[WHY&HOW]
dc_post_update_surfaces_to_stream needs to be called after a full update
completes in order to optimize clocks and watermarks for power. Add
missing calls before idle entry is requested to ensure optimal power.
Reviewed-by: Aurabindo Pillai <aurabindo.pillai@amd.com>
Signed-off-by: Dillon Varone <Dillon.Varone@amd.com>
Signed-off-by: Ivan Lipski <ivan.lipski@amd.com>
Tested-by: Dan Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[WHY&HOW]
Reprogramming watermarks with stutter allowed can cause instability on
some ASICs. Disable it prior to raising watermarks (prepare bandwidth),
then re-enable after lowering (optimize bandwidth).
Reviewed-by: Alvin Lee <Alvin.Lee2@amd.com>
Signed-off-by: Dillon Varone <Dillon.Varone@amd.com>
Signed-off-by: Ivan Lipski <ivan.lipski@amd.com>
Tested-by: Dan Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Align vcn5_fw_shared structure with FW
Signed-off-by: Sonny Jiang <sonny.jiang@amd.com>
Reviewed-by: Leo Liu <leo.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The MES set resources packet has an optional bit 'lr_compute_wa'
which can be used for preventing MES hangs on long compute jobs.
Set this bit by default.
Co-developed-by: Yifan Zhang <yifan1.zhang@amd.com>
Signed-off-by: Yifan Zhang <yifan1.zhang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
we dont need to allocate local array of pages to hold
the pages returned by the hmm, instead we could use
the hmm_range structure itself to get to hmm_pfn
and get the required pages directly.
This avoids call to alloc/free quite a lot.
Signed-off-by: Sunil Khatri <sunil.khatri@amd.com>
Suggested-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Acked-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Place the smu 13.0.0 pptable header in the correct folder
Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Mangesh Gadre <mangesh.gadre@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
It's not necessary that the build string and atom header section has a
difference of 32 bytes. Use the remaining bytes in the section as copy
limit.
Fixes: d6fa802661 ("drm/amdgpu: Add vbios build number interface")
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Make sure the headers are sorted alphabetically to ensure consistent
code.
Signed-off-by: Luca Weiss <luca.weiss@fairphone.com>
Reviewed-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Javier Martinez Canillas <javierm@redhat.com>
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://lore.kernel.org/r/20250623-simple-drm-fb-icc-v2-4-f69b86cd3d7d@fairphone.com
Make sure the headers are sorted alphabetically to ensure consistent
code.
Signed-off-by: Luca Weiss <luca.weiss@fairphone.com>
Reviewed-by: Javier Martinez Canillas <javierm@redhat.com>
Reviewed-by: Thomas Zimmermann <tzimmermann@suse.de>
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Several fbdev drivers select FB_MODE_HELPER in the Kconfig. The
setting controls some helper functions and has no meaning to the
user. Make it an internal option.
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
Link: https://lore.kernel.org/r/20250909132047.152612-3-tzimmermann@suse.de
The option CONFIG_FB_TILEBLITTING is controlled by the user. Selecting
it from drivers can lead to cyclic dependencies within the config. In
fbcon, there's special handling for tile blitting, which currently
cannot be disabled without first disabling the relevant fbdev drivers.
Fix the Kconfig dependency to make it work.
Some guidelines for using select can be found in the kernel docs at [1].
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://elixir.bootlin.com/linux/v6.16/source/Documentation/kbuild/kconfig-language.rst#L147 # [1]
Acked-by: Arnd Bergmann <arnd@arndb.de>
Link: https://lore.kernel.org/r/20250909132047.152612-2-tzimmermann@suse.de
In the ACPI debugger interface, the helper functions for read and write
operations use "int" as the length parameter data type. When a large
"size_t count" is passed from the file operations, this cast to "int"
results in truncation and a negative value due to signed integer
representation.
Logically, this negative number propagates to the min() calculation,
where it is selected over the positive buffer space value, leading to
unexpected behavior. Subsequently, when this negative value is used in
copy_to_user() or copy_from_user(), it is interpreted as a large positive
value due to the unsigned nature of the size parameter in these functions,
causing the copy operations to attempt handling sizes far beyond the
intended buffer limits.
Address the issue by:
- Changing the length parameters in acpi_aml_read_user() and
acpi_aml_write_user() from "int" to "size_t", aligning with the
expected unsigned size semantics.
