu_offset to catch anyone using non-static From a94a14720eaf55c5f06d6ca7ecbe3f87f6864fc6 Mon Sep 17 00:00:00 2001 From: Vasant Hegde Date: Tue, 7 May 2013 16:54:47 +0000 Subject: [PATCH 134/198] powerpc/rtas_flash: Fix validate_flash buffer overflow issue ibm,validate-flash-image RTAS call output buffer contains 150 - 200 bytes of data on latest system. Presently we have output buffer size as 64 bytes and we use sprintf to copy data from RTAS buffer to local buffer. This causes kernel oops (see below call trace). This patch increases local buffer size to 256 and also uses snprintf instead of sprintf to copy data from RTAS buffer. Kernel call trace : ------------------- Oops: Kernel access of bad area, sig: 11 [#1] SMP NR_CPUS=1024 NUMA pSeries Modules linked in: nfs fscache lockd auth_rpcgss nfs_acl sunrpc fuse loop dm_mod ipv6 ipv6_lib usb_storage ehea(X) sr_mod qlge ses cdrom enclosure st be2net sg ext3 jbd mbcache usbhid hid ohci_hcd ehci_hcd usbcore qla2xxx usb_common sd_mod crc_t10dif scsi_dh_hp_sw scsi_dh_rdac scsi_dh_alua scsi_dh_emc scsi_dh lpfc scsi_transport_fc scsi_tgt ipr(X) libata scsi_mod Supported: Yes NIP: 4520323031333130 LR: 4520323031333130 CTR: 0000000000000000 REGS: c0000001b91779b0 TRAP: 0400 Tainted: G X (3.0.13-0.27-ppc64) MSR: 8000000040009032 CR: 44022488 XER: 20000018 TASK = c0000001bca1aba0[4736] 'cat' THREAD: c0000001b9174000 CPU: 36 GPR00: 4520323031333130 c0000001b9177c30 c000000000f87c98 000000000000009b GPR04: c0000001b9177c4a 000000000000000b 3520323031333130 2032303133313031 GPR08: 3133313031350a4d 000000000000009b 0000000000000000 c0000000003664a4 GPR12: 0000000022022448 c000000003ee6c00 0000000000000002 00000000100e8a90 GPR16: 00000000100cb9d8 0000000010093370 000000001001d310 0000000000000000 GPR20: 0000000000008000 00000000100fae60 000000000000005e 0000000000000000 GPR24: 0000000010129350 46573738302e3030 2046573738302e30 300a4d4720323031 GPR28: 333130313520554e 4b4e4f574e0a4d47 2032303133313031 3520323031333130 NIP [4520323031333130] 0x4520323031333130 LR [4520323031333130] 0x4520323031333130 Call Trace: [c0000001b9177c30] [4520323031333130] 0x4520323031333130 (unreliable) Instruction dump: XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX Signed-off-by: Vasant Hegde Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/kernel/rtas_flash.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/arch/powerpc/kernel/rtas_flash.c b/arch/powerpc/kernel/rtas_flash.c index 5b3022470126..2f3cdb01506d 100644 --- a/arch/powerpc/kernel/rtas_flash.c +++ b/arch/powerpc/kernel/rtas_flash.c @@ -89,6 +89,7 @@ /* Array sizes */ #define VALIDATE_BUF_SIZE 4096 +#define VALIDATE_MSG_LEN 256 #define RTAS_MSG_MAXLEN 64 /* Quirk - RTAS requires 4k list length and block size */ @@ -466,7 +467,7 @@ static void validate_flash(struct rtas_validate_flash_t *args_buf) } static int get_validate_flash_msg(struct rtas_validate_flash_t *args_buf, - char *msg) + char *msg, int msglen) { int n; @@ -474,7 +475,8 @@ static int get_validate_flash_msg(struct rtas_validate_flash_t *args_buf, n = sprintf(msg, "%d\n", args_buf->update_results); if ((args_buf->update_results >= VALIDATE_CUR_UNKNOWN) || (args_buf->update_results == VALIDATE_TMP_UPDATE)) - n += sprintf(msg + n, "%s\n", args_buf->buf); + n += snprintf(msg + n, msglen - n, "%s\n", + args_buf->buf); } else { n = sprintf(msg, "%d\n", args_buf->status); } @@ -486,11 +488,11 @@ static ssize_t validate_flash_read(struct file *file, char __user *buf, { struct rtas_validate_flash_t *const args_buf = &rtas_validate_flash_data; - char msg[RTAS_MSG_MAXLEN]; + char msg[VALIDATE_MSG_LEN]; int msglen; mutex_lock(&rtas_validate_flash_mutex); - msglen = get_validate_flash_msg(args_buf, msg); + msglen = get_validate_flash_msg(args_buf, msg, VALIDATE_MSG_LEN); mutex_unlock(&rtas_validate_flash_mutex); return simple_read_from_buffer(buf, count, ppos, msg, msglen); From 120496ac2d2d60aee68d3123a68169502a85f4b5 Mon Sep 17 00:00:00 2001 From: Robert Jennings Date: Tue, 7 May 2013 04:34:11 +0000 Subject: [PATCH 135/198] powerpc: Bring all threads online prior to migration/hibernation This patch brings online all threads which are present but not online prior to migration/hibernation. After migration/hibernation those threads are taken back offline. During migration/hibernation all online CPUs must call H_JOIN, this is required by the hypervisor. Without this patch, threads that are offline (H_CEDE'd) will not be woken to make the H_JOIN call and the OS will be deadlocked (all threads either JOIN'd or CEDE'd). Cc: Signed-off-by: Robert Jennings Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/include/asm/rtas.h | 2 + arch/powerpc/kernel/rtas.c | 113 +++++++++++++++++++++++ arch/powerpc/platforms/pseries/suspend.c | 22 +++++ 3 files changed, 137 insertions(+) diff --git a/arch/powerpc/include/asm/rtas.h b/arch/powerpc/include/asm/rtas.h index a8bc2bb4adc9..34fd70488d83 100644 --- a/arch/powerpc/include/asm/rtas.h +++ b/arch/powerpc/include/asm/rtas.h @@ -264,6 +264,8 @@ extern void rtas_progress(char *s, unsigned short hex); extern void rtas_initialize(void); extern int rtas_suspend_cpu(struct rtas_suspend_me_data *data); extern int rtas_suspend_last_cpu(struct rtas_suspend_me_data *data); +extern int rtas_online_cpus_mask(cpumask_var_t cpus); +extern int rtas_offline_cpus_mask(cpumask_var_t cpus); extern int rtas_ibm_suspend_me(struct rtas_args *); struct rtc_time; diff --git a/arch/powerpc/kernel/rtas.c b/arch/powerpc/kernel/rtas.c index 1fd6e7b2f390..52add6f3e201 100644 --- a/arch/powerpc/kernel/rtas.c +++ b/arch/powerpc/kernel/rtas.c @@ -19,6 +19,7 @@ #include #include #include +#include #include #include #include @@ -807,6 +808,95 @@ static void rtas_percpu_suspend_me(void *info) __rtas_suspend_cpu((struct rtas_suspend_me_data *)info, 1); } +enum rtas_cpu_state { + DOWN, + UP, +}; + +#ifndef CONFIG_SMP +static int rtas_cpu_state_change_mask(enum rtas_cpu_state state, + cpumask_var_t cpus) +{ + if (!cpumask_empty(cpus)) { + cpumask_clear(cpus); + return -EINVAL; + } else + return 0; +} +#else +/* On return cpumask will be altered to indicate CPUs changed. + * CPUs with states changed will be set in the mask, + * CPUs with status unchanged will be unset in the mask. */ +static int rtas_cpu_state_change_mask(enum rtas_cpu_state state, + cpumask_var_t cpus) +{ + int cpu; + int cpuret = 0; + int ret = 0; + + if (cpumask_empty(cpus)) + return 0; + + for_each_cpu(cpu, cpus) { + switch (state) { + case DOWN: + cpuret = cpu_down(cpu); + break; + case UP: + cpuret = cpu_up(cpu); + break; + } + if (cpuret) { + pr_debug("%s: cpu_%s for cpu#%d returned %d.\n", + __func__, + ((state == UP) ? "up" : "down"), + cpu, cpuret); + if (!ret) + ret = cpuret; + if (state == UP) { + /* clear bits for unchanged cpus, return */ + cpumask_shift_right(cpus, cpus, cpu); + cpumask_shift_left(cpus, cpus, cpu); + break; + } else { + /* clear bit for unchanged cpu, continue */ + cpumask_clear_cpu(cpu, cpus); + } + } + } + + return ret; +} +#endif + +int rtas_online_cpus_mask(cpumask_var_t cpus) +{ + int ret; + + ret = rtas_cpu_state_change_mask(UP, cpus); + + if (ret) { + cpumask_var_t tmp_mask; + + if (!alloc_cpumask_var(&tmp_mask, GFP_TEMPORARY)) + return ret; + + /* Use tmp_mask to preserve cpus mask from first failure */ + cpumask_copy(tmp_mask, cpus); + rtas_offline_cpus_mask(tmp_mask); + free_cpumask_var(tmp_mask); + } + + return ret; +} +EXPORT_SYMBOL(rtas_online_cpus_mask); + +int rtas_offline_cpus_mask(cpumask_var_t cpus) +{ + return rtas_cpu_state_change_mask(DOWN, cpus); +} +EXPORT_SYMBOL(rtas_offline_cpus_mask); + int rtas_ibm_suspend_me(struct rtas_args *args) { long state; @@ -814,6 +904,8 @@ int rtas_ibm_suspend_me(struct rtas_args *args) unsigned long retbuf[PLPAR_HCALL_BUFSIZE]; struct rtas_suspend_me_data data; DECLARE_COMPLETION_ONSTACK(done); + cpumask_var_t offline_mask; + int cpuret; if (!rtas_service_present("ibm,suspend-me")) return -ENOSYS; @@ -837,11 +929,24 @@ int rtas_ibm_suspend_me(struct rtas_args *args) return 0; } + if (!alloc_cpumask_var(&offline_mask, GFP_TEMPORARY)) + return -ENOMEM; + atomic_set(&data.working, 0); atomic_set(&data.done, 0); atomic_set(&data.error, 0); data.token = rtas_token("ibm,suspend-me"); data.complete = &done; + + /* All present CPUs must be online */ + cpumask_andnot(offline_mask, cpu_present_mask, cpu_online_mask); + cpuret = rtas_online_cpus_mask(offline_mask); + if (cpuret) { + pr_err("%s: Could not bring present CPUs online.\n", __func__); + atomic_set(&data.error, cpuret); + goto out; + } + stop_topology_update(); /* Call function on all CPUs. One of us will make the @@ -857,6 +962,14 @@ int rtas_ibm_suspend_me(struct rtas_args *args) start_topology_update(); + /* Take down CPUs not online prior to suspend */ + cpuret = rtas_offline_cpus_mask(offline_mask); + if (cpuret) + pr_warn("%s: Could not restore CPUs to offline state.\n", + __func__); + +out: + free_cpumask_var(offline_mask); return atomic_read(&data.error); } #else /* CONFIG_PPC_PSERIES */ diff --git a/arch/powerpc/platforms/pseries/suspend.c b/arch/powerpc/platforms/pseries/suspend.c index 47226e04126d..5f997e79d570 100644 --- a/arch/powerpc/platforms/pseries/suspend.c +++ b/arch/powerpc/platforms/pseries/suspend.c @@ -16,6 +16,7 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ +#include #include #include #include @@ -126,11 +127,15 @@ static ssize_t store_hibernate(struct device *dev, struct device_attribute *attr, const char *buf, size_t count) { + cpumask_var_t offline_mask; int rc; if (!capable(CAP_SYS_ADMIN)) return -EPERM; + if (!alloc_cpumask_var(&offline_mask, GFP_TEMPORARY)) + return -ENOMEM; + stream_id = simple_strtoul(buf, NULL, 16); do { @@ -140,15 +145,32 @@ static ssize_t store_hibernate(struct device *dev, } while (rc == -EAGAIN); if (!rc) { + /* All present CPUs must be online */ + cpumask_andnot(offline_mask, cpu_present_mask, + cpu_online_mask); + rc = rtas_online_cpus_mask(offline_mask); + if (rc) { + pr_err("%s: Could not bring present CPUs online.\n", + __func__); + goto out; + } + stop_topology_update(); rc = pm_suspend(PM_SUSPEND_MEM); start_topology_update(); + + /* Take down CPUs not online prior to suspend */ + if (!rtas_offline_cpus_mask(offline_mask)) + pr_warn("%s: Could not restore CPUs to offline " + "state.\n", __func__); } stream_id = 0; if (!rc) rc = count; +out: + free_cpumask_var(offline_mask); return rc; } From b80ec3dc8e100f6c8b27c02cbb6079d401c3a054 Mon Sep 17 00:00:00 2001 From: Michael Ellerman Date: Mon, 6 May 2013 18:43:39 +0000 Subject: [PATCH 136/198] powerpc: Make CONFIG_RTAS_PROC depend on CONFIG_PROC_FS We are getting build errors with CONFIG_PROC_FS=n: arch/powerpc/kernel/rtas_flash.c In function 'rtas_flash_init': 745:33: error: unused variable 'f' [-Werror=unused-variable] But rtas_flash.c should not be built when CONFIG_PROC_FS=n, beacause all it does is provide a /proc interface to the RTAS flash routines. CONFIG_RTAS_FLASH already depends on CONFIG_RTAS_PROC, to indicate that it depends on the RTAS proc support, but CONFIG_RTAS_PROC does not depend on CONFIG_PROC_FS. So fix that. Signed-off-by: Michael Ellerman Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/platforms/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/powerpc/platforms/Kconfig b/arch/powerpc/platforms/Kconfig index 34d224be93ba..632444f15550 100644 --- a/arch/powerpc/platforms/Kconfig +++ b/arch/powerpc/platforms/Kconfig @@ -128,7 +128,7 @@ config PPC_RTAS_DAEMON config RTAS_PROC bool "Proc interface to RTAS" - depends on PPC_RTAS + depends on PPC_RTAS && PROC_FS default y config RTAS_FLASH From af945cf4bfc6586f0739b8c7f67af75576dec25e Mon Sep 17 00:00:00 2001 From: Li Zhong Date: Mon, 6 May 2013 22:44:41 +0000 Subject: [PATCH 137/198] powerpc: Fix MAX_STACK_TRACE_ENTRIES too low warning again Saw this warning again, and this time from the ret_from_fork path. It seems we could clear the back chain earlier in copy_thread(), which could cover both path, and also fix potential lockdep usage in schedule_tail(), or exception occurred before we clear the back chain. Signed-off-by: Li Zhong Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/kernel/entry_32.S | 2 -- arch/powerpc/kernel/entry_64.S | 2 -- arch/powerpc/kernel/process.c | 1 + 3 files changed, 1 insertion(+), 4 deletions(-) diff --git a/arch/powerpc/kernel/entry_32.S b/arch/powerpc/kernel/entry_32.S index e514de57a125..d22e73e4618b 100644 --- a/arch/powerpc/kernel/entry_32.S +++ b/arch/powerpc/kernel/entry_32.S @@ -439,8 +439,6 @@ ret_from_fork: ret_from_kernel_thread: REST_NVGPRS(r1) bl schedule_tail - li r3,0 - stw r3,0(r1) mtlr r14 mr r3,r15 PPC440EP_ERR42 diff --git a/arch/powerpc/kernel/entry_64.S b/arch/powerpc/kernel/entry_64.S index 3fe5259e2fea..48e8a86e7cb2 100644 --- a/arch/powerpc/kernel/entry_64.S +++ b/arch/powerpc/kernel/entry_64.S @@ -376,8 +376,6 @@ _GLOBAL(ret_from_fork) _GLOBAL(ret_from_kernel_thread) bl .schedule_tail REST_NVGPRS(r1) - li r3,0 - std r3,0(r1) ld r14, 0(r14) mtlr r14 mr r3,r15 diff --git a/arch/powerpc/kernel/process.c b/arch/powerpc/kernel/process.c index ceb4e7b62cf4..21981a89e394 100644 --- a/arch/powerpc/kernel/process.c +++ b/arch/powerpc/kernel/process.c @@ -971,6 +971,7 @@ int copy_thread(unsigned long clone_flags, unsigned long usp, * do some house keeping and then return from the fork or clone * system call, using the stack frame created above. */ + ((unsigned long *)sp)[0] = 0; sp -= sizeof(struct pt_regs); kregs = (struct pt_regs *) sp; sp -= STACK_FRAME_OVERHEAD; From 75b93da43aa1132bf23dafbb4badb028ccf78129 Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Tue, 14 May 2013 15:10:02 +1000 Subject: [PATCH 138/198] powerpc/powernv: Detect OPAL v3 API version Future firmwares will support that new version Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/include/asm/firmware.h | 4 +++- arch/powerpc/include/asm/opal.h | 3 ++- arch/powerpc/platforms/powernv/opal.c | 6 +++++- arch/powerpc/platforms/powernv/setup.c | 4 +++- 4 files changed, 13 insertions(+), 4 deletions(-) diff --git a/arch/powerpc/include/asm/firmware.h b/arch/powerpc/include/asm/firmware.h index 0df54646f968..681bc0314b6b 100644 --- a/arch/powerpc/include/asm/firmware.h +++ b/arch/powerpc/include/asm/firmware.h @@ -52,6 +52,7 @@ #define FW_FEATURE_BEST_ENERGY ASM_CONST(0x0000000080000000) #define FW_FEATURE_TYPE1_AFFINITY ASM_CONST(0x0000000100000000) #define FW_FEATURE_PRRN ASM_CONST(0x0000000200000000) +#define FW_FEATURE_OPALv3 ASM_CONST(0x0000000400000000) #ifndef __ASSEMBLY__ @@ -69,7 +70,8 @@ enum { FW_FEATURE_SET_MODE | FW_FEATURE_BEST_ENERGY | FW_FEATURE_TYPE1_AFFINITY | FW_FEATURE_PRRN, FW_FEATURE_PSERIES_ALWAYS = 0, - FW_FEATURE_POWERNV_POSSIBLE = FW_FEATURE_OPAL | FW_FEATURE_OPALv2, + FW_FEATURE_POWERNV_POSSIBLE = FW_FEATURE_OPAL | FW_FEATURE_OPALv2 | + FW_FEATURE_OPALv3, FW_FEATURE_POWERNV_ALWAYS = 0, FW_FEATURE_PS3_POSSIBLE = FW_FEATURE_LPAR | FW_FEATURE_PS3_LV1, FW_FEATURE_PS3_ALWAYS = FW_FEATURE_LPAR | FW_FEATURE_PS3_LV1, diff --git a/arch/powerpc/include/asm/opal.h b/arch/powerpc/include/asm/opal.h index b2906adb89d3..cbb9305ab15a 100644 --- a/arch/powerpc/include/asm/opal.h +++ b/arch/powerpc/include/asm/opal.h @@ -243,7 +243,8 @@ enum OpalMCE_TlbErrorType { enum OpalThreadStatus { OPAL_THREAD_INACTIVE = 0x0, - OPAL_THREAD_STARTED = 0x1 + OPAL_THREAD_STARTED = 0x1, + OPAL_THREAD_UNAVAILABLE = 0x2 /* opal-v3 */ }; enum OpalPciBusCompare { diff --git a/arch/powerpc/platforms/powernv/opal.c b/arch/powerpc/platforms/powernv/opal.c index 27907cb18b8a..628c564ceadb 100644 --- a/arch/powerpc/platforms/powernv/opal.c +++ b/arch/powerpc/platforms/powernv/opal.c @@ -56,7 +56,11 @@ int __init early_init_dt_scan_opal(unsigned long node, opal.entry, entryp, entrysz); powerpc_firmware_features |= FW_FEATURE_OPAL; - if (of_flat_dt_is_compatible(node, "ibm,opal-v2")) { + if (of_flat_dt_is_compatible(node, "ibm,opal-v3")) { + powerpc_firmware_features |= FW_FEATURE_OPALv2; + powerpc_firmware_features |= FW_FEATURE_OPALv3; + printk("OPAL V3 detected !\n"); + } else if (of_flat_dt_is_compatible(node, "ibm,opal-v2")) { powerpc_firmware_features |= FW_FEATURE_OPALv2; printk("OPAL V2 detected !\n"); } else { diff --git a/arch/powerpc/platforms/powernv/setup.c b/arch/powerpc/platforms/powernv/setup.c index c20381caaa3e..d4459bfc92f7 100644 --- a/arch/powerpc/platforms/powernv/setup.c +++ b/arch/powerpc/platforms/powernv/setup.c @@ -78,7 +78,9 @@ static void pnv_show_cpuinfo(struct seq_file *m) if (root) model = of_get_property(root, "model", NULL); seq_printf(m, "machine\t\t: PowerNV %s\n", model); - if (firmware_has_feature(FW_FEATURE_OPALv2)) + if (firmware_has_feature(FW_FEATURE_OPALv3)) + seq_printf(m, "firmware\t: OPAL v3\n"); + else if (firmware_has_feature(FW_FEATURE_OPALv2)) seq_printf(m, "firmware\t: OPAL v2\n"); else if (firmware_has_feature(FW_FEATURE_OPAL)) seq_printf(m, "firmware\t: OPAL v1\n"); From b2b48584dfc69a5ebc70ee7976524d7122e5df5f Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Tue, 14 May 2013 15:12:31 +1000 Subject: [PATCH 139/198] powerpc/powernv: Fix starting of secondary CPUs on OPALv2 and v3 The current code fails to handle kexec on OPALv2. This fixes it and adds code to improve the situation on OPALv3 where we can query the CPU status from the firmware and decide what to do based on that. Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/platforms/powernv/smp.c | 62 +++++++++++++++++++++++++--- 1 file changed, 56 insertions(+), 6 deletions(-) diff --git a/arch/powerpc/platforms/powernv/smp.c b/arch/powerpc/platforms/powernv/smp.c index 6a3ecca5b725..88c9459c3e07 100644 --- a/arch/powerpc/platforms/powernv/smp.c +++ b/arch/powerpc/platforms/powernv/smp.c @@ -71,18 +71,68 @@ int pnv_smp_kick_cpu(int nr) BUG_ON(nr < 0 || nr >= NR_CPUS); - /* On OPAL v2 the CPU are still spinning inside OPAL itself, - * get them back now + /* + * If we already started or OPALv2 is not supported, we just + * kick the CPU via the PACA */ - if (!paca[nr].cpu_start && firmware_has_feature(FW_FEATURE_OPALv2)) { - pr_devel("OPAL: Starting CPU %d (HW 0x%x)...\n", nr, pcpu); - rc = opal_start_cpu(pcpu, start_here); + if (paca[nr].cpu_start || !firmware_has_feature(FW_FEATURE_OPALv2)) + goto kick; + + /* + * At this point, the CPU can either be spinning on the way in + * from kexec or be inside OPAL waiting to be started for the + * first time. OPAL v3 allows us to query OPAL to know if it + * has the CPUs, so we do that + */ + if (firmware_has_feature(FW_FEATURE_OPALv3)) { + uint8_t status; + + rc = opal_query_cpu_status(pcpu, &status); if (rc != OPAL_SUCCESS) { - pr_warn("OPAL Error %ld starting CPU %d\n", + pr_warn("OPAL Error %ld querying CPU %d state\n", rc, nr); return -ENODEV; } + + /* + * Already started, just kick it, probably coming from + * kexec and spinning + */ + if (status == OPAL_THREAD_STARTED) + goto kick; + + /* + * Available/inactive, let's kick it + */ + if (status == OPAL_THREAD_INACTIVE) { + pr_devel("OPAL: Starting CPU %d (HW 0x%x)...\n", + nr, pcpu); + rc = opal_start_cpu(pcpu, start_here); + if (rc != OPAL_SUCCESS) { + pr_warn("OPAL Error %ld starting CPU %d\n", + rc, nr); + return -ENODEV; + } + } else { + /* + * An unavailable CPU (or any other unknown status) + * shouldn't be started. It should also + * not be in the possible map but currently it can + * happen + */ + pr_devel("OPAL: CPU %d (HW 0x%x) is unavailable" + " (status %d)...\n", nr, pcpu, status); + return -ENODEV; + } + } else { + /* + * On OPAL v2, we just kick it and hope for the best, + * we must not test the error from opal_start_cpu() or + * we would fail to get CPUs from kexec. + */ + opal_start_cpu(pcpu, start_here); } + kick: return smp_generic_kick_cpu(nr); } From ca9d7aea5952c5e90bfe28ed738a7e962c0adad2 Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Mon, 13 May 2013 00:23:38 +0000 Subject: [PATCH 140/198] powerpc: Provide __bswapdi2 Some versions of GCC apparently expect this to be provided by libgcc. Updates from Mikey to fix 32 bit version and adding "r" to registers. Signed-off-by: David Woodhouse Signed-off-by: Michael Neuling Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/kernel/misc_32.S | 11 +++++++++++ arch/powerpc/kernel/misc_64.S | 11 +++++++++++ arch/powerpc/kernel/ppc_ksyms.c | 3 ++- 3 files changed, 24 insertions(+), 1 deletion(-) diff --git a/arch/powerpc/kernel/misc_32.S b/arch/powerpc/kernel/misc_32.S index 19e096bd0e73..e469f30e6eeb 100644 --- a/arch/powerpc/kernel/misc_32.S +++ b/arch/powerpc/kernel/misc_32.S @@ -657,6 +657,17 @@ _GLOBAL(__ucmpdi2) li r3,2 blr +_GLOBAL(__bswapdi2) + rotlwi r9,r4,8 + rotlwi r10,r3,8 + rlwimi r9,r4,24,0,7 + rlwimi r10,r3,24,0,7 + rlwimi r9,r4,24,16,23 + rlwimi r10,r3,24,16,23 + mr r3,r9 + mr r4,r10 + blr + _GLOBAL(abs) srawi r4,r3,31 xor r3,r3,r4 diff --git a/arch/powerpc/kernel/misc_64.S b/arch/powerpc/kernel/misc_64.S index 5cfa8008693b..6820e45f557b 100644 --- a/arch/powerpc/kernel/misc_64.S +++ b/arch/powerpc/kernel/misc_64.S @@ -234,6 +234,17 @@ _GLOBAL(__flush_dcache_icache) isync blr +_GLOBAL(__bswapdi2) + srdi r8,r3,32 + rlwinm r7,r3,8,0xffffffff + rlwimi r7,r3,24,0,7 + rlwinm r9,r8,8,0xffffffff + rlwimi r7,r3,24,16,23 + rlwimi r9,r8,24,0,7 + rlwimi r9,r8,24,16,23 + sldi r7,r7,32 + or r3,r7,r9 + blr #if defined(CONFIG_PPC_PMAC) || defined(CONFIG_PPC_MAPLE) /* diff --git a/arch/powerpc/kernel/ppc_ksyms.c b/arch/powerpc/kernel/ppc_ksyms.c index 78b8766fd79e..c29666586998 100644 --- a/arch/powerpc/kernel/ppc_ksyms.c +++ b/arch/powerpc/kernel/ppc_ksyms.c @@ -143,7 +143,8 @@ EXPORT_SYMBOL(__lshrdi3); int __ucmpdi2(unsigned long long, unsigned long long); EXPORT_SYMBOL(__ucmpdi2); #endif - +long long __bswapdi2(long long); +EXPORT_SYMBOL(__bswapdi2); EXPORT_SYMBOL(memcpy); EXPORT_SYMBOL(memset); EXPORT_SYMBOL(memmove); From dcb615aef988b57deef3d9b4557ff20f681a82b0 Mon Sep 17 00:00:00 2001 From: Alexander Gordeev Date: Mon, 13 May 2013 00:57:49 +0000 Subject: [PATCH 141/198] powerpc: Fix irq_set_affinity() return values Signed-off-by: Alexander Gordeev Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/platforms/wsp/ics.c | 2 +- arch/powerpc/sysdev/ehv_pic.c | 2 +- arch/powerpc/sysdev/mpic.c | 2 +- arch/powerpc/sysdev/xics/ics-opal.c | 2 +- 4 files changed, 4 insertions(+), 4 deletions(-) diff --git a/arch/powerpc/platforms/wsp/ics.c b/arch/powerpc/platforms/wsp/ics.c index 97fe82ee8633..2d3b1dd9571d 100644 --- a/arch/powerpc/platforms/wsp/ics.c +++ b/arch/powerpc/platforms/wsp/ics.c @@ -361,7 +361,7 @@ static int wsp_chip_set_affinity(struct irq_data *d, xive = xive_set_server(xive, get_irq_server(ics, hw_irq)); wsp_ics_set_xive(ics, hw_irq, xive); - return 0; + return IRQ_SET_MASK_OK; } static struct irq_chip wsp_irq_chip = { diff --git a/arch/powerpc/sysdev/ehv_pic.c b/arch/powerpc/sysdev/ehv_pic.c index 6e0e1005227f..9cd0e60716fe 100644 --- a/arch/powerpc/sysdev/ehv_pic.c +++ b/arch/powerpc/sysdev/ehv_pic.c @@ -81,7 +81,7 @@ int ehv_pic_set_affinity(struct irq_data *d, const struct cpumask *dest, ev_int_set_config(src, config, prio, cpuid); spin_unlock_irqrestore(&ehv_pic_lock, flags); - return 0; + return IRQ_SET_MASK_OK; } static unsigned int ehv_pic_type_to_vecpri(unsigned int type) diff --git a/arch/powerpc/sysdev/mpic.c b/arch/powerpc/sysdev/mpic.c index ee21b5e71aec..0a13ecb270c7 100644 --- a/arch/powerpc/sysdev/mpic.c +++ b/arch/powerpc/sysdev/mpic.c @@ -836,7 +836,7 @@ int mpic_set_affinity(struct irq_data *d, const struct cpumask *cpumask, mpic_physmask(mask)); } - return 0; + return IRQ_SET_MASK_OK; } static unsigned int mpic_type_to_vecpri(struct mpic *mpic, unsigned int type) diff --git a/arch/powerpc/sysdev/xics/ics-opal.c b/arch/powerpc/sysdev/xics/ics-opal.c index f7e8609df0d5..39d72212655e 100644 --- a/arch/powerpc/sysdev/xics/ics-opal.c +++ b/arch/powerpc/sysdev/xics/ics-opal.c @@ -148,7 +148,7 @@ static int ics_opal_set_affinity(struct irq_data *d, __func__, d->irq, hw_irq, server, rc); return -1; } - return 0; + return IRQ_SET_MASK_OK; } static struct irq_chip ics_opal_irq_chip = { From 6cecf76b47ba6bea3c81d170afc2e0b244e5849c Mon Sep 17 00:00:00 2001 From: Scott Wood Date: Mon, 13 May 2013 14:14:53 +0000 Subject: [PATCH 142/198] powerpc/booke64: Fix kernel hangs at kernel_dbg_exc MSR_DE is not cleared on entry to the kernel, and we don't clear it explicitly outside of debug code. If we have MSR_DE set in prime_debug_regs(), and the new thread has events enabled in DBCR0 (e.g. ICMP is set in thread->dbsr0, even though it was cleared in the real DBCR0 when the thread got scheduled out), we'll end up taking a debug exception in the kernel when DBCR0 is loaded. DSRR0 will not point to an exception vector, and the kernel ends up hanging at kernel_dbg_exc. Fix this by always clearing MSR_DE when we load new debug state. Another observed source of kernel_dbg_exc hangs is with the branch taken event. If this event is active, but we take a non-debug trap (e.g. a TLB miss or an asynchronous interrupt) before the next branch. We end up taking a branch-taken debug exception on the initial branch instruction of the exception vector, but because the debug exception is DBSR_BT rather than DBSR_IC we branch to kernel_dbg_exc before even checking the DSRR0 address. Fix this by checking for DBSR_BT as well as DBSR_IC, which is what 32-bit does and what the comments suggest was intended in the 64-bit code as well. Signed-off-by: Scott Wood Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/kernel/exceptions-64e.S | 8 ++++---- arch/powerpc/kernel/process.c | 7 +++++++ 2 files changed, 11 insertions(+), 4 deletions(-) diff --git a/arch/powerpc/kernel/exceptions-64e.S b/arch/powerpc/kernel/exceptions-64e.S index 42a756eec9ff..645170a07ada 100644 --- a/arch/powerpc/kernel/exceptions-64e.S +++ b/arch/powerpc/kernel/exceptions-64e.S @@ -489,7 +489,7 @@ END_FTR_SECTION_IFSET(CPU_FTR_ALTIVEC) */ mfspr r14,SPRN_DBSR /* check single-step/branch taken */ - andis. r15,r14,DBSR_IC@h + andis. r15,r14,(DBSR_IC|DBSR_BT)@h beq+ 1f LOAD_REG_IMMEDIATE(r14,interrupt_base_book3e) @@ -500,7 +500,7 @@ END_FTR_SECTION_IFSET(CPU_FTR_ALTIVEC) bge+ cr1,1f /* here it looks like we got an inappropriate debug exception. */ - lis r14,DBSR_IC@h /* clear the IC event */ + lis r14,(DBSR_IC|DBSR_BT)@h /* clear the event */ rlwinm r11,r11,0,~MSR_DE /* clear DE in the CSRR1 value */ mtspr SPRN_DBSR,r14 mtspr SPRN_CSRR1,r11 @@ -555,7 +555,7 @@ kernel_dbg_exc: */ mfspr r14,SPRN_DBSR /* check single-step/branch taken */ - andis. r15,r14,DBSR_IC@h + andis. r15,r14,(DBSR_IC|DBSR_BT)@h beq+ 1f LOAD_REG_IMMEDIATE(r14,interrupt_base_book3e) @@ -566,7 +566,7 @@ kernel_dbg_exc: bge+ cr1,1f /* here it looks like we got an inappropriate debug exception. */ - lis r14,DBSR_IC@h /* clear the IC event */ + lis r14,(DBSR_IC|DBSR_BT)@h /* clear the event */ rlwinm r11,r11,0,~MSR_DE /* clear DE in the DSRR1 value */ mtspr SPRN_DBSR,r14 mtspr SPRN_DSRR1,r11 diff --git a/arch/powerpc/kernel/process.c b/arch/powerpc/kernel/process.c index 21981a89e394..a902723fdc69 100644 --- a/arch/powerpc/kernel/process.c +++ b/arch/powerpc/kernel/process.c @@ -339,6 +339,13 @@ static void set_debug_reg_defaults(struct thread_struct *thread) static void prime_debug_regs(struct thread_struct *thread) { + /* + * We could have inherited MSR_DE from userspace, since + * it doesn't get cleared on exception entry. Make sure + * MSR_DE is clear before we enable any debug events. + */ + mtmsr(mfmsr() & ~MSR_DE); + mtspr(SPRN_IAC1, thread->iac1); mtspr(SPRN_IAC2, thread->iac2); #if CONFIG_PPC_ADV_DEBUG_IACS > 2 From 22ecbe8dcef4f6bf80ed4358600e8f6e03105c5c Mon Sep 17 00:00:00 2001 From: Li Zhong Date: Mon, 13 May 2013 16:16:40 +0000 Subject: [PATCH 143/198] powerpc: Syscall hooks for context tracking subsystem This is the syscall slow path hooks for context tracking subsystem, corresponding to [PATCH] x86: Syscall hooks for userspace RCU extended QS commit bf5a3c13b939813d28ce26c01425054c740d6731 TIF_MEMDIE is moved to the second 16-bits (with value 17), as it seems there is no asm code using it. TIF_NOHZ is added to _TIF_SYCALL_T_OR_A, so it is better for it to be in the same 16 bits with others in the group, so in the asm code, andi. with this group could work. Signed-off-by: Li Zhong Acked-by: Frederic Weisbecker Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/include/asm/thread_info.h | 7 +++++-- arch/powerpc/kernel/ptrace.c | 5 +++++ 2 files changed, 10 insertions(+), 2 deletions(-) diff --git a/arch/powerpc/include/asm/thread_info.h b/arch/powerpc/include/asm/thread_info.h index 8ceea14d6fe4..ba7b1973866e 100644 --- a/arch/powerpc/include/asm/thread_info.h +++ b/arch/powerpc/include/asm/thread_info.h @@ -97,7 +97,7 @@ static inline struct thread_info *current_thread_info(void) #define TIF_PERFMON_CTXSW 6 /* perfmon needs ctxsw calls */ #define TIF_SYSCALL_AUDIT 7 /* syscall auditing active */ #define TIF_SINGLESTEP 8 /* singlestepping active */ -#define TIF_MEMDIE 9 /* is terminating due to OOM killer */ +#define TIF_NOHZ 9 /* in adaptive nohz mode */ #define TIF_SECCOMP 10 /* secure computing */ #define TIF_RESTOREALL 11 /* Restore all regs (implies NOERROR) */ #define TIF_NOERROR 12 /* Force successful syscall return */ @@ -106,6 +106,7 @@ static inline struct thread_info *current_thread_info(void) #define TIF_SYSCALL_TRACEPOINT 15 /* syscall tracepoint instrumentation */ #define TIF_EMULATE_STACK_STORE 16 /* Is an instruction emulation for stack store? */ +#define TIF_MEMDIE 17 /* is terminating due to OOM killer */ /* as above, but as bit values */ #define _TIF_SYSCALL_TRACE (1< #include #include +#include #include #include @@ -1788,6 +1789,8 @@ long do_syscall_trace_enter(struct pt_regs *regs) { long ret = 0; + user_exit(); + secure_computing_strict(regs->gpr[0]); if (test_thread_flag(TIF_SYSCALL_TRACE) && @@ -1832,4 +1835,6 @@ void do_syscall_trace_leave(struct pt_regs *regs) step = test_thread_flag(TIF_SINGLESTEP); if (step || test_thread_flag(TIF_SYSCALL_TRACE)) tracehook_report_syscall_exit(regs, step); + + user_enter(); } From ba12eedee321eeb5baecaada285daeb3462c35f5 Mon Sep 17 00:00:00 2001 From: Li Zhong Date: Mon, 13 May 2013 16:16:41 +0000 Subject: [PATCH 144/198] powerpc: Exception hooks for context tracking subsystem This is the exception hooks for context tracking subsystem, including data access, program check, single step, instruction breakpoint, machine check, alignment, fp unavailable, altivec assist, unknown exception, whose handlers might use RCU. This patch corresponds to [PATCH] x86: Exception hooks for userspace RCU extended QS commit 6ba3c97a38803883c2eee489505796cb0a727122 But after the exception handling moved to generic code, and some changes in following two commits: 56dd9470d7c8734f055da2a6bac553caf4a468eb context_tracking: Move exception handling to generic code 6c1e0256fad84a843d915414e4b5973b7443d48d context_tracking: Restore correct previous context state on exception exit it is able for exception hooks to use the generic code above instead of a redundant arch implementation. Signed-off-by: Li Zhong Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/kernel/traps.c | 80 ++++++++++++++++++++++++--------- arch/powerpc/mm/fault.c | 41 +++++++++++------ arch/powerpc/mm/hash_utils_64.c | 36 +++++++++++---- 3 files changed, 112 insertions(+), 45 deletions(-) diff --git a/arch/powerpc/kernel/traps.c b/arch/powerpc/kernel/traps.c index 83efa2f7d926..a7a648f6b750 100644 --- a/arch/powerpc/kernel/traps.c +++ b/arch/powerpc/kernel/traps.c @@ -35,6 +35,7 @@ #include #include #include +#include #include #include @@ -667,6 +668,7 @@ int machine_check_generic(struct pt_regs *regs) void machine_check_exception(struct pt_regs *regs) { + enum ctx_state prev_state = exception_enter(); int recover = 0; __get_cpu_var(irq_stat).mce_exceptions++; @@ -683,7 +685,7 @@ void machine_check_exception(struct pt_regs *regs) recover = cur_cpu_spec->machine_check(regs); if (recover > 0) - return; + goto bail; #if defined(CONFIG_8xx) && defined(CONFIG_PCI) /* the qspan pci read routines can cause machine checks -- Cort @@ -693,20 +695,23 @@ void machine_check_exception(struct pt_regs *regs) * -- BenH */ bad_page_fault(regs, regs->dar, SIGBUS); - return; + goto bail; #endif if (debugger_fault_handler(regs)) - return; + goto bail; if (check_io_access(regs)) - return; + goto bail; die("Machine check", regs, SIGBUS); /* Must die if the interrupt is not recoverable */ if (!(regs->msr & MSR_RI)) panic("Unrecoverable Machine check"); + +bail: + exception_exit(prev_state); } void SMIException(struct pt_regs *regs) @@ -716,20 +721,29 @@ void SMIException(struct pt_regs *regs) void unknown_exception(struct pt_regs *regs) { + enum ctx_state prev_state = exception_enter(); + printk("Bad trap at PC: %lx, SR: %lx, vector=%lx\n", regs->nip, regs->msr, regs->trap); _exception(SIGTRAP, regs, 0, 0); + + exception_exit(prev_state); } void instruction_breakpoint_exception(struct pt_regs *regs) { + enum ctx_state prev_state = exception_enter(); + if (notify_die(DIE_IABR_MATCH, "iabr_match", regs, 5, 5, SIGTRAP) == NOTIFY_STOP) - return; + goto bail; if (debugger_iabr_match(regs)) - return; + goto bail; _exception(SIGTRAP, regs, TRAP_BRKPT, regs->nip); + +bail: + exception_exit(prev_state); } void RunModeException(struct pt_regs *regs) @@ -739,15 +753,20 @@ void RunModeException(struct pt_regs *regs) void __kprobes single_step_exception(struct pt_regs *regs) { + enum ctx_state prev_state = exception_enter(); + clear_single_step(regs); if (notify_die(DIE_SSTEP, "single_step", regs, 5, 5, SIGTRAP) == NOTIFY_STOP) - return; + goto bail; if (debugger_sstep(regs)) - return; + goto bail; _exception(SIGTRAP, regs, TRAP_TRACE, regs->nip); + +bail: + exception_exit(prev_state); } /* @@ -1005,6 +1024,7 @@ int is_valid_bugaddr(unsigned long addr) void __kprobes program_check_exception(struct pt_regs *regs) { + enum ctx_state prev_state = exception_enter(); unsigned int reason = get_reason(regs); extern int do_mathemu(struct pt_regs *regs); @@ -1014,26 +1034,26 @@ void __kprobes program_check_exception(struct pt_regs *regs) if (reason & REASON_FP) { /* IEEE FP exception */ parse_fpe(regs); - return; + goto bail; } if (reason & REASON_TRAP) { /* Debugger is first in line to stop recursive faults in * rcu_lock, notify_die, or atomic_notifier_call_chain */ if (debugger_bpt(regs)) - return; + goto bail; /* trap exception */ if (notify_die(DIE_BPT, "breakpoint", regs, 5, 5, SIGTRAP) == NOTIFY_STOP) - return; + goto bail; if (!(regs->msr & MSR_PR) && /* not user-mode */ report_bug(regs->nip, regs) == BUG_TRAP_TYPE_WARN) { regs->nip += 4; - return; + goto bail; } _exception(SIGTRAP, regs, TRAP_BRKPT, regs->nip); - return; + goto bail; } #ifdef CONFIG_PPC_TRANSACTIONAL_MEM if (reason & REASON_TM) { @@ -1049,7 +1069,7 @@ void __kprobes program_check_exception(struct pt_regs *regs) if (!user_mode(regs) && report_bug(regs->nip, regs) == BUG_TRAP_TYPE_WARN) { regs->nip += 4; - return; + goto bail; } /* If usermode caused this, it's done something illegal and * gets a SIGILL slap on the wrist. We call it an illegal @@ -1059,7 +1079,7 @@ void __kprobes program_check_exception(struct pt_regs *regs) */ if (user_mode(regs)) { _exception(SIGILL, regs, ILL_ILLOPN, regs->nip); - return; + goto bail; } else { printk(KERN_EMERG "Unexpected TM Bad Thing exception " "at %lx (msr 0x%x)\n", regs->nip, reason); @@ -1083,16 +1103,16 @@ void __kprobes program_check_exception(struct pt_regs *regs) switch (do_mathemu(regs)) { case 0: emulate_single_step(regs); - return; + goto bail; case 1: { int code = 0; code = __parse_fpscr(current->thread.fpscr.val); _exception(SIGFPE, regs, code, regs->nip); - return; + goto bail; } case -EFAULT: _exception(SIGSEGV, regs, SEGV_MAPERR, regs->nip); - return; + goto bail; } /* fall through on any other errors */ #endif /* CONFIG_MATH_EMULATION */ @@ -1103,10 +1123,10 @@ void __kprobes program_check_exception(struct pt_regs *regs) case 0: regs->nip += 4; emulate_single_step(regs); - return; + goto bail; case -EFAULT: _exception(SIGSEGV, regs, SEGV_MAPERR, regs->nip); - return; + goto bail; } } @@ -1114,10 +1134,14 @@ void __kprobes program_check_exception(struct pt_regs *regs) _exception(SIGILL, regs, ILL_PRVOPC, regs->nip); else _exception(SIGILL, regs, ILL_ILLOPC, regs->nip); + +bail: + exception_exit(prev_state); } void alignment_exception(struct pt_regs *regs) { + enum ctx_state prev_state = exception_enter(); int sig, code, fixed = 0; /* We restore the interrupt state now */ @@ -1131,7 +1155,7 @@ void alignment_exception(struct pt_regs *regs) if (fixed == 1) { regs->nip += 4; /* skip over emulated instruction */ emulate_single_step(regs); - return; + goto bail; } /* Operand address was bad */ @@ -1146,6 +1170,9 @@ void alignment_exception(struct pt_regs *regs) _exception(sig, regs, code, regs->dar); else bad_page_fault(regs, regs->dar, sig); + +bail: + exception_exit(prev_state); } void StackOverflow(struct pt_regs *regs) @@ -1174,23 +1201,32 @@ void trace_syscall(struct pt_regs *regs) void kernel_fp_unavailable_exception(struct pt_regs *regs) { + enum ctx_state prev_state = exception_enter(); + printk(KERN_EMERG "Unrecoverable FP Unavailable Exception " "%lx at %lx\n", regs->trap, regs->nip); die("Unrecoverable FP Unavailable Exception", regs, SIGABRT); + + exception_exit(prev_state); } void altivec_unavailable_exception(struct pt_regs *regs) { + enum ctx_state prev_state = exception_enter(); + if (user_mode(regs)) { /* A user program has executed an altivec instruction, but this kernel doesn't support altivec. */ _exception(SIGILL, regs, ILL_ILLOPC, regs->nip); - return; + goto bail; } printk(KERN_EMERG "Unrecoverable VMX/Altivec Unavailable Exception " "%lx at %lx\n", regs->trap, regs->nip); die("Unrecoverable VMX/Altivec Unavailable Exception", regs, SIGABRT); + +bail: + exception_exit(prev_state); } void vsx_unavailable_exception(struct pt_regs *regs) diff --git a/arch/powerpc/mm/fault.c b/arch/powerpc/mm/fault.c index 229951ffc351..8726779e1409 100644 --- a/arch/powerpc/mm/fault.c +++ b/arch/powerpc/mm/fault.c @@ -32,6 +32,7 @@ #include #include #include +#include #include #include @@ -196,6 +197,7 @@ static int mm_fault_error(struct pt_regs *regs, unsigned long addr, int fault) int __kprobes do_page_fault(struct pt_regs *regs, unsigned long address, unsigned long error_code) { + enum ctx_state prev_state = exception_enter(); struct vm_area_struct * vma; struct mm_struct *mm = current->mm; unsigned int flags = FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_KILLABLE; @@ -204,6 +206,7 @@ int __kprobes do_page_fault(struct pt_regs *regs, unsigned long address, int trap = TRAP(regs); int is_exec = trap == 0x400; int fault; + int rc = 0; #if !(defined(CONFIG_4xx) || defined(CONFIG_BOOKE)) /* @@ -230,28 +233,30 @@ int __kprobes do_page_fault(struct pt_regs *regs, unsigned long address, * look at it */ if (error_code & ICSWX_DSI_UCT) { - int rc = acop_handle_fault(regs, address, error_code); + rc = acop_handle_fault(regs, address, error_code); if (rc) - return rc; + goto bail; } #endif /* CONFIG_PPC_ICSWX */ if (notify_page_fault(regs)) - return 0; + goto bail; if (unlikely(debugger_fault_handler(regs))) - return 0; + goto bail; /* On a kernel SLB miss we can only check for a valid exception entry */ - if (!user_mode(regs) && (address >= TASK_SIZE)) - return SIGSEGV; + if (!user_mode(regs) && (address >= TASK_SIZE)) { + rc = SIGSEGV; + goto bail; + } #if !(defined(CONFIG_4xx) || defined(CONFIG_BOOKE) || \ defined(CONFIG_PPC_BOOK3S_64)) if (error_code & DSISR_DABRMATCH) { /* breakpoint match */ do_break(regs, address, error_code); - return 0; + goto bail; } #endif @@ -260,8 +265,10 @@ int __kprobes do_page_fault(struct pt_regs *regs, unsigned long address, local_irq_enable(); if (in_atomic() || mm == NULL) { - if (!user_mode(regs)) - return SIGSEGV; + if (!user_mode(regs)) { + rc = SIGSEGV; + goto bail; + } /* in_atomic() in user mode is really bad, as is current->mm == NULL. */ printk(KERN_EMERG "Page fault in user mode with " @@ -417,9 +424,11 @@ good_area: */ fault = handle_mm_fault(mm, vma, address, flags); if (unlikely(fault & (VM_FAULT_RETRY|VM_FAULT_ERROR))) { - int rc = mm_fault_error(regs, address, fault); + rc = mm_fault_error(regs, address, fault); if (rc >= MM_FAULT_RETURN) - return rc; + goto bail; + else + rc = 0; } /* @@ -454,7 +463,7 @@ good_area: } up_read(&mm->mmap_sem); - return 0; + goto bail; bad_area: up_read(&mm->mmap_sem); @@ -463,7 +472,7 @@ bad_area_nosemaphore: /* User mode accesses cause a SIGSEGV */ if (user_mode(regs)) { _exception(SIGSEGV, regs, code, address); - return 0; + goto bail; } if (is_exec && (error_code & DSISR_PROTFAULT)) @@ -471,7 +480,11 @@ bad_area_nosemaphore: " page (%lx) - exploit attempt? (uid: %d)\n", address, from_kuid(&init_user_ns, current_uid())); - return SIGSEGV; + rc = SIGSEGV; + +bail: + exception_exit(prev_state); + return rc; } diff --git a/arch/powerpc/mm/hash_utils_64.c b/arch/powerpc/mm/hash_utils_64.c index 88ac0eeaadde..e303a6d74e3a 100644 --- a/arch/powerpc/mm/hash_utils_64.c +++ b/arch/powerpc/mm/hash_utils_64.c @@ -33,6 +33,7 @@ #include #include #include +#include #include #include @@ -954,6 +955,7 @@ void hash_failure_debug(unsigned long ea, unsigned long access, */ int hash_page(unsigned long ea, unsigned long access, unsigned long trap) { + enum ctx_state prev_state = exception_enter(); pgd_t *pgdir; unsigned long vsid; struct mm_struct *mm; @@ -973,7 +975,8 @@ int hash_page(unsigned long ea, unsigned long access, unsigned long trap) mm = current->mm; if (! mm) { DBG_LOW(" user region with no mm !\n"); - return 1; + rc = 1; + goto bail; } psize = get_slice_psize(mm, ea); ssize = user_segment_size(ea); @@ -992,19 +995,23 @@ int hash_page(unsigned long ea, unsigned long access, unsigned long trap) /* Not a valid range * Send the problem up to do_page_fault */ - return 1; + rc = 1; + goto bail; } DBG_LOW(" mm=%p, mm->pgdir=%p, vsid=%016lx\n", mm, mm->pgd, vsid); /* Bad address. */ if (!vsid) { DBG_LOW("Bad address!\n"); - return 1; + rc = 1; + goto bail; } /* Get pgdir */ pgdir = mm->pgd; - if (pgdir == NULL) - return 1; + if (pgdir == NULL) { + rc = 1; + goto bail; + } /* Check CPU locality */ tmp = cpumask_of(smp_processor_id()); @@ -1027,7 +1034,8 @@ int hash_page(unsigned long ea, unsigned long access, unsigned long trap) ptep = find_linux_pte_or_hugepte(pgdir, ea, &hugeshift); if (ptep == NULL || !pte_present(*ptep)) { DBG_LOW(" no PTE !\n"); - return 1; + rc = 1; + goto bail; } /* Add _PAGE_PRESENT to the required access perm */ @@ -1038,13 +1046,16 @@ int hash_page(unsigned long ea, unsigned long access, unsigned long trap) */ if (access & ~pte_val(*ptep)) { DBG_LOW(" no access !\n"); - return 1; + rc = 1; + goto bail; } #ifdef CONFIG_HUGETLB_PAGE - if (hugeshift) - return __hash_page_huge(ea, access, vsid, ptep, trap, local, + if (hugeshift) { + rc = __hash_page_huge(ea, access, vsid, ptep, trap, local, ssize, hugeshift, psize); + goto bail; + } #endif /* CONFIG_HUGETLB_PAGE */ #ifndef CONFIG_PPC_64K_PAGES @@ -1124,6 +1135,9 @@ int hash_page(unsigned long ea, unsigned long access, unsigned long trap) pte_val(*(ptep + PTRS_PER_PTE))); #endif DBG_LOW(" -> rc=%d\n", rc); + +bail: + exception_exit(prev_state); return rc; } EXPORT_SYMBOL_GPL(hash_page); @@ -1259,6 +1273,8 @@ void flush_hash_range(unsigned long number, int local) */ void low_hash_fault(struct pt_regs *regs, unsigned long address, int rc) { + enum ctx_state prev_state = exception_enter(); + if (user_mode(regs)) { #ifdef CONFIG_PPC_SUBPAGE_PROT if (rc == -2) @@ -1268,6 +1284,8 @@ void low_hash_fault(struct pt_regs *regs, unsigned long address, int rc) _exception(SIGBUS, regs, BUS_ADRERR, address); } else bad_page_fault(regs, address, SIGBUS); + + exception_exit(prev_state); } long hpte_insert_repeating(unsigned long hash, unsigned long vpn, From 106ed886abc46723c290e82e5f161d9b40820203 Mon Sep 17 00:00:00 2001 From: Li Zhong Date: Mon, 13 May 2013 16:16:42 +0000 Subject: [PATCH 145/198] powerpc: Exit user context on notify resume This patch allows RCU usage in do_notify_resume, e.g. signal handling. It corresponds to [PATCH] x86: Exit RCU extended QS on notify resume commit edf55fda35c7dc7f2d9241c3abaddaf759b457c6 Signed-off-by: Li Zhong Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/kernel/signal.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/arch/powerpc/kernel/signal.c b/arch/powerpc/kernel/signal.c index cf12eae02de5..d63b5024c563 100644 --- a/arch/powerpc/kernel/signal.c +++ b/arch/powerpc/kernel/signal.c @@ -13,6 +13,7 @@ #include #include #include +#include #include #include #include @@ -159,6 +160,8 @@ static int do_signal(struct pt_regs *regs) void do_notify_resume(struct pt_regs *regs, unsigned long thread_info_flags) { + user_exit(); + if (thread_info_flags & _TIF_UPROBE) uprobe_notify_resume(regs); @@ -169,4 +172,6 @@ void do_notify_resume(struct pt_regs *regs, unsigned long thread_info_flags) clear_thread_flag(TIF_NOTIFY_RESUME); tracehook_notify_resume(regs); } + + user_enter(); } From 5d1c57451154047ab31aadfbc712bb8aa174bfb3 Mon Sep 17 00:00:00 2001 From: Li Zhong Date: Mon, 13 May 2013 16:16:43 +0000 Subject: [PATCH 146/198] powerpc: Use the new schedule_user API on userspace preemption This patch corresponds to [PATCH] x86: Use the new schedule_user API on userspace preemption commit 0430499ce9d78691f3985962021b16bf8f8a8048 Signed-off-by: Li Zhong Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/include/asm/context_tracking.h | 10 ++++++++++ arch/powerpc/kernel/entry_64.S | 3 ++- 2 files changed, 12 insertions(+), 1 deletion(-) create mode 100644 arch/powerpc/include/asm/context_tracking.h diff --git a/arch/powerpc/include/asm/context_tracking.h b/arch/powerpc/include/asm/context_tracking.h new file mode 100644 index 000000000000..b6f5a33b8ee2 --- /dev/null +++ b/arch/powerpc/include/asm/context_tracking.h @@ -0,0 +1,10 @@ +#ifndef _ASM_POWERPC_CONTEXT_TRACKING_H +#define _ASM_POWERPC_CONTEXT_TRACKING_H + +#ifdef CONFIG_CONTEXT_TRACKING +#define SCHEDULE_USER bl .schedule_user +#else +#define SCHEDULE_USER bl .schedule +#endif + +#endif diff --git a/arch/powerpc/kernel/entry_64.S b/arch/powerpc/kernel/entry_64.S index 48e8a86e7cb2..794889b588cd 100644 --- a/arch/powerpc/kernel/entry_64.S +++ b/arch/powerpc/kernel/entry_64.S @@ -33,6 +33,7 @@ #include #include #include +#include /* * System calls. @@ -632,7 +633,7 @@ _GLOBAL(ret_from_except_lite) andi. r0,r4,_TIF_NEED_RESCHED beq 1f bl .restore_interrupts - bl .schedule + SCHEDULE_USER b .ret_from_except_lite 1: bl .save_nvgprs From a1797b2fd2051515689fd54a9db8fd20ebd5e5f7 Mon Sep 17 00:00:00 2001 From: Li Zhong Date: Mon, 13 May 2013 16:16:44 +0000 Subject: [PATCH 147/198] powerpc: select HAVE_CONTEXT_TRACKING for pSeries Start context tracking support from pSeries. Signed-off-by: Li Zhong Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/platforms/pseries/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/powerpc/platforms/pseries/Kconfig b/arch/powerpc/platforms/pseries/Kconfig index 9a0941bc4d31..023b288f895b 100644 --- a/arch/powerpc/platforms/pseries/Kconfig +++ b/arch/powerpc/platforms/pseries/Kconfig @@ -18,6 +18,7 @@ config PPC_PSERIES select PPC_PCI_CHOICE if EXPERT select ZLIB_DEFLATE select PPC_DOORBELL + select HAVE_CONTEXT_TRACKING default y config PPC_SPLPAR From d52f2dc40b52201700001e868093c5ec827a8f33 Mon Sep 17 00:00:00 2001 From: Michael Neuling Date: Mon, 13 May 2013 18:44:56 +0000 Subject: [PATCH 148/198] powerpc/perf: Move BHRB code into CONFIG_PPC64 region The new Branch History Rolling buffer (BHRB) code is only useful on 64bit processors, so move it into the #ifdef CONFIG_PPC64 region. This avoids code bloat on 32bit systems. Signed-off-by: Michael Neuling Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/perf/core-book3s.c | 248 ++++++++++++++++---------------- 1 file changed, 127 insertions(+), 121 deletions(-) diff --git a/arch/powerpc/perf/core-book3s.c b/arch/powerpc/perf/core-book3s.c index c627843c5b2e..843bb8be8380 100644 --- a/arch/powerpc/perf/core-book3s.c +++ b/arch/powerpc/perf/core-book3s.c @@ -100,6 +100,10 @@ static inline int siar_valid(struct pt_regs *regs) return 1; } +static inline void power_pmu_bhrb_enable(struct perf_event *event) {} +static inline void power_pmu_bhrb_disable(struct perf_event *event) {} +void power_pmu_flush_branch_stack(void) {} +static inline void power_pmu_bhrb_read(struct cpu_hw_events *cpuhw) {} #endif /* CONFIG_PPC32 */ static bool regs_use_siar(struct pt_regs *regs) @@ -308,6 +312,129 @@ static inline int siar_valid(struct pt_regs *regs) return 1; } + +/* Reset all possible BHRB entries */ +static void power_pmu_bhrb_reset(void) +{ + asm volatile(PPC_CLRBHRB); +} + +static void power_pmu_bhrb_enable(struct perf_event *event) +{ + struct cpu_hw_events *cpuhw = &__get_cpu_var(cpu_hw_events); + + if (!ppmu->bhrb_nr) + return; + + /* Clear BHRB if we changed task context to avoid data leaks */ + if (event->ctx->task && cpuhw->bhrb_context != event->ctx) { + power_pmu_bhrb_reset(); + cpuhw->bhrb_context = event->ctx; + } + cpuhw->bhrb_users++; +} + +static void power_pmu_bhrb_disable(struct perf_event *event) +{ + struct cpu_hw_events *cpuhw = &__get_cpu_var(cpu_hw_events); + + if (!ppmu->bhrb_nr) + return; + + cpuhw->bhrb_users--; + WARN_ON_ONCE(cpuhw->bhrb_users < 0); + + if (!cpuhw->disabled && !cpuhw->bhrb_users) { + /* BHRB cannot be turned off when other + * events are active on the PMU. + */ + + /* avoid stale pointer */ + cpuhw->bhrb_context = NULL; + } +} + +/* Called from ctxsw to prevent one process's branch entries to + * mingle with the other process's entries during context switch. + */ +void power_pmu_flush_branch_stack(void) +{ + if (ppmu->bhrb_nr) + power_pmu_bhrb_reset(); +} + + +/* Processing BHRB entries */ +static void power_pmu_bhrb_read(struct cpu_hw_events *cpuhw) +{ + u64 val; + u64 addr; + int r_index, u_index, target, pred; + + r_index = 0; + u_index = 0; + while (r_index < ppmu->bhrb_nr) { + /* Assembly read function */ + val = read_bhrb(r_index); + + /* Terminal marker: End of valid BHRB entries */ + if (val == 0) { + break; + } else { + /* BHRB field break up */ + addr = val & BHRB_EA; + pred = val & BHRB_PREDICTION; + target = val & BHRB_TARGET; + + /* Probable Missed entry: Not applicable for POWER8 */ + if ((addr == 0) && (target == 0) && (pred == 1)) { + r_index++; + continue; + } + + /* Real Missed entry: Power8 based missed entry */ + if ((addr == 0) && (target == 1) && (pred == 1)) { + r_index++; + continue; + } + + /* Reserved condition: Not a valid entry */ + if ((addr == 0) && (target == 1) && (pred == 0)) { + r_index++; + continue; + } + + /* Is a target address */ + if (val & BHRB_TARGET) { + /* First address cannot be a target address */ + if (r_index == 0) { + r_index++; + continue; + } + + /* Update target address