The EAX of the CPUID Leaf 023H enumerates the mask of valid sub-leaves.
To tell the availability of the sub-leaf 1 (enumerate the counter mask),
perf should check the bit 1 (0x2) of EAS, rather than bit 0 (0x1).
The error is not user-visible on bare metal. Because the sub-leaf 0 and
the sub-leaf 1 are always available. However, it may bring issues in a
virtualization environment when a VMM only enumerates the sub-leaf 0.
Introduce the cpuid35_e?x to replace the macros, which makes the
implementation style consistent.
Fixes: eb467aaac2 ("perf/x86/intel: Support Architectural PerfMon Extension leaf")
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20250129154820.3755948-3-kan.liang@linux.intel.com
The PEBS-via-PT feature is exposed for the e-core of some hybrid
platforms, e.g., ADL and MTL. But it never works.
$ dmesg | grep PEBS
[ 1.793888] core: cpu_atom PMU driver: PEBS-via-PT
$ perf record -c 1000 -e '{intel_pt/branch=0/,
cpu_atom/cpu-cycles,aux-output/pp}' -C8
Error:
The sys_perf_event_open() syscall returned with 22 (Invalid argument)
for event (cpu_atom/cpu-cycles,aux-output/pp).
"dmesg | grep -i perf" may provide additional information.
The "PEBS-via-PT" is printed if the corresponding bit of per-PMU
capabilities is set. Since the feature is supported by the e-core HW,
perf sets the bit for e-core. However, for Intel PT, if a feature is not
supported on all CPUs, it is not supported at all. The PEBS-via-PT event
cannot be created successfully.
The PEBS-via-PT is no longer enumerated on the latest hybrid platform. It
will be deprecated on future platforms with Arch PEBS. Let's remove it
from the existing hybrid platforms.
Fixes: d9977c43bf ("perf/x86: Register hybrid PMUs")
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/20250129154820.3755948-2-kan.liang@linux.intel.com
Pull performance events updates from Ingo Molnar:
"Seqlock optimizations that arose in a perf context and were merged
into the perf tree:
- seqlock: Add raw_seqcount_try_begin (Suren Baghdasaryan)
- mm: Convert mm_lock_seq to a proper seqcount (Suren Baghdasaryan)
- mm: Introduce mmap_lock_speculate_{try_begin|retry} (Suren
Baghdasaryan)
- mm/gup: Use raw_seqcount_try_begin() (Peter Zijlstra)
Core perf enhancements:
- Reduce 'struct page' footprint of perf by mapping pages in advance
(Lorenzo Stoakes)
- Save raw sample data conditionally based on sample type (Yabin Cui)
- Reduce sampling overhead by checking sample_type in
perf_sample_save_callchain() and perf_sample_save_brstack() (Yabin
Cui)
- Export perf_exclude_event() (Namhyung Kim)
Uprobes scalability enhancements: (Andrii Nakryiko)
- Simplify find_active_uprobe_rcu() VMA checks
- Add speculative lockless VMA-to-inode-to-uprobe resolution
- Simplify session consumer tracking
- Decouple return_instance list traversal and freeing
- Ensure return_instance is detached from the list before freeing
- Reuse return_instances between multiple uretprobes within task
- Guard against kmemdup() failing in dup_return_instance()
AMD core PMU driver enhancements:
- Relax privilege filter restriction on AMD IBS (Namhyung Kim)
AMD RAPL energy counters support: (Dhananjay Ugwekar)
- Introduce topology_logical_core_id() (K Prateek Nayak)
- Remove the unused get_rapl_pmu_cpumask() function
- Remove the cpu_to_rapl_pmu() function
- Rename rapl_pmu variables
- Make rapl_model struct global
- Add arguments to the init and cleanup functions
- Modify the generic variable names to *_pkg*
- Remove the global variable rapl_msrs
- Move the cntr_mask to rapl_pmus struct
- Add core energy counter support for AMD CPUs
Intel core PMU driver enhancements:
- Support RDPMC 'metrics clear mode' feature (Kan Liang)
- Clarify adaptive PEBS processing (Kan Liang)
- Factor out functions for PEBS records processing (Kan Liang)
- Simplify the PEBS records processing for adaptive PEBS (Kan Liang)
Intel uncore driver enhancements: (Kan Liang)
- Convert buggy pmu->func_id use to pmu->registered
- Support more units on Granite Rapids"
* tag 'perf-core-2025-01-20' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (33 commits)
perf: map pages in advance
perf/x86/intel/uncore: Support more units on Granite Rapids
perf/x86/intel/uncore: Clean up func_id
perf/x86/intel: Support RDPMC metrics clear mode
uprobes: Guard against kmemdup() failing in dup_return_instance()
perf/x86: Relax privilege filter restriction on AMD IBS
perf/