The 'running' variable is only used in the P4 PMU. Current perf sets the
variable in the critical function x86_pmu_start(), which wastes cycles
for everybody not running on P4.
Move cpuc->running into the P4 specific p4_pmu_enable_event().
Add a static per-CPU 'p4_running' variable to replace the 'running'
variable in the struct cpu_hw_events. Saves space for the generic
structure.
The p4_pmu_enable_all() also invokes the p4_pmu_enable_event(), but it
should not set cpuc->running. Factor out __p4_pmu_enable_event() for
p4_pmu_enable_all().
Suggested-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/1618410990-21383-1-git-send-email-kan.liang@linux.intel.com
Add a kselftest for testing process-wide perf events with synchronous
SIGTRAP on events (using breakpoints). In particular, we want to test
that changes to the event propagate to all children, and the SIGTRAPs
are in fact synchronously sent to the thread where the event occurred.
Note: The "signal_stress" test case is also added later in the series to
perf tool's built-in tests. The test here is more elaborate in that
respect, which on one hand avoids bloating the perf tool unnecessarily,
but we also benefit from structured tests with TAP-compliant output that
the kselftest framework provides.
Signed-off-by: Marco Elver <elver@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20210408103605.1676875-8-elver@google.com
Adds bit perf_event_attr::sigtrap, which can be set to cause events to
send SIGTRAP (with si_code TRAP_PERF) to the task where the event
occurred. The primary motivation is to support synchronous signals on
perf events in the task where an event (such as breakpoints) triggered.
To distinguish perf events based on the event type, the type is set in
si_errno. For events that are associated with an address, si_addr is
copied from perf_sample_data.
The new field perf_event_attr::sig_data is copied to si_perf, which
allows user space to disambiguate which event (of the same type)
triggered the signal. For example, user space could encode the relevant
information it cares about in sig_data.
We note that the choice of an opaque u64 provides the simplest and most
flexible option. Alternatives where a reference to some user space data
is passed back suffer from the problem that modification of referenced
data (be it the event fd, or the perf_event_attr) can race with the
signal being delivered (of course, the same caveat applies if user space
decides to store a pointer in sig_data, but the ABI explicitly avoids
prescribing such a design).
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Marco Elver <elver@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Dmitry Vyukov <dvyukov@google.com>
Link: https://lore.kernel.org/lkml/YBv3rAT566k+6zjg@hirez.programming.kicks-ass.net/
Introduces the TRAP_PERF si_code, and associated siginfo_t field
si_perf. These will be used by the perf event subsystem to send signals
(if requested) to the task where an event occurred.
Signed-off-by: Marco Elver <elver@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Geert Uytterhoeven <geert@linux-m68k.org> # m68k
Acked-by: Arnd Bergmann <arnd@arndb.de> # asm-generic
Link: https://lkml.kernel.org/r/20210408103605.1676875-6-elver@google.com
Adds bit perf_event_attr::remove_on_exec, to support removing an event
from a task on exec.
This option supports the case where an event is supposed to be
process-wide only, and should not propagate beyond exec, to limit
monitoring to the original process image only.
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Marco Elver <elver@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20210408103605.1676875-5-elver@google.com
Adds bit perf_event_attr::inherit_thread, to restricting inheriting
events only if the child was cloned with CLONE_THREAD.
This option supports the case where an event is supposed to be
process-wide only (including subthreads), but should not propagate
beyond the current process's shared environment.
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Marco Elver <elver@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/lkml/YBvj6eJR%2FDY2TsEB@hirez.programming.kicks-ass.net/
As with other ioctls (such as PERF_EVENT_IOC_{ENABLE,DISABLE}), fix up
handling of PERF_EVENT_IOC_MODIFY_ATTRIBUTES to also apply to children.