for the previous entry */ + cpuhw->bhrb_entries[u_index - 1].to = addr; + cpuhw->bhrb_entries[u_index - 1].mispred = pred; + cpuhw->bhrb_entries[u_index - 1].predicted = ~pred; + + /* Dont increment u_index */ + r_index++; + } else { + /* Update address, flags for current entry */ + cpuhw->bhrb_entries[u_index].from = addr; + cpuhw->bhrb_entries[u_index].mispred = pred; + cpuhw->bhrb_entries[u_index].predicted = ~pred; + + /* Successfully popullated one entry */ + u_index++; + r_index++; + } + } + } + cpuhw->bhrb_stack.nr = u_index; + return; +} + #endif /* CONFIG_PPC64 */ static void perf_event_interrupt(struct pt_regs *regs); @@ -904,47 +1031,6 @@ static int collect_events(struct perf_event *group, int max_count, return n; } -/* Reset all possible BHRB entries */ -static void power_pmu_bhrb_reset(void) -{ - asm volatile(PPC_CLRBHRB); -} - -void power_pmu_bhrb_enable(struct perf_event *event) -{ - struct cpu_hw_events *cpuhw = &__get_cpu_var(cpu_hw_events); - - if (!ppmu->bhrb_nr) - return; - - /* Clear BHRB if we changed task context to avoid data leaks */ - if (event->ctx->task && cpuhw->bhrb_context != event->ctx) { - power_pmu_bhrb_reset(); - cpuhw->bhrb_context = event->ctx; - } - cpuhw->bhrb_users++; -} - -void power_pmu_bhrb_disable(struct perf_event *event) -{ - struct cpu_hw_events *cpuhw = &__get_cpu_var(cpu_hw_events); - - if (!ppmu->bhrb_nr) - return; - - cpuhw->bhrb_users--; - WARN_ON_ONCE(cpuhw->bhrb_users < 0); - - if (!cpuhw->disabled && !cpuhw->bhrb_users) { - /* BHRB cannot be turned off when other - * events are active on the PMU. - */ - - /* avoid stale pointer */ - cpuhw->bhrb_context = NULL; - } -} - /* * Add a event to the PMU. * If all events are not already frozen, then we disable and @@ -1180,15 +1266,6 @@ int power_pmu_commit_txn(struct pmu *pmu) return 0; } -/* Called from ctxsw to prevent one process's branch entries to - * mingle with the other process's entries during context switch. - */ -void power_pmu_flush_branch_stack(void) -{ - if (ppmu->bhrb_nr) - power_pmu_bhrb_reset(); -} - /* * Return 1 if we might be able to put event on a limited PMC, * or 0 if not. @@ -1458,77 +1535,6 @@ struct pmu power_pmu = { .flush_branch_stack = power_pmu_flush_branch_stack, }; -/* Processing BHRB entries */ -void power_pmu_bhrb_read(struct cpu_hw_events *cpuhw) -{ - u64 val; - u64 addr; - int r_index, u_index, target, pred; - - r_index = 0; - u_index = 0; - while (r_index < ppmu->bhrb_nr) { - /* Assembly read function */ - val = read_bhrb(r_index); - - /* Terminal marker: End of valid BHRB entries */ - if (val == 0) { - break; - } else { - /* BHRB field break up */ - addr = val & BHRB_EA; - pred = val & BHRB_PREDICTION; - target = val & BHRB_TARGET; - - /* Probable Missed entry: Not applicable for POWER8 */ - if ((addr == 0) && (target == 0) && (pred == 1)) { - r_index++; - continue; - } - - /* Real Missed entry: Power8 based missed entry */ - if ((addr == 0) && (target == 1) && (pred == 1)) { - r_index++; - continue; - } - - /* Reserved condition: Not a valid entry */ - if ((addr == 0) && (target == 1) && (pred == 0)) { - r_index++; - continue; - } - - /* Is a target address */ - if (val & BHRB_TARGET) { - /* First address cannot be a target address */ - if (r_index == 0) { - r_index++; - continue; - } - - /* Update target address for the previous entry */ - cpuhw->bhrb_entries[u_index - 1].to = addr; - cpuhw->bhrb_entries[u_index - 1].mispred = pred; - cpuhw->bhrb_entries[u_index - 1].predicted = ~pred; - - /* Dont increment u_index */ - r_index++; - } else { - /* Update address, flags for current entry */ - cpuhw->bhrb_entries[u_index].from = addr; - cpuhw->bhrb_entries[u_index].mispred = pred; - cpuhw->bhrb_entries[u_index].predicted = ~pred; - - /* Successfully popullated one entry */ - u_index++; - r_index++; - } - } - } - cpuhw->bhrb_stack.nr = u_index; - return; -} - /* * A counter has overflowed; update its count and record * things if requested. Note that interrupts are hard-disabled From 506e70d13236dfdb0bc15e0914298ab0a7b7f4df Mon Sep 17 00:00:00 2001 From: Michael Neuling Date: Mon, 13 May 2013 18:44:57 +0000 Subject: [PATCH 149/198] powerpc/pmu: Fix order of interpreting BHRB target entries The current Branch History Rolling Buffer (BHRB) code misinterprets the order of entries in the hardware buffer. It assumes that a branch target address will be read _after_ its corresponding branch. In reality the branch target comes before (lower mfbhrb entry) it's corresponding branch. This is a rewrite of the code to take this into account. Signed-off-by: Michael Neuling Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/perf/core-book3s.c | 93 +++++++++++++++++---------------- 1 file changed, 48 insertions(+), 45 deletions(-) diff --git a/arch/powerpc/perf/core-book3s.c b/arch/powerpc/perf/core-book3s.c index 843bb8be8380..3fdfe4575b8f 100644 --- a/arch/powerpc/perf/core-book3s.c +++ b/arch/powerpc/perf/core-book3s.c @@ -363,72 +363,75 @@ void power_pmu_flush_branch_stack(void) power_pmu_bhrb_reset(); } - /* Processing BHRB entries */ -static void power_pmu_bhrb_read(struct cpu_hw_events *cpuhw) +void power_pmu_bhrb_read(struct cpu_hw_events *cpuhw) { u64 val; u64 addr; - int r_index, u_index, target, pred; + int r_index, u_index, pred; r_index = 0; u_index = 0; while (r_index < ppmu->bhrb_nr) { /* Assembly read function */ - val = read_bhrb(r_index); - - /* Terminal marker: End of valid BHRB entries */ - if (val == 0) { + val = read_bhrb(r_index++); + if (!val) + /* Terminal marker: End of valid BHRB entries */ break; - } else { - /* BHRB field break up */ + else { addr = val & BHRB_EA; pred = val & BHRB_PREDICTION; - target = val & BHRB_TARGET; - /* Probable Missed entry: Not applicable for POWER8 */ - if ((addr == 0) && (target == 0) && (pred == 1)) { - r_index++; + if (!addr) + /* invalid entry */ continue; - } - /* Real Missed entry: Power8 based missed entry */ - if ((addr == 0) && (target == 1) && (pred == 1)) { - r_index++; - continue; - } + /* Branches are read most recent first (ie. mfbhrb 0 is + * the most recent branch). + * There are two types of valid entries: + * 1) a target entry which is the to address of a + * computed goto like a blr,bctr,btar. The next + * entry read from the bhrb will be branch + * corresponding to this target (ie. the actual + * blr/bctr/btar instruction). + * 2) a from address which is an actual branch. If a + * target entry proceeds this, then this is the + * matching branch for that target. If this is not + * following a target entry, then this is a branch + * where the target is given as an immediate field + * in the instruction (ie. an i or b form branch). + * In this case we need to read the instruction from + * memory to determine the target/to address. + */ - /* Reserved condition: Not a valid entry */ - if ((addr == 0) && (target == 1) && (pred == 0)) { - r_index++; - continue; - } - - /* Is a target address */ if (val & BHRB_TARGET) { - /* First address cannot be a target address */ - if (r_index == 0) { - r_index++; - continue; - } - - /* Update target address for the previous entry */ - cpuhw->bhrb_entries[u_index - 1].to = addr; - cpuhw->bhrb_entries[u_index - 1].mispred = pred; - cpuhw->bhrb_entries[u_index - 1].predicted = ~pred; - - /* Dont increment u_index */ - r_index++; - } else { - /* Update address, flags for current entry */ - cpuhw->bhrb_entries[u_index].from = addr; + /* Target branches use two entries + * (ie. computed gotos/XL form) + */ + cpuhw->bhrb_entries[u_index].to = addr; cpuhw->bhrb_entries[u_index].mispred = pred; cpuhw->bhrb_entries[u_index].predicted = ~pred; - /* Successfully popullated one entry */ - u_index++; - r_index++; + /* Get from address in next entry */ + val = read_bhrb(r_index++); + addr = val & BHRB_EA; + if (val & BHRB_TARGET) { + /* Shouldn't have two targets in a + row.. Reset index and try again */ + r_index--; + addr = 0; + } + cpuhw->bhrb_entries[u_index].from = addr; + } else { + /* Branches to immediate field + (ie I or B form) */ + cpuhw->bhrb_entries[u_index].from = addr; + cpuhw->bhrb_entries[u_index].to = 0; + cpuhw->bhrb_entries[u_index].mispred = pred; + cpuhw->bhrb_entries[u_index].predicted = ~pred; } + u_index++; + } } cpuhw->bhrb_stack.nr = u_index; From 691231846cebfe1fbbcf898c8af17a569dbb5463 Mon Sep 17 00:00:00 2001 From: Michael Neuling Date: Mon, 13 May 2013 18:44:58 +0000 Subject: [PATCH 150/198] powerpc/perf: Fix setting of "to" addresses for BHRB Currently we only set the "to" address in the branch stack when the CPU explicitly gives us a value. Unfortunately it only does this for XL form branches (eg blr, bctr, bctar) and not I and B form branches (eg b, bc). Fortunately if we read the instruction from memory we can extract the offset of a branch and calculate the target address. This adds a function power_pmu_bhrb_to() to calculate the target/to address of the corresponding I and B form branches. It handles branches in both user and kernel spaces. It also plumbs this into the perf brhb reading code. Signed-off-by: Michael Neuling Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/perf/core-book3s.c | 31 ++++++++++++++++++++++++++++++- 1 file changed, 30 insertions(+), 1 deletion(-) diff --git a/arch/powerpc/perf/core-book3s.c b/arch/powerpc/perf/core-book3s.c index 3fdfe4575b8f..426180b84978 100644 --- a/arch/powerpc/perf/core-book3s.c +++ b/arch/powerpc/perf/core-book3s.c @@ -13,11 +13,13 @@ #include #include #include +#include #include #include #include #include #include +#include #define BHRB_MAX_ENTRIES 32 #define BHRB_TARGET 0x0000000000000002 @@ -362,6 +364,32 @@ void power_pmu_flush_branch_stack(void) if (ppmu->bhrb_nr) power_pmu_bhrb_reset(); } +/* Calculate the to address for a branch */ +static __u64 power_pmu_bhrb_to(u64 addr) +{ + unsigned int instr; + int ret; + __u64 target; + + if (is_kernel_addr(addr)) + return branch_target((unsigned int *)addr); + + /* Userspace: need copy instruction here then translate it */ + pagefault_disable(); + ret = __get_user_inatomic(instr, (unsigned int __user *)addr); + if (ret) { + pagefault_enable(); + return 0; + } + pagefault_enable(); + + target = branch_target(&instr); + if ((!target) || (instr & BRANCH_ABSOLUTE)) + return target; + + /* Translate relative branch target from kernel to user address */ + return target - (unsigned long)&instr + addr; +} /* Processing BHRB entries */ void power_pmu_bhrb_read(struct cpu_hw_events *cpuhw) @@ -426,7 +454,8 @@ void power_pmu_bhrb_read(struct cpu_hw_events *cpuhw) /* Branches to immediate field (ie I or B form) */ cpuhw->bhrb_entries[u_index].from = addr; - cpuhw->bhrb_entries[u_index].to = 0; + cpuhw->bhrb_entries[u_index].to = + power_pmu_bhrb_to(addr); cpuhw->bhrb_entries[u_index].mispred = pred; cpuhw->bhrb_entries[u_index].predicted = ~pred; } From e34166ad63eac4d0fa98b4c4ed7a98202a18faef Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Tue, 14 May 2013 17:02:11 +1000 Subject: [PATCH 151/198] powerpc: Set show_unhandled_signals to 1 by default Just like other architectures Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/kernel/signal.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/powerpc/kernel/signal.c b/arch/powerpc/kernel/signal.c index d63b5024c563..577a8aa69c6e 100644 --- a/arch/powerpc/kernel/signal.c +++ b/arch/powerpc/kernel/signal.c @@ -25,7 +25,7 @@ * through debug.exception-trace sysctl. */ -int show_unhandled_signals = 0; +int show_unhandled_signals = 1; /* * Allocate space for the signal frame From f7ea0fd639c2c48d3c61b6eec75362be290c6874 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Mon, 13 May 2013 21:40:27 +0200 Subject: [PATCH 152/198] tick: Don't invoke tick_nohz_stop_sched_tick() if the cpu is offline commit 5b39939a4 (nohz: Move ts->idle_calls incrementation into strict idle logic) moved code out of tick_nohz_stop_sched_tick() and missed to bail out when the cpu is offline. That's causing subsequent failures as an offline CPU is supposed to die and not to fiddle with nohz magic. Return false in can_stop_idle_tick() if the cpu is offline. Reported-and-tested-by: Jiri Kosina Reported-and-tested-by: Prarit Bhargava Cc: Frederic Weisbecker Cc: Borislav Petkov Cc: Tony Luck Cc: x86@kernel.org Link: http://lkml.kernel.org/r/alpine.LFD.2.02.1305132138160.2863@ionos Signed-off-by: Thomas Gleixner --- kernel/time/tick-sched.c | 1 + 1 file changed, 1 insertion(+) diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c index 0eed1db2d792..012142187db9 100644 --- a/kernel/time/tick-sched.c +++ b/kernel/time/tick-sched.c @@ -469,6 +469,7 @@ static bool can_stop_idle_tick(int cpu, struct tick_sched *ts) if (unlikely(!cpu_online(cpu))) { if (cpu == tick_do_timer_cpu) tick_do_timer_cpu = TICK_DO_TIMER_NONE; + return false; } if (unlikely(ts->nohz_mode == NOHZ_MODE_INACTIVE)) From b47430d3adbedbfdb5979ba4874f5dadf94f16b1 Mon Sep 17 00:00:00 2001 From: "Srivatsa S. Bhat" Date: Tue, 14 May 2013 04:01:27 +0530 Subject: [PATCH 153/198] rcu/idle: Wrap cpu-idle poll mode within rcu_idle_enter/exit MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Bjørn Mork reported the following warning when running powertop. [ 49.289034] ------------[ cut here ]------------ [ 49.289055] WARNING: at kernel/rcutree.c:502 rcu_eqs_exit_common.isra.48+0x3d/0x125() [ 49.289244] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 3.10.0-bisect-rcu-warn+ #107 [ 49.289251] ffffffff8157d8c8 ffffffff81801e28 ffffffff8137e4e3 ffffffff81801e68 [ 49.289260] ffffffff8103094f ffffffff81801e68 0000000000000000 ffff88023afcd9b0 [ 49.289268] 0000000000000000 0140000000000000 ffff88023bee7700 ffffffff81801e78 [ 49.289276] Call Trace: [ 49.289285] [] dump_stack+0x19/0x1b [ 49.289293] [] warn_slowpath_common+0x62/0x7b [ 49.289300] [] warn_slowpath_null+0x15/0x17 [ 49.289306] [] rcu_eqs_exit_common.isra.48+0x3d/0x125 [ 49.289314] [] ? trace_hardirqs_off_caller+0x37/0xa6 [ 49.289320] [] rcu_idle_exit+0x85/0xa8 [ 49.289327] [] trace_cpu_idle_rcuidle+0xae/0xff [ 49.289334] [] cpu_startup_entry+0x72/0x115 [ 49.289341] [] rest_init+0x149/0x150 [ 49.289347] [] ? csum_partial_copy_generic+0x16c/0x16c [ 49.289355] [] start_kernel+0x3f0/0x3fd [ 49.289362] [] ? repair_env_string+0x5a/0x5a [ 49.289368] [] x86_64_start_reservations+0x2a/0x2c [ 49.289375] [] x86_64_start_kernel+0xcd/0xd1 [ 49.289379] ---[ end trace 07a1cc95e29e9036 ]--- The warning is that 'rdtp->dynticks' has an unexpected value, which roughly translates to - the calls to rcu_idle_enter() and rcu_idle_exit() were not made in the correct order, or otherwise messed up. And Bjørn's painstaking debugging indicated that this happens when the idle loop enters the poll mode. Looking at the poll function cpu_idle_poll(), and the implementation of trace_cpu_idle_rcuidle(), the problem becomes very clear: cpu_idle_poll() lacks calls to rcu_idle_enter/exit(), and trace_cpu_idle_rcuidle() calls them in the reverse order - first rcu_idle_exit(), and then rcu_idle_enter(). Hence the even/odd alternative sequencing of rdtp->dynticks goes for a toss. And powertop readily triggers this because powertop uses the idle-tracing infrastructure extensively. So, to fix this, wrap the code in cpu_idle_poll() within rcu_idle_enter/exit(), so that it blends properly with the calls inside trace_cpu_idle_rcuidle() and thus get the function ordering right. Reported-and-tested-by: Bjørn Mork Cc: Paul McKenney Cc: Steven Rostedt Cc: Dipankar Sarma Signed-off-by: Srivatsa S. Bhat Link: http://lkml.kernel.org/r/519169BF.4080208@linux.vnet.ibm.com Signed-off-by: Thomas Gleixner --- kernel/cpu/idle.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/kernel/cpu/idle.c b/kernel/cpu/idle.c index 8b86c0c68edf..d5585f5e038e 100644 --- a/kernel/cpu/idle.c +++ b/kernel/cpu/idle.c @@ -40,11 +40,13 @@ __setup("hlt", cpu_idle_nopoll_setup); static inline int cpu_idle_poll(void) { + rcu_idle_enter(); trace_cpu_idle_rcuidle(0, smp_processor_id()); local_irq_enable(); while (!need_resched()) cpu_relax(); trace_cpu_idle_rcuidle(PWR_EVENT_EXIT, smp_processor_id()); + rcu_idle_exit(); return 1; } From 42a5cf46cd56f46267d2a9fcf2655f4078cd3042 Mon Sep 17 00:00:00 2001 From: Tirupathi Reddy Date: Tue, 14 May 2013 13:59:02 +0530 Subject: [PATCH 154/198] timer: Don't reinitialize the cpu base lock during CPU_UP_PREPARE An inactive timer's base can refer to a offline cpu's base. In the current code, cpu_base's lock is blindly reinitialized each time a CPU is brought up. If a CPU is brought online during the period that another thread is trying to modify an inactive timer on that CPU with holding its timer base lock, then the lock will be reinitialized under its feet. This leads to following SPIN_BUG(). <0> BUG: spinlock already unlocked on CPU#3, kworker/u:3/1466 <0> lock: 0xe3ebe000, .magic: dead4ead, .owner: kworker/u:3/1466, .owner_cpu: 1 <4> [] (unwind_backtrace+0x0/0x11c) from [] (do_raw_spin_unlock+0x40/0xcc) <4> [] (do_raw_spin_unlock+0x40/0xcc) from [] (_raw_spin_unlock+0x8/0x30) <4> [] (_raw_spin_unlock+0x8/0x30) from [] (mod_timer+0x294/0x310) <4> [] (mod_timer+0x294/0x310) from [] (queue_delayed_work_on+0x104/0x120) <4> [] (queue_delayed_work_on+0x104/0x120) from [] (sdhci_msm_bus_voting+0x88/0x9c) <4> [] (sdhci_msm_bus_voting+0x88/0x9c) from [] (sdhci_disable+0x40/0x48) <4> [] (sdhci_disable+0x40/0x48) from [] (mmc_release_host+0x4c/0xb0) <4> [] (mmc_release_host+0x4c/0xb0) from [] (mmc_sd_detect+0x90/0xfc) <4> [] (mmc_sd_detect+0x90/0xfc) from [] (mmc_rescan+0x7c/0x2c4) <4> [] (mmc_rescan+0x7c/0x2c4) from [] (process_one_work+0x27c/0x484) <4> [] (process_one_work+0x27c/0x484) from [] (worker_thread+0x210/0x3b0) <4> [] (worker_thread+0x210/0x3b0) from [] (kthread+0x80/0x8c) <4> [] (kthread+0x80/0x8c) from [] (kernel_thread_exit+0x0/0x8) As an example, this particular crash occurred when CPU #3 is executing mod_timer() on an inactive timer whose base is refered to offlined CPU #2. The code locked the timer_base corresponding to CPU #2. Before it could proceed, CPU #2 came online and reinitialized the spinlock corresponding to its base. Thus now CPU #3 held a lock which was reinitialized. When CPU #3 finally ended up unlocking the old cpu_base corresponding to CPU #2, we hit the above SPIN_BUG(). CPU #0 CPU #3 CPU #2 ------ ------- ------- ..... ...... mod_timer() lock_timer_base spin_lock_irqsave(&base->lock) cpu_up(2) ..... ...... init_timers_cpu() .... ..... spin_lock_init(&base->lock) ..... spin_unlock_irqrestore(&base->lock) ...... Allocation of per_cpu timer vector bases is done only once under "tvec_base_done[]" check. In the current code, spinlock_initialization of base->lock isn't under this check. When a CPU is up each time the base lock is reinitialized. Move base spinlock initialization under the check. Signed-off-by: Tirupathi Reddy Cc: stable@vger.kernel.org Link: http://lkml.kernel.org/r/1368520142-4136-1-git-send-email-tirupath@codeaurora.org Signed-off-by: Thomas Gleixner --- kernel/timer.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/timer.c b/kernel/timer.c index 09bca8ce9771..7376589adc28 100644 --- a/kernel/timer.c +++ b/kernel/timer.c @@ -1539,12 +1539,12 @@ static int __cpuinit init_timers_cpu(int cpu) boot_done = 1; base = &boot_tvec_bases; } + spin_lock_init(&base->lock); tvec_base_done[cpu] = 1; } else { base = per_cpu(tvec_bases, cpu); } - spin_lock_init(&base->lock); for (j = 0; j < TVN_SIZE; j++) { INIT_LIST_HEAD(base->tv5.vec + j); From b4f711ee03d28f776fd2324fd0bd999cc428e4d2 Mon Sep 17 00:00:00 2001 From: John Stultz Date: Wed, 24 Apr 2013 11:32:56 -0700 Subject: [PATCH 155/198] time: Revert ALWAYS_USE_PERSISTENT_CLOCK compile time optimizaitons Kay Sievers noted that the ALWAYS_USE_PERSISTENT_CLOCK config, which enables some minor compile time optimization to avoid uncessary code in mostly the suspend/resume path could cause problems for userland. In particular, the dependency for RTC_HCTOSYS on !ALWAYS_USE_PERSISTENT_CLOCK, which avoids setting the time twice and simplifies suspend/resume, has the side effect of causing the /sys/class/rtc/rtcN/hctosys flag to always be zero, and this flag is commonly used by udev to setup the /dev/rtc symlink to /dev/rtcN, which can cause pain for older applications. While the udev rules could use some work to be less fragile, breaking userland should strongly be avoided. Additionally the compile time optimizations are fairly minor, and the code being optimized is likely to be reworked in the future, so lets revert this change. Reported-by: Kay Sievers Signed-off-by: John Stultz Cc: stable #3.9 Cc: Feng Tang Cc: Jason Gunthorpe Link: http://lkml.kernel.org/r/1366828376-18124-1-git-send-email-john.stultz@linaro.org Signed-off-by: Thomas Gleixner --- arch/x86/Kconfig | 1 - drivers/rtc/Kconfig | 2 -- include/linux/time.h | 4 ---- kernel/time/Kconfig | 5 ----- 4 files changed, 12 deletions(-) diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 5db2117ae288..45c41249321f 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -108,7 +108,6 @@ config X86 select GENERIC_CLOCKEVENTS_BROADCAST if X86_64 || (X86_32 && X86_LOCAL_APIC) select GENERIC_TIME_VSYSCALL if X86_64 select KTIME_SCALAR if X86_32 - select ALWAYS_USE_PERSISTENT_CLOCK select GENERIC_STRNCPY_FROM_USER select GENERIC_STRNLEN_USER select HAVE_CONTEXT_TRACKING if X86_64 diff --git a/drivers/rtc/Kconfig b/drivers/rtc/Kconfig index 0c81915b1997..b9838130a7b0 100644 --- a/drivers/rtc/Kconfig +++ b/drivers/rtc/Kconfig @@ -20,7 +20,6 @@ if RTC_CLASS config RTC_HCTOSYS bool "Set system time from RTC on startup and resume" default y - depends on !ALWAYS_USE_PERSISTENT_CLOCK help If you say yes here, the system time (wall clock) will be set using the value read from a specified RTC device. This is useful to avoid @@ -29,7 +28,6 @@ config RTC_HCTOSYS config RTC_SYSTOHC bool "Set the RTC time based on NTP synchronization" default y - depends on !ALWAYS_USE_PERSISTENT_CLOCK help If you say yes here, the system time (wall clock) will be stored in the RTC specified by RTC_HCTOSYS_DEVICE approximately every 11 diff --git a/include/linux/time.h b/include/linux/time.h index 22d81b3c955b..d5d229b2e5af 100644 --- a/include/linux/time.h +++ b/include/linux/time.h @@ -117,14 +117,10 @@ static inline bool timespec_valid_strict(const struct timespec *ts) extern bool persistent_clock_exist; -#ifdef ALWAYS_USE_PERSISTENT_CLOCK -#define has_persistent_clock() true -#else static inline bool has_persistent_clock(void) { return persistent_clock_exist; } -#endif extern void read_persistent_clock(struct timespec *ts); extern void read_boot_clock(struct timespec *ts); diff --git a/kernel/time/Kconfig b/kernel/time/Kconfig index 24510d84efd7..b69692250af4 100644 --- a/kernel/time/Kconfig +++ b/kernel/time/Kconfig @@ -12,11 +12,6 @@ config CLOCKSOURCE_WATCHDOG config ARCH_CLOCKSOURCE_DATA bool -# Platforms has a persistent clock -config ALWAYS_USE_PERSISTENT_CLOCK - bool - default n - # Timekeeping vsyscall support config GENERIC_TIME_VSYSCALL bool From 60705c89460fdc7227f2d153b68b3f34814738a4 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (Red Hat)" Date: Tue, 14 May 2013 15:40:48 -0400 Subject: [PATCH 156/198] tracing: Fix leaks of filter preds Special preds are created when folding a series of preds that can be done in serial. These are allocated in an ops field of the pred structure. But they were never freed, causing memory leaks. This was discovered using the kmemleak checker: unreferenced object 0xffff8800797fd5e0 (size 32): comm "swapper/0", pid 1, jiffies 4294690605 (age 104.608s) hex dump (first 32 bytes): 00 00 01 00 03 00 05 00 07 00 09 00 0b 00 0d 00 ................ 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ backtrace: [] kmemleak_alloc+0x73/0x98 [] kmemleak_alloc_recursive.constprop.42+0x16/0x18 [] __kmalloc+0xd7/0x125 [] kcalloc.constprop.24+0x2d/0x2f [] fold_pred_tree_cb+0xa9/0xf4 [] walk_pred_tree+0x47/0xcc [] replace_preds.isra.20+0x6f8/0x72f [] create_filter+0x4e/0x8b [] ftrace_test_event_filter+0x5a/0x155 [] do_one_initcall+0xa0/0x137 [] kernel_init_freeable+0x14d/0x1dc [] kernel_init+0xe/0xdb [] ret_from_fork+0x7c/0xb0 [] 0xffffffffffffffff Cc: Tom Zanussi Cc: stable@vger.kernel.org # 2.6.39+ Signed-off-by: Steven Rostedt --- kernel/trace/trace_events_filter.