component has io_mutex, but it had been used at
snd_soc_component_update_bits_legacy() only which does read and write.
static int snd_soc_component_update_bits_legacy(...)
{
...
=> mutex_lock(&component->io_mutex);
...
old = snd_soc_component_read(...);
...
ret = snd_soc_component_write(...);
...
=> mutex_unlock(&component->io_mutex);
...
}
It is pointless if it is not used with both read and write functions.
This patch uses io_mutex correctly with read/write.
Here, xxx_no_lock() is local functions.
static int snd_soc_component_read(...)
{
...
=> mutex_lock(&component->io_mutex);
val = soc_component_read_no_lock(...);
=> mutex_unlock(&component->io_mutex);
...
}
static int snd_soc_component_write(...)
{
...
=> mutex_lock(&component->io_mutex);
ret = soc_component_write_no_lock(...);
=> mutex_unlock(&component->io_mutex);
...
}
static int snd_soc_component_update_bits_legacy(...)
{
...
=> mutex_lock(&component->io_mutex);
...
old = soc_component_read_no_lock(...);
...
ret = soc_component_write_no_lock(...);
...
=> mutex_unlock(&component->io_mutex);
...
}
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87r1uf4mfa.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
We had read/write function for Codec, Platform, etc,
but these has been merged into snd_soc_component_read/write().
Internally, it is using regmap or driver function.
In read case, each styles are like below
regmap
ret = regmap_read(..., reg, &val);
driver function
val = xxx->read(..., reg);
Because of this kind of different style, to keep same read style,
when we merged each read function into snd_soc_component_read(),
we created snd_soc_component_read32(), like below.
commit 738b49efe6 ("ASoC: add snd_soc_component_read32")
(1) val = snd_soc_component_read32(component, reg);
(2) ret = snd_soc_component_read(component, reg, &val);
Many drivers are using snd_soc_component_read32(), and
some drivers are using snd_soc_component_read() today.
In generally, we don't check read function successes,
because, we will have many other issues at initial timing
if read function didn't work.
Now we can use soc_component_err() when error case.
This means, it is easy to notice if error occurred.
This patch aggressively merge snd_soc_component_read() and _read32(),
and makes snd_soc_component_read/write() as generally style.
This patch do
1) merge snd_soc_component_read() and snd_soc_component_read32()
2) it uses soc_component_err() when error case (easy to notice)
3) keeps read32 for now by #define
4) update snd_soc_component_read() for all drivers
Because _read() user drivers are not too many, this patch changes
all user drivers.
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/87sgev4mfl.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
When the extended advertisement feature is enabled, a hardcoded min and
max interval of 0x8000 is used. This patch fixes this issue by using
the configured min/max value.
This was validated by setting min/max in main.conf and making sure the
right setting is applied:
< HCI Command: LE Set Extended Advertising Parameters (0x08|0x0036) plen
25 #93 [hci0] 10.953011
…
Min advertising interval: 181.250 msec (0x0122)
Max advertising interval: 181.250 msec (0x0122)
…
Signed-off-by: Alain Michaud <alainm@chromium.org>
Reviewed-by: Abhishek Pandit-Subedi <abhishekpandit@chromium.org>
Reviewed-by: Daniel Winkler <danielwinkler@google.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
If SVE is enabled then 'ret' can be assigned the return value of
kvm_vcpu_enable_sve() which may be 0 causing future "goto out" sites to
erroneously return 0 on failure rather than -EINVAL as expected.
Remove the initialisation of 'ret' and make setting the return value
explicit to avoid this situation in the future.
Fixes: 9a3cdf26e3 ("KVM: arm64/sve: Allow userspace to enable SVE for vcpus")
Cc: stable@vger.kernel.org
Reported-by: James Morse <james.morse@arm.com>
Signed-off-by: Steven Price <steven.price@arm.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20200617105456.28245-1-steven.price@arm.com
The "inline" keyword is a hint for the compiler to inline a function. The
functions system_uses_irq_prio_masking() and gic_write_pmr() are used by
the code running at EL2 on a non-VHE system, so mark them as
__always_inline to make sure they'll always be part of the .hyp.text
section.