Suggested-by: Dmitry Vyukov <dvyukov@google.com>
Signed-off-by: Marco Elver <elver@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Dmitry Vyukov <dvyukov@google.com>
Link: https://lkml.kernel.org/r/20210408103605.1676875-3-elver@google.com
Make perf_event_exit_event() more robust, such that we can use it from
other contexts. Specifically the up and coming remove_on_exec.
For this to work we need to address a few issues. Remove_on_exec will
not destroy the entire context, so we cannot rely on TASK_TOMBSTONE to
disable event_function_call() and we thus have to use
perf_remove_from_context().
When using perf_remove_from_context(), there's two races to consider.
The first is against close(), where we can have concurrent tear-down
of the event. The second is against child_list iteration, which should
not find a half baked event.
To address this, teach perf_remove_from_context() to special case
!ctx->is_active and about DETACH_CHILD.
[ elver@google.com: fix racing parent/child exit in sync_child_event(). ]
Signed-off-by: Marco Elver <elver@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20210408103605.1676875-2-elver@google.com
Turns out, the default setting of attr.aux_watermark to half of the total
buffer size is not very useful, especially with smaller buffers. The
problem is that, after half of the buffer is filled up, the kernel updates
->aux_head and sets up the next "transaction", while observing that
->aux_tail is still zero (as userspace haven't had the chance to update
it), meaning that the trace will have to stop at the end of this second
"transaction". This means, for example, that the second PERF_RECORD_AUX in
every trace comes with TRUNCATED flag set.
Setting attr.aux_watermark to quarter of the buffer gives enough space for
the ->aux_tail update to be observed and prevents the data loss.
The obligatory before/after showcase:
> # perf_before record -e intel_pt//u -m,8 uname
> Linux
> [ perf record: Woken up 6 times to write data ]
> Warning:
> AUX data lost 4 times out of 10!
>
> [ perf record: Captured and wrote 0.099 MB perf.data ]
> # perf record -e intel_pt//u -m,8 uname
> Linux
> [ perf record: Woken up 4 times to write data ]
> [ perf record: Captured and wrote 0.039 MB perf.data ]
The effect is still visible with large workloads and large buffers,
although less pronounced.
Signed-off-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20210414154955.49603-3-alexander.shishkin@linux.intel.com
Currently, we start allocating AUX pages half the size of the total
requested AUX buffer size, ignoring the attr.aux_watermark setting. This,
in turn, makes intel_pt driver disregard the watermark also, as it uses
page order for its SG (ToPA) configuration.
Now, this can be fixed in the intel_pt PMU driver, but seeing as it's the
only one currently making use of high order allocations, there is no
reason not to fix the allocator instead. This way, any other driver
wishing to add this support would not have to worry about this.
Signed-off-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20210414154955.49603-2-alexander.shishkin@linux.intel.com
When snd-hda-codec-hdmi is used with ASoC HDA controller like SOF (acomp
used for ELD notifications), display connection change done during suspend,
can be lost due to following sequence of events:
1. system in S3 suspend
2. DP/HDMI receiver connected
3. system resumed
4. HDA controller resumed, but card->deferred_resume_work not complete
5. acomp eld_notify callback
6. eld_notify ignored as power state is not CTL_POWER_D0
7. HDA resume deferred work completed, power state set to CTL_POWER_D0
This results in losing the notification, and the jack state reported to
user-space is not correct.
The check on step 6 was added in commit 8ae743e82f ("ALSA: hda - Skip
ELD notification during system suspend"). It would seem with the deferred
resume logic in ASoC core, this check is not safe.
Fix the issue by modifying the check to use "dev.power.power_state.event"
instead of ALSA specific card power state variable.
BugLink: https://github.com/thesofproject/linux/issues/2825
Suggested-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20210416131157.1881366-1-kai.vehmanen@linux.intel.com
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The pci_bus->bridge reference may no longer be valid after
pci_bus_remove() resulting in passing a bad value to device_unregister()
for the associated bridge device.
Store the host_bridge reference in a separate variable prior to
pci_bus_remove().