- Updating return types and local variables in acpi_aml_read() and
acpi_aml_write() to "ssize_t" for consistency with kernel file
operation conventions.
- Using "size_t" for the "n" variable to ensure calculations remain
unsigned.
- Using min_t() for circ_count_to_end() and circ_space_to_end() to
ensure type-safe comparisons and prevent integer overflow.
Signed-off-by: Amir Mohammad Jahangirzad <a.jahangirzad@gmail.com>
Link: https://patch.msgid.link/20250923013113.20615-1-a.jahangirzad@gmail.com
[ rjw: Changelog tweaks, local variable definitions ordering adjustments ]
Fixes: 8cfb0cdf07 ("ACPI / debugger: Add IO interface to access debugger functionalities")
Cc: 4.5+ <stable@vger.kernel.org> # 4.5+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
TRACE_SEQ_BUFFER_SIZE is dependent on the architecture for its size. on 64-bit
systems, it is 8148 bytes. forced 8-byte alignment in size_t and seq_buf means
that trace_seq is 8200 bytes on 64-bit systems. moving the buffer to the end
of the struct fixes the issue. there shouldn't be any side effects, i.e.
pointer arithmetic on trace_seq
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Link: https://lore.kernel.org/20250821053917.23301-1-git@elijahs.space
Signed-off-by: Elijah Wright <git@elijahs.space>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Replace the opencoded for_each_cpu(cpu, cpu_online_mask) loop with the
more readable and equivalent for_each_online_cpu(cpu) macro.
Link: https://lore.kernel.org/20250811064158.2456-1-wangfushuai@baidu.com
Signed-off-by: Fushuai Wang <wangfushuai@baidu.com>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Move the cpuidle driver check from __acpi_processor_start() to
acpi_processor_power_init() which allows variable acpi_idle_driver to
become static.
No intentional functional impact.
Signed-off-by: Huisong Li <lihuisong@huawei.com>
Link: https://patch.msgid.link/20250923032428.2656329-2-lihuisong@huawei.com
[ rjw: Subject tweak, new changelog, adjustment of a new comment ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Currently the syscall trace events show each value as hexadecimal, but
without adding "0x" it can be confusing:
sys_write(fd: 4, buf: 0x55c4a1fa9270, count: 44)
Looks like the above write wrote 44 bytes, when in reality it wrote 68
bytes.
Add a "0x" for all values greater or equal to 10 to remove the ambiguity.
For values less than 10, leave off the "0x" as that just adds noise to the
output.
Also change the iterator to check if "i" is nonzero and print the ", "
delimiter at the start, then adding the logic to the trace_seq_printf() at
the end.
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Takaya Saeki <takayas@google.com>
Cc: Tom Zanussi <zanussi@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ian Rogers <irogers@google.com>
Cc: Douglas Raillard <douglas.raillard@arm.com>
Link: https://lore.kernel.org/20250923130713.764558957@kernel.org
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
The syscall events are pseudo events that hook to the raw syscalls. The
ftrace_syscall_enter/exit() callback is called by the raw_syscall
enter/exit tracepoints respectively whenever any of the syscall events are
enabled.
The trace_array has an array of syscall "files" that correspond to the
system calls based on their __NR_SYSCALL number. The array is read and if
there's a pointer to a trace_event_file then it is considered enabled and
if it is NULL that syscall event is considered disabled.
Currently it uses an rcu_dereference_sched() to get this pointer and a
rcu_assign_ptr() or RCU_INIT_POINTER() to write to it. This is unnecessary
as the file pointer will not go away outside the synchronization of the
tracepoint logic itself. And this code adds no extra RCU synchronization
that uses this.
Replace these functions with a simple READ_ONCE() and WRITE_ONCE() which
is all they need. This will also allow this code to not depend on
preemption being disabled as system call tracepoints are now allowed to
fault.
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Takaya Saeki <takayas@google.com>
Cc: Tom Zanussi <zanussi@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ian Rogers <irogers@google.com>
Cc: Douglas Raillard <douglas.raillard@arm.com>
Link: https://lore.kernel.org/20250923130713.594320290@kernel.org
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
For some coresight components like CTI and TPDM, there could be
numerous of them. From the node name, we can only get the type and
register address of the component. We can't identify the HW or the
system the component belongs to. Add label sysfs node support for
showing the intuitive name of the device.