This fixes the following splat when trying to run a VM:
[ 47.625273] Kernel panic - not syncing: HYP panic:
[ 47.625273] PS:a00003c9 PC:0000ca0b42049fc4 ESR:86000006
[ 47.625273] FAR:0000ca0b42049fc4 HPFAR:0000000010001000 PAR:0000000000000000
[ 47.625273] VCPU:0000000000000000
[ 47.647261] CPU: 1 PID: 217 Comm: kvm-vcpu-0 Not tainted 5.8.0-rc1-ARCH+ #61
[ 47.654508] Hardware name: Globalscale Marvell ESPRESSOBin Board (DT)
[ 47.661139] Call trace:
[ 47.663659] dump_backtrace+0x0/0x1cc
[ 47.667413] show_stack+0x18/0x24
[ 47.670822] dump_stack+0xb8/0x108
[ 47.674312] panic+0x124/0x2f4
[ 47.677446] panic+0x0/0x2f4
[ 47.680407] SMP: stopping secondary CPUs
[ 47.684439] Kernel Offset: disabled
[ 47.688018] CPU features: 0x240402,20002008
[ 47.692318] Memory Limit: none
[ 47.695465] ---[ end Kernel panic - not syncing: HYP panic:
[ 47.695465] PS:a00003c9 PC:0000ca0b42049fc4 ESR:86000006
[ 47.695465] FAR:0000ca0b42049fc4 HPFAR:0000000010001000 PAR:0000000000000000
[ 47.695465] VCPU:0000000000000000 ]---
The instruction abort was caused by the code running at EL2 trying to fetch
an instruction which wasn't mapped in the EL2 translation tables. Using
objdump showed the two functions as separate symbols in the .text section.
Fixes: 85738e05dc ("arm64: kvm: Unmask PMR before entering guest")
Cc: stable@vger.kernel.org
Signed-off-by: Alexandru Elisei <alexandru.elisei@arm.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Acked-by: James Morse <james.morse@arm.com>
Link: https://lore.kernel.org/r/20200618171254.1596055-1-alexandru.elisei@arm.com
The RPC client currently doesn't handle ERR_CHUNK replies correctly.
rpcrdma_complete_rqst() incorrectly passes a negative number to
xprt_complete_rqst() as the number of bytes copied. Instead, set
task->tk_status to the error value, and return zero bytes copied.
In these cases, return -EIO rather than -EREMOTEIO. The RPC client's
finite state machine doesn't know what to do with -EREMOTEIO.
Additional clean ups:
- Don't double-count RDMA_ERROR replies
- Remove a stale comment
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Cc: <stable@kernel.vger.org>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
1. Ensure that only rpcrdma_cm_event_handler() modifies
ep->re_connect_status to avoid racy changes to that field.
2. Ensure that xprt_force_disconnect() is invoked only once as a
transport is closed or destroyed.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Clean up: Sometimes creating a fresh rpcrdma_ep can fail. That's why
xprt_rdma_connect() always checks if the r_xprt->rx_ep pointer is
valid before dereferencing it. Instead, xprt_rdma_connect() can
simply check rpcrdma_xprt_connect()'s return value.
Also, there's no need to set re_connect_status to zero just after
the rpcrdma_ep is created, since it is allocated with kzalloc.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
r_xprt->rx_ep is known to be good while the transport's send lock is
held. Otherwise additional references on rx_ep must be held when it
is used outside of that lock's critical sections.
For now, bump the rx_ep reference count once whenever there is at
least one outstanding Receive WR. This avoids the memory bandwidth
overhead of taking and releasing the reference count for every
ib_post_recv() and Receive completion.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
If shared interrupt comes late, during probe error path or device remove
(could be triggered with CONFIG_DEBUG_SHIRQ), the interrupt handler
dspi_interrupt() will access registers with the clock being disabled.
This leads to external abort on non-linefetch on Toradex Colibri VF50
module (with Vybrid VF5xx):
$ echo 4002d000.spi > /sys/devices/platform/soc/40000000.bus/4002d000.spi/driver/unbind
Unhandled fault: external abort on non-linefetch (0x1008) at 0x8887f02c
Internal error: : 1008 [#1] ARM
Hardware name: Freescale Vybrid VF5xx/VF6xx (Device Tree)
Backtrace:
(regmap_mmio_read32le)
(regmap_mmio_read)
(_regmap_bus_reg_read)
(_regmap_read)
(regmap_read)
(dspi_interrupt)
(free_irq)
(devm_irq_release)
(release_nodes)
(devres_release_all)
(device_release_driver_internal)
The resource-managed framework should not be used for shared interrupt
handling, because the interrupt handler might be called after releasing
other resources and disabling clocks.
Similar bug could happen during suspend - the shared interrupt handler
could be invoked after suspending the device. Each device sharing this
interrupt line should disable the IRQ during suspend so handler will be
invoked only in following cases:
1. None suspended,
2. All devices resumed.
Fixes: 349ad66c0a ("spi:Add Freescale DSPI driver for Vybrid VF610 platform")
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20200622110543.5035-3-krzk@kernel.org
Signed-off-by: Mark Brown <broonie@kernel.org>
During device removal, the driver should unregister the SPI controller
and stop the hardware. Otherwise the dspi_transfer_one_message() could
wait on completion infinitely.