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/kernel/trace/trace_events_filter.c b/kernel/trace/trace_events_filter.c index a6361178de5a..e1b653f7e1ca 100644 --- a/kernel/trace/trace_events_filter.c +++ b/kernel/trace/trace_events_filter.c @@ -750,7 +750,11 @@ static int filter_set_pred(struct event_filter *filter, static void __free_preds(struct event_filter *filter) { + int i; + if (filter->preds) { + for (i = 0; i < filter->n_preds; i++) + kfree(filter->preds[i].ops); kfree(filter->preds); filter->preds = NULL; } From c02c7e65d9b13670e34bc523744cf4f6e99c198a Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Mon, 13 May 2013 20:58:34 +0900 Subject: [PATCH 157/198] tracing/kprobes: Use rcu_dereference_raw for tp->files Use rcu_dereference_raw() for accessing tp->files. Because the write-side uses rcu_assign_pointer() for memory barrier, the read-side also has to use rcu_dereference_raw() with read memory barrier. Link: http://lkml.kernel.org/r/20130513115834.6545.17022.stgit@mhiramat-M0-7522 Cc: Srikar Dronamraju Cc: Oleg Nesterov Cc: Tom Zanussi Cc: Frederic Weisbecker Cc: Ingo Molnar Signed-off-by: Masami Hiramatsu Signed-off-by: Steven Rostedt --- kernel/trace/trace_kprobe.c | 47 +++++++++++++++++++++++++++++-------- 1 file changed, 37 insertions(+), 10 deletions(-) diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 636d45fe69b3..0a3d8d5c483d 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -185,9 +185,14 @@ static struct trace_probe *find_trace_probe(const char *event, static int trace_probe_nr_files(struct trace_probe *tp) { - struct ftrace_event_file **file = tp->files; + struct ftrace_event_file **file; int ret = 0; + /* + * Since all tp->files updater is protected by probe_enable_lock, + * we don't need to lock an rcu_read_lock. + */ + file = rcu_dereference_raw(tp->files); if (file) while (*(file++)) ret++; @@ -209,9 +214,10 @@ enable_trace_probe(struct trace_probe *tp, struct ftrace_event_file *file) mutex_lock(&probe_enable_lock); if (file) { - struct ftrace_event_file **new, **old = tp->files; + struct ftrace_event_file **new, **old; int n = trace_probe_nr_files(tp); + old = rcu_dereference_raw(tp->files); /* 1 is for new one and 1 is for stopper */ new = kzalloc((n + 2) * sizeof(struct ftrace_event_file *), GFP_KERNEL); @@ -251,11 +257,17 @@ enable_trace_probe(struct trace_probe *tp, struct ftrace_event_file *file) static int trace_probe_file_index(struct trace_probe *tp, struct ftrace_event_file *file) { + struct ftrace_event_file **files; int i; - if (tp->files) { - for (i = 0; tp->files[i]; i++) - if (tp->files[i] == file) + /* + * Since all tp->files updater is protected by probe_enable_lock, + * we don't need to lock an rcu_read_lock. + */ + files = rcu_dereference_raw(tp->files); + if (files) { + for (i = 0; files[i]; i++) + if (files[i] == file) return i; } @@ -274,10 +286,11 @@ disable_trace_probe(struct trace_probe *tp, struct ftrace_event_file *file) mutex_lock(&probe_enable_lock); if (file) { - struct ftrace_event_file **new, **old = tp->files; + struct ftrace_event_file **new, **old; int n = trace_probe_nr_files(tp); int i, j; + old = rcu_dereference_raw(tp->files); if (n == 0 || trace_probe_file_index(tp, file) < 0) { ret = -EINVAL; goto out_unlock; @@ -872,9 +885,16 @@ __kprobe_trace_func(struct trace_probe *tp, struct pt_regs *regs, static __kprobes void kprobe_trace_func(struct trace_probe *tp, struct pt_regs *regs) { - struct ftrace_event_file **file = tp->files; + /* + * Note: preempt is already disabled around the kprobe handler. + * However, we still need an smp_read_barrier_depends() corresponding + * to smp_wmb() in rcu_assign_pointer() to access the pointer. + */ + struct ftrace_event_file **file = rcu_dereference_raw(tp->files); + + if (unlikely(!file)) + return; - /* Note: preempt is already disabled around the kprobe handler */ while (*file) { __kprobe_trace_func(tp, regs, *file); file++; @@ -925,9 +945,16 @@ static __kprobes void kretprobe_trace_func(struct trace_probe *tp, struct kretprobe_instance *ri, struct pt_regs *regs) { - struct ftrace_event_file **file = tp->files; + /* + * Note: preempt is already disabled around the kprobe handler. + * However, we still need an smp_read_barrier_depends() corresponding + * to smp_wmb() in rcu_assign_pointer() to access the pointer. + */ + struct ftrace_event_file **file = rcu_dereference_raw(tp->files); + + if (unlikely(!file)) + return; - /* Note: preempt is already disabled around the kprobe handler */ while (*file) { __kretprobe_trace_func(tp, ri, regs, *file); file++; From 3d1fc7b0880c4db612a3d3211a808659e28af588 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Mon, 13 May 2013 20:58:37 +0900 Subject: [PATCH 158/198] tracing/kprobes: Fix a sparse warning for incorrect type in assignment Fix a sparse warning about the rcu operated pointer is defined without __rcu address space. Link: http://lkml.kernel.org/r/20130513115837.6545.23322.stgit@mhiramat-M0-7522 Cc: Srikar Dronamraju Cc: Oleg Nesterov Cc: Tom Zanussi Cc: Frederic Weisbecker Cc: Ingo Molnar Signed-off-by: Masami Hiramatsu Signed-off-by: Steven Rostedt --- kernel/trace/trace_kprobe.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 0a3d8d5c483d..81c5109b9d00 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -35,7 +35,7 @@ struct trace_probe { const char *symbol; /* symbol name */ struct ftrace_event_class class; struct ftrace_event_call call; - struct ftrace_event_file **files; + struct ftrace_event_file * __rcu *files; ssize_t size; /* trace entry size */ unsigned int nr_args; struct probe_arg args[]; From b62fdd97fcae17e483b005bafd13fadbd9840672 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Mon, 13 May 2013 20:58:39 +0900 Subject: [PATCH 159/198] tracing/kprobes: Make print_*probe_event static According to sparse warning, print_*probe_event static because those functions are not directly called from outside. Link: http://lkml.kernel.org/r/20130513115839.6545.83067.stgit@mhiramat-M0-7522 Cc: Srikar Dronamraju Cc: Oleg Nesterov Cc: Tom Zanussi Cc: Frederic Weisbecker Cc: Ingo Molnar Signed-off-by: Masami Hiramatsu Signed-off-by: Steven Rostedt --- kernel/trace/trace_kprobe.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 81c5109b9d00..9f46e98ba8f2 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -962,7 +962,7 @@ kretprobe_trace_func(struct trace_probe *tp, struct kretprobe_instance *ri, } /* Event entry printers */ -enum print_line_t +static enum print_line_t print_kprobe_event(struct trace_iterator *iter, int flags, struct trace_event *event) { @@ -998,7 +998,7 @@ partial: return TRACE_TYPE_PARTIAL_LINE; } -enum print_line_t +static enum print_line_t print_kretprobe_event(struct trace_iterator *iter, int flags, struct trace_event *event) { From fc1f7d5606487ae28d6c84e95401952927d7379e Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Thu, 25 Apr 2013 20:31:43 +0000 Subject: [PATCH 160/198] clocksource: apb_timer: Remove unsused function Signed-off-by: Thomas Gleixner Acked-by: John Stultz Cc: Magnus Damm Acked-by: Jamie Iles Link: http://lkml.kernel.org/r/20130425143435.558006195@linutronix.de Signed-off-by: Thomas Gleixner --- drivers/clocksource/dw_apb_timer.c | 12 ------------ include/linux/dw_apb_timer.h | 1 - 2 files changed, 13 deletions(-) diff --git a/drivers/clocksource/dw_apb_timer.c b/drivers/clocksource/dw_apb_timer.c index 8c2a35f26d9b..e54ca1062d8e 100644 --- a/drivers/clocksource/dw_apb_timer.c +++ b/drivers/clocksource/dw_apb_timer.c @@ -387,15 +387,3 @@ cycle_t dw_apb_clocksource_read(struct dw_apb_clocksource *dw_cs) { return (cycle_t)~apbt_readl(&dw_cs->timer, APBTMR_N_CURRENT_VALUE); } - -/** - * dw_apb_clocksource_unregister() - unregister and free a clocksource. - * - * @dw_cs: The clocksource to unregister/free. - */ -void dw_apb_clocksource_unregister(struct dw_apb_clocksource *dw_cs) -{ - clocksource_unregister(&dw_cs->cs); - - kfree(dw_cs); -} diff --git a/include/linux/dw_apb_timer.h b/include/linux/dw_apb_timer.h index dd755ce2a5eb..b1cd9597c241 100644 --- a/include/linux/dw_apb_timer.h +++ b/include/linux/dw_apb_timer.h @@ -51,7 +51,6 @@ dw_apb_clocksource_init(unsigned rating, const char *name, void __iomem *base, void dw_apb_clocksource_register(struct dw_apb_clocksource *dw_cs); void dw_apb_clocksource_start(struct dw_apb_clocksource *dw_cs); cycle_t dw_apb_clocksource_read(struct dw_apb_clocksource *dw_cs); -void dw_apb_clocksource_unregister(struct dw_apb_clocksource *dw_cs); extern void dw_apb_timer_init(void); #endif /* __DW_APB_TIMER_H__ */ From 5d33b883aed81c6fbcd09c6f7c3619eee850a7e2 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Thu, 25 Apr 2013 20:31:43 +0000 Subject: [PATCH 161/198] clocksource: Always verify highres capability If a clocksource has a (wrong) high rating, but can't be used as a timebase for oneshot tick mode, it is unconditionally selected even when the system is already in oneshot tick mode. This causes full system failure. Verify the clocksource selection against the oneshot mode. Signed-off-by: Thomas Gleixner Acked-by: John Stultz Cc: Magnus Damm Link: http://lkml.kernel.org/r/20130425143435.635040849@linutronix.de Signed-off-by: Thomas Gleixner --- kernel/time/clocksource.c | 31 ++++++++++++++++++++++++++----- 1 file changed, 26 insertions(+), 5 deletions(-) diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c index c9583382141a..dda5c7130d93 100644 --- a/kernel/time/clocksource.c +++ b/kernel/time/clocksource.c @@ -553,6 +553,26 @@ static u64 clocksource_max_deferment(struct clocksource *cs) #ifndef CONFIG_ARCH_USES_GETTIMEOFFSET +static struct clocksource *clocksource_find_best(bool oneshot) +{ + struct clocksource *cs; + + if (!finished_booting || list_empty(&clocksource_list)) + return NULL; + + /* + * We pick the clocksource with the highest rating. If oneshot + * mode is active, we pick the highres valid clocksource with + * the best rating. + */ + list_for_each_entry(cs, &clocksource_list, list) { + if (oneshot && !(cs->flags & CLOCK_SOURCE_VALID_FOR_HRES)) + continue; + return cs; + } + return NULL; +} + /** * clocksource_select - Select the best clocksource available * @@ -563,12 +583,14 @@ static u64 clocksource_max_deferment(struct clocksource *cs) */ static void clocksource_select(void) { + bool oneshot = tick_oneshot_mode_active(); struct clocksource *best, *cs; - if (!finished_booting || list_empty(&clocksource_list)) + /* Find the best suitable clocksource */ + best = clocksource_find_best(oneshot); + if (!best) return; - /* First clocksource on the list has the best rating. */ - best = list_first_entry(&clocksource_list, struct clocksource, list); + /* Check for the override clocksource. */ list_for_each_entry(cs, &clocksource_list, list) { if (strcmp(cs->name, override_name) != 0) @@ -578,8 +600,7 @@ static void clocksource_select(void) * capable clocksource if the tick code is in oneshot * mode (highres or nohz) */ - if (!(cs->flags & CLOCK_SOURCE_VALID_FOR_HRES) && - tick_oneshot_mode_active()) { + if (!(cs->flags & CLOCK_SOURCE_VALID_FOR_HRES) && oneshot) { /* Override clocksource cannot be used. */ printk(KERN_WARNING "Override clocksource %s is not " "HRT compatible. Cannot switch while in " From ba919d1caa2e624eb8c6cae1f2ce0a253e697d45 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Thu, 25 Apr 2013 20:31:44 +0000 Subject: [PATCH 162/198] clocksource: Let timekeeping_notify return success/error timekeeping_notify() can fail due cs->enable() failure. Though the caller does not notice and happily keeps the wrong clocksource as the current one. Let the caller know about failure, so the current clocksource will be shown correctly in sysfs. Signed-off-by: Thomas Gleixner Acked-by: John Stultz Cc: Magnus Damm Link: http://lkml.kernel.org/r/20130425143435.696321912@linutronix.de Signed-off-by: Thomas Gleixner --- include/linux/clocksource.h | 2 +- kernel/time/clocksource.c | 6 +++--- kernel/time/timekeeping.c | 5 +++-- 3 files changed, 7 insertions(+), 6 deletions(-) diff --git a/include/linux/clocksource.h b/include/linux/clocksource.h index 7279b94c01da..aa6ba44e75d5 100644 --- a/include/linux/clocksource.h +++ b/include/linux/clocksource.h @@ -321,7 +321,7 @@ static inline void __clocksource_updatefreq_khz(struct clocksource *cs, u32 khz) } -extern void timekeeping_notify(struct clocksource *clock); +extern int timekeeping_notify(struct clocksource *clock); extern cycle_t clocksource_mmio_readl_up(struct clocksource *); extern cycle_t clocksource_mmio_readl_down(struct clocksource *); diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c index dda5c7130d93..1923a340bd91 100644 --- a/kernel/time/clocksource.c +++ b/kernel/time/clocksource.c @@ -611,10 +611,10 @@ static void clocksource_select(void) best = cs; break; } - if (curr_clocksource != best) { - printk(KERN_INFO "Switching to clocksource %s\n", best->name); + + if (curr_clocksource != best && !timekeeping_notify(best)) { + pr_info("Switched to clocksource %s\n", best->name); curr_clocksource = best; - timekeeping_notify(curr_clocksource); } } diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c index 98cd470bbe49..da6e10c7a378 100644 --- a/kernel/time/timekeeping.c +++ b/kernel/time/timekeeping.c @@ -648,14 +648,15 @@ static int change_clocksource(void *data) * This function is called from clocksource.c after a new, better clock * source has been registered. The caller holds the clocksource_mutex. */ -void timekeeping_notify(struct clocksource *clock) +int timekeeping_notify(struct clocksource *clock) { struct timekeeper *tk = &timekeeper; if (tk->clock == clock) - return; + return 0; stop_machine(change_clocksource, clock, NULL); tick_clock_notify(); + return tk->clock == clock ? 0 : -1; } /** From 09ac369c825d9d593404306d59062d854b321e9b Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Thu, 25 Apr 2013 20:31:44 +0000 Subject: [PATCH 163/198] clocksource: Add module refcount Add a module refcount, so the current clocksource cannot be removed unconditionally. Signed-off-by: Thomas Gleixner Cc: John Stultz Cc: Magnus Damm Link: http://lkml.kernel.org/r/20130425143435.762417789@linutronix.de Signed-off-by: Thomas Gleixner --- include/linux/clocksource.h | 3 +++ kernel/time/timekeeping.c | 19 ++++++++++++++----- 2 files changed, 17 insertions(+), 5 deletions(-) diff --git a/include/linux/clocksource.h b/include/linux/clocksource.h index aa6ba44e75d5..32a895b114d8 100644 --- a/include/linux/clocksource.h +++ b/include/linux/clocksource.h @@ -21,6 +21,7 @@ /* clocksource cycle base type */ typedef u64 cycle_t; struct clocksource; +struct module; #ifdef CONFIG_ARCH_CLOCKSOURCE_DATA #include @@ -162,6 +163,7 @@ extern u64 timecounter_cyc2time(struct timecounter *tc, * @suspend: suspend function for the clocksource, if necessary * @resume: resume function for the clocksource, if necessary * @cycle_last: most recent cycle counter value seen by ::read() + * @owner: module reference, must be set by clocksource in modules */ struct clocksource { /* @@ -195,6 +197,7 @@ struct clocksource { cycle_t cs_last; cycle_t wd_last; #endif + struct module *owner; } ____cacheline_aligned; /* diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c index da6e10c7a378..933efa4071c3 100644 --- a/kernel/time/timekeeping.c +++ b/kernel/time/timekeeping.c @@ -627,11 +627,20 @@ static int change_clocksource(void *data) write_seqcount_begin(&timekeeper_seq); timekeeping_forward_now(tk); - if (!new->enable || new->enable(new) == 0) { - old = tk->clock; - tk_setup_internals(tk, new); - if (old->disable) - old->disable(old); + /* + * If the cs is in module, get a module reference. Succeeds + * for built-in code (owner == NULL) as well. + */ + if (try_module_get(new->owner)) { + if (!new->enable || new->enable(new) == 0) { + old = tk->clock; + tk_setup_internals(tk, new); + if (old->disable) + old->disable(old); + module_put(old->owner); + } else { + module_put(new->owner); + } } timekeeping_update(tk, true, true); From f5a2e34375a5e2b711aea488ac3ae50eeba6d57c Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Thu, 25 Apr 2013 20:31:45 +0000 Subject: [PATCH 164/198] clocksource: Allow clocksource select to skip current clocksource Preparatory patch for clocksource unbind support. Split out code from clocksource_select and modify it, so it skips the current clocksource on request and tries to find a fallback clocksource. Convert all existing users. No functional change. Signed-off-by: Thomas Gleixner Cc: John Stultz Cc: Magnus Damm Link: http://lkml.kernel.org/r/20130425143435.834965397@linutronix.de Signed-off-by: Thomas Gleixner --- kernel/time/clocksource.c | 31 ++++++++++++++++++++----------- 1 file changed, 20 insertions(+), 11 deletions(-) diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c index 1923a340bd91..9782997cb6cf 100644 --- a/kernel/time/clocksource.c +++ b/kernel/time/clocksource.c @@ -553,7 +553,7 @@ static u64 clocksource_max_deferment(struct clocksource *cs) #ifndef CONFIG_ARCH_USES_GETTIMEOFFSET -static struct clocksource *clocksource_find_best(bool oneshot) +static struct clocksource *clocksource_find_best(bool oneshot, bool skipcur) { struct clocksource *cs; @@ -566,6 +566,8 @@ static struct clocksource *clocksource_find_best(bool oneshot) * the best rating. */ list_for_each_entry(cs, &clocksource_list, list) { + if (skipcur && cs == curr_clocksource) + continue; if (oneshot && !(cs->flags & CLOCK_SOURCE_VALID_FOR_HRES)) continue; return cs; @@ -573,26 +575,20 @@ static struct clocksource *clocksource_find_best(bool oneshot) return NULL; } -/** - * clocksource_select - Select the best clocksource available - * - * Private function. Must hold clocksource_mutex when called. - * - * Select the clocksource with the best rating, or the clocksource, - * which is selected by userspace override. - */ -static void clocksource_select(void) +static void __clocksource_select(bool skipcur) { bool oneshot = tick_oneshot_mode_active(); struct clocksource *best, *cs; /* Find the best suitable clocksource */ - best = clocksource_find_best(oneshot); + best = clocksource_find_best(oneshot, skipcur); if (!best) return; /* Check for the override clocksource. */ list_for_each_entry(cs, &clocksource_list, list) { + if (skipcur && cs == curr_clocksource) + continue; if (strcmp(cs->name, override_name) != 0) continue; /* @@ -618,6 +614,19 @@ static void clocksource_select(void) } } +/** + * clocksource_select - Select the best clocksource available + * + * Private function. Must hold clocksource_mutex when called. + * + * Select the clocksource with the best rating, or the clocksource, + * which is selected by userspace override. + */ +static void clocksource_select(void) +{ + return __clocksource_select(false); +} + #else /* !CONFIG_ARCH_USES_GETTIMEOFFSET */ static inline void clocksource_select(void) { } From 29b5407819f59731c9423238fae03b756822708c Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Thu, 25 Apr 2013 20:31:45 +0000 Subject: [PATCH 165/198] clocksource: Split out user string input Split out the user string input for clocksource override. Preparatory patch for unbind. [ jstultz: Fix an off by one error ] Signed-off-by: Thomas Gleixner Cc: John Stultz Cc: Magnus Damm Link: http://lkml.kernel.org/r/20130425143435.895851338@linutronix.de Signed-off-by: Thomas Gleixner --- kernel/time/clocksource.c | 37 +++++++++++++++++++++++-------------- 1 file changed, 23 insertions(+), 14 deletions(-) diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c index 9782997cb6cf..d7f1a45c2fa5 100644 --- a/kernel/time/clocksource.c +++ b/kernel/time/clocksource.c @@ -174,7 +174,8 @@ clocks_calc_mult_shift(u32 *mult, u32 *shift, u32 from, u32 to, u32 maxsec) static struct clocksource *curr_clocksource; static LIST_HEAD(clocksource_list); static DEFINE_MUTEX(clocksource_mutex); -static char override_name[32]; +#define CS_NAME_LEN 32 +static char override_name[CS_NAME_LEN]; static int finished_booting; #ifdef CONFIG_CLOCKSOURCE_WATCHDOG @@ -838,6 +839,23 @@ sysfs_show_current_clocksources(struct device *dev, return count; } +static size_t clocksource_get_uname(const char *buf, char *dst, size_t cnt) +{ + size_t ret = cnt; + + /* strings from sysfs write are not 0 terminated! */ + if (!cnt || cnt >= CS_NAME_LEN) + return -EINVAL; + + /* strip of \n: */ + if (buf[cnt-1] == '\n') + cnt--; + if (cnt > 0) + memcpy(dst, buf, cnt); + dst[cnt] = 0; + return ret; +} + /** * sysfs_override_clocksource - interface for manually overriding clocksource * @dev: unused @@ -852,22 +870,13 @@ static ssize_t sysfs_override_clocksource(struct device *dev, struct device_attribute *attr, const char *buf, size_t count) { - size_t ret = count; - - /* strings from sysfs write are not 0 terminated! */ - if (count >= sizeof(override_name)) - return -EINVAL; - - /* strip of \n: */ - if (buf[count-1] == '\n') - count--; + size_t ret; mutex_lock(&clocksource_mutex); - if (count > 0) - memcpy(override_name, buf, count); - override_name[count] = 0; - clocksource_select(); + ret = clocksource_get_uname(buf, override_name, count); + if (ret >= 0) + clocksource_select(); mutex_unlock(&clocksource_mutex); From 7eaeb34305dee26634f7c98ae62646da5cebe91d Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Thu, 25 Apr 2013 20:31:46 +0000 Subject: [PATCH 166/198] clocksource: Provide unbind interface in sysfs With the module refcount held for the current clocksource there is no way to unload the module. Provide a sysfs interface which allows to unbind the clocksource. One could argue that the clocksource override could be (ab)used to do so, but the clocksource override cannot be used from the kernel itself, while an unbind function can be used to programmatically check whether a clocksource can be shutdown or not. The unbind functionality uses the new skip current feature of clocksource_select and verifies that a fallback clocksource has been installed. If the clocksource which should be unbound is the current clocksource and no fallback can be found, unbind returns -EBUSY. This does not support the unbinding of a clocksource which is used as the watchdog clocksource. No point in fostering crappy hardware. Signed-off-by: Thomas Gleixner Cc: John Stultz Cc: Magnus Damm Link: http://lkml.kernel.org/r/20130425143435.964218245@linutronix.de Signed-off-by: Thomas Gleixner --- kernel/time/clocksource.c | 73 +++++++++++++++++++++++++++++++++++++++ 1 file changed, 73 insertions(+) diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c index d7f1a45c2fa5..791d1aeb17ac 100644 --- a/kernel/time/clocksource.c +++ b/kernel/time/clocksource.c @@ -440,6 +440,11 @@ static int clocksource_watchdog_kthread(void *data) return 0; } +static bool clocksource_is_watchdog(struct clocksource *cs) +{ + return cs == watchdog; +} + #else /* CONFIG_CLOCKSOURCE_WATCHDOG */ static void clocksource_enqueue_watchdog(struct clocksource *cs) @@ -451,6 +456,7 @@ static void clocksource_enqueue_watchdog(struct clocksource *cs) static inline void clocksource_dequeue_watchdog(struct clocksource *cs) { } static inline void clocksource_resume_watchdog(void) { } static inline int clocksource_watchdog_kthread(void *data) { return 0; } +static bool clocksource_is_watchdog(struct clocksource *cs) { return false; } #endif /* CONFIG_CLOCKSOURCE_WATCHDOG */ @@ -628,6 +634,11 @@ static void clocksource_select(void) return __clocksource_select(false); } +static void clocksource_select_fallback(void) +{ + return __clocksource_select(true); +} + #else /* !CONFIG_ARCH_USES_GETTIMEOFFSET */ static inline void clocksource_select(void) { } @@ -803,6 +814,29 @@ void clocksource_change_rating(struct clocksource *cs, int rating) } EXPORT_SYMBOL(clocksource_change_rating); +/* + * Unbind clocksource @cs. Called with clocksource_mutex held + */ +static int clocksource_unbind(struct clocksource *cs) +{ + /* + * I really can't convince myself to support this on hardware + * designed by lobotomized monkeys. + */ + if (clocksource_is_watchdog(cs)) + return -EBUSY; + + if (cs == curr_clocksource) { + /* Select and try to install a replacement clock source */ + clocksource_select_fallback(); + if (curr_clocksource == cs) + return -EBUSY; + } + clocksource_dequeue_watchdog(cs); + list_del_init(&cs->list); + return 0; +} + /** * clocksource_unregister - remove a registered clocksource * @cs: clocksource to be unregistered @@ -883,6 +917,40 @@ static ssize_t sysfs_override_clocksource(struct device *dev, return ret; } +/** + * sysfs_unbind_current_clocksource - interface for manually unbinding clocksource + * @dev: unused + * @attr: unused + * @buf: unused + * @count: length of buffer + * + * Takes input from sysfs interface for manually unbinding a clocksource. + */ +static ssize_t sysfs_unbind_clocksource(struct device *dev, + struct device_attribute *attr, + const char *buf, size_t count) +{ + struct clocksource *cs; + char name[CS_NAME_LEN]; + size_t ret; + + ret = clocksource_get_uname(buf, name, count); + if (ret < 0) + return ret; + + ret = -ENODEV; + mutex_lock(&clocksource_mutex); + list_for_each_entry(cs, &clocksource_list, list) { + if (strcmp(cs->name, name)) + continue; + ret = clocksource_unbind(cs); + break; + } + mutex_unlock(&clocksource_mutex); + + return ret ? ret : count; +} + /** * sysfs_show_available_clocksources - sysfs interface for listing clocksource * @dev: unused @@ -925,6 +993,8 @@ sysfs_show_available_clocksources(struct device *dev, static DEVICE_ATTR(current_clocksource, 0644, sysfs_show_current_clocksources, sysfs_override_clocksource); +static DEVICE_ATTR(unbind_clocksource, 0200, NULL, sysfs_unbind_clocksource); + static DEVICE_ATTR(available_clocksource, 0444, sysfs_show_available_clocksources, NULL); @@ -948,6 +1018,9 @@ static int __init init_clocksource_sysfs(void) error = device_create_file( &device_clocksource, &dev_attr_current_clocksource); + if (!error) + error = device_create_file(&device_clocksource, + &dev_attr_unbind_clocksource); if (!error) error = device_create_file( &device_clocksource, From a89c7edbe7d7aa80f507915f3dd801211b116b79 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Thu, 25 Apr 2013 20:31:46 +0000 Subject: [PATCH 167/198] clocksource: Let clocksource_unregister() return success/error The unregister call can fail, if the clocksource is the current one and there is no replacement clocksource available. It can also fail, if the clocksource is the watchdog clocksource and I'm not going to provide support for this. Signed-off-by: Thomas Gleixner Cc: John Stultz Cc: Magnus Damm Link: http://lkml.kernel.org/r/20130425143436.029915527@linutronix.de Signed-off-by: Thomas Gleixner --- include/linux/clocksource.h | 2 +- kernel/time/clocksource.c | 33 ++++++++++++--------------------- 2 files changed, 13 insertions(+), 22 deletions(-) diff --git a/include/linux/clocksource.h b/include/linux/clocksource.h index 32a895b114d8..2f39a4911668 100644 --- a/include/linux/clocksource.h +++ b/include/linux/clocksource.h @@ -282,7 +282,7 @@ static inline s64 clocksource_cyc2ns(cycle_t cycles, u32 mult, u32 shift) extern int clocksource_register(struct clocksource*); -extern void clocksource_unregister(struct clocksource*); +extern int clocksource_unregister(struct clocksource*); extern void clocksource_touch_watchdog(void); extern struct clocksource* clocksource_get_next(void); extern void clocksource_change_rating(struct clocksource *cs, int rating); diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c index 791d1aeb17ac..31b90332f47b 100644 --- a/kernel/time/clocksource.c +++ b/kernel/time/clocksource.c @@ -389,28 +389,17 @@ static void clocksource_enqueue_watchdog(struct clocksource *cs) static void clocksource_dequeue_watchdog(struct clocksource *cs) { - struct clocksource *tmp; unsigned long flags; spin_lock_irqsave(&watchdog_lock, flags); - if (cs->flags & CLOCK_SOURCE_MUST_VERIFY) { - /* cs is a watched clocksource. */ - list_del_init(&cs->wd_list); - } else if (cs == watchdog) { - /* Reset watchdog cycles */ - clocksource_reset_watchdog(); - /* Current watchdog is removed. Find an alternative. */ - watchdog = NULL; - list_for_each_entry(tmp, &clocksource_list, list) { - if (tmp == cs || tmp->flags & CLOCK_SOURCE_MUST_VERIFY) - continue; - if (!watchdog || tmp->rating > watchdog->rating) - watchdog = tmp; + if (cs != watchdog) { + if (cs->flags & CLOCK_SOURCE_MUST_VERIFY) { + /* cs is a watched clocksource. */ + list_del_init(&cs->wd_list); + /* Check if the watchdog timer needs to be stopped. */ + clocksource_stop_watchdog(); } } - cs->flags &= ~CLOCK_SOURCE_WATCHDOG; - /* Check if the watchdog timer needs to be stopped. */ - clocksource_stop_watchdog(); spin_unlock_irqrestore(&watchdog_lock, flags); } @@ -841,13 +830,15 @@ static int clocksource_unbind(struct clocksource *cs) * clocksource_unregister - remove a registered clocksource * @cs: clocksource to be unregistered */ -void clocksource_unregister(struct clocksource *cs) +int clocksource_unregister(struct clocksource *cs) { + int ret = 0; + mutex_lock(&clocksource_mutex); - clocksource_dequeue_watchdog(cs); - list_del(&cs->list); - clocksource_select(); + if (!list_empty(&cs->list)) + ret = clocksource_unbind(cs); mutex_unlock(&clocksource_mutex); + return ret; } EXPORT_SYMBOL(clocksource_unregister); From 7172a286ced0c1f4f239a0fa09db54ed37d3ead2 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Thu, 25 Apr 2013 20:31:47 +0000 Subject: [PATCH 168/198] clockevents: Get rid of the notifier chain 7+ years and still a single user. Kill it. Signed-off-by: Thomas Gleixner Cc: John Stultz Cc: Magnus Damm Link: http://lkml.kernel.org/r/20130425143436.098520211@linutronix.de Signed-off-by: Thomas Gleixner --- include/linux/clockchips.h | 1 - kernel/time/clockevents.c | 35 +++-------------------------------- kernel/time/tick-broadcast.c | 5 ++--- kernel/time/tick-common.c | 30 +++++------------------------- kernel/time/tick-internal.h | 7 ++++--- 5 files changed, 14 insertions(+), 64 deletions(-) diff --git a/include/linux/clockchips.h b/include/linux/clockchips.h index 963d71431388..2f498f66c1bb 100644 --- a/include/linux/clockchips.h +++ b/include/linux/clockchips.h @@ -150,7 +150,6 @@ extern void clockevents_exchange_device(struct clock_event_device *old, struct clock_event_device *new); extern void clockevents_set_mode(struct clock_event_device *dev, enum clock_event_mode mode); -extern int clockevents_register_notifier(struct notifier_block *nb); extern int clockevents_program_event(struct clock_event_device *dev, ktime_t expires, bool force); diff --git a/kernel/time/clockevents.c b/kernel/time/clockevents.c index c6d6400ee137..dd70b4842c62 100644 --- a/kernel/time/clockevents.c +++ b/kernel/time/clockevents.c @@ -15,7 +15,6 @@ #include #include #include -#include #include #include "tick-internal.h" @@ -23,10 +22,6 @@ /* The registered clock event devices */ static LIST_HEAD(clockevent_devices); static LIST_HEAD(clockevents_released); - -/* Notification for clock events */ -static RAW_NOTIFIER_HEAD(clockevents_chain); - /* Protection for the above */ static DEFINE_RAW_SPINLOCK(clockevents_lock); @@ -232,30 +227,6 @@ int clockevents_program_event(struct clock_event_device *dev, ktime_t expires, return (rc && force) ? clockevents_program_min_delta(dev) : rc; } -/** - * clockevents_register_notifier - register a clock events change listener - */ -int clockevents_register_notifier(struct notifier_block *nb) -{ - unsigned long flags; - int ret; - - raw_spin_lock_irqsave(&clockevents_lock, flags); - ret = raw_notifier_chain_register(&clockevents_chain, nb); - raw_spin_unlock_irqrestore(&clockevents_lock, flags); - - return ret; -} - -/* - * Notify about a clock event change. Called with clockevents_lock - * held. - */ -static void clockevents_do_notify(unsigned long reason, void *dev) -{ - raw_notifier_call_chain(&clockevents_chain, reason, dev); -} - /* * Called after a notify add to make devices available which were * released from the notifier call. @@ -269,7 +240,7 @@ static void clockevents_notify_released(void) struct clock_event_device, list); list_del(&dev->list); list_add(&dev->list, &clockevent_devices); - clockevents_do_notify(CLOCK_EVT_NOTIFY_ADD, dev); + tick_check_new_device(dev); } } @@ -290,7 +261,7 @@ void clockevents_register_device(struct clock_event_device *dev) raw_spin_lock_irqsave(&clockevents_lock, flags); list_add(&dev->list, &clockevent_devices); - clockevents_do_notify(CLOCK_EVT_NOTIFY_ADD, dev); + tick_check_new_device(dev); clockevents_notify_released(); raw_spin_unlock_irqrestore(&clockevents_lock, flags); @@ -433,7 +404,7 @@ void clockevents_notify(unsigned long reason, void *arg) int cpu; raw_spin_lock_irqsave(&clockevents_lock, flags); - clockevents_do_notify(reason, arg); + tick_notify(reason, arg); switch (reason) { case CLOCK_EVT_NOTIFY_CPU_DEAD: diff --git a/kernel/time/tick-broadcast.c b/kernel/time/tick-broadcast.c index 24938d577669..3500caaa0bfd 100644 --- a/kernel/time/tick-broadcast.c +++ b/kernel/time/tick-broadcast.c @@ -64,7 +64,7 @@ static void tick_broadcast_start_periodic(struct clock_event_device *bc) /* * Check, if the device can be utilized as broadcast device: */ -int tick_check_broadcast_device(struct clock_event_device *dev) +void tick_install_broadcast_device(struct clock_event_device *dev) { struct clock_event_device *cur = tick_broadcast_device.evtdev; @@ -72,7 +72,7 @@ int tick_check_broadcast_device(struct clock_event_device *dev) (tick_broadcast_device.evtdev && tick_broadcast_device.evtdev->rating >= dev->rating) || (dev->features & CLOCK_EVT_FEAT_C3STOP)) - return 0; + return; clockevents_exchange_device(tick_broadcast_device.evtdev, dev); if (cur) @@ -90,7 +90,6 @@ int tick_check_broadcast_device(struct clock_event_device *dev) */ if (dev->features & CLOCK_EVT_FEAT_ONESHOT) tick_clock_notify(); - return 1; } /* diff --git a/kernel/time/tick-common.c b/kernel/time/tick-common.c index 5d3fb100bc06..dbf4e18d5101 100644 --- a/kernel/time/tick-common.c +++ b/kernel/time/tick-common.c @@ -208,11 +208,11 @@ static void tick_setup_device(struct tick_device *td, /* * Check, if the new registered device should be used. */ -static int tick_check_new_device(struct clock_event_device *newdev) +void tick_check_new_device(struct clock_event_device *newdev) { struct clock_event_device *curdev; struct tick_device *td; - int cpu, ret = NOTIFY_OK; + int cpu; unsigned long flags; raw_spin_lock_irqsave(&tick_device_lock, flags); @@ -275,18 +275,14 @@ static int tick_check_new_device(struct clock_event_device *newdev) tick_oneshot_notify(); raw_spin_unlock_irqrestore(&tick_device_lock, flags); - return NOTIFY_STOP; + return; out_bc: /* * Can the new device be used as a broadcast device ? */ - if (tick_check_broadcast_device(newdev)) - ret = NOTIFY_STOP; - + tick_install_broadcast_device(newdev); raw_spin_unlock_irqrestore(&tick_device_lock, flags); - - return ret; } /* @@ -360,17 +356,10 @@ static void tick_resume(void) raw_spin_unlock_irqrestore(&tick_device_lock, flags); } -/* - * Notification about clock event devices - */ -static int tick_notify(struct notifier_block *nb, unsigned long reason, - void *dev) +void tick_notify(unsigned long reason, void *dev) { switch (reason) { - case CLOCK_EVT_NOTIFY_ADD: - return tick_check_new_device(dev); - case CLOCK_EVT_NOTIFY_BROADCAST_ON: case CLOCK_EVT_NOTIFY_BROADCAST_OFF: case CLOCK_EVT_NOTIFY_BROADCAST_FORCE: @@ -404,21 +393,12 @@ static int tick_notify(struct notifier_block *nb, unsigned long reason, default: break; } - - return NOTIFY_OK; } -static struct notifier_block tick_notifier = { - .notifier_call = tick_notify, -}; - /** * tick_init - initialize the tick control - * - * Register the notifier with the clockevents framework */ void __init tick_init(void) { - clockevents_register_notifier(&tick_notifier); tick_broadcast_init(); } diff --git a/kernel/time/tick-internal.h b/kernel/time/tick-internal.h index f0299eae4602..60742fe6f63d 100644 --- a/kernel/time/tick-internal.h +++ b/kernel/time/tick-internal.h @@ -18,6 +18,8 @@ extern int tick_do_timer_cpu __read_mostly; extern void tick_setup_periodic(struct clock_event_device *dev, int broadcast); extern void tick_handle_periodic(struct clock_event_device *dev); +extern void tick_notify(unsigned long reason, void *dev); +extern void tick_check_new_device(struct clock_event_device *dev); extern void clockevents_shutdown(struct clock_event_device *dev); @@ -90,7 +92,7 @@ static inline bool tick_broadcast_oneshot_available(void) { return false; } */ #ifdef CONFIG_GENERIC_CLOCKEVENTS_BROADCAST extern int tick_device_uses_broadcast(struct clock_event_device *dev, int cpu); -extern int tick_check_broadcast_device(struct clock_event_device *dev); +extern void tick_install_broadcast_device(struct clock_event_device *dev); extern int tick_is_broadcast_device(struct clock_event_device *dev); extern void tick_broadcast_on_off(unsigned long reason, int *oncpu); extern void tick_shutdown_broadcast(unsigned int *cpup); @@ -102,9 +104,8 @@ tick_set_periodic_handler(struct clock_event_device *dev, int broadcast); #else /* !BROADCAST */ -static inline int tick_check_broadcast_device(struct clock_event_device *dev) +static inline void tick_install_broadcast_device(struct clock_event_device *dev) { - return 0; } static inline int tick_is_broadcast_device(struct clock_event_device *dev) From 7126cac426137633e470167524e7bcb590fd49b3 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Thu, 25 Apr 2013 20:31:48 +0000 Subject: [PATCH 169/198] clockevents: Simplify locking Now that the notifier chain is gone there are no other users and it's pointless to nest tick_device_lock inside of clockevents_lock because there is no other use case. Signed-off-by: Thomas Gleixner Cc: John Stultz Cc: Magnus Damm Link: http://lkml.kernel.org/r/20130425143436.162888472@linutronix.de Signed-off-by: Thomas Gleixner --- kernel/time/tick-common.c | 22 +++++----------------- 1 file changed, 5 insertions(+), 17 deletions(-) diff --git a/kernel/time/tick-common.c b/kernel/time/tick-common.c index dbf4e18d5101..170a4bdfa99e 100644 --- a/kernel/time/tick-common.c +++ b/kernel/time/tick-common.c @@ -33,7 +33,6 @@ DEFINE_PER_CPU(struct tick_device, tick_cpu_device); ktime_t tick_next_period; ktime_t tick_period; int tick_do_timer_cpu __read_mostly = TICK_DO_TIMER_BOOT; -static DEFINE_RAW_SPINLOCK(tick_device_lock); /* * Debugging: see timer_list.c @@ -206,16 +205,14 @@ static void tick_setup_device(struct tick_device *td, } /* - * Check, if the new registered device should be used. + * Check, if the new registered device should be used. Called with + * clockevents_lock held and interrupts disabled. */ void tick_check_new_device(struct clock_event_device *newdev) { struct clock_event_device *curdev; struct tick_device *td; int cpu; - unsigned long flags; - - raw_spin_lock_irqsave(&tick_device_lock, flags); cpu = smp_processor_id(); if (!cpumask_test_cpu(cpu, newdev->cpumask)) @@ -273,8 +270,6 @@ void tick_check_new_device(struct clock_event_device *newdev) tick_setup_device(td, newdev, cpu, cpumask_of(cpu)); if (newdev->features & CLOCK_EVT_FEAT_ONESHOT) tick_oneshot_notify(); - - raw_spin_unlock_irqrestore(&tick_device_lock, flags); return; out_bc: @@ -282,7 +277,6 @@ out_bc: * Can the new device be used as a broadcast device ? */ tick_install_broadcast_device(newdev); - raw_spin_unlock_irqrestore(&tick_device_lock, flags); } /* @@ -311,9 +305,7 @@ static void tick_shutdown(unsigned int *cpup) { struct tick_device *td = &per_cpu(tick_cpu_device, *cpup); struct clock_event_device *dev = td->evtdev; - unsigned long flags; - raw_spin_lock_irqsave(&tick_device_lock, flags); td->mode = TICKDEV_MODE_PERIODIC; if (dev) { /* @@ -325,26 +317,20 @@ static void tick_shutdown(unsigned int *cpup) dev->event_handler = clockevents_handle_noop; td->evtdev = NULL; } - raw_spin_unlock_irqrestore(&tick_device_lock, flags); } static void tick_suspend(void) { struct tick_device *td = &__get_cpu_var(tick_cpu_device); - unsigned long flags; - raw_spin_lock_irqsave(&tick_device_lock, flags); clockevents_shutdown(td->evtdev); - raw_spin_unlock_irqrestore(&tick_device_lock, flags); } static void tick_resume(void) { struct tick_device *td = &__get_cpu_var(tick_cpu_device); - unsigned long flags; int broadcast = tick_resume_broadcast(); - raw_spin_lock_irqsave(&tick_device_lock, flags); clockevents_set_mode(td->evtdev, CLOCK_EVT_MODE_RESUME); if (!broadcast) { @@ -353,9 +339,11 @@ static void tick_resume(void) else tick_resume_oneshot(); } - raw_spin_unlock_irqrestore(&tick_device_lock, flags); } +/* + * Called with clockevents_lock held and interrupts disabled + */ void tick_notify(unsigned long reason, void *dev) { switch (reason) { From 8c53daf63f56791ed47fc585206ef3049489612f Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Thu, 25 Apr 2013 20:31:48 +0000 Subject: [PATCH 170/198] clockevents: Move the tick_notify() switch case to clockevents_notify() No need to call another function and have duplicated cases. Signed-off-by: Thomas Gleixner Cc: John Stultz Cc: Magnus Damm Link: http://lkml.kernel.org/r/20130425143436.235746557@linutronix.de Signed-off-by: Thomas Gleixner --- kernel/time/clockevents.c | 28 ++++++++++++++++++++- kernel/time/tick-common.c | 50 +++---------------------------------- kernel/time/tick-internal.h | 5 +++- 3 files changed, 35 insertions(+), 48 deletions(-) diff --git a/kernel/time/clockevents.c b/kernel/time/clockevents.c index dd70b4842c62..0e3a8448e115 100644 --- a/kernel/time/clockevents.c +++ b/kernel/time/clockevents.c @@ -404,10 +404,36 @@ void clockevents_notify(unsigned long reason, void *arg) int cpu; raw_spin_lock_irqsave(&clockevents_lock, flags); - tick_notify(reason, arg); switch (reason) { + case CLOCK_EVT_NOTIFY_BROADCAST_ON: + case CLOCK_EVT_NOTIFY_BROADCAST_OFF: + case CLOCK_EVT_NOTIFY_BROADCAST_FORCE: + tick_broadcast_on_off(reason, arg); + break; + + case CLOCK_EVT_NOTIFY_BROADCAST_ENTER: + case CLOCK_EVT_NOTIFY_BROADCAST_EXIT: + tick_broadcast_oneshot_control(reason); + break; + + case CLOCK_EVT_NOTIFY_CPU_DYING: + tick_handover_do_timer(arg); + break; + + case CLOCK_EVT_NOTIFY_SUSPEND: + tick_suspend(); + tick_suspend_broadcast(); + break; + + case CLOCK_EVT_NOTIFY_RESUME: + tick_resume(); + break; + case CLOCK_EVT_NOTIFY_CPU_DEAD: + tick_shutdown_broadcast_oneshot(arg); + tick_shutdown_broadcast(arg); + tick_shutdown(arg); /* * Unregister the clock event devices which were * released from the users in the notify chain. diff --git a/kernel/time/tick-common.c b/kernel/time/tick-common.c index 170a4bdfa99e..84c7cfca4d7d 100644 --- a/kernel/time/tick-common.c +++ b/kernel/time/tick-common.c @@ -284,7 +284,7 @@ out_bc: * * Called with interrupts disabled. */ -static void tick_handover_do_timer(int *cpup) +void tick_handover_do_timer(int *cpup) { if (*cpup == tick_do_timer_cpu) { int cpu = cpumask_first(cpu_online_mask); @@ -301,7 +301,7 @@ static void tick_handover_do_timer(int *cpup) * access the hardware device itself. * We just set the mode and remove it from the lists. */ -static void tick_shutdown(unsigned int *cpup) +void tick_shutdown(unsigned int *cpup) { struct tick_device *td = &per_cpu(tick_cpu_device, *cpup); struct clock_event_device *dev = td->evtdev; @@ -319,14 +319,14 @@ static void tick_shutdown(unsigned int *cpup) } } -static void tick_suspend(void) +void tick_suspend(void) { struct tick_device *td = &__get_cpu_var(tick_cpu_device); clockevents_shutdown(td->evtdev); } -static void tick_resume(void) +void tick_resume(void) { struct tick_device *td = &__get_cpu_var(tick_cpu_device); int broadcast = tick_resume_broadcast(); @@ -341,48 +341,6 @@ static void tick_resume(void) } } -/* - * Called with clockevents_lock held and interrupts disabled - */ -void tick_notify(unsigned long reason, void *dev) -{ - switch (reason) { - - case CLOCK_EVT_NOTIFY_BROADCAST_ON: - case CLOCK_EVT_NOTIFY_BROADCAST_OFF: - case CLOCK_EVT_NOTIFY_BROADCAST_FORCE: - tick_broadcast_on_off(reason, dev); - break; - - case CLOCK_EVT_NOTIFY_BROADCAST_ENTER: - case CLOCK_EVT_NOTIFY_BROADCAST_EXIT: - tick_broadcast_oneshot_control(reason); - break; - - case CLOCK_EVT_NOTIFY_CPU_DYING: - tick_handover_do_timer(dev); - break; - - case CLOCK_EVT_NOTIFY_CPU_DEAD: - tick_shutdown_broadcast_oneshot(dev); - tick_shutdown_broadcast(dev); - tick_shutdown(dev); - break; - - case CLOCK_EVT_NOTIFY_SUSPEND: - tick_suspend(); - tick_suspend_broadcast(); - break; - - case CLOCK_EVT_NOTIFY_RESUME: - tick_resume(); - break; - - default: - break; - } -} - /** * tick_init - initialize the tick control */ diff --git a/kernel/time/tick-internal.h b/kernel/time/tick-internal.h index 60742fe6f63d..06bfc8802dfb 100644 --- a/kernel/time/tick-internal.h +++ b/kernel/time/tick-internal.h @@ -18,8 +18,11 @@ extern int tick_do_timer_cpu __read_mostly; extern void tick_setup_periodic(struct clock_event_device *dev, int broadcast); extern void tick_handle_periodic(struct clock_event_device *dev); -extern void tick_notify(unsigned long reason, void *dev); extern void tick_check_new_device(struct clock_event_device *dev); +extern void tick_handover_do_timer(int *cpup); +extern void tick_shutdown(unsigned int *cpup); +extern void tick_suspend(void); +extern void tick_resume(void); extern void clockevents_shutdown(struct clock_event_device *dev); From ccf33d6880f39a35158fff66db13000ae4943fac Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Thu, 25 Apr 2013 20:31:49 +0000 Subject: [PATCH 171/198] clockevents: Add module refcount We want to be able to remove clockevent modules as well. Add a refcount so we don't remove a module with an active clock event device. Signed-off-by: Thomas Gleixner Cc: John Stultz Cc: Magnus Damm Link: http://lkml.kernel.org/r/20130425143436.307435149@linutronix.de Signed-off-by: Thomas Gleixner --- include/linux/clockchips.h | 3 +++ kernel/time/clockevents.c | 1 + kernel/time/tick-broadcast.c | 3 +++ kernel/time/tick-common.c | 4 ++++ 4 files changed, 11 insertions(+) diff --git a/include/linux/clockchips.h b/include/linux/clockchips.h index 2f498f66c1bb..ae1193bcf074 100644 --- a/include/linux/clockchips.h +++ b/include/linux/clockchips.h @@ -30,6 +30,7 @@ enum clock_event_nofitiers { #include struct clock_event_device; +struct module; /* Clock event mode commands */ enum clock_event_mode { @@ -83,6 +84,7 @@ enum clock_event_mode { * @irq: IRQ number (only for non CPU local devices) * @cpumask: cpumask to indicate for which CPUs this device works * @list: list head for the management code + * @owner: module reference */ struct clock_event_device { void (*event_handler)(struct clock_event_device *); @@ -112,6 +114,7 @@ struct clock_event_device { int irq; const struct cpumask *cpumask; struct list_head list; + struct module *owner; } ____cacheline_aligned; /* diff --git a/kernel/time/clockevents.c b/kernel/time/clockevents.c index 0e3a8448e115..89e394caa769 100644 --- a/kernel/time/clockevents.c +++ b/kernel/time/clockevents.c @@ -357,6 +357,7 @@ void clockevents_exchange_device(struct clock_event_device *old, * released list and do a notify add later. */ if (old) { + module_put(old->owner); clockevents_set_mode(old, CLOCK_EVT_MODE_UNUSED); list_del(&old->list); list_add(&old->list, &clockevents_released); diff --git a/kernel/time/tick-broadcast.c b/kernel/time/tick-broadcast.c index 3500caaa0bfd..0e374cd2e0ef 100644 --- a/kernel/time/tick-broadcast.c +++ b/kernel/time/tick-broadcast.c @@ -19,6 +19,7 @@ #include #include #include +#include #include "tick-internal.h" @@ -73,6 +74,8 @@ void tick_install_broadcast_device(struct clock_event_device *dev) tick_broadcast_device.evtdev->rating >= dev->rating) || (dev->features & CLOCK_EVT_FEAT_C3STOP)) return; + if (!try_module_get(dev->owner)) + return; clockevents_exchange_device(tick_broadcast_device.evtdev, dev); if (cur) diff --git a/kernel/time/tick-common.c b/kernel/time/tick-common.c index 84c7cfca4d7d..433a1e11d13b 100644 --- a/kernel/time/tick-common.c +++ b/kernel/time/tick-common.c @@ -18,6 +18,7 @@ #include #include #include +#include #include @@ -257,6 +258,9 @@ void tick_check_new_device(struct clock_event_device *newdev) goto out_bc; } + if (!try_module_get(newdev->owner)) + return; + /* * Replace the eventually existing device by the new * device. If the current device is the broadcast device, do From 501f867064e95f9a6f540e60705be0937280e7ec Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Thu, 25 Apr 2013 20:31:49 +0000 Subject: [PATCH 172/198] clockevents: Provide sysfs interface Provide a simple sysfs interface for the clockevent devices. Show the current active clockevent device. Signed-off-by: Thomas Gleixner Cc: John Stultz Cc: Magnus Damm Link: http://lkml.kernel.org/r/20130425143436.371634778@linutronix.de Signed-off-by: Thomas Gleixner --- kernel/time/clockevents.c | 86 +++++++++++++++++++++++++++++++++++++++ 1 file changed, 86 insertions(+) diff --git a/kernel/time/clockevents.c b/kernel/time/clockevents.c index 89e394caa769..0a23f4f29934 100644 --- a/kernel/time/clockevents.c +++ b/kernel/time/clockevents.c @@ -16,6 +16,7 @@ #include #include #include +#include #include "tick-internal.h" @@ -460,4 +461,89 @@ void clockevents_notify(unsigned long reason, void *arg) raw_spin_unlock_irqrestore(&clockevents_lock, flags); } EXPORT_SYMBOL_GPL(clockevents_notify); + +#ifdef CONFIG_SYSFS +struct bus_type clockevents_subsys = { + .name = "clockevents", + .dev_name = "clockevent", +}; + +static DEFINE_PER_CPU(struct device, tick_percpu_dev); +static struct tick_device *tick_get_tick_dev(struct device *dev); + +static ssize_t sysfs_show_current_tick_dev(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct tick_device *td; + ssize_t count = 0; + + raw_spin_lock_irq(&clockevents_lock); + td = tick_get_tick_dev(dev); + if (td && td->evtdev) + count = snprintf(buf, PAGE_SIZE, "%s\n", td->evtdev->name); + raw_spin_unlock_irq(&clockevents_lock); + return count; +} +static DEVICE_ATTR(current_device, 0444, sysfs_show_current_tick_dev, NULL); + +#ifdef CONFIG_GENERIC_CLOCKEVENTS_BROADCAST +static struct device tick_bc_dev = { + .init_name = "broadcast", + .id = 0, + .bus = &clockevents_subsys, +}; + +static struct tick_device *tick_get_tick_dev(struct device *dev) +{ + return dev == &tick_bc_dev ? tick_get_broadcast_device() : + &per_cpu(tick_cpu_device, dev->id); +} + +static __init int tick_broadcast_init_sysfs(void) +{ + int err = device_register(&tick_bc_dev); + + if (!err) + err = device_create_file(&tick_bc_dev, &dev_attr_current_device); + return err; +} +#else +static struct tick_device *tick_get_tick_dev(struct device *dev) +{ + return &per_cpu(tick_cpu_device, dev->id); +} +static inline int tick_broadcast_init_sysfs(void) { return 0; } #endif + +static int __init tick_init_sysfs(void) +{ + int cpu; + + for_each_possible_cpu(cpu) { + struct device *dev = &per_cpu(tick_percpu_dev, cpu); + int err; + + dev->id = cpu; + dev->bus = &clockevents_subsys; + err = device_register(dev); + if (!err) + err = device_create_file(dev, &dev_attr_current_device); + if (err) + return err; + } + return tick_broadcast_init_sysfs(); +} + +static int __init clockevents_init_sysfs(void) +{ + int err = subsys_system_register(&clockevents_subsys, NULL); + + if (!err) + err = tick_init_sysfs(); + return err; +} +device_initcall(clockevents_init_sysfs); +#endif /* SYSFS */ + +#endif /* GENERIC_CLOCK_EVENTS */ From 45cb8e01b2ecef1c2afb18333e95793fa1a90281 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Thu, 25 Apr 2013 20:31:50 +0000 Subject: [PATCH 173/198] clockevents: Split out selection logic Split out the clockevent device selection logic. Preparatory patch to allow unbinding active clockevent devices. Signed-off-by: Thomas Gleixner Cc: John Stultz Cc: Magnus Damm Link: http://lkml.kernel.org/r/20130425143436.431796247@linutronix.de Signed-off-by: Thomas Gleixner --- kernel/time/tick-broadcast.c | 25 ++++++++++--- kernel/time/tick-common.c | 69 +++++++++++++++++++----------------- 2 files changed, 56 insertions(+), 38 deletions(-) diff --git a/kernel/time/tick-broadcast.c b/kernel/time/tick-broadcast.c index 0e374cd2e0ef..d067c01586f5 100644 --- a/kernel/time/tick-broadcast.c +++ b/kernel/time/tick-broadcast.c @@ -65,19 +65,34 @@ static void tick_broadcast_start_periodic(struct clock_event_device *bc) /* * Check, if the device can be utilized as broadcast device: */ +static bool tick_check_broadcast_device(struct clock_event_device *curdev, + struct clock_event_device *newdev) +{ + if ((newdev->features & CLOCK_EVT_FEAT_DUMMY) || + (newdev->features & CLOCK_EVT_FEAT_C3STOP)) + return false; + + if (tick_broadcast_device.mode == TICKDEV_MODE_ONESHOT && + !