Fixes: 7340056567 ("powerpc/pci: Reorder pci bus/bridge unregistration during PHB removal")
Signed-off-by: Tyrel Datwyler <tyreld@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20210211182435.47968-1-tyreld@linux.ibm.com
Now that __fake_sleep is static, we get a warning about it being unused
in some configurations:
drivers/macintosh/via-pmu.c:190:12: warning: '__fake_sleep' defined but not used
190 | static int __fake_sleep;
Move it inside the ifdef where it's used to avoid the warning.
Fixes: 95d1439233 ("macintosh/via-pmu: Make some symbols static")
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20210416114139.772236-1-mpe@ellerman.id.au
soc_pcm_params_symmetry() checks CPU / Codec symmetry.
Unfortunately there was bug on it (= A) which didn't check Codec.
But is back by (B).
A: v5.7: commit c840f7698d ("ASoC: soc-pcm: Merge for_each_rtd_cpu/codec_dais()")
B: v5.12: commit 3a90672111 ("ASoC: soc-pcm: cleanup soc_pcm_params_symmetry()")
In total,
old - v5.6 (= Generation-1):
symmetric_rate : DAI_Link / CPU / Codec
symmetric_channels : DAI_Link / CPU / Codec
symmetric_sample_bits : DAI_Link / CPU / Codec
v5.7 - v5.11 (= Generation-2): (= because of bug by (A))
symmetric_rate : DAI_Link / CPU
symmetric_channels : DAI_Link / CPU / Codec
symmetric_sample_bits : DAI_Link / CPU / Codec
v5.12 - (= Generation-3): (= back by (B))
symmetric_rate : DAI_Link / CPU / Codec
symmetric_channels : DAI_Link / CPU / Codec
symmetric_sample_bits : DAI_Link / CPU / Codec
OTOH, we can use DPCM which is configured by FE / BE.
Both FE / BE uses dummy-DAI.
FE: CPU <-> dummy-DAI
BE: dummy-DAI <-> Codec
One note is that we can use .be_hw_params_fixup in DPCM case.
This means BE settings might be fixuped/updated by FE.
This feature is used for example on MIXer case.
It can be happen not only for rate, but for channels/sample_bits too.
Because of these reasons, below issue happen on
Generation-1 / Generation-3, if...
1) Sound Card used DPCM
2) It exchanges rate to 48kHz by using .be_hw_params_fixup()
3) Codec had symmetric_rate = 1
I didn't confirm, but maybe same things happen
if it exchanged channels/sample_bits at Generation-1/2/3 too.
# aplay 44100.wav
# aplay 44100.wav
=> [kernel] be.ak4613-hifi: ASoC: unmatched rate symmetry: snd-soc-dummy-dai:44100 - soc_pcm_params_symmetry:48000
[kernel] be.ak4613-hifi: ASoC: hw_params BE failed -22
[kernel] fe.rsnd-dai.0: ASoC: hw_params BE failed -22
aplay: set_params:1407: Unable to install hw params:
ACCESS: RW_INTERLEAVED
FORMAT: S16_LE
SUBFORMAT: STD
SAMPLE_BITS: 16
FRAME_BITS: 32
CHANNELS: 2
RATE: 44100
PERIOD_TIME: (23219 23220)
PERIOD_SIZE: 1024
PERIOD_BYTES: 4096
PERIODS: 4
BUFFER_TIME: (92879 92880)
BUFFER_SIZE: 4096
BUFFER_BYTES: 16384
TICK_TIME: 0
soc_pcm_params_symmetry() checks by below
if (symmetry)
for_each_rtd_cpu_dais(rtd, i, cpu_dai)
if (cpu_dai->xxx && cpu_dai->xxx != d.xxx) {
dev_err(rtd->dev, "...");
return -EINVAL;
}
Because of above reason 3) (= Codec had symmetric_rate = 1)
BE can't ignore "if (symmetric)".