Signed-off-by: Mao Jinlong <quic_jinlmao@quicinc.com>
Reviewed-by: Mike Leach <mike.leach@linaro.org>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Link: https://lore.kernel.org/r/20250816072529.3716968-3-quic_jinlmao@quicinc.com
Current name of coresight component's folder consists of prefix of
the device and the id in the device list. When run 'ls' command,
we can get the register address of the device. Take CTI for example,
if we want to set the config for modem CTI, but we can't know which
CTI is modem CTI from all current information.
cti_sys0 -> ../../../devices/platform/soc@0/138f0000.cti/cti_sys0
cti_sys1 -> ../../../devices/platform/soc@0/13900000.cti/cti_sys1
Add label to show hardware context information of each coresight
device. There will be a sysfs node label in each device folder.
cat /sys/bus/coresight/devices/cti_sys0/label
Signed-off-by: Mao Jinlong <quic_jinlmao@quicinc.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Reviewed-by: Mike Leach <mike.leach@linaro.org>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Link: https://lore.kernel.org/r/20250816072529.3716968-2-quic_jinlmao@quicinc.com
The return value of devm_kzalloc could be an null pointer,
use "!desc.pdata" to fix incorrect handling return value
of devm_kzalloc.
Fixes: 4277f035d2 ("coresight: trbe: Add a representative coresight_platform_data for TRBE")
Signed-off-by: Lin Yujun <linyujun809@h-partners.com>
Reviewed-by: James Clark <james.clark@linaro.org>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Link: https://lore.kernel.org/r/20250908122022.1315399-1-linyujun809@h-partners.com
Some TPDM devices support both CMB and DSB datasets, requiring
the system to enable the port with both corresponding element sizes.
Currently, the logic treats tpdm_read_element_size as successful if
the CMB element size is retrieved correctly, regardless of whether
the DSB element size is obtained. This behavior causes issues
when parsing data from TPDM devices that depend on both element sizes.
To address this, the function should explicitly fail if the DSB
element size cannot be read correctly.
Fixes: e6d7f5252f ("coresight-tpda: Add support to configure CMB element")
Reviewed-by: James Clark <james.clark@linaro.org>
Signed-off-by: Jie Gan <jie.gan@oss.qualcomm.com>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Link: https://lore.kernel.org/r/20250906-fix_element_size_issue-v2-1-dbb0ac2541a9@oss.qualcomm.com
When the TRBE driver fails to allocate a buffer, it currently returns
the error code "-ENOMEM". However, the caller etm_setup_aux() only
checks for a NULL pointer, so it misses the error. As a result, the
driver continues and eventually causes a kernel panic.
Fix this by returning a NULL pointer from arm_trbe_alloc_buffer() on
allocation failures. This allows that the callers can properly handle
the failure.
Fixes: 3fbf7f011f ("coresight: sink: Add TRBE driver")
Reported-by: Tamas Zsoldos <tamas.zsoldos@arm.com>
Signed-off-by: Leo Yan <leo.yan@arm.com>
Reviewed-by: James Clark <james.clark@linaro.org>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Link: https://lore.kernel.org/r/20250904-cs_etm_auxsetup_fix_error_handling-v2-1-a502d0bafb95@arm.com
The validation for driver data pointers and clock pointers are redundant
in the runtime PM callbacks. After a driver's probing, its driver data
and clocks have been initialized successfully, this ensures it is safe
to access driver data and clocks in the runtime PM callbacks. A corner
case is a clock pointer is NULL, in this case, the clock core layer can
handle it properly. So remove these redundant checking.
In runtime resume, respect values returned from clock function and add
error handling.
Reviewed-by: Anshuman Khandual <anshuman.khandual@arm.com>
Reviewed-by: Yeoreum Yun <yeoreum.yun@arm.com>
Tested-by: James Clark <james.clark@linaro.org>
Signed-off-by: Leo Yan <leo.yan@arm.com>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Link: https://lore.kernel.org/r/20250731-arm_cs_fix_clock_v4-v6-10-1dfe10bb3f6f@arm.com
Since atclk is enabled after pclk during the probe phase, this commit
maintains the same sequence for the runtime resume flow.