(newdev->features & CLOCK_EVT_FEAT_ONESHOT)) + return false; + + return !curdev || newdev->rating > curdev->rating; +} + +/* + * Conditionally install/replace broadcast device + */ void tick_install_broadcast_device(struct clock_event_device *dev) { struct clock_event_device *cur = tick_broadcast_device.evtdev; - if ((dev->features & CLOCK_EVT_FEAT_DUMMY) || - (tick_broadcast_device.evtdev && - tick_broadcast_device.evtdev->rating >= dev->rating) || - (dev->features & CLOCK_EVT_FEAT_C3STOP)) + if (!tick_check_broadcast_device(cur, dev)) return; + if (!try_module_get(dev->owner)) return; - clockevents_exchange_device(tick_broadcast_device.evtdev, dev); + clockevents_exchange_device(cur, dev); if (cur) cur->event_handler = clockevents_handle_noop; tick_broadcast_device.evtdev = dev; diff --git a/kernel/time/tick-common.c b/kernel/time/tick-common.c index 433a1e11d13b..c34021650348 100644 --- a/kernel/time/tick-common.c +++ b/kernel/time/tick-common.c @@ -205,6 +205,37 @@ static void tick_setup_device(struct tick_device *td, tick_setup_oneshot(newdev, handler, next_event); } +static bool tick_check_percpu(struct clock_event_device *curdev, + struct clock_event_device *newdev, int cpu) +{ + if (!cpumask_test_cpu(cpu, newdev->cpumask)) + return false; + if (cpumask_equal(newdev->cpumask, cpumask_of(cpu))) + return true; + /* Check if irq affinity can be set */ + if (newdev->irq >= 0 && !irq_can_set_affinity(newdev->irq)) + return false; + /* Prefer an existing cpu local device */ + if (curdev && cpumask_equal(curdev->cpumask, cpumask_of(cpu))) + return false; + return true; +} + +static bool tick_check_preferred(struct clock_event_device *curdev, + struct clock_event_device *newdev) +{ + /* Prefer oneshot capable device */ + if (!(newdev->features & CLOCK_EVT_FEAT_ONESHOT)) { + if (curdev && (curdev->features & CLOCK_EVT_FEAT_ONESHOT)) + return false; + if (tick_oneshot_mode_active()) + return false; + } + + /* Use the higher rated one */ + return !curdev || newdev->rating > curdev->rating; +} + /* * Check, if the new registered device should be used. Called with * clockevents_lock held and interrupts disabled. @@ -223,40 +254,12 @@ void tick_check_new_device(struct clock_event_device *newdev) curdev = td->evtdev; /* cpu local device ? */ - if (!cpumask_equal(newdev->cpumask, cpumask_of(cpu))) { + if (!tick_check_percpu(curdev, newdev, cpu)) + goto out_bc; - /* - * If the cpu affinity of the device interrupt can not - * be set, ignore it. - */ - if (!irq_can_set_affinity(newdev->irq)) - goto out_bc; - - /* - * If we have a cpu local device already, do not replace it - * by a non cpu local device - */ - if (curdev && cpumask_equal(curdev->cpumask, cpumask_of(cpu))) - goto out_bc; - } - - /* - * If we have an active device, then check the rating and the oneshot - * feature. - */ - if (curdev) { - /* - * Prefer one shot capable devices ! - */ - if ((curdev->features & CLOCK_EVT_FEAT_ONESHOT) && - !(newdev->features & CLOCK_EVT_FEAT_ONESHOT)) - goto out_bc; - /* - * Check the rating - */ - if (curdev->rating >= newdev->rating) - goto out_bc; - } + /* Preference decision */ + if (!tick_check_preferred(curdev, newdev)) + goto out_bc; if (!try_module_get(newdev->owner)) return; From 03e13cf5ee60584fe0c831682c67212effb7fca4 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Thu, 25 Apr 2013 20:31:50 +0000 Subject: [PATCH 174/198] clockevents: Implement unbind functionality Provide a sysfs interface to allow unbinding of clockevent devices. The device is unbound if it is unused or if there is a replacement device available. Unbinding of broadcast devices is not supported as we don't want to foster that nonsense. If no replacement device is available the unbind returns -EBUSY. Unbind is available from the kernel and through sysfs, which is necessary to drop the module refcount. Signed-off-by: Thomas Gleixner Cc: John Stultz Cc: Magnus Damm Link: http://lkml.kernel.org/r/20130425143436.499216659@linutronix.de Signed-off-by: Thomas Gleixner --- include/linux/clockchips.h | 1 + kernel/time/clockevents.c | 125 ++++++++++++++++++++++++++++++++++++ kernel/time/clocksource.c | 9 +-- kernel/time/tick-common.c | 24 +++++++ kernel/time/tick-internal.h | 7 ++ 5 files changed, 162 insertions(+), 4 deletions(-) diff --git a/include/linux/clockchips.h b/include/linux/clockchips.h index ae1193bcf074..0857922e8ad0 100644 --- a/include/linux/clockchips.h +++ b/include/linux/clockchips.h @@ -141,6 +141,7 @@ static inline unsigned long div_sc(unsigned long ticks, unsigned long nsec, extern u64 clockevent_delta2ns(unsigned long latch, struct clock_event_device *evt); extern void clockevents_register_device(struct clock_event_device *dev); +extern int clockevents_unbind_device(struct clock_event_device *ced, int cpu); extern void clockevents_config(struct clock_event_device *dev, u32 freq); extern void clockevents_config_and_register(struct clock_event_device *dev, diff --git a/kernel/time/clockevents.c b/kernel/time/clockevents.c index 0a23f4f29934..38959c866789 100644 --- a/kernel/time/clockevents.c +++ b/kernel/time/clockevents.c @@ -25,6 +25,13 @@ static LIST_HEAD(clockevent_devices); static LIST_HEAD(clockevents_released); /* Protection for the above */ static DEFINE_RAW_SPINLOCK(clockevents_lock); +/* Protection for unbind operations */ +static DEFINE_MUTEX(clockevents_mutex); + +struct ce_unbind { + struct clock_event_device *ce; + int res; +}; /** * clockevents_delta2ns - Convert a latch value (device ticks) to nanoseconds @@ -245,6 +252,90 @@ static void clockevents_notify_released(void) } } +/* + * Try to install a replacement clock event device + */ +static int clockevents_replace(struct clock_event_device *ced) +{ + struct clock_event_device *dev, *newdev = NULL; + + list_for_each_entry(dev, &clockevent_devices, list) { + if (dev == ced || dev->mode != CLOCK_EVT_MODE_UNUSED) + continue; + + if (!tick_check_replacement(newdev, dev)) + continue; + + if (!try_module_get(dev->owner)) + continue; + + if (newdev) + module_put(newdev->owner); + newdev = dev; + } + if (newdev) { + tick_install_replacement(newdev); + list_del_init(&ced->list); + } + return newdev ? 0 : -EBUSY; +} + +/* + * Called with clockevents_mutex and clockevents_lock held + */ +static int __clockevents_try_unbind(struct clock_event_device *ced, int cpu) +{ + /* Fast track. Device is unused */ + if (ced->mode == CLOCK_EVT_MODE_UNUSED) { + list_del_init(&ced->list); + return 0; + } + + return ced == per_cpu(tick_cpu_device, cpu).evtdev ? -EAGAIN : -EBUSY; +} + +/* + * SMP function call to unbind a device + */ +static void __clockevents_unbind(void *arg) +{ + struct ce_unbind *cu = arg; + int res; + + raw_spin_lock(&clockevents_lock); + res = __clockevents_try_unbind(cu->ce, smp_processor_id()); + if (res == -EAGAIN) + res = clockevents_replace(cu->ce); + cu->res = res; + raw_spin_unlock(&clockevents_lock); +} + +/* + * Issues smp function call to unbind a per cpu device. Called with + * clockevents_mutex held. + */ +static int clockevents_unbind(struct clock_event_device *ced, int cpu) +{ + struct ce_unbind cu = { .ce = ced, .res = -ENODEV }; + + smp_call_function_single(cpu, __clockevents_unbind, &cu, 1); + return cu.res; +} + +/* + * Unbind a clockevents device. + */ +int clockevents_unbind_device(struct clock_event_device *ced, int cpu) +{ + int ret; + + mutex_lock(&clockevents_mutex); + ret = clockevents_unbind(ced, cpu); + mutex_unlock(&clockevents_mutex); + return ret; +} +EXPORT_SYMBOL_GPL(clockevents_unbind); + /** * clockevents_register_device - register a clock event device * @dev: device to register @@ -487,6 +578,38 @@ static ssize_t sysfs_show_current_tick_dev(struct device *dev, } static DEVICE_ATTR(current_device, 0444, sysfs_show_current_tick_dev, NULL); +/* We don't support the abomination of removable broadcast devices */ +static ssize_t sysfs_unbind_tick_dev(struct device *dev, + struct device_attribute *attr, + const char *buf, size_t count) +{ + char name[CS_NAME_LEN]; + size_t ret = sysfs_get_uname(buf, name, count); + struct clock_event_device *ce; + + if (ret < 0) + return ret; + + ret = -ENODEV; + mutex_lock(&clockevents_mutex); + raw_spin_lock_irq(&clockevents_lock); + list_for_each_entry(ce, &clockevent_devices, list) { + if (!strcmp(ce->name, name)) { + ret = __clockevents_try_unbind(ce, dev->id); + break; + } + } + raw_spin_unlock_irq(&clockevents_lock); + /* + * We hold clockevents_mutex, so ce can't go away + */ + if (ret == -EAGAIN) + ret = clockevents_unbind(ce, dev->id); + mutex_unlock(&clockevents_mutex); + return ret ? ret : count; +} +static DEVICE_ATTR(unbind_device, 0200, NULL, sysfs_unbind_tick_dev); + #ifdef CONFIG_GENERIC_CLOCKEVENTS_BROADCAST static struct device tick_bc_dev = { .init_name = "broadcast", @@ -529,6 +652,8 @@ static int __init tick_init_sysfs(void) err = device_register(dev); if (!err) err = device_create_file(dev, &dev_attr_current_device); + if (!err) + err = device_create_file(dev, &dev_attr_unbind_device); if (err) return err; } diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c index 31b90332f47b..6d05b00410cc 100644 --- a/kernel/time/clocksource.c +++ b/kernel/time/clocksource.c @@ -31,6 +31,8 @@ #include #include +#include "tick-internal.h" + void timecounter_init(struct timecounter *tc, const struct cyclecounter *cc, u64 start_tstamp) @@ -174,7 +176,6 @@ clocks_calc_mult_shift(u32 *mult, u32 *shift, u32 from, u32 to, u32 maxsec) static struct clocksource *curr_clocksource; static LIST_HEAD(clocksource_list); static DEFINE_MUTEX(clocksource_mutex); -#define CS_NAME_LEN 32 static char override_name[CS_NAME_LEN]; static int finished_booting; @@ -864,7 +865,7 @@ sysfs_show_current_clocksources(struct device *dev, return count; } -static size_t clocksource_get_uname(const char *buf, char *dst, size_t cnt) +size_t sysfs_get_uname(const char *buf, char *dst, size_t cnt) { size_t ret = cnt; @@ -899,7 +900,7 @@ static ssize_t sysfs_override_clocksource(struct device *dev, mutex_lock(&clocksource_mutex); - ret = clocksource_get_uname(buf, override_name, count); + ret = sysfs_get_uname(buf, override_name, count); if (ret >= 0) clocksource_select(); @@ -925,7 +926,7 @@ static ssize_t sysfs_unbind_clocksource(struct device *dev, char name[CS_NAME_LEN]; size_t ret; - ret = clocksource_get_uname(buf, name, count); + ret = sysfs_get_uname(buf, name, count); if (ret < 0) return ret; diff --git a/kernel/time/tick-common.c b/kernel/time/tick-common.c index c34021650348..5edfb4806032 100644 --- a/kernel/time/tick-common.c +++ b/kernel/time/tick-common.c @@ -205,6 +205,17 @@ static void tick_setup_device(struct tick_device *td, tick_setup_oneshot(newdev, handler, next_event); } +void tick_install_replacement(struct clock_event_device *newdev) +{ + struct tick_device *td = &__get_cpu_var(tick_cpu_device); + int cpu = smp_processor_id(); + + clockevents_exchange_device(td->evtdev, newdev); + tick_setup_device(td, newdev, cpu, cpumask_of(cpu)); + if (newdev->features & CLOCK_EVT_FEAT_ONESHOT) + tick_oneshot_notify(); +} + static bool tick_check_percpu(struct clock_event_device *curdev, struct clock_event_device *newdev, int cpu) { @@ -236,6 +247,19 @@ static bool tick_check_preferred(struct clock_event_device *curdev, return !curdev || newdev->rating > curdev->rating; } +/* + * Check whether the new device is a better fit than curdev. curdev + * can be NULL ! + */ +bool tick_check_replacement(struct clock_event_device *curdev, + struct clock_event_device *newdev) +{ + if (tick_check_percpu(curdev, newdev, smp_processor_id())) + return false; + + return tick_check_preferred(curdev, newdev); +} + /* * Check, if the new registered device should be used. Called with * clockevents_lock held and interrupts disabled. diff --git a/kernel/time/tick-internal.h b/kernel/time/tick-internal.h index 06bfc8802dfb..be1690eaecff 100644 --- a/kernel/time/tick-internal.h +++ b/kernel/time/tick-internal.h @@ -11,6 +11,8 @@ extern seqlock_t jiffies_lock; #define TICK_DO_TIMER_NONE -1 #define TICK_DO_TIMER_BOOT -2 +#define CS_NAME_LEN 32 + DECLARE_PER_CPU(struct tick_device, tick_cpu_device); extern ktime_t tick_next_period; extern ktime_t tick_period; @@ -23,9 +25,14 @@ extern void tick_handover_do_timer(int *cpup); extern void tick_shutdown(unsigned int *cpup); extern void tick_suspend(void); extern void tick_resume(void); +extern bool tick_check_replacement(struct clock_event_device *curdev, + struct clock_event_device *newdev); +extern void tick_install_replacement(struct clock_event_device *dev); extern void clockevents_shutdown(struct clock_event_device *dev); +extern size_t sysfs_get_uname(const char *buf, char *dst, size_t cnt); + /* * NO_HZ / high resolution timer shared code */ From c7e99fc75de8882bc4104455ace366d9d3599a96 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Tue, 28 May 2013 09:28:02 +0200 Subject: [PATCH 175/198] clockevents: Define CS_NAME_LEN unconditionally Unbreak architectures which do not use clockevents, but require to build some of the core timekeeping infrastructure Reported-by: Ingo Molnar Signed-off-by: Thomas Gleixner --- kernel/time/tick-internal.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/kernel/time/tick-internal.h b/kernel/time/tick-internal.h index be1690eaecff..bc906cad709b 100644 --- a/kernel/time/tick-internal.h +++ b/kernel/time/tick-internal.h @@ -6,13 +6,13 @@ extern seqlock_t jiffies_lock; +#define CS_NAME_LEN 32 + #ifdef CONFIG_GENERIC_CLOCKEVENTS_BUILD #define TICK_DO_TIMER_NONE -1 #define TICK_DO_TIMER_BOOT -2 -#define CS_NAME_LEN 32 - DECLARE_PER_CPU(struct tick_device, tick_cpu_device); extern ktime_t tick_next_period; extern ktime_t tick_period; From 1eaff67266b6b6c97bbd33cf2c20577822836413 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Tue, 28 May 2013 09:48:46 +0200 Subject: [PATCH 176/198] clocksource: Implement clocksource_select_fallback() for CONFIG_ARCH_USES_GETTIMEOFFSET=y commit 7eaeb34305 (clocksource: Provide unbind interface in sysfs) implemented clocksource_select_fallback() which is not defined for CONFIG_ARCH_USES_GETTIMEOFFSET=y. Add an empty inline function for that. Reported-by: Ingo Molnar Reported-by: fengguang.wu@intel.com Signed-off-by: Thomas Gleixner --- kernel/time/clocksource.c | 1 + 1 file changed, 1 insertion(+) diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c index 6d05b00410cc..e713ef7d19a7 100644 --- a/kernel/time/clocksource.c +++ b/kernel/time/clocksource.c @@ -632,6 +632,7 @@ static void clocksource_select_fallback(void) #else /* !CONFIG_ARCH_USES_GETTIMEOFFSET */ static inline void clocksource_select(void) { } +static inline void clocksource_select_fallback(void) { } #endif From a44b8bd6072ac82fc49af941de8c1ac85d94852d Mon Sep 17 00:00:00 2001 From: Liu Ying Date: Tue, 7 May 2013 16:38:40 +0800 Subject: [PATCH 177/198] ktime: Use macro NSEC_PER_USEC where appropriate We've got the macro NSEC_PER_USEC defined in header file include/linux/time.h. To make the code decent, this patch replaces the immediate number 1000 to convert bewteen a time value in microseconds and one in nanoseconds with the macro NSEC_PER_USEC. Signed-off-by: Liu Ying Cc: David S. Miller Cc: Daniel Borkmann Cc: Thomas Gleixner Cc: John Stultz Signed-off-by: John Stultz --- include/linux/ktime.h | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/include/linux/ktime.h b/include/linux/ktime.h index bbca12804d12..608728a164a9 100644 --- a/include/linux/ktime.h +++ b/include/linux/ktime.h @@ -229,7 +229,8 @@ static inline ktime_t timespec_to_ktime(const struct timespec ts) static inline ktime_t timeval_to_ktime(const struct timeval tv) { return (ktime_t) { .tv = { .sec = (s32)tv.tv_sec, - .nsec = (s32)tv.tv_usec * 1000 } }; + .nsec = (s32)(tv.tv_usec * + NSEC_PER_USEC) } }; } /** @@ -320,12 +321,12 @@ static inline s64 ktime_us_delta(const ktime_t later, const ktime_t earlier) static inline ktime_t ktime_add_us(const ktime_t kt, const u64 usec) { - return ktime_add_ns(kt, usec * 1000); + return ktime_add_ns(kt, usec * NSEC_PER_USEC); } static inline ktime_t ktime_sub_us(const ktime_t kt, const u64 usec) { - return ktime_sub_ns(kt, usec * 1000); + return ktime_sub_ns(kt, usec * NSEC_PER_USEC); } extern ktime_t ktime_add_safe(const ktime_t lhs, const ktime_t rhs); From 35b210855069ae34450eeaafc0f1b5bfa4f71c87 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Thu, 16 May 2013 15:47:49 +0200 Subject: [PATCH 178/198] ktime: Add __must_check prefix to ktime_to_timespec_cond The function is currently mainly used in the networking code and if others start using it, they must check the result, otherwise it cannot be determined if the timespec conversion suceeded. Currently no user lacks this check, but make future users aware of a possible misusage. Signed-off-by: Daniel Borkmann Signed-off-by: John Stultz --- include/linux/ktime.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/include/linux/ktime.h b/include/linux/ktime.h index 608728a164a9..fc66b301b648 100644 --- a/include/linux/ktime.h +++ b/include/linux/ktime.h @@ -339,7 +339,8 @@ extern ktime_t ktime_add_safe(const ktime_t lhs, const ktime_t rhs); * * Returns true if there was a successful conversion, false if kt was 0. */ -static inline bool ktime_to_timespec_cond(const ktime_t kt, struct timespec *ts) +static inline __must_check bool ktime_to_timespec_cond(const ktime_t kt, + struct timespec *ts) { if (kt.tv64) { *ts = ktime_to_timespec(kt); From 55a68c23e0a675b2b8ac2656fd6edbf98b78e4c6 Mon Sep 17 00:00:00 2001 From: Pavel Machek Date: Tue, 7 May 2013 22:11:26 +0200 Subject: [PATCH 179/198] dw_apb_timer_of.c: Remove parts that were picoxcell-specific It seems we made a mistake when creating dw_apb_timer_of.c: picoxcell sched_clock had parts that were not related to dw_apb_timer, yet we moved them to dw_apb_timer_of, and tried to use them on socfpga. This results in system where user/system time is not measured properly, as demonstrated by time dd if=/dev/urandom of=/dev/zero bs=100000 count=100 So this patch switches sched_clock to hardware that exists on both platforms, and adds missing of_node_put() in dw_apb_timer_init(). Signed-off-by: Pavel Machek Acked-by: Jamie Iles Signed-off-by: John Stultz --- arch/arm/mach-picoxcell/common.h | 2 - drivers/clocksource/dw_apb_timer.c | 6 --- drivers/clocksource/dw_apb_timer_of.c | 54 ++++++++++++--------------- include/linux/dw_apb_timer.h | 6 +++ 4 files changed, 30 insertions(+), 38 deletions(-) diff --git a/arch/arm/mach-picoxcell/common.h b/arch/arm/mach-picoxcell/common.h index 481b42a4ef15..237fb3bcbd04 100644 --- a/arch/arm/mach-picoxcell/common.h +++ b/arch/arm/mach-picoxcell/common.h @@ -12,6 +12,4 @@ #include -extern void dw_apb_timer_init(void); - #endif /* __PICOXCELL_COMMON_H__ */ diff --git a/drivers/clocksource/dw_apb_timer.c b/drivers/clocksource/dw_apb_timer.c index e54ca1062d8e..e7042bc5c7d2 100644 --- a/drivers/clocksource/dw_apb_timer.c +++ b/drivers/clocksource/dw_apb_timer.c @@ -21,12 +21,6 @@ #define APBT_MIN_PERIOD 4 #define APBT_MIN_DELTA_USEC 200 -#define APBTMR_N_LOAD_COUNT 0x00 -#define APBTMR_N_CURRENT_VALUE 0x04 -#define APBTMR_N_CONTROL 0x08 -#define APBTMR_N_EOI 0x0c -#define APBTMR_N_INT_STATUS 0x10 - #define APBTMRS_INT_STATUS 0xa0 #define APBTMRS_EOI 0xa4 #define APBTMRS_RAW_INT_STATUS 0xa8 diff --git a/drivers/clocksource/dw_apb_timer_of.c b/drivers/clocksource/dw_apb_timer_of.c index ab09ed3742ee..44a3b9163c46 100644 --- a/drivers/clocksource/dw_apb_timer_of.c +++ b/drivers/clocksource/dw_apb_timer_of.c @@ -57,6 +57,15 @@ static void add_clockevent(struct device_node *event_timer) dw_apb_clockevent_register(ced); } +static void __iomem *sched_io_base; + +/* This is actually same as __apbt_read_clocksource(), but with + different interface */ +static u32 read_sched_clock_sptimer(void) +{ + return ~__raw_readl(sched_io_base + APBTMR_N_CURRENT_VALUE); +} + static void add_clocksource(struct device_node *source_timer) { void __iomem *iobase; @@ -71,42 +80,28 @@ static void add_clocksource(struct device_node *source_timer) dw_apb_clocksource_start(cs); dw_apb_clocksource_register(cs); -} -static void __iomem *sched_io_base; - -static u32 read_sched_clock(void) -{ - return __raw_readl(sched_io_base); -} - -static const struct of_device_id sptimer_ids[] __initconst = { - { .compatible = "picochip,pc3x2-rtc" }, - { .compatible = "snps,dw-apb-timer-sp" }, - { /* Sentinel */ }, -}; - -static void init_sched_clock(void) -{ - struct device_node *sched_timer; - u32 rate; - - sched_timer = of_find_matching_node(NULL, sptimer_ids); - if (!sched_timer) - panic("No RTC for sched clock to use"); - - timer_get_base_and_rate(sched_timer, &sched_io_base, &rate); - of_node_put(sched_timer); - - setup_sched_clock(read_sched_clock, 32, rate); + sched_io_base = iobase; + setup_sched_clock(read_sched_clock_sptimer, 32, rate); } static const struct of_device_id osctimer_ids[] __initconst = { { .compatible = "picochip,pc3x2-timer" }, { .compatible = "snps,dw-apb-timer-osc" }, - {}, + { .compatible = "snps,dw-apb-timer-sp" }, + { /* Sentinel */ }, }; +/* + You don't have to use dw_apb_timer for scheduler clock, + this should also work fine on arm: + + twd_local_timer_of_register(); + arch_timer_of_register(); + arch_timer_sched_clock_init(); +*/ + + void __init dw_apb_timer_init(void) { struct device_node *event_timer, *source_timer; @@ -121,7 +116,6 @@ void __init dw_apb_timer_init(void) panic("No timer for clocksource"); add_clocksource(source_timer); + of_node_put(event_timer); of_node_put(source_timer); - - init_sched_clock(); } diff --git a/include/linux/dw_apb_timer.h b/include/linux/dw_apb_timer.h index b1cd9597c241..de0904e38f33 100644 --- a/include/linux/dw_apb_timer.h +++ b/include/linux/dw_apb_timer.h @@ -17,6 +17,12 @@ #include #include +#define APBTMR_N_LOAD_COUNT 0x00 +#define APBTMR_N_CURRENT_VALUE 0x04 +#define APBTMR_N_CONTROL 0x08 +#define APBTMR_N_EOI 0x0c +#define APBTMR_N_INT_STATUS 0x10 + #define APBTMRS_REG_SIZE 0x14 struct dw_apb_timer { From 0a0a7e66fa269de78975ea8d4e825a66d92b8d70 Mon Sep 17 00:00:00 2001 From: Baruch Siach Date: Sun, 26 May 2013 15:12:46 +0300 Subject: [PATCH 180/198] clocksource: dw_apb: Remove unused header The time.h header seems not to be used by current code. Removing this include allows the driver to build on other architecture that do not have this header. Cc: Ingo Molnar Cc: Peter Zijlstra Cc: John Stultz Cc: Thomas Gleixner Cc: Jamie Iles Cc: Dinh Nguyen Acked-by: Jamie Iles Signed-off-by: Baruch Siach [tweaked commit message and header] Signed-off-by: John Stultz --- drivers/clocksource/dw_apb_timer_of.c | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/clocksource/dw_apb_timer_of.c b/drivers/clocksource/dw_apb_timer_of.c index 44a3b9163c46..8d2f809a0967 100644 --- a/drivers/clocksource/dw_apb_timer_of.c +++ b/drivers/clocksource/dw_apb_timer_of.c @@ -21,7 +21,6 @@ #include #include -#include #include static void timer_get_base_and_rate(struct device_node *np, From 3565184ed0c1ea46bea5b792da5f72a83c43e49b Mon Sep 17 00:00:00 2001 From: David Vrabel Date: Mon, 13 May 2013 18:56:06 +0100 Subject: [PATCH 181/198] x86: Increase precision of x86_platform.get/set_wallclock() All the virtualized platforms (KVM, lguest and Xen) have persistent wallclocks that have more than one second of precision. read_persistent_wallclock() and update_persistent_wallclock() allow for nanosecond precision but their implementation on x86 with x86_platform.get/set_wallclock() only allows for one second precision. This means guests may see a wallclock time that is off by up to 1 second. Make set_wallclock() and get_wallclock() take a struct timespec parameter (which allows for nanosecond precision) so KVM and Xen guests may start with a more accurate wallclock time and a Xen dom0 can maintain a more accurate wallclock for guests. Signed-off-by: David Vrabel Signed-off-by: John Stultz --- arch/x86/include/asm/mc146818rtc.h | 4 ++-- arch/x86/include/asm/x86_init.h | 6 ++++-- arch/x86/kernel/kvmclock.c | 9 +++------ arch/x86/kernel/rtc.c | 17 +++++++---------- arch/x86/lguest/boot.c | 4 ++-- arch/x86/platform/efi/efi.c | 10 ++++++---- arch/x86/xen/time.c | 19 ++++++------------- include/linux/efi.h | 4 ++-- 8 files changed, 32 insertions(+), 41 deletions(-) diff --git a/arch/x86/include/asm/mc146818rtc.h b/arch/x86/include/asm/mc146818rtc.h index d354fb781c57..a55c7efcc4ed 100644 --- a/arch/x86/include/asm/mc146818rtc.h +++ b/arch/x86/include/asm/mc146818rtc.h @@ -95,8 +95,8 @@ static inline unsigned char current_lock_cmos_reg(void) unsigned char rtc_cmos_read(unsigned char addr); void rtc_cmos_write(unsigned char val, unsigned char addr); -extern int mach_set_rtc_mmss(unsigned long nowtime); -extern unsigned long mach_get_cmos_time(void); +extern int mach_set_rtc_mmss(const struct timespec *now); +extern void mach_get_cmos_time(struct