Additionally, calling spi_unregister_controller() first in device
removal reverse-matches the probe function, where SPI controller is
registered at the end.
Fixes: 05209f4570 ("spi: fsl-dspi: add missing clk_disable_unprepare() in dspi_remove()")
Reported-by: Vladimir Oltean <olteanv@gmail.com>
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20200622110543.5035-1-krzk@kernel.org
Signed-off-by: Mark Brown <broonie@kernel.org>
This separate helper only existed to guarantee the mutual exclusivity of
CLONE_PIDFD and CLONE_PARENT_SETTID for legacy clone since CLONE_PIDFD
abuses the parent_tid field to return the pidfd. But we can actually handle
this uniformely thus removing the helper. For legacy clone we can detect
that CLONE_PIDFD is specified in conjunction with CLONE_PARENT_SETTID
because they will share the same memory which is invalid and for clone3()
setting the separate pidfd and parent_tid fields to the same memory is
bogus as well. So fold that helper directly into _do_fork() by detecting
this case.
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: "Peter Zijlstra (Intel)" <peterz@infradead.org>
Cc: linux-m68k@lists.linux-m68k.org
Cc: x86@kernel.org
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
qcom_swrm_port_ops and qcom_swrm_ops are not modified and can be made
const to allow the compiler to put them in read-only memory.
Before:
text data bss dec hex filename
18266 3056 256 21578 544a drivers/soundwire/qcom.o
After:
text data bss dec hex filename
18426 2896 256 21578 544a drivers/soundwire/qcom.o
Signed-off-by: Rikard Falkeborn <rikard.falkeborn@gmail.com>
Link: https://lore.kernel.org/r/20200609230029.69802-1-rikard.falkeborn@gmail.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
With CONFIG_DEBUG_VIRTUAL=y, __pa() checks for addr value and if it's
less than PAGE_OFFSET it leads to a BUG().
#define __pa(x)
({
VIRTUAL_BUG_ON((unsigned long)(x) < PAGE_OFFSET);
(unsigned long)(x) & 0x0fffffffffffffffUL;
})
kernel BUG at arch/powerpc/kvm/book3s_64_mmu_radix.c:43!
cpu 0x70: Vector: 700 (Program Check) at [c0000018a2187360]
pc: c000000000161b30: __kvmhv_copy_tofrom_guest_radix+0x130/0x1f0
lr: c000000000161d5c: kvmhv_copy_from_guest_radix+0x3c/0x80
...
kvmhv_copy_from_guest_radix+0x3c/0x80
kvmhv_load_from_eaddr+0x48/0xc0
kvmppc_ld+0x98/0x1e0
kvmppc_load_last_inst+0x50/0x90
kvmppc_hv_emulate_mmio+0x288/0x2b0
kvmppc_book3s_radix_page_fault+0xd8/0x2b0
kvmppc_book3s_hv_page_fault+0x37c/0x1050
kvmppc_vcpu_run_hv+0xbb8/0x1080
kvmppc_vcpu_run+0x34/0x50
kvm_arch_vcpu_ioctl_run+0x2fc/0x410
kvm_vcpu_ioctl+0x2b4/0x8f0
ksys_ioctl+0xf4/0x150
sys_ioctl+0x28/0x80
system_call_exception+0x104/0x1d0
system_call_common+0xe8/0x214
kvmhv_copy_tofrom_guest_radix() uses a NULL value for to/from to
indicate direction of copy.
Avoid calling __pa() if the value is NULL to avoid the BUG().
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
[mpe: Massage change log a bit to mention CONFIG_DEBUG_VIRTUAL]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20200611120159.680284-1-aneesh.kumar@linux.ibm.com
The debugfs_create_dir() function never returns NULL and anyway the
correct behavior is to ignore errors in this situation. The
debugfs_create_file() will become a no-op if "emc->debugfs.root" is an
error pointer.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
The suspend/resume functions have no callers depending on
configuration, so they must be marked __maybe_unused to
avoid these harmless warnings:
drivers/memory/tegra/tegra186.c:1578:12: error: 'tegra186_mc_resume' defined but not used [-Werror=unused-function]
1578 | static int tegra186_mc_resume(struct device *dev)
| ^~~~~~~~~~~~~~~~~~
drivers/memory/tegra/tegra186.c:1573:12: error: 'tegra186_mc_suspend' defined but not used [-Werror=unused-function]
1573 | static int tegra186_mc_suspend(struct device *dev)
| ^~~~~~~~~~~~~~~~~~~
Fixes: 177602b006 ("memory: tegra: Add system sleep support")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
The Tegra EMC scaling support code is not a clock provider, but merely a
clock consumer, and thus does not need to include
<linux/clk-provider.h>.