Reviewed-by: Anshuman Khandual <anshuman.khandual@arm.com>
Reviewed-by: Yeoreum Yun <yeoreum.yun@arm.com>
Tested-by: James Clark <james.clark@linaro.org>
Signed-off-by: Leo Yan <leo.yan@arm.com>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Link: https://lore.kernel.org/r/20250731-arm_cs_fix_clock_v4-v6-9-1dfe10bb3f6f@arm.com
The driver data no longer needs to be allocated separately in the static
and dynamic probes. Moved the allocation into the low-level functions to
avoid code duplication.
Reviewed-by: Anshuman Khandual <anshuman.khandual@arm.com>
Reviewed-by: Yeoreum Yun <yeoreum.yun@arm.com>
Tested-by: James Clark <james.clark@linaro.org>
Signed-off-by: Leo Yan <leo.yan@arm.com>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Link: https://lore.kernel.org/r/20250731-arm_cs_fix_clock_v4-v6-8-1dfe10bb3f6f@arm.com
CoreSight drivers enable pclk and atclk conditionally. For example,
pclk is only enabled in the static probe, while atclk is an optional
clock that it is enabled for both dynamic and static probes, if it is
present. In the current CoreSight drivers, these two clocks are
initialized separately. This causes complex and duplicate codes.
CoreSight drivers are refined so that clocks are initialized in one go.
For this purpose, this commit renames coresight_get_enable_apb_pclk() to
coresight_get_enable_clocks() and encapsulates clock initialization
logic:
- If a clock is initialized successfully, its clock pointer is assigned
to the double pointer passed as an argument.
- For ACPI devices, clocks are controlled by firmware, directly bail
out.
- Skip enabling pclk for an AMBA device.
- If atclk is not found, the corresponding double pointer is set to
NULL. The function returns Success (0) to guide callers can proceed
with no error.
- Otherwise, an error number is returned for failures.
The function became complex, move it from the header to the CoreSight
core layer and the symbol is exported. Added comments for recording
details.
Suggested-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Reviewed-by: Anshuman Khandual <anshuman.khandual@arm.com>
Reviewed-by: Yeoreum Yun <yeoreum.yun@arm.com>
Tested-by: James Clark <james.clark@linaro.org>
Signed-off-by: Leo Yan <leo.yan@arm.com>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Link: https://lore.kernel.org/r/20250731-arm_cs_fix_clock_v4-v6-7-1dfe10bb3f6f@arm.com
The programming clock is enabled by AMBA bus driver before a dynamic
probe. As a result, a CoreSight driver may redundantly enable the same
clock.
To avoid this, add a check for device type and skip enabling the
programming clock for AMBA devices. The returned NULL pointer will be
tolerated by the drivers.
Fixes: 73d779a03a ("coresight: etm4x: Change etm4_platform_driver driver for MMIO devices")
Reviewed-by: Anshuman Khandual <anshuman.khandual@arm.com>
Reviewed-by: Yeoreum Yun <yeoreum.yun@arm.com>
Tested-by: James Clark <james.clark@linaro.org>
Signed-off-by: Leo Yan <leo.yan@arm.com>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Link: https://lore.kernel.org/r/20250731-arm_cs_fix_clock_v4-v6-6-1dfe10bb3f6f@arm.com
Some CoreSight components have trace bus clocks 'atclk' and are enabled
using clk_prepare_enable(). These clocks are not disabled when modules
exit.
As atclk is optional, use devm_clk_get_optional_enabled() to manage it.
The benefit is the driver model layer can automatically disable and
release clocks.
Check the returned value with IS_ERR() to detect errors but leave the
NULL pointer case if the clock is not found. And remove the error
handling codes which are no longer needed.
Fixes: d1839e6877 ("coresight: etm: retrieve and handle atclk")
Reviewed-by: Anshuman Khandual <anshuman.khandual@arm.com>
Reviewed-by: Yeoreum Yun <yeoreum.yun@arm.com>
Tested-by: James Clark <james.clark@linaro.org>
Signed-off-by: Leo Yan <leo.yan@arm.com>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Link: https://lore.kernel.org/r/20250731-arm_cs_fix_clock_v4-v6-5-1dfe10bb3f6f@arm.com
Some CoreSight components have programming clocks (pclk) and are enabled
using clk_get() and clk_prepare_enable(). However, in many cases, these
clocks are not disabled when modules exit and only released by clk_put().
To fix the issue, this commit refactors programming clock by replacing
clk_get() and clk_prepare_enable() with devm_clk_get_optional_enabled()
for enabling APB clock. If the "apb_pclk" clock is not found, a NULL
pointer is returned, and the function proceeds to attempt enabling the
"apb" clock.