timespec *now); #define RTC_IRQ 8 diff --git a/arch/x86/include/asm/x86_init.h b/arch/x86/include/asm/x86_init.h index d8d99222b36a..828a1565ba57 100644 --- a/arch/x86/include/asm/x86_init.h +++ b/arch/x86/include/asm/x86_init.h @@ -142,6 +142,8 @@ struct x86_cpuinit_ops { void (*fixup_cpu_id)(struct cpuinfo_x86 *c, int node); }; +struct timespec; + /** * struct x86_platform_ops - platform specific runtime functions * @calibrate_tsc: calibrate TSC @@ -156,8 +158,8 @@ struct x86_cpuinit_ops { */ struct x86_platform_ops { unsigned long (*calibrate_tsc)(void); - unsigned long (*get_wallclock)(void); - int (*set_wallclock)(unsigned long nowtime); + void (*get_wallclock)(struct timespec *ts); + int (*set_wallclock)(const struct timespec *ts); void (*iommu_shutdown)(void); bool (*is_untracked_pat_range)(u64 start, u64 end); void (*nmi_init)(void); diff --git a/arch/x86/kernel/kvmclock.c b/arch/x86/kernel/kvmclock.c index d2c381280e3c..0db81ab511cc 100644 --- a/arch/x86/kernel/kvmclock.c +++ b/arch/x86/kernel/kvmclock.c @@ -48,10 +48,9 @@ static struct pvclock_wall_clock wall_clock; * have elapsed since the hypervisor wrote the data. So we try to account for * that with system time */ -static unsigned long kvm_get_wallclock(void) +static void kvm_get_wallclock(struct timespec *now) { struct pvclock_vcpu_time_info *vcpu_time; - struct timespec ts; int low, high; int cpu; @@ -64,14 +63,12 @@ static unsigned long kvm_get_wallclock(void) cpu = smp_processor_id(); vcpu_time = &hv_clock[cpu].pvti; - pvclock_read_wallclock(&wall_clock, vcpu_time, &ts); + pvclock_read_wallclock(&wall_clock, vcpu_time, now); preempt_enable(); - - return ts.tv_sec; } -static int kvm_set_wallclock(unsigned long now) +static int kvm_set_wallclock(const struct timespec *now) { return -1; } diff --git a/arch/x86/kernel/rtc.c b/arch/x86/kernel/rtc.c index 198eb201ed3b..0aa29394ed6f 100644 --- a/arch/x86/kernel/rtc.c +++ b/arch/x86/kernel/rtc.c @@ -38,8 +38,9 @@ EXPORT_SYMBOL(rtc_lock); * jump to the next second precisely 500 ms later. Check the Motorola * MC146818A or Dallas DS12887 data sheet for details. */ -int mach_set_rtc_mmss(unsigned long nowtime) +int mach_set_rtc_mmss(const struct timespec *now) { + unsigned long nowtime = now->tv_sec; struct rtc_time tm; int retval = 0; @@ -58,7 +59,7 @@ int mach_set_rtc_mmss(unsigned long nowtime) return retval; } -unsigned long mach_get_cmos_time(void) +void mach_get_cmos_time(struct timespec *now) { unsigned int status, year, mon, day, hour, min, sec, century = 0; unsigned long flags; @@ -107,7 +108,8 @@ unsigned long mach_get_cmos_time(void) } else year += CMOS_YEARS_OFFS; - return mktime(year, mon, day, hour, min, sec); + now->tv_sec = mktime(year, mon, day, hour, min, sec); + now->tv_nsec = 0; } /* Routines for accessing the CMOS RAM/RTC. */ @@ -135,18 +137,13 @@ EXPORT_SYMBOL(rtc_cmos_write); int update_persistent_clock(struct timespec now) { - return x86_platform.set_wallclock(now.tv_sec); + return x86_platform.set_wallclock(&now); } /* not static: needed by APM */ void read_persistent_clock(struct timespec *ts) { - unsigned long retval; - - retval = x86_platform.get_wallclock(); - - ts->tv_sec = retval; - ts->tv_nsec = 0; + x86_platform.get_wallclock(ts); } diff --git a/arch/x86/lguest/boot.c b/arch/x86/lguest/boot.c index 7114c63f047d..8424d5adcfa2 100644 --- a/arch/x86/lguest/boot.c +++ b/arch/x86/lguest/boot.c @@ -882,9 +882,9 @@ int lguest_setup_irq(unsigned int irq) * It would be far better for everyone if the Guest had its own clock, but * until then the Host gives us the time on every interrupt. */ -static unsigned long lguest_get_wallclock(void) +static void lguest_get_wallclock(struct timespec *now) { - return lguest_data.time.tv_sec; + *now = lguest_data.time; } /* diff --git a/arch/x86/platform/efi/efi.c b/arch/x86/platform/efi/efi.c index 55856b2310d3..dd3b82530145 100644 --- a/arch/x86/platform/efi/efi.c +++ b/arch/x86/platform/efi/efi.c @@ -352,8 +352,9 @@ static efi_status_t __init phys_efi_get_time(efi_time_t *tm, return status; } -int efi_set_rtc_mmss(unsigned long nowtime) +int efi_set_rtc_mmss(const struct timespec *now) { + unsigned long nowtime = now->tv_sec; efi_status_t status; efi_time_t eft; efi_time_cap_t cap; @@ -388,7 +389,7 @@ int efi_set_rtc_mmss(unsigned long nowtime) return 0; } -unsigned long efi_get_time(void) +void efi_get_time(struct timespec *now) { efi_status_t status; efi_time_t eft; @@ -398,8 +399,9 @@ unsigned long efi_get_time(void) if (status != EFI_SUCCESS) pr_err("Oops: efitime: can't read time!\n"); - return mktime(eft.year, eft.month, eft.day, eft.hour, - eft.minute, eft.second); + now->tv_sec = mktime(eft.year, eft.month, eft.day, eft.hour, + eft.minute, eft.second); + now->tv_nsec = 0; } /* diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c index 3d88bfdf9e1c..a1947ac2da82 100644 --- a/arch/x86/xen/time.c +++ b/arch/x86/xen/time.c @@ -191,32 +191,25 @@ static void xen_read_wallclock(struct timespec *ts) put_cpu_var(xen_vcpu); } -static unsigned long xen_get_wallclock(void) +static void xen_get_wallclock(struct timespec *now) { - struct timespec ts; - - xen_read_wallclock(&ts); - return ts.tv_sec; + xen_read_wallclock(now); } -static int xen_set_wallclock(unsigned long now) +static int xen_set_wallclock(const struct timespec *now) { struct xen_platform_op op; - int rc; /* do nothing for domU */ if (!xen_initial_domain()) return -1; op.cmd = XENPF_settime; - op.u.settime.secs = now; - op.u.settime.nsecs = 0; + op.u.settime.secs = now->tv_sec; + op.u.settime.nsecs = now->tv_nsec; op.u.settime.system_time = xen_clocksource_read(); - rc = HYPERVISOR_dom0_op(&op); - WARN(rc != 0, "XENPF_settime failed: now=%ld\n", now); - - return rc; + return HYPERVISOR_dom0_op(&op); } static struct clocksource xen_clocksource __read_mostly = { diff --git a/include/linux/efi.h b/include/linux/efi.h index 2bc0ad78d058..0068bba6f8b6 100644 --- a/include/linux/efi.h +++ b/include/linux/efi.h @@ -594,8 +594,8 @@ extern u64 efi_mem_attribute (unsigned long phys_addr, unsigned long size); extern int __init efi_uart_console_only (void); extern void efi_initialize_iomem_resources(struct resource *code_resource, struct resource *data_resource, struct resource *bss_resource); -extern unsigned long efi_get_time(void); -extern int efi_set_rtc_mmss(unsigned long nowtime); +extern void efi_get_time(struct timespec *now); +extern int efi_set_rtc_mmss(const struct timespec *now); extern void efi_reserve_boot_services(void); extern struct efi_memory_map memmap; From 6cffe00f7d4e24679eae6b7aae4caaf915288256 Mon Sep 17 00:00:00 2001 From: Todd Poynor Date: Wed, 15 May 2013 14:38:11 -0700 Subject: [PATCH 182/198] alarmtimer: Add functions for timerfd support Add functions needed for hooking up alarmtimer to timerfd: * alarm_restart: Similar to hrtimer_restart, restart an alarmtimer after the expires time has already been updated (as with alarm_forward). * alarm_forward_now: Similar to hrtimer_forward_now, move the expires time forward to an interval from the current time of the associated clock. * alarm_start_relative: Start an alarmtimer with an expires time relative to the current time of the associated clock. * alarm_expires_remaining: Similar to hrtimer_expires_remaining, return the amount of time remaining until alarm expiry. Signed-off-by: Todd Poynor Signed-off-by: John Stultz --- include/linux/alarmtimer.h | 4 ++++ kernel/time/alarmtimer.c | 39 +++++++++++++++++++++++++++++++++++++- 2 files changed, 42 insertions(+), 1 deletion(-) diff --git a/include/linux/alarmtimer.h b/include/linux/alarmtimer.h index 9069694e70eb..a899402a5a0e 100644 --- a/include/linux/alarmtimer.h +++ b/include/linux/alarmtimer.h @@ -44,10 +44,14 @@ struct alarm { void alarm_init(struct alarm *alarm, enum alarmtimer_type type, enum alarmtimer_restart (*function)(struct alarm *, ktime_t)); int alarm_start(struct alarm *alarm, ktime_t start); +int alarm_start_relative(struct alarm *alarm, ktime_t start); +void alarm_restart(struct alarm *alarm); int alarm_try_to_cancel(struct alarm *alarm); int alarm_cancel(struct alarm *alarm); u64 alarm_forward(struct alarm *alarm, ktime_t now, ktime_t interval); +u64 alarm_forward_now(struct alarm *alarm, ktime_t interval); +ktime_t alarm_expires_remaining(const struct alarm *alarm); /* Provide way to access the rtc device being used by alarmtimers */ struct rtc_device *alarmtimer_get_rtcdev(void); diff --git a/kernel/time/alarmtimer.c b/kernel/time/alarmtimer.c index f11d83b12949..3e5cba274475 100644 --- a/kernel/time/alarmtimer.c +++ b/kernel/time/alarmtimer.c @@ -199,6 +199,12 @@ static enum hrtimer_restart alarmtimer_fired(struct hrtimer *timer) } +ktime_t alarm_expires_remaining(const struct alarm *alarm) +{ + struct alarm_base *base = &alarm_bases[alarm->type]; + return ktime_sub(alarm->node.expires, base->gettime()); +} + #ifdef CONFIG_RTC_CLASS /** * alarmtimer_suspend - Suspend time callback @@ -305,7 +311,7 @@ void alarm_init(struct alarm *alarm, enum alarmtimer_type type, } /** - * alarm_start - Sets an alarm to fire + * alarm_start - Sets an absolute alarm to fire * @alarm: ptr to alarm to set * @start: time to run the alarm */ @@ -324,6 +330,31 @@ int alarm_start(struct alarm *alarm, ktime_t start) return ret; } +/** + * alarm_start_relative - Sets a relative alarm to fire + * @alarm: ptr to alarm to set + * @start: time relative to now to run the alarm + */ +int alarm_start_relative(struct alarm *alarm, ktime_t start) +{ + struct alarm_base *base = &alarm_bases[alarm->type]; + + start = ktime_add(start, base->gettime()); + return alarm_start(alarm, start); +} + +void alarm_restart(struct alarm *alarm) +{ + struct alarm_base *base = &alarm_bases[alarm->type]; + unsigned long flags; + + spin_lock_irqsave(&base->lock, flags); + hrtimer_set_expires(&alarm->timer, alarm->node.expires); + hrtimer_restart(&alarm->timer); + alarmtimer_enqueue(base, alarm); + spin_unlock_irqrestore(&base->lock, flags); +} + /** * alarm_try_to_cancel - Tries to cancel an alarm timer * @alarm: ptr to alarm to be canceled @@ -394,6 +425,12 @@ u64 alarm_forward(struct alarm *alarm, ktime_t now, ktime_t interval) return overrun; } +u64 alarm_forward_now(struct alarm *alarm, ktime_t interval) +{ + struct alarm_base *base = &alarm_bases[alarm->type]; + + return alarm_forward(alarm, base->gettime(), interval); +} From 11ffa9d6065f344a9bd769a2452f26f2f671e5f8 Mon Sep 17 00:00:00 2001 From: Todd Poynor Date: Wed, 15 May 2013 14:38:12 -0700 Subject: [PATCH 183/198] timerfd: Add alarm timers Add support for clocks CLOCK_REALTIME_ALARM and CLOCK_BOOTTIME_ALARM, thereby enabling wakeup alarm timers via file descriptors. Signed-off-by: Todd Poynor Signed-off-by: John Stultz --- fs/timerfd.c | 131 ++++++++++++++++++++++++++++++++++++++++++--------- 1 file changed, 108 insertions(+), 23 deletions(-) diff --git a/fs/timerfd.c b/fs/timerfd.c index 32b644f03690..929312180dd0 100644 --- a/fs/timerfd.c +++ b/fs/timerfd.c @@ -8,6 +8,7 @@ * */ +#include #include #include #include @@ -26,7 +27,10 @@ #include struct timerfd_ctx { - struct hrtimer tmr; + union { + struct hrtimer tmr; + struct alarm alarm; + } t; ktime_t tintv; ktime_t moffs; wait_queue_head_t wqh; @@ -41,14 +45,19 @@ struct timerfd_ctx { static LIST_HEAD(cancel_list); static DEFINE_SPINLOCK(cancel_lock); +static inline bool isalarm(struct timerfd_ctx *ctx) +{ + return ctx->clockid == CLOCK_REALTIME_ALARM || + ctx->clockid == CLOCK_BOOTTIME_ALARM; +} + /* * This gets called when the timer event triggers. We set the "expired" * flag, but we do not re-arm the timer (in case it's necessary, * tintv.tv64 != 0) until the timer is accessed. */ -static enum hrtimer_restart timerfd_tmrproc(struct hrtimer *htmr) +static void timerfd_triggered(struct timerfd_ctx *ctx) { - struct timerfd_ctx *ctx = container_of(htmr, struct timerfd_ctx, tmr); unsigned long flags; spin_lock_irqsave(&ctx->wqh.lock, flags); @@ -56,10 +65,25 @@ static enum hrtimer_restart timerfd_tmrproc(struct hrtimer *htmr) ctx->ticks++; wake_up_locked(&ctx->wqh); spin_unlock_irqrestore(&ctx->wqh.lock, flags); +} +static enum hrtimer_restart timerfd_tmrproc(struct hrtimer *htmr) +{ + struct timerfd_ctx *ctx = container_of(htmr, struct timerfd_ctx, + t.tmr); + timerfd_triggered(ctx); return HRTIMER_NORESTART; } +static enum alarmtimer_restart timerfd_alarmproc(struct alarm *alarm, + ktime_t now) +{ + struct timerfd_ctx *ctx = container_of(alarm, struct timerfd_ctx, + t.alarm); + timerfd_triggered(ctx); + return ALARMTIMER_NORESTART; +} + /* * Called when the clock was set to cancel the timers in the cancel * list. This will wake up processes waiting on these timers. The @@ -107,8 +131,9 @@ static bool timerfd_canceled(struct timerfd_ctx *ctx) static void timerfd_setup_cancel(struct timerfd_ctx *ctx, int flags) { - if (ctx->clockid == CLOCK_REALTIME && (flags & TFD_TIMER_ABSTIME) && - (flags & TFD_TIMER_CANCEL_ON_SET)) { + if ((ctx->clockid == CLOCK_REALTIME || + ctx->clockid == CLOCK_REALTIME_ALARM) && + (flags & TFD_TIMER_ABSTIME) && (flags & TFD_TIMER_CANCEL_ON_SET)) { if (!ctx->might_cancel) { ctx->might_cancel = true; spin_lock(&cancel_lock); @@ -124,7 +149,11 @@ static ktime_t timerfd_get_remaining(struct timerfd_ctx *ctx) { ktime_t remaining; - remaining = hrtimer_expires_remaining(&ctx->tmr); + if (isalarm(ctx)) + remaining = alarm_expires_remaining(&ctx->t.alarm); + else + remaining = hrtimer_expires_remaining(&ctx->t.tmr); + return remaining.tv64 < 0 ? ktime_set(0, 0): remaining; } @@ -142,11 +171,28 @@ static int timerfd_setup(struct timerfd_ctx *ctx, int flags, ctx->expired = 0; ctx->ticks = 0; ctx->tintv = timespec_to_ktime(ktmr->it_interval); - hrtimer_init(&ctx->tmr, clockid, htmode); - hrtimer_set_expires(&ctx->tmr, texp); - ctx->tmr.function = timerfd_tmrproc; + + if (isalarm(ctx)) { + alarm_init(&ctx->t.alarm, + ctx->clockid == CLOCK_REALTIME_ALARM ? + ALARM_REALTIME : ALARM_BOOTTIME, + timerfd_alarmproc); + } else { + hrtimer_init(&ctx->t.tmr, clockid, htmode); + hrtimer_set_expires(&ctx->t.tmr, texp); + ctx->t.tmr.function = timerfd_tmrproc; + } + if (texp.tv64 != 0) { - hrtimer_start(&ctx->tmr, texp, htmode); + if (isalarm(ctx)) { + if (flags & TFD_TIMER_ABSTIME) + alarm_start(&ctx->t.alarm, texp); + else + alarm_start_relative(&ctx->t.alarm, texp); + } else { + hrtimer_start(&ctx->t.tmr, texp, htmode); + } + if (timerfd_canceled(ctx)) return -ECANCELED; } @@ -158,7 +204,11 @@ static int timerfd_release(struct inode *inode, struct file *file) struct timerfd_ctx *ctx = file->private_data; timerfd_remove_cancel(ctx); - hrtimer_cancel(&ctx->tmr); + + if (isalarm(ctx)) + alarm_cancel(&ctx->t.alarm); + else + hrtimer_cancel(&ctx->t.tmr); kfree_rcu(ctx, rcu); return 0; } @@ -215,9 +265,15 @@ static ssize_t timerfd_read(struct file *file, char __user *buf, size_t count, * callback to avoid DoS attacks specifying a very * short timer period. */ - ticks += hrtimer_forward_now(&ctx->tmr, - ctx->tintv) - 1; - hrtimer_restart(&ctx->tmr); + if (isalarm(ctx)) { + ticks += alarm_forward_now( + &ctx->t.alarm, ctx->tintv) - 1; + alarm_restart(&ctx->t.alarm); + } else { + ticks += hrtimer_forward_now(&ctx->t.tmr, + ctx->tintv) - 1; + hrtimer_restart(&ctx->t.tmr); + } } ctx->expired = 0; ctx->ticks = 0; @@ -259,7 +315,9 @@ SYSCALL_DEFINE2(timerfd_create, int, clockid, int, flags) if ((flags & ~TFD_CREATE_FLAGS) || (clockid != CLOCK_MONOTONIC && - clockid != CLOCK_REALTIME)) + clockid != CLOCK_REALTIME && + clockid != CLOCK_REALTIME_ALARM && + clockid != CLOCK_BOOTTIME_ALARM)) return -EINVAL; ctx = kzalloc(sizeof(*ctx), GFP_KERNEL); @@ -268,7 +326,15 @@ SYSCALL_DEFINE2(timerfd_create, int, clockid, int, flags) init_waitqueue_head(&ctx->wqh); ctx->clockid = clockid; - hrtimer_init(&ctx->tmr, clockid, HRTIMER_MODE_ABS); + + if (isalarm(ctx)) + alarm_init(&ctx->t.alarm, + ctx->clockid == CLOCK_REALTIME_ALARM ? + ALARM_REALTIME : ALARM_BOOTTIME, + timerfd_alarmproc); + else + hrtimer_init(&ctx->t.tmr, clockid, HRTIMER_MODE_ABS); + ctx->moffs = ktime_get_monotonic_offset(); ufd = anon_inode_getfd("[timerfd]", &timerfd_fops, ctx, @@ -305,8 +371,14 @@ static int do_timerfd_settime(int ufd, int flags, */ for (;;) { spin_lock_irq(&ctx->wqh.lock); - if (hrtimer_try_to_cancel(&ctx->tmr) >= 0) - break; + + if (isalarm(ctx)) { + if (alarm_try_to_cancel(&ctx->t.alarm) >= 0) + break; + } else { + if (hrtimer_try_to_cancel(&ctx->t.tmr) >= 0) + break; + } spin_unlock_irq(&ctx->wqh.lock); cpu_relax(); } @@ -317,8 +389,12 @@ static int do_timerfd_settime(int ufd, int flags, * We do not update "ticks" and "expired" since the timer will be * re-programmed again in the following timerfd_setup() call. */ - if (ctx->expired && ctx->tintv.tv64) - hrtimer_forward_now(&ctx->tmr, ctx->tintv); + if (ctx->expired && ctx->tintv.tv64) { + if (isalarm(ctx)) + alarm_forward_now(&ctx->t.alarm, ctx->tintv); + else + hrtimer_forward_now(&ctx->t.tmr, ctx->tintv); + } old->it_value = ktime_to_timespec(timerfd_get_remaining(ctx)); old->it_interval = ktime_to_timespec(ctx->tintv); @@ -345,9 +421,18 @@ static int do_timerfd_gettime(int ufd, struct itimerspec *t) spin_lock_irq(&ctx->wqh.lock); if (ctx->expired && ctx->tintv.tv64) { ctx->expired = 0; - ctx->ticks += - hrtimer_forward_now(&ctx->tmr, ctx->tintv) - 1; - hrtimer_restart(&ctx->tmr); + + if (isalarm(ctx)) { + ctx->ticks += + alarm_forward_now( + &ctx->t.alarm, ctx->tintv) - 1; + alarm_restart(&ctx->t.alarm); + } else { + ctx->ticks += + hrtimer_forward_now(&ctx->t.tmr, ctx->tintv) + - 1; + hrtimer_restart(&ctx->t.tmr); + } } t->it_value = ktime_to_timespec(timerfd_get_remaining(ctx)); t->it_interval = ktime_to_timespec(ctx->tintv); From 5c83545f24ab3dd67e0ae0e2b795fea750f08c34 Mon Sep 17 00:00:00 2001 From: Colin Cross Date: Tue, 21 May 2013 22:32:14 -0700 Subject: [PATCH 184/198] power: Add option to log time spent in suspend Below is a patch from android kernel that maintains a histogram of suspend times. Please review and provide feedback. Statistices on the time spent in suspend are kept in /sys/kernel/debug/sleep_time. Cc: Android Kernel Team Cc: Colin Cross Cc: Todd Poynor Cc: San Mehat Cc: Benoit Goby Cc: John Stultz Cc: Thomas Gleixner Signed-off-by: Colin Cross Signed-off-by: Todd Poynor [zoran.markovic@linaro.org: Re-formatted suspend time table to better fit expected values. Moved accounting of suspend time into timekeeping core. Removed CONFIG_SUSPEND_TIME flag and made the feature conditional on CONFIG_DEBUG_FS. Changed the file name to sleep_time to better fit terminology in timekeeping core. Changed seq_printf to seq_puts. Tweaked commit message] Signed-off-by: Zoran Markovic Signed-off-by: John Stultz --- kernel/time/Makefile | 1 + kernel/time/timekeeping.c | 2 + kernel/time/timekeeping_debug.c | 72 ++++++++++++++++++++++++++++++ kernel/time/timekeeping_internal.h | 14 ++++++ 4 files changed, 89 insertions(+) create mode 100644 kernel/time/timekeeping_debug.c create mode 100644 kernel/time/timekeeping_internal.h diff --git a/kernel/time/Makefile b/kernel/time/Makefile index ff7d9d2ab504..d52ac8bf0006 100644 --- a/kernel/time/Makefile +++ b/kernel/time/Makefile @@ -7,3 +7,4 @@ obj-$(CONFIG_GENERIC_CLOCKEVENTS_BROADCAST) += tick-broadcast.o obj-$(CONFIG_TICK_ONESHOT) += tick-oneshot.o obj-$(CONFIG_TICK_ONESHOT) += tick-sched.o obj-$(CONFIG_TIMER_STATS) += timer_stats.o +obj-$(CONFIG_DEBUG_FS) += timekeeping_debug.o diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c index 933efa4071c3..838fc0777b68 100644 --- a/kernel/time/timekeeping.c +++ b/kernel/time/timekeeping.c @@ -25,6 +25,7 @@ #include "tick-internal.h" #include "ntp_internal.h" +#include "timekeeping_internal.h" static struct timekeeper timekeeper; static DEFINE_RAW_SPINLOCK(timekeeper_lock); @@ -851,6 +852,7 @@ static void __timekeeping_inject_sleeptime(struct timekeeper *tk, tk_xtime_add(tk, delta); tk_set_wall_to_mono(tk, timespec_sub(tk->wall_to_monotonic, *delta)); tk_set_sleep_time(tk, timespec_add(tk->total_sleep_time, *delta)); + tk_debug_account_sleep_time(delta); } /** diff --git a/kernel/time/timekeeping_debug.c b/kernel/time/timekeeping_debug.c new file mode 100644 index 000000000000..802433a4f5eb --- /dev/null +++ b/kernel/time/timekeeping_debug.c @@ -0,0 +1,72 @@ +/* + * debugfs file to track time spent in suspend + * + * Copyright (c) 2011, Google, Inc. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + */ + +#include +#include +#include +#include +#include +#include + +static unsigned int sleep_time_bin[32] = {0}; + +static int tk_debug_show_sleep_time(struct seq_file *s, void *data) +{ + unsigned int bin; + seq_puts(s, " time (secs) count\n"); + seq_puts(s, "------------------------------\n"); + for (bin = 0; bin < 32; bin++) { + if (sleep_time_bin[bin] == 0) + continue; + seq_printf(s, "%10u - %-10u %4u\n", + bin ? 1 << (bin - 1) : 0, 1 << bin, + sleep_time_bin[bin]); + } + return 0; +} + +static int tk_debug_sleep_time_open(struct inode *inode, struct file *file) +{ + return single_open(file, tk_debug_show_sleep_time, NULL); +} + +static const struct file_operations tk_debug_sleep_time_fops = { + .open = tk_debug_sleep_time_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +static int __init tk_debug_sleep_time_init(void) +{ + struct dentry *d; + + d = debugfs_create_file("sleep_time", 0444, NULL, NULL, + &tk_debug_sleep_time_fops); + if (!d) { + pr_err("Failed to create sleep_time debug file\n"); + return -ENOMEM; + } + + return 0; +} +late_initcall(tk_debug_sleep_time_init); + +void tk_debug_account_sleep_time(struct timespec *t) +{ + sleep_time_bin[fls(t->tv_sec)]++; +} + diff --git a/kernel/time/timekeeping_internal.h b/kernel/time/timekeeping_internal.h new file mode 100644 index 000000000000..13323ea08ffa --- /dev/null +++ b/kernel/time/timekeeping_internal.h @@ -0,0 +1,14 @@ +#ifndef _TIMEKEEPING_INTERNAL_H +#define _TIMEKEEPING_INTERNAL_H +/* + * timekeeping debug functions + */ +#include + +#ifdef CONFIG_DEBUG_FS +extern void tk_debug_account_sleep_time(struct timespec *t); +#else +#define tk_debug_account_sleep_time(x) +#endif + +#endif /* _TIMEKEEPING_INTERNAL_H */ From ce0b098981544d9f4e910ea48f3af3e726ca6d6b Mon Sep 17 00:00:00 2001 From: John Stultz Date: Tue, 28 May 2013 17:03:09 -0700 Subject: [PATCH 185/198] x86: Fix vrtc_get_time/set_mmss to use new timespec interface The patch "x86: Increase precision of x86_platform.get/set_wallclock" changed the x86 platform set_wallclock/get_wallclock interfaces to use nsec granular timespecs instead of a second granular interface. However, that patch missed converting the vrtc code, so this patch converts those functions to use timespecs. Many thanks to the kbuild test robot for finding this! Reported-by: kbuild test robot Signed-off-by: John Stultz --- arch/x86/include/asm/mrst-vrtc.h | 4 ++-- arch/x86/platform/mrst/vrtc.c | 11 ++++++----- 2 files changed, 8 insertions(+), 7 deletions(-) diff --git a/arch/x86/include/asm/mrst-vrtc.h b/arch/x86/include/asm/mrst-vrtc.h index 73668abdbedf..1e69a75412a4 100644 --- a/arch/x86/include/asm/mrst-vrtc.h +++ b/arch/x86/include/asm/mrst-vrtc.h @@ -3,7 +3,7 @@ extern unsigned char vrtc_cmos_read(unsigned char reg); extern void vrtc_cmos_write(unsigned char val, unsigned char reg); -extern unsigned long vrtc_get_time(void); -extern int vrtc_set_mmss(unsigned long nowtime); +extern void vrtc_get_time(struct timespec *now); +extern int vrtc_set_mmss(const struct timespec *now); #endif diff --git a/arch/x86/platform/mrst/vrtc.c b/arch/x86/platform/mrst/vrtc.c index d62b0a3b5c14..5e355b134ba4 100644 --- a/arch/x86/platform/mrst/vrtc.c +++ b/arch/x86/platform/mrst/vrtc.c @@ -56,7 +56,7 @@ void vrtc_cmos_write(unsigned char val, unsigned char reg) } EXPORT_SYMBOL_GPL(vrtc_cmos_write); -unsigned long vrtc_get_time(void) +void vrtc_get_time(struct timespec *now) { u8 sec, min, hour, mday, mon; unsigned long flags; @@ -82,17 +82,18 @@ unsigned long vrtc_get_time(void) printk(KERN_INFO "vRTC: sec: %d min: %d hour: %d day: %d " "mon: %d year: %d\n", sec, min, hour, mday, mon, year); - return mktime(year, mon, mday, hour, min, sec); + now->tv_sec = mktime(year, mon, mday, hour, min, sec); + now->tv_nsec = 0; } -int vrtc_set_mmss(unsigned long nowtime) +int vrtc_set_mmss(const struct timespec *now) { unsigned long flags; struct rtc_time tm; int year; int retval = 0; - rtc_time_to_tm(nowtime, &tm); + rtc_time_to_tm(now->tv_sec, &tm); if (!rtc_valid_tm(&tm) && tm.tm_year >= 72) { /* * tm.year is the number of years since 1900, and the @@ -110,7 +111,7 @@ int vrtc_set_mmss(unsigned long nowtime) } else { printk(KERN_ERR "%s: Invalid vRTC value: write of %lx to vRTC failed\n", - __FUNCTION__, nowtime); + __FUNCTION__, now->tv_sec); retval = -EINVAL; } return retval; From 1a33bd2be705cbb3f57d7223b60baea441039307 Mon Sep 17 00:00:00 2001 From: Baruch Siach Date: Wed, 29 May 2013 10:11:17 +0200 Subject: [PATCH 186/198] clocksource: dw_apb: Fix error check irq_of_parse_and_map() returns 0 on error, while the code checks for NO_IRQ. This breaks on platforms that have NO_IRQ != 0. Cc: Signed-off-by: Baruch Siach Signed-off-by: Daniel Lezcano --- drivers/clocksource/dw_apb_timer_of.