At 1st aplay, soc_pcm_params_symmetry() ignores it,
because dummy-DAI->rate is 0.
After this check, each DAI sets/keep settings.
In above sample case, BE gets 48000 and FE gets 44100,
and it happen BE -> FE order.
Because DPCM is sharing *same* dummy-DAI,
dummy-DAI sets as 48000 by BE, and is overwrote by 44100 by FE.
This settings never be cleaned (= a) after 1st aplay,
because dummy-DAI is used from FE/BE, never be last user (b).
static int soc_pcm_hw_clean(...)
{
...
for_each_rtd_dais(rtd, i, dai) {
...
(b) if (snd_soc_dai_active(dai) == 1)
(a) soc_pcm_set_dai_params(dai, NULL);
...
}
...
}
At 2nd aplay, BE gets 48000 but dummy-DAI is keeping 44100,
soc_pcm_params_symmetry() checks will fail.
To solve this issue, this patch ignores dummy-DAI
at soc_pcm_params_symmetry()
Link: https://lore.kernel.org/r/87a6q0z4xt.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87y2djxa2n.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
It indicates unmatched symmetry value, but not indicates on which DAI.
This patch indicates it.
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/871rbbyono.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
__soc_pcm_params_symmetry() macro is using "name" as parameter
which will be exchanged to rate/channles/sample_bit, like below
dai->name => dai->rate
dai->name => dai->channels
dai->name => dai->sample_bit
But, dai itself has "name". This means
1) It is very confusable naming
2) It can't use dai->name
This patch use "xxx" instead of "name"
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/8735vryoob.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
.set_ios() is called from .resume() as well. For SDIO device which sets
keep-power-in-suspend, nothing should be changed after resuming, as well
as sample tuning value, since this value is tuned already. So we should
not overwrite it with the default value.
Signed-off-by: Shawn Lin <shawn.lin@rock-chips.com>
Link: https://lore.kernel.org/r/1618539454-182170-1-git-send-email-shawn.lin@rock-chips.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
If the pNFS layout segment is marked with the NFS_LSEG_LAYOUTRETURN
flag, then the assumption is that it has some reporting requirement
to perform through a layoutreturn (e.g. flexfiles layout stats or error
information).
Fixes: e0b7d420f7 ("pNFS: Don't discard layout segments that are marked for return")
Cc: stable@vger.kernel.org
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
If the NFS super block is being unmounted, then we currently may end up
telling the server that we've forgotten the layout while it is actually
still in use by the client.
In that case, just assume that the client will soon return the layout
anyway, and so return NFS4ERR_DELAY in response to the layout recall.
Fixes: 58ac3e5923 ("NFSv4/pnfs: Clean up nfs_layout_find_inode()")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
The spi controller supports 44-bit address space on AXI in DMA mode,
so set dma_addr_t width to 44-bit to avoid using a swiotlb mapping.
In addition, if dma_map_single fails, it should return immediately
instead of continuing doing the DMA operation which bases on invalid
address.
This fixes the following crash which occurs in reading a big block
from flash:
[ 123.633577] zynqmp-qspi ff0f0000.spi: swiotlb buffer is full (sz: 4194304 bytes), total 32768 (slots), used 0 (slots)
[ 123.644230] zynqmp-qspi ff0f0000.spi: ERR:rxdma:memory not mapped
[ 123.784625] Unable to handle kernel paging request at virtual address 00000000003fffc0
[ 123.792536] Mem abort info:
[ 123.795313] ESR = 0x96000145
[ 123.798351] EC = 0x25: DABT (current EL), IL = 32 bits
[ 123.803655] SET = 0, FnV = 0
[ 123.806693] EA = 0, S1PTW = 0
[ 123.809818] Data abort info:
[ 123.812683] ISV = 0, ISS = 0x00000145
[ 123.816503] CM = 1, WnR = 1
[ 123.819455] user pgtable: 4k pages, 48-bit VAs, pgdp=0000000805047000
[ 123.825887] [00000000003fffc0] pgd=0000000803b45003, p4d=0000000803b45003, pud=0000000000000000
[ 123.834586] Internal error: Oops: 96000145 [#1] PREEMPT SMP
Fixes: 1c26372e5a ("spi: spi-zynqmp-gqspi: Update driver to use spi-mem framework")
Signed-off-by: Quanyang Wang <quanyang.wang@windriver.com>
Link: https://lore.kernel.org/r/20210416004652.2975446-6-quanyang.wang@windriver.com
Signed-off-by: Mark Brown <broonie@kernel.org>
When handling op->addr, it is using the buffer "tmpbuf" which has been
freed. This will trigger a use-after-free KASAN warning. Let's use
temporary variables to store op->addr.val and op->cmd.opcode to fix
this issue.