Since ACPI platforms rely on firmware to manage clocks, returning a NULL
pointer in this case leaves clock management to the firmware rather than
the driver. This effectively avoids a clock imbalance issue during
module removal - where the clock could be disabled twice: once during
the ACPI runtime suspend and again during the devm resource release.
Callers are updated to reuse the returned error value.
With the change, programming clocks are managed as resources in driver
model layer, allowing clock cleanup to be handled automatically. As a
result, manual cleanup operations are no longer needed and are removed
from the Coresight drivers.
Fixes: 73d779a03a ("coresight: etm4x: Change etm4_platform_driver driver for MMIO devices")
Reviewed-by: Yeoreum Yun <yeoreum.yun@arm.com>
Tested-by: James Clark <james.clark@linaro.org>
Signed-off-by: Leo Yan <leo.yan@arm.com>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Link: https://lore.kernel.org/r/20250731-arm_cs_fix_clock_v4-v6-4-1dfe10bb3f6f@arm.com
The atclk is an optional clock for the CoreSight ETMv4, but the driver
misses to initialize it.
This change enables atclk in probe of the ETMv4 driver, and dynamically
control the clock during suspend and resume.
No need to check the driver data and clock pointer in the runtime
suspend and resume, so remove checks. And add error handling in the
resume function.
Add a minor fix to the comment format when adding the atclk field.
Fixes: 2e1cdfe184 ("coresight-etm4x: Adding CoreSight ETM4x driver")
Reviewed-by: Anshuman Khandual <anshuman.khandual@arm.com>
Reviewed-by: Yeoreum Yun <yeoreum.yun@arm.com>
Tested-by: James Clark <james.clark@linaro.org>
Signed-off-by: Leo Yan <leo.yan@arm.com>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Link: https://lore.kernel.org/r/20250731-arm_cs_fix_clock_v4-v6-3-1dfe10bb3f6f@arm.com
The atclk is an optional clock for the CoreSight CATU, but the driver
misses to initialize it.
This change enables atclk in probe of the CATU driver, and dynamically
control the clock during suspend and resume.
The checks for driver data and clocks in suspend and resume are not
needed, remove them. Add error handling in the resume function.
Fixes: fcacb5c154 ("coresight: Introduce support for Coresight Address Translation Unit")
Reviewed-by: Anshuman Khandual <anshuman.khandual@arm.com>
Reviewed-by: Yeoreum Yun <yeoreum.yun@arm.com>
Tested-by: James Clark <james.clark@linaro.org>
Signed-off-by: Leo Yan <leo.yan@arm.com>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Link: https://lore.kernel.org/r/20250731-arm_cs_fix_clock_v4-v6-2-1dfe10bb3f6f@arm.com
The atclk is an optional clock for the CoreSight TMC, but the driver
misses to initialize it. In most cases, TMC shares the atclk clock with
other CoreSight components. Since these components enable the clock
before the TMC device is initialized, the TMC continues properly,
which is why we don’t observe any lockup issues.
This change enables atclk in probe of the TMC driver. Given the clock
is optional, it is possible to return NULL if the clock does not exist.
IS_ERR() is tolerant for this case.
Dynamically disable and enable atclk during suspend and resume. The
clock pointers will never be error values if the driver has successfully
probed, and the case of a NULL pointer case will be handled by the clock
core layer. The driver data is always valid after probe. Therefore,
remove the related checks. Also in the resume flow adds error handling.
Fixes: bc4bf7fe98 ("coresight-tmc: add CoreSight TMC driver")
Reviewed-by: Anshuman Khandual <anshuman.khandual@arm.com>
Reviewed-by: Yeoreum Yun <yeoreum.yun@arm.com>
Tested-by: James Clark <james.clark@linaro.org>
Signed-off-by: Leo Yan <leo.yan@arm.com>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Link: https://lore.kernel.org/r/20250731-arm_cs_fix_clock_v4-v6-1-1dfe10bb3f6f@arm.com
The TRCEXTINSELR is only implemented if TRCIDR5.NUMEXTINSEL > 0.
To avoid invalid accesses, introduce a check on numextinsel
(derived from TRCIDR5[11:9]) before reading or writing to this register.