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/clocksource/dw_apb_timer_of.c b/drivers/clocksource/dw_apb_timer_of.c index ab09ed3742ee..6b02eddc5f50 100644 --- a/drivers/clocksource/dw_apb_timer_of.c +++ b/drivers/clocksource/dw_apb_timer_of.c @@ -44,7 +44,7 @@ static void add_clockevent(struct device_node *event_timer) u32 irq, rate; irq = irq_of_parse_and_map(event_timer, 0); - if (irq == NO_IRQ) + if (irq == 0) panic("No IRQ for clock event timer"); timer_get_base_and_rate(event_timer, &iobase, &rate); From c19672492d233e0012b60fbfa460ffac1381ee26 Mon Sep 17 00:00:00 2001 From: Jingchang Lu Date: Wed, 29 May 2013 10:12:17 +0200 Subject: [PATCH 187/198] clocksource: Add Freescale Vybrid pit timer support Add Freescale Vybrid Family period interrupt timer support. Signed-off-by: Jingchang Lu Reviewed-by: Daniel Lezcano Acked-by: Shawn Guo Signed-off-by: Daniel Lezcano --- drivers/clocksource/Kconfig | 5 + drivers/clocksource/Makefile | 1 + drivers/clocksource/vf_pit_timer.c | 194 +++++++++++++++++++++++++++++ 3 files changed, 200 insertions(+) create mode 100644 drivers/clocksource/vf_pit_timer.c diff --git a/drivers/clocksource/Kconfig b/drivers/clocksource/Kconfig index f151c6cf27c3..0a04257edf65 100644 --- a/drivers/clocksource/Kconfig +++ b/drivers/clocksource/Kconfig @@ -85,3 +85,8 @@ config CLKSRC_SAMSUNG_PWM Samsung S3C, S5P and Exynos SoCs, replacing an earlier driver for all devicetree enabled platforms. This driver will be needed only on systems that do not have the Exynos MCT available. + +config VF_PIT_TIMER + bool + help + Support for Period Interrupt Timer on Freescale Vybrid Family SoCs. diff --git a/drivers/clocksource/Makefile b/drivers/clocksource/Makefile index 8d979c72aa94..36a9ac105d55 100644 --- a/drivers/clocksource/Makefile +++ b/drivers/clocksource/Makefile @@ -26,6 +26,7 @@ obj-$(CONFIG_ARCH_BCM) += bcm_kona_timer.o obj-$(CONFIG_CADENCE_TTC_TIMER) += cadence_ttc_timer.o obj-$(CONFIG_CLKSRC_EXYNOS_MCT) += exynos_mct.o obj-$(CONFIG_CLKSRC_SAMSUNG_PWM) += samsung_pwm_timer.o +obj-$(CONFIG_VF_PIT_TIMER) += vf_pit_timer.o obj-$(CONFIG_ARM_ARCH_TIMER) += arm_arch_timer.o obj-$(CONFIG_CLKSRC_METAG_GENERIC) += metag_generic.o diff --git a/drivers/clocksource/vf_pit_timer.c b/drivers/clocksource/vf_pit_timer.c new file mode 100644 index 000000000000..598399d57fc5 --- /dev/null +++ b/drivers/clocksource/vf_pit_timer.c @@ -0,0 +1,194 @@ +/* + * Copyright 2012-2013 Freescale Semiconductor, Inc. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version 2 + * of the License, or (at your option) any later version. + */ + +#include +#include +#include +#include +#include +#include + +/* + * Each pit takes 0x10 Bytes register space + */ +#define PITMCR 0x00 +#define PIT0_OFFSET 0x100 +#define PITn_OFFSET(n) (PIT0_OFFSET + 0x10 * (n)) +#define PITLDVAL 0x00 +#define PITCVAL 0x04 +#define PITTCTRL 0x08 +#define PITTFLG 0x0c + +#define PITMCR_MDIS (0x1 << 1) + +#define PITTCTRL_TEN (0x1 << 0) +#define PITTCTRL_TIE (0x1 << 1) +#define PITCTRL_CHN (0x1 << 2) + +#define PITTFLG_TIF 0x1 + +static void __iomem *clksrc_base; +static void __iomem *clkevt_base; +static unsigned long cycle_per_jiffy; + +static inline void pit_timer_enable(void) +{ + __raw_writel(PITTCTRL_TEN | PITTCTRL_TIE, clkevt_base + PITTCTRL); +} + +static inline void pit_timer_disable(void) +{ + __raw_writel(0, clkevt_base + PITTCTRL); +} + +static inline void pit_irq_acknowledge(void) +{ + __raw_writel(PITTFLG_TIF, clkevt_base + PITTFLG); +} + +static unsigned int pit_read_sched_clock(void) +{ + return __raw_readl(clksrc_base + PITCVAL); +} + +static int __init pit_clocksource_init(unsigned long rate) +{ + /* set the max load value and start the clock source counter */ + __raw_writel(0, clksrc_base + PITTCTRL); + __raw_writel(~0UL, clksrc_base + PITLDVAL); + __raw_writel(PITTCTRL_TEN, clksrc_base + PITTCTRL); + + setup_sched_clock(pit_read_sched_clock, 32, rate); + return clocksource_mmio_init(clksrc_base + PITCVAL, "vf-pit", rate, + 300, 32, clocksource_mmio_readl_down); +} + +static int pit_set_next_event(unsigned long delta, + struct clock_event_device *unused) +{ + /* + * set a new value to PITLDVAL register will not restart the timer, + * to abort the current cycle and start a timer period with the new + * value, the timer must be disabled and enabled again. + * and the PITLAVAL should be set to delta minus one according to pit + * hardware requirement. + */ + pit_timer_disable(); + __raw_writel(delta - 1, clkevt_base + PITLDVAL); + pit_timer_enable(); + + return 0; +} + +static void pit_set_mode(enum clock_event_mode mode, + struct clock_event_device *evt) +{ + switch (mode) { + case CLOCK_EVT_MODE_PERIODIC: + pit_set_next_event(cycle_per_jiffy, evt); + break; + default: + break; + } +} + +static irqreturn_t pit_timer_interrupt(int irq, void *dev_id) +{ + struct clock_event_device *evt = dev_id; + + pit_irq_acknowledge(); + + /* + * pit hardware doesn't support oneshot, it will generate an interrupt + * and reload the counter value from PITLDVAL when PITCVAL reach zero, + * and start the counter again. So software need to disable the timer + * to stop the counter loop in ONESHOT mode. + */ + if (likely(evt->mode == CLOCK_EVT_MODE_ONESHOT)) + pit_timer_disable(); + + evt->event_handler(evt); + + return IRQ_HANDLED; +} + +static struct clock_event_device clockevent_pit = { + .name = "VF pit timer", + .features = CLOCK_EVT_FEAT_PERIODIC | CLOCK_EVT_FEAT_ONESHOT, + .set_mode = pit_set_mode, + .set_next_event = pit_set_next_event, + .rating = 300, +}; + +static struct irqaction pit_timer_irq = { + .name = "VF pit timer", + .flags = IRQF_TIMER | IRQF_IRQPOLL, + .handler = pit_timer_interrupt, + .dev_id = &clockevent_pit, +}; + +static int __init pit_clockevent_init(unsigned long rate, int irq) +{ + __raw_writel(0, clkevt_base + PITTCTRL); + __raw_writel(PITTFLG_TIF, clkevt_base + PITTFLG); + + BUG_ON(setup_irq(irq, &pit_timer_irq)); + + clockevent_pit.cpumask = cpumask_of(0); + clockevent_pit.irq = irq; + /* + * The value for the LDVAL register trigger is calculated as: + * LDVAL trigger = (period / clock period) - 1 + * The pit is a 32-bit down count timer, when the conter value + * reaches 0, it will generate an interrupt, thus the minimal + * LDVAL trigger value is 1. And then the min_delta is + * minimal LDVAL trigger value + 1, and the max_delta is full 32-bit. + */ + clockevents_config_and_register(&clockevent_pit, rate, 2, 0xffffffff); + + return 0; +} + +static void __init pit_timer_init(struct device_node *np) +{ + struct clk *pit_clk; + void __iomem *timer_base; + unsigned long clk_rate; + int irq; + + timer_base = of_iomap(np, 0); + BUG_ON(!timer_base); + + /* + * PIT0 and PIT1 can be chained to build a 64-bit timer, + * so choose PIT2 as clocksource, PIT3 as clockevent device, + * and leave PIT0 and PIT1 unused for anyone else who needs them. + */ + clksrc_base = timer_base + PITn_OFFSET(2); + clkevt_base = timer_base + PITn_OFFSET(3); + + irq = irq_of_parse_and_map(np, 0); + BUG_ON(irq <= 0); + + pit_clk = of_clk_get(np, 0); + BUG_ON(IS_ERR(pit_clk)); + + BUG_ON(clk_prepare_enable(pit_clk)); + + clk_rate = clk_get_rate(pit_clk); + cycle_per_jiffy = clk_rate / (HZ); + + /* enable the pit module */ + __raw_writel(~PITMCR_MDIS, timer_base + PITMCR); + + BUG_ON(pit_clocksource_init(clk_rate)); + + pit_clockevent_init(clk_rate, irq); +} +CLOCKSOURCE_OF_DECLARE(vf610, "fsl,vf610-pit", pit_timer_init); From 77ba83bb1bb1cdabd522d32536f8eee65a870145 Mon Sep 17 00:00:00 2001 From: Daniel Tang Date: Sat, 1 Jun 2013 16:02:37 +1000 Subject: [PATCH 188/198] clocksource: Add TI-Nspire timer support This patch adds a clocksource/clockevent driver for the timer found on some models in the TI-Nspire calculator series. The timer has two 16bit subtimers within its memory mapped I/O interface but only the first can generate interrupts. The first subtimer is used to generate clockevents but only if an interrupt number and register is given. The interrupt acknowledgement mechanism is a little strange because the interrupt mask and acknowledge registers are located in another memory mapped I/O peripheral. The address of this register is passed to the driver through device tree bindings. The second subtimer is used as a clocksource because it isn't capable of generating an interrupt. This subtimer is always added. Reviewed-by: Linus Walleij Signed-off-by: Daniel Tang Signed-off-by: Daniel Lezcano --- .../bindings/timer/lsi,zevio-timer.txt | 33 +++ drivers/clocksource/Makefile | 1 + drivers/clocksource/zevio-timer.c | 215 ++++++++++++++++++ 3 files changed, 249 insertions(+) create mode 100644 Documentation/devicetree/bindings/timer/lsi,zevio-timer.txt create mode 100644 drivers/clocksource/zevio-timer.c diff --git a/Documentation/devicetree/bindings/timer/lsi,zevio-timer.txt b/Documentation/devicetree/bindings/timer/lsi,zevio-timer.txt new file mode 100644 index 000000000000..b2d07ad90e9a --- /dev/null +++ b/Documentation/devicetree/bindings/timer/lsi,zevio-timer.txt @@ -0,0 +1,33 @@ +TI-NSPIRE timer + +Required properties: + +- compatible : should be "lsi,zevio-timer". +- reg : The physical base address and size of the timer (always first). +- clocks: phandle to the source clock. + +Optional properties: + +- interrupts : The interrupt number of the first timer. +- reg : The interrupt acknowledgement registers + (always after timer base address) + +If any of the optional properties are not given, the timer is added as a +clock-source only. + +Example: + +timer { + compatible = "lsi,zevio-timer"; + reg = <0x900D0000 0x1000>, <0x900A0020 0x8>; + interrupts = <19>; + clocks = <&timer_clk>; +}; + +Example (no clock-events): + +timer { + compatible = "lsi,zevio-timer"; + reg = <0x900D0000 0x1000>; + clocks = <&timer_clk>; +}; diff --git a/drivers/clocksource/Makefile b/drivers/clocksource/Makefile index 36a9ac105d55..4853ea0f8fd5 100644 --- a/drivers/clocksource/Makefile +++ b/drivers/clocksource/Makefile @@ -22,6 +22,7 @@ obj-$(CONFIG_ARCH_PRIMA2) += timer-prima2.o obj-$(CONFIG_SUN4I_TIMER) += sun4i_timer.o obj-$(CONFIG_ARCH_TEGRA) += tegra20_timer.o obj-$(CONFIG_VT8500_TIMER) += vt8500_timer.o +obj-$(CONFIG_ARCH_NSPIRE) += zevio-timer.o obj-$(CONFIG_ARCH_BCM) += bcm_kona_timer.o obj-$(CONFIG_CADENCE_TTC_TIMER) += cadence_ttc_timer.o obj-$(CONFIG_CLKSRC_EXYNOS_MCT) += exynos_mct.o diff --git a/drivers/clocksource/zevio-timer.c b/drivers/clocksource/zevio-timer.c new file mode 100644 index 000000000000..ca81809d159d --- /dev/null +++ b/drivers/clocksource/zevio-timer.c @@ -0,0 +1,215 @@ +/* + * linux/drivers/clocksource/zevio-timer.c + * + * Copyright (C) 2013 Daniel Tang + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2, as + * published by the Free Software Foundation. + * + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#define IO_CURRENT_VAL 0x00 +#define IO_DIVIDER 0x04 +#define IO_CONTROL 0x08 + +#define IO_TIMER1 0x00 +#define IO_TIMER2 0x0C + +#define IO_MATCH_BEGIN 0x18 +#define IO_MATCH(x) (IO_MATCH_BEGIN + ((x) << 2)) + +#define IO_INTR_STS 0x00 +#define IO_INTR_ACK 0x00 +#define IO_INTR_MSK 0x04 + +#define CNTL_STOP_TIMER (1 << 4) +#define CNTL_RUN_TIMER (0 << 4) + +#define CNTL_INC (1 << 3) +#define CNTL_DEC (0 << 3) + +#define CNTL_TOZERO 0 +#define CNTL_MATCH(x) ((x) + 1) +#define CNTL_FOREVER 7 + +/* There are 6 match registers but we only use one. */ +#define TIMER_MATCH 0 + +#define TIMER_INTR_MSK (1 << (TIMER_MATCH)) +#define TIMER_INTR_ALL 0x3F + +struct zevio_timer { + void __iomem *base; + void __iomem *timer1, *timer2; + void __iomem *interrupt_regs; + + struct clk *clk; + struct clock_event_device clkevt; + struct irqaction clkevt_irq; + + char clocksource_name[64]; + char clockevent_name[64]; +}; + +static int zevio_timer_set_event(unsigned long delta, + struct clock_event_device *dev) +{ + struct zevio_timer *timer = container_of(dev, struct zevio_timer, + clkevt); + + writel(delta, timer->timer1 + IO_CURRENT_VAL); + writel(CNTL_RUN_TIMER | CNTL_DEC | CNTL_MATCH(TIMER_MATCH), + timer->timer1 + IO_CONTROL); + + return 0; +} + +static void zevio_timer_set_mode(enum clock_event_mode mode, + struct clock_event_device *dev) +{ + struct zevio_timer *timer = container_of(dev, struct zevio_timer, + clkevt); + + switch (mode) { + case CLOCK_EVT_MODE_RESUME: + case CLOCK_EVT_MODE_ONESHOT: + /* Enable timer interrupts */ + writel(TIMER_INTR_MSK, timer->interrupt_regs + IO_INTR_MSK); + writel(TIMER_INTR_ALL, timer->interrupt_regs + IO_INTR_ACK); + break; + case CLOCK_EVT_MODE_SHUTDOWN: + case CLOCK_EVT_MODE_UNUSED: + /* Disable timer interrupts */ + writel(0, timer->interrupt_regs + IO_INTR_MSK); + writel(TIMER_INTR_ALL, timer->interrupt_regs + IO_INTR_ACK); + /* Stop timer */ + writel(CNTL_STOP_TIMER, timer->timer1 + IO_CONTROL); + break; + case CLOCK_EVT_MODE_PERIODIC: + default: + /* Unsupported */ + break; + } +} + +static irqreturn_t zevio_timer_interrupt(int irq, void *dev_id) +{ + struct zevio_timer *timer = dev_id; + u32 intr; + + intr = readl(timer->interrupt_regs + IO_INTR_ACK); + if (!(intr & TIMER_INTR_MSK)) + return IRQ_NONE; + + writel(TIMER_INTR_MSK, timer->interrupt_regs + IO_INTR_ACK); + writel(CNTL_STOP_TIMER, timer->timer1 + IO_CONTROL); + + if (timer->clkevt.event_handler) + timer->clkevt.event_handler(&timer->clkevt); + + return IRQ_HANDLED; +} + +static int __init zevio_timer_add(struct device_node *node) +{ + struct zevio_timer *timer; + struct resource res; + int irqnr, ret; + + timer = kzalloc(sizeof(*timer), GFP_KERNEL); + if (!timer) + return -ENOMEM; + + timer->base = of_iomap(node, 0); + if (!timer->base) { + ret = -EINVAL; + goto error_free; + } + timer->timer1 = timer->base + IO_TIMER1; + timer->timer2 = timer->base + IO_TIMER2; + + timer->clk = of_clk_get(node, 0); + if (IS_ERR(timer->clk)) { + ret = PTR_ERR(timer->clk); + pr_err("Timer clock not found! (error %d)\n", ret); + goto error_unmap; + } + + timer->interrupt_regs = of_iomap(node, 1); + irqnr = irq_of_parse_and_map(node, 0); + + of_address_to_resource(node, 0, &res); + scnprintf(timer->clocksource_name, sizeof(timer->clocksource_name), + "%llx.%s_clocksource", + (unsigned long long)res.start, node->name); + + scnprintf(timer->clockevent_name, sizeof(timer->clockevent_name), + "%llx.%s_clockevent", + (unsigned long long)res.start, node->name); + + if (timer->interrupt_regs && irqnr) { + timer->clkevt.name = timer->clockevent_name; + timer->clkevt.set_next_event = zevio_timer_set_event; + timer->clkevt.set_mode = zevio_timer_set_mode; + timer->clkevt.rating = 200; + timer->clkevt.cpumask = cpu_all_mask; + timer->clkevt.features = CLOCK_EVT_FEAT_ONESHOT; + timer->clkevt.irq = irqnr; + + writel(CNTL_STOP_TIMER, timer->timer1 + IO_CONTROL); + writel(0, timer->timer1 + IO_DIVIDER); + + /* Start with timer interrupts disabled */ + writel(0, timer->interrupt_regs + IO_INTR_MSK); + writel(TIMER_INTR_ALL, timer->interrupt_regs + IO_INTR_ACK); + + /* Interrupt to occur when timer value matches 0 */ + writel(0, timer->base + IO_MATCH(TIMER_MATCH)); + + timer->clkevt_irq.name = timer->clockevent_name; + timer->clkevt_irq.handler = zevio_timer_interrupt; + timer->clkevt_irq.dev_id = timer; + timer->clkevt_irq.flags = IRQF_TIMER | IRQF_IRQPOLL; + + setup_irq(irqnr, &timer->clkevt_irq); + + clockevents_config_and_register(&timer->clkevt, + clk_get_rate(timer->clk), 0x0001, 0xffff); + pr_info("Added %s as clockevent\n", timer->clockevent_name); + } + + writel(CNTL_STOP_TIMER, timer->timer2 + IO_CONTROL); + writel(0, timer->timer2 + IO_CURRENT_VAL); + writel(0, timer->timer2 + IO_DIVIDER); + writel(CNTL_RUN_TIMER | CNTL_FOREVER | CNTL_INC, + timer->timer2 + IO_CONTROL); + + clocksource_mmio_init(timer->timer2 + IO_CURRENT_VAL, + timer->clocksource_name, + clk_get_rate(timer->clk), + 200, 16, + clocksource_mmio_readw_up); + + pr_info("Added %s as clocksource\n", timer->clocksource_name); + + return 0; +error_unmap: + iounmap(timer->base); +error_free: + kfree(timer); + return ret; +} + +CLOCKSOURCE_OF_DECLARE(zevio_timer, "lsi,zevio-timer", zevio_timer_add); From 11682a41618f8094cb7a9330b4b6a12ffaef5774 Mon Sep 17 00:00:00 2001 From: Marcus Gelderie Date: Tue, 4 Jun 2013 09:32:09 +0200 Subject: [PATCH 189/198] alarmtimer: Export symbols of functions declared in linux/alarmtimer.h Export symbols so they can be used by drivers/staging/android/alarm-dev.c if it is built as a module. So far alarm-dev is built-in but module support is planned (see drivers/staging/android/TODO). Signed-off-by: Marcus Gelderie [jstultz: tweaked commit message, also export newly added functions] Signed-off-by: John Stultz --- kernel/time/alarmtimer.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/kernel/time/alarmtimer.c b/kernel/time/alarmtimer.c index 3e5cba274475..eec50fcef9e4 100644 --- a/kernel/time/alarmtimer.c +++ b/kernel/time/alarmtimer.c @@ -204,6 +204,7 @@ ktime_t alarm_expires_remaining(const struct alarm *alarm) struct alarm_base *base = &alarm_bases[alarm->type]; return ktime_sub(alarm->node.expires, base->gettime()); } +EXPORT_SYMBOL_GPL(alarm_expires_remaining); #ifdef CONFIG_RTC_CLASS /** @@ -309,6 +310,7 @@ void alarm_init(struct alarm *alarm, enum alarmtimer_type type, alarm->type = type; alarm->state = ALARMTIMER_STATE_INACTIVE; } +EXPORT_SYMBOL_GPL(alarm_init); /** * alarm_start - Sets an absolute alarm to fire @@ -329,6 +331,7 @@ int alarm_start(struct alarm *alarm, ktime_t start) spin_unlock_irqrestore(&base->lock, flags); return ret; } +EXPORT_SYMBOL_GPL(alarm_start); /** * alarm_start_relative - Sets a relative alarm to fire @@ -342,6 +345,7 @@ int alarm_start_relative(struct alarm *alarm, ktime_t start) start = ktime_add(start, base->gettime()); return alarm_start(alarm, start); } +EXPORT_SYMBOL_GPL(alarm_start_relative); void alarm_restart(struct alarm *alarm) { @@ -354,6 +358,7 @@ void alarm_restart(struct alarm *alarm) alarmtimer_enqueue(base, alarm); spin_unlock_irqrestore(&base->lock, flags); } +EXPORT_SYMBOL_GPL(alarm_restart); /** * alarm_try_to_cancel - Tries to cancel an alarm timer @@ -375,6 +380,7 @@ int alarm_try_to_cancel(struct alarm *alarm) spin_unlock_irqrestore(&base->lock, flags); return ret; } +EXPORT_SYMBOL_GPL(alarm_try_to_cancel); /** @@ -392,6 +398,7 @@ int alarm_cancel(struct alarm *alarm) cpu_relax(); } } +EXPORT_SYMBOL_GPL(alarm_cancel); u64 alarm_forward(struct alarm *alarm, ktime_t now, ktime_t interval) @@ -424,6 +431,7 @@ u64 alarm_forward(struct alarm *alarm, ktime_t now, ktime_t interval) alarm->node.expires = ktime_add(alarm->node.expires, interval); return overrun; } +EXPORT_SYMBOL_GPL(alarm_forward); u64 alarm_forward_now(struct alarm *alarm, ktime_t interval) { @@ -431,7 +439,7 @@ u64 alarm_forward_now(struct alarm *alarm, ktime_t interval) return alarm_forward(alarm, base->gettime(), interval); } - +EXPORT_SYMBOL_GPL(alarm_forward_now); /** From 5a9b5855c248d0298eac4d5490c7bc11c2f1983b Mon Sep 17 00:00:00 2001 From: Stephen Boyd Date: Sat, 1 Jun 2013 23:39:38 -0700 Subject: [PATCH 190/198] ARM: sched_clock: Remove unused needs_suspend member The needs_suspend member is unused now that we always do the suspend/resume handling (see 6a4dae5 (ARM: 7565/1: sched: stop sched_clock() during suspend, 2012-10-23)). Acked-by: Will Deacon Signed-off-by: Stephen Boyd Signed-off-by: John Stultz --- arch/arm/kernel/sched_clock.c | 1 - 1 file changed, 1 deletion(-) diff --git a/arch/arm/kernel/sched_clock.c b/arch/arm/kernel/sched_clock.c index e8edcaa0e432..45efe86e8b08 100644 --- a/arch/arm/kernel/sched_clock.c +++ b/arch/arm/kernel/sched_clock.c @@ -24,7 +24,6 @@ struct clock_data { u32 mult; u32 shift; bool suspended; - bool needs_suspend; }; static void sched_clock_poll(unsigned long wrap_ticks); From ffbfb5e316f0db486798ccf1db36a577ffe79637 Mon Sep 17 00:00:00 2001 From: Stephen Boyd Date: Sat, 1 Jun 2013 23:39:39 -0700 Subject: [PATCH 191/198] ARM: sched_clock: Return suspended count earlier If we're suspended and sched_clock() is called we're going to read the hardware one more time and throw away that value and return back the cached value we saved during the suspend callback. This is wasteful. Let's short circuit all that and return the cached value as early as possible if we're suspended. Acked-by: Will Deacon Signed-off-by: Stephen Boyd Signed-off-by: John Stultz --- arch/arm/kernel/sched_clock.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/arch/arm/kernel/sched_clock.c b/arch/arm/kernel/sched_clock.c index 45efe86e8b08..a781c59b93c0 100644 --- a/arch/arm/kernel/sched_clock.c +++ b/arch/arm/kernel/sched_clock.c @@ -55,9 +55,6 @@ static unsigned long long notrace cyc_to_sched_clock(u32 cyc, u32 mask) u64 epoch_ns; u32 epoch_cyc; - if (cd.suspended) - return cd.epoch_ns; - /* * Load the epoch_cyc and epoch_ns atomically. We do this by * ensuring that we always write epoch_cyc, epoch_ns and @@ -174,6 +171,9 @@ unsigned long long __read_mostly (*sched_clock_func)(void) = sched_clock_32; unsigned long long notrace sched_clock(void) { + if (cd.suspended) + return cd.epoch_ns; + return sched_clock_func(); } From 38ff87f77af0b5a93fc8581cff1d6e5692ab8970 Mon Sep 17 00:00:00 2001 From: Stephen Boyd Date: Sat, 1 Jun 2013 23:39:40 -0700 Subject: [PATCH 192/198] sched_clock: Make ARM's sched_clock generic for all architectures Nothing about the sched_clock implementation in the ARM port is specific to the architecture. Generalize the code so that other architectures can use it by selecting GENERIC_SCHED_CLOCK. Signed-off-by: Stephen Boyd [jstultz: Merge minor collisions with other patches in my tree] Signed-off-by: John Stultz --- arch/arm/Kconfig | 1 + arch/arm/common/timer-sp.c | 2 +- arch/arm/kernel/Makefile | 2 +- arch/arm/kernel/arch_timer.c | 2 +- arch/arm/kernel/time.c | 4 +--- arch/arm/mach-davinci/time.c | 2 +- arch/arm/mach-imx/time.c | 2 +- arch/arm/mach-integrator/integrator_ap.c | 2 +- arch/arm/mach-ixp4xx/common.c | 2 +- arch/arm/mach-mmp/time.c | 2 +- arch/arm/mach-msm/timer.c | 2 +- arch/arm/mach-omap1/time.c | 2 +- arch/arm/mach-omap2/timer.c | 2 +- arch/arm/mach-pxa/time.c | 2 +- arch/arm/mach-sa1100/time.c | 2 +- arch/arm/mach-u300/timer.c | 2 +- arch/arm/plat-iop/time.c | 2 +- arch/arm/plat-omap/counter_32k.c | 2 +- arch/arm/plat-orion/time.c | 2 +- arch/arm/plat-samsung/samsung-time.c | 2 +- arch/arm/plat-versatile/sched-clock.c | 2 +- drivers/clocksource/bcm2835_timer.c | 2 +- drivers/clocksource/clksrc-dbx500-prcmu.c | 3 +-- drivers/clocksource/dw_apb_timer_of.c | 3 +-- drivers/clocksource/mxs_timer.c | 2 +- drivers/clocksource/nomadik-mtu.c | 2 +- drivers/clocksource/samsung_pwm_timer.c | 2 +- drivers/clocksource/tegra20_timer.c | 2 +- drivers/clocksource/time-armada-370-xp.c | 2 +- drivers/clocksource/timer-marco.c | 2 +- drivers/clocksource/timer-prima2.c | 2 +- {arch/arm/include/asm => include/linux}/sched_clock.h | 9 +++++++-- init/Kconfig | 3 +++ init/main.c | 2 ++ kernel/time/Makefile | 1 + {arch/arm/kernel => kernel/time}/sched_clock.c | 3 +-- 36 files changed, 45 insertions(+), 38 deletions(-) rename {arch/arm/include/asm => include/linux}/sched_clock.h (75%) rename {arch/arm/kernel => kernel/time}/sched_clock.c (99%) diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index 49d993cee512..53d3a356f61f 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -14,6 +14,7 @@ config ARM select GENERIC_IRQ_PROBE select GENERIC_IRQ_SHOW select GENERIC_PCI_IOMAP + select GENERIC_SCHED_CLOCK select GENERIC_SMP_IDLE_THREAD select GENERIC_IDLE_POLL_SETUP select GENERIC_STRNCPY_FROM_USER diff --git a/arch/arm/common/timer-sp.c b/arch/arm/common/timer-sp.c index ddc740769601..023ee63827a2 100644 --- a/arch/arm/common/timer-sp.c +++ b/arch/arm/common/timer-sp.c @@ -28,8 +28,8 @@ #include #include #include +#include -#include #include #include diff --git a/arch/arm/kernel/Makefile b/arch/arm/kernel/Makefile index 5f3338eacad2..97cb0576d07c 100644 --- a/arch/arm/kernel/Makefile +++ b/arch/arm/kernel/Makefile @@ -16,7 +16,7 @@ CFLAGS_REMOVE_return_address.o = -pg # Object file lists. obj-y := elf.o entry-armv.o entry-common.o irq.o opcodes.o \ - process.o ptrace.o return_address.o sched_clock.o \ + process.o ptrace.o return_address.o \ setup.o signal.o stacktrace.o sys_arm.o time.o traps.o obj-$(CONFIG_ATAGS) += atags_parse.o diff --git a/arch/arm/kernel/arch_timer.c b/arch/arm/kernel/arch_timer.c index 59dcdced6e30..221f07b11ccb 100644 --- a/arch/arm/kernel/arch_timer.c +++ b/arch/arm/kernel/arch_timer.c @@ -11,9 +11,9 @@ #include #include #include +#include #include -#include #include diff --git a/arch/arm/kernel/time.c b/arch/arm/kernel/time.c index abff4e9aaee0..98aee3258398 100644 --- a/arch/arm/kernel/time.c +++ b/arch/arm/kernel/time.c @@ -24,9 +24,9 @@ #include #include #include +#include #include -#include #include #include #include @@ -120,6 +120,4 @@ void __init time_init(void) machine_desc->init_time(); else clocksource_of_init(); - - sched_clock_postinit(); } diff --git a/arch/arm/mach-davinci/time.c b/arch/arm/mach-davinci/time.c index bad361ec1666..7a55b5c95971 100644 --- a/arch/arm/mach-davinci/time.c +++ b/arch/arm/mach-davinci/time.c @@ -18,8 +18,8 @@ #include #include #include +#include -#include #include #include diff --git a/arch/arm/mach-imx/time.c b/arch/arm/mach-imx/time.c index fea91313678b..cd46529e9eaa 100644 --- a/arch/arm/mach-imx/time.c +++ b/arch/arm/mach-imx/time.c @@ -26,8 +26,8 @@ #include #include #include +#include -#include #include #include "common.h" diff --git a/arch/arm/mach-integrator/integrator_ap.c b/arch/arm/mach-integrator/integrator_ap.c index b23c8e4f28e8..aa4346227c41 100644 --- a/arch/arm/mach-integrator/integrator_ap.c +++ b/arch/arm/mach-integrator/integrator_ap.c @@ -41,6 +41,7 @@ #include #include #include +#include #include