Fixes: 1c26372e5a ("spi: spi-zynqmp-gqspi: Update driver to use spi-mem framework")
Signed-off-by: Quanyang Wang <quanyang.wang@windriver.com>
Link: https://lore.kernel.org/r/20210416004652.2975446-5-quanyang.wang@windriver.com
Signed-off-by: Mark Brown <broonie@kernel.org>
During a transfer the driver filled the fifo with 4bytes,
even if the data that needs to be transfer is less that 4bytes.
This resulted in slab-out-of-bounds bug in KernelAddressSanitizer.
This patch resolves slab-out-of-bounds bug by filling the fifo
with the number of bytes that needs to transferred.
Signed-off-by: Amit Kumar Mahapatra <amit.kumar-mahapatra@xilinx.com>
Link: https://lore.kernel.org/r/20210416004652.2975446-4-quanyang.wang@windriver.com
Signed-off-by: Mark Brown <broonie@kernel.org>
After calling platform_set_drvdata(pdev, xqspi) in probe, the return
value of dev_get_drvdata(dev) is a pointer to struct zynqmp_qspi but
not struct spi_controller. A wrong structure type passing to the
functions spi_controller_suspend/resume will hang the system.
And we should check the return value of spi_controller_suspend, if
an error is returned, return it to PM subsystem to stop suspend.
Also, GQSPI_EN_MASK should be written to GQSPI_EN_OFST to enable
the spi controller in zynqmp_qspi_resume since it was disabled in
zynqmp_qspi_suspend before.
Fixes: 1c26372e5a ("spi: spi-zynqmp-gqspi: Update driver to use spi-mem framework")
Signed-off-by: Quanyang Wang <quanyang.wang@windriver.com>
Link: https://lore.kernel.org/r/20210416004652.2975446-3-quanyang.wang@windriver.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The clks "pclk" and "ref_clk" are enabled twice during the probe. The
first time is in the function zynqmp_qspi_probe and the second time is
in zynqmp_qspi_setup_op which is called by devm_spi_register_controller.
Then calling zynqmp_qspi_remove (rmmod this module) to disable these clks
will trigger a warning as below:
[ 309.124604] Unpreparing enabled qspi_ref
[ 309.128641] WARNING: CPU: 1 PID: 537 at drivers/clk/clk.c:824 clk_core_unprepare+0x108/0x110
Since pm_runtime works now, clks can be enabled/disabled by calling
zynqmp_runtime_suspend/resume. So we don't need to enable these clks
explicitly in zynqmp_qspi_setup_op. Remove them to fix this issue.
And remove clk enabling/disabling in zynqmp_qspi_resume because there is
no spi transfer operation so enabling ref_clk is redundant meanwhile pclk
is not disabled for it is shared with other peripherals.
Furthermore replace clk_enable/disable with clk_prepare_enable and
clk_disable_unprepare in runtime_suspend/resume functions.
Fixes: 1c26372e5a ("spi: spi-zynqmp-gqspi: Update driver to use spi-mem framework")
Signed-off-by: Quanyang Wang <quanyang.wang@windriver.com>
Link: https://lore.kernel.org/r/20210416004652.2975446-2-quanyang.wang@windriver.com
Signed-off-by: Mark Brown <broonie@kernel.org>
A DAI link has 3 components:
* CPU
* platform
* codec(s)
A component is specified via:
* name
* of_node
* dai_name
In order to avoid confusion when building a sound card we disallow
matching by both name and of_node (1).
soc_check_tplg_fes allows overriding certain BE links by overriding
BE link name. This doesn't work well if BE link was specified via DT,
because we end up with a link with both name and of_node specified
which is conflicting with (1).
In order to fix this we need to:
* override of_node if component was specified via DT
* override name, otherwise.
Signed-off-by: Daniel Baluta <daniel.baluta@nxp.com>
Link: https://lore.kernel.org/r/20210414101212.65573-1-daniel.baluta@oss.nxp.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Add quirks for jack detection, rt711 DAI and DMIC
Reviewed-by: Bard Liao <bard.liao@intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Signed-off-by: Vamshi Krishna Gopal <vamshi.krishna.gopal@intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20210415175013.192862-6-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Add one configuration with no RT711.
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Reviewed-by: Libin Yang <libin.yang@intel.com>
Link: https://lore.kernel.org/r/20210415175013.192862-3-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Add configurations ported over from TGL.
The topology names need to include link information given all the
hardware permutations.
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Reviewed-by: Libin Yang <libin.yang@intel.com>
Link: https://lore.kernel.org/r/20210415175013.192862-2-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Since commit 01e99aeca3 'blk-mq: insert passthrough request into
hctx->dispatch directly', passthrough request should not appear in
IO-scheduler any more, so blk_rq_is_passthrough checking in addon IO
schedulers is redundant.
(Notes: this patch passes generic IO load test with hdds under SAS
controller and hdds under AHCI controller but obviously not covers all.
Not sure if passthrough request can still escape into IO scheduler from
blk_mq_sched_insert_requests, which is used by blk_mq_flush_plug_list and
has lots of indirect callers.)
Signed-off-by: Lin Feng <linf@wangsu.com>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Commit 01e99aeca3 ("blk-mq: insert passthrough request into
hctx->dispatch directly") gives high priority to passthrough requests and
bypass underlying IO scheduler. But as we allocate tag for such request it
still runs io-scheduler's callback limit_depth, while we really want is to
give full sbitmap-depth capabity to such request for acquiring available
tag.
blktrace shows PC requests(dmraid -s -c -i) hit bfq's limit_depth:
8,0 2 0 0.000000000 39952 1,0 m N bfq [bfq_limit_depth] wr_busy 0 sync 0 depth 8
8,0 2 1 0.000008134 39952 D R 4 [dmraid]
8,0 2 2 0.000021538 24 C R [0]
8,0 2 0 0.000035442 39952 1,0 m N bfq [bfq_limit_depth] wr_busy 0 sync 0 depth 8
8,0 2 3 0.000038813 39952 D R 24 [dmraid]
8,0 2 4 0.000044356 24 C R [0]
This patch introduce a new wrapper to make code not that ugly.
Signed-off-by: Lin Feng <linf@wangsu.com>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/20210415033920.213963-1-linf@wangsu.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Use SPDX-License-Identifier: GPL-2.0-only, instead of hand writing it.
This also removes a reference to http://www.xyratex.com which seems to be
down.
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
'hpre_cfg_by_dsm' has checked and printed error path internally. It is not
necessary to do it here, so remove it.
It should return error immediately when return value of 'hpre_cfg_by_dsm'
is non-zero, and no need to execute the remaining sentences.
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
When the received sqe is abnormal, the error message in the sqe written
back by the hardware is printed to help to analyze the abnormal causes.
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The return value of 'le32_to_cpu' is unsigned, so change the
variable type from 'int' to 'unsigned int'.
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
There are two spaces after return, just keep one.