Fixes: f5bd523690 ("coresight: etm4x: Convert all register accesses")
Signed-off-by: Yuanfang Zhang <yuanfang.zhang@oss.qualcomm.com>
Reviewed-by: James Clark <james.clark@linaro.org>
Reviewed-by: Mike Leach <mike.leach@linaro.org>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Link: https://lore.kernel.org/r/20250812-trcextinselr_issue-v2-1-e6eb121dfcf4@oss.qualcomm.com
The devm_ioremap_resource() function returns error pointers on error.
It never returns NULL. Update the error checking to match.
Fixes: e54a52a28a ("coresight: add coresight Trace Network On Chip driver")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Reviewed-by: Mike Leach <mike.leach@linaro.org>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Link: https://lore.kernel.org/r/7bd9fae8-a15f-412a-8800-ce47acf0b5ce@sabinyo.mountain
Jan Kara reported that the shared ILOCK held across the journal
flush during fdatasync operations slows down O_DSYNC DIO on
unwritten extents significantly. The underlying issue is that
unwritten extent conversion needs the ILOCK exclusive, whilst the
datasync operation after the extent conversion holds it shared.
Hence we cannot be flushing the journal for one IO completion whilst
at the same time doing unwritten extent conversion on another IO
completion on the same inode. This means that IO completions
lock-step, and IO performance is dependent on the journal flush
latency.
Jan demonstrated that reducing the ifdatasync lock hold time can
improve O_DSYNC DIO to unwritten extents performance by 2.5x.
Discussion on that patch found issues with the method, and we
came to the conclusion that separately tracking datasync flush
sequences was the best approach to solving the problem.
The fsync code uses the ILOCK to serialise against concurrent
modifications in the transaction commit phase. In a transaction
commit, there are several disjoint updates to inode log item state
that need to be considered atomically by the fsync code. These
operations are all done under ILOCK_EXCL context:
1. ili_fsync_flags is updated in ->iop_precommit
2. i_pincount is updated in ->iop_pin before it is added to the CIL
3. ili_commit_seq is updated in ->iop_committing, after it has been
added to the CIL
In fsync, we need to:
1. check that the inode is dirty in the journal (ipincount)
2. check that ili_fsync_flags is set
3. grab the ili_commit_seq if a journal flush is needed
4. clear the ili_fsync_flags to ensure that new modifications that
require fsync are tracked in ->iop_precommit correctly
The serialisation of ipincount/ili_commit_seq is needed
to ensure that we don't try to unnecessarily flush the journal.
The serialisation of ili_fsync_flags being set in
->iop_precommit and cleared in fsync post journal flush is
required for correctness.
Hence holding the ILOCK_SHARED in xfs_file_fsync() performs all this
serialisation for us. Ideally, we want to remove the need to hold
the ILOCK_SHARED in xfs_file_fsync() for best performance.
We start with the observation that fsync/fdatasync() only need to
wait for operations that have been completed. Hence operations that
are still being committed have not completed and datasync operations
do not need to wait for them.
This means we can use a single point in time in the commit process
to signal "this modification is complete". This is what
->iop_committing is supposed to provide - it is the point at which
the object is unlocked after the modification has been recorded in
the CIL. Hence we could use ili_commit_seq to determine if we should
flush the journal.
In theory, we can already do this. However, in practice this will
expose an internal global CIL lock to the IO path. The ipincount()
checks optimise away the need to take this lock - if the inode is
not pinned, then it is not in the CIL and we don't need to check if
a journal flush at ili_commit_seq needs to be performed.
The reason this is needed is that the ili_commit_seq is never
cleared. Once it is set, it remains set even once the journal has
been committed and the object has been unpinned. Hence we have to
look that journal internal commit sequence state to determine if
ili_commit_seq needs to be acted on or not.
We can solve this by clearing ili_commit_seq when the inode is
unpinned. If we clear it atomically with the last unpin going away,
then we are guaranteed that new modifications will order correctly
as they add a new pin counts and we won't clear a sequence number
for an active modification in the CIL.
Further, we can then allow the per-transaction flag state to
propagate into ->iop_committing (instead of clearing it in
->iop_precommit) and that will allow us to determine if the
modification needs a full fsync or just a datasync, and so we can
record a separate datasync sequence number (Jan's idea!) and then
use that in the fdatasync path instead of the full fsync sequence
number.
With this infrastructure in place, we no longer need the
ILOCK_SHARED in the fsync path. All serialisation is done against
the commit sequence numbers - if the sequence number is set, then we
have to flush the journal. If it is not set, then we have nothing to
do. This greatly simplifies the fsync implementation....
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Tested-by: Jan Kara <jack@suse.cz>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Carlos Maiolino <cem@kernel.org>
There are similar extsize checks and updates done inside and outside
the inode item lock, which could all be done under a single top
level logic branch outside the ili_lock. The COW extsize fixup can
potentially miss updating the XFS_ILOG_CORE in ili_fsync_fields, so
moving this code up above the ili_fsync_fields update could also be
considered a fix.
Further, to make the next change a bit cleaner, move where we
calculate the on-disk flag mask to after we attach the cluster
buffer to the the inode log item.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Carlos Maiolino <cem@kernel.org>
The function is unused; remove it.
Instead of relying on a general blit helper, drivers should pick a blit
function by themselves from their list of supported color formats.
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Javier Martinez Canillas <javierm@redhat.com>
Link: https://lore.kernel.org/r/20250918154207.84714-4-tzimmermann@suse.de
Some configurations of sysfb outputs require format conversion from
framebuffer to scanout buffer. It is a driver bug if the conversion
helper is missing, yet it might happen on odd scanout formats. The old
code, based on drm_fb_blit(), only detects this situation during the
commit's hardware update, which is too late to abort the update.
Lookup the correct blit helper as part of the check phase. Then store
it in the sysfb plane state. Allows for detection of a missing helper
before the commit has started. Also avoids drm_fb_blit()'s large switch
statement on each updated scanline. Only a single lookup has to be done.
The lookup is in drm_sysfb_get_blit_func(), which only tracks formats
supported by sysfb drivers.
The lookup happens in sysfb's begin_fb_access helper instead of its
atomic_check helper. This allows vesadrm, and possibly other drivers,
to implement their own atomic_check without interfering with blit
lookups. Vesadrm implements XRGB8888 on top of R8 formats with the
help of the atomic_check. Doing the blit lookup in begin_fb_access then
always uses the correct CRTC format on all drivers.
v2:
- vesadrm: use drm_sysfb_plane_helper_begin_fb_access()
- fix type in commit description (Javier)
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Javier Martinez Canillas <javierm@redhat.com>
Link: https://lore.kernel.org/r/20250918154207.84714-3-tzimmermann@suse.de
The plane-state type struct drm_sysfb_plane_state will store the
helper for blitting to the scanout buffer.
v2:
- add variable for duplicated shadow-plane state (Javier)
- fix build error
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Javier Martinez Canillas <javierm@redhat.com>
Link: https://lore.kernel.org/r/20250918154207.84714-2-tzimmermann@suse.de
The dm-raid code was using hardcoded integer values to represent the
read-only/read-write state of RAID arrays instead of the proper
enumeration constants defined in the md_ro_state enumerator type.
Changes:
- Replace hardcoded integers with the appropriate md_ro_state enumerator
values
- Add the missing MD_RDONLY setting in the post_suspend function
(no failures have been attributed to this inconsistency, the fix
ensures correct state transitions for completeness)
This improves code clarity and maintainability by using the defined
enumeration constants rather than magic numbers, ensuring the code
properly conforms to the established API interface.
Signed-off-by: Heinz Mauelshagen <heinzm@redhat.com>
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
The previous patch preferred async interface for the purpose of testing.
However, the synchronous interface is faster, so it should be preferred.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Make sure that the "data" argument passed to integrity_sector_checksum is
always page-aligned and add an "offset" argument that specifies the
offset from the start of the page. This will enable us to use the
asynchronous hash interface later.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Allocate the recalculate buffer with kmalloc rather than vmalloc. This
will be needed later, for the simplification of the asynchronous hash
interface.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Similar to how we annotate dma-fences, add lockep annotation to
the pm_block completion to ensure we don't wait for it while holding
locks that are needed in the pm notifier or in the device
suspend / resume callbacks.
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://lore.kernel.org/r/20250918142848.21807-3-thomas.hellstrom@linux.intel.com
During pm notifier eviction we may still race with validations.
Ensure those are blocked out during eviction to ensure we have
access to as much system memory as possible.
During the suspend operation itself, we run single-threaded so that
shouldn't be a problem.
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://lore.kernel.org/r/20250918142848.21807-2-thomas.hellstrom@linux.intel.com