Fixes: 0bf368c5b2cf ("memory: tegra: Add EMC scaling support code for Tegra210")
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Thierry Reding <treding@nvidia.com>
The call to tegra_bpmp_get() must be balanced by a call to
tegra_bpmp_put() in case of error, as already done in the remove
function.
Add an error handling path and corresponding goto.
Fixes: 52d15dd23f ("memory: tegra: Support DVFS on Tegra186 and later")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Thierry Reding <treding@nvidia.com>
The memory clock-rate change could be running on a non-boot CPU, while the
boot CPU handles the EMC interrupt. This introduces an unnecessary latency
since boot CPU should handle the interrupt and then notify the sibling CPU
about clock-rate change completion. In some rare cases boot CPU could be
in uninterruptible state for a significant time (like in a case of KASAN +
NFS root), it could get to the point that completion timeouts before boot
CPU gets a chance to handle interrupt. The solution is to get rid of the
completion and replace it with interrupt-status polling.
Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
The memory clock-rate change could be running on a non-boot CPU, while the
boot CPU handles the EMC interrupt. This introduces an unnecessary latency
since boot CPU should handle the interrupt and then notify the sibling CPU
about clock-rate change completion. In some rare cases boot CPU could be
in uninterruptible state for a significant time (like in a case of KASAN +
NFS root), it could get to the point that completion timeouts before boot
CPU gets a chance to handle interrupt. The solution is to get rid of the
completion and replace it with interrupt-status polling.
Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
Derated timings are used to ensure that the memory chips keep operating
correctly at high temperatures. This adds code to support polling of the
chip operating state when high temperatures are measured on the chip and
change the refresh mode accordingly. Under very high temperatures, the
driver will switch to the derated tables to ensure proper operation of
the memory chips.
Signed-off-by: Thierry Reding <treding@nvidia.com>
This patch includes the sequence for clock tuning and the dynamic
training mechanism for the clock above 800MHz.
And historically there have been different sequences to change the EMC
clock. The sequence to be used is specified in the EMC table.
However, for the currently supported upstreaming platform, only the most
recent sequence is used. So only support that in this patch.
Based on the work of Peter De Schrijver <pdeschrijver@nvidia.com>.
Signed-off-by: Joseph Lo <josephl@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
This is the initial patch for Tegra210 EMC frequency scaling. It has the
code to program various aspects of the EMC that are standardized, but it
does not yet include the specific programming sequence needed for clock
scaling.
The driver is designed to support LPDDR4 SDRAM. Devices that use LPDDR4
need to perform training of the RAM before it can be used. Firmware will
perform this training during early boot and pass a table of supported
frequencies to the kernel via device tree.
For the frequencies above 800 MHz, periodic retraining is needed to
compensate for changes in timing. This periodic training will have to be
performed until the frequency drops back to or below 800 MHz.
This driver provides helpers used during this runtime retraining that
will be used by the sequence specific code in a follow-up patch.
Based on work by Peter De Schrijver <pdeschrijver@nvidia.com>.
Signed-off-by: Joseph Lo <josephl@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
Replace the symbolic permissions with octals in order to make them
readable.
Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
Rather than a plain-vanilla init/exit, this patch provides 3 steps in
the initialization needed for driver selection, machine driver
selection and deal with power rail dependencies.
- ACPI scan: this step is done at a very early stage to detect the
presence of a SoundWire Controller and enabled links at the BIOS
level. This step may be called from the legacy HDaudio driver, which
will abort its probe to let the Sound Open Firmware (SOF) handle the
hardware.
- probe: this step allocates all the required memory and will add a
sdw_bus, which in turn will result in identifying all possible Slaves
listed below the Controller ACPI companion device. All the information
is reported to the parent PCI driver which will select the relevant
machine driver.
- startup: this last step starts the bus reset, which results in Slave
devices reporting as ATTACHED and being enumerated. This step is only
done during the card creation stage, after the DSP is powered to
account for internal power rail dependencies.
These 3 steps are already supported in the Sound Open firmware
drivers and upstream.
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Link: https://lore.kernel.org/r/20200531182102.27840-7-yung-chuan.liao@linux.intel.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
ASoC: Fixes for v5.8
This is a collection of mostly small fixes, mostly fixing fallout from
some of the DPCM changes that went in last time around which shook out
some issues on i.MX and Qualcomm platforms. The addition of a managed
version of snd_soc_register_dai() is to fix resource leaks.
There's also a few new device IDs for x86 systems.