Signed-off-by: Hui Tang <tanghui20@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Debian's clang carries a patch that makes the default FPU mode
'vfp3-d16' instead of 'neon' for 'armv7-a' to avoid generating NEON
instructions on hardware that does not support them:
5a61ca6f21/debian/patches/clang-arm-default-vfp3-on-armv7a.patchhttps://bugs.debian.org/841474https://bugs.debian.org/842142https://bugs.debian.org/914268
This results in the following build error when clang's integrated
assembler is used because the '.arch' directive overrides the '.fpu'
directive:
arch/arm/crypto/curve25519-core.S:25:2: error: instruction requires: NEON
vmov.i32 q0, #1
^
arch/arm/crypto/curve25519-core.S:26:2: error: instruction requires: NEON
vshr.u64 q1, q0, #7
^
arch/arm/crypto/curve25519-core.S:27:2: error: instruction requires: NEON
vshr.u64 q0, q0, #8
^
arch/arm/crypto/curve25519-core.S:28:2: error: instruction requires: NEON
vmov.i32 d4, #19
^
Shuffle the order of the '.arch' and '.fpu' directives so that the code
builds regardless of the default FPU mode. This has been tested against
both clang with and without Debian's patch and GCC.
Cc: stable@vger.kernel.org
Fixes: d8f1308a02 ("crypto: arm/curve25519 - wire up NEON implementation")
Link: https://github.com/ClangBuiltLinux/continuous-integration2/issues/118
Reported-by: Arnd Bergmann <arnd@arndb.de>
Suggested-by: Arnd Bergmann <arnd@arndb.de>
Suggested-by: Jessica Clarke <jrtc27@jrtc27.com>
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Acked-by: Jason A. Donenfeld <Jason@zx2c4.com>
Reviewed-by: Nick Desaulniers <ndesaulniers@google.com>
Tested-by: Nick Desaulniers <ndesaulniers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Enable the detection of hangs by setting watchdog timers (WDTs) on
generations that supports that feature.
The default timeout value comes from HW specs. WTDs are reset each time
an accelerator wins arbitration and is able to send/read a command to/from
an accelerator.
The value has added significant margin to make sure there are no spurious
timeouts. The scope of watchdog is per QAT device.
If a timeout is detected, the firmware resets the accelerator and
returns a response descriptor with an appropriate error code.
Signed-off-by: Wojciech Ziemba <wojciech.ziemba@intel.com>
Reviewed-by: Giovanni Cabiddu <giovanni.cabiddu@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The buffer of the hardware sge needs to be initialized by
soft sgl.
Signed-off-by: Kai Ye <yekai13@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This seems "32" and "31" is obfuscating, It might be better to add a comment,
which explain it.
Signed-off-by: Kai Ye <yekai13@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
According to Documentation/core-api/printk-formats.rst, Use
the correct print format. Printing an unsigned int value should use %u
instead of %d.
Signed-off-by: Kai Ye <yekai13@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
According to Documentation/core-api/printk-formats.rst, Use
the correct print format. Printing an unsigned int value should use %u
instead of %d.
Signed-off-by: Kai Ye <yekai13@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
pm_runtime_get_sync will increment pm usage counter even it failed.
Forgetting to putting operation will result in reference leak here.
Fix it by replacing it with pm_runtime_resume_and_get to keep usage
counter balanced.
Signed-off-by: Shixin Liu <liushixin2@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
pm_runtime_get_sync will increment pm usage counter even it failed.
Forgetting to putting operation will result in reference leak here.
Fix it by replacing it with pm_runtime_resume_and_get to keep usage
counter balanced.
Signed-off-by: Shixin Liu <liushixin2@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
pm_runtime_get_sync will increment pm usage counter even it failed.
Forgetting to putting operation will result in reference leak here.
Fix it by replacing it with pm_runtime_resume_and_get to keep usage
counter balanced.
Signed-off-by: Shixin Liu <liushixin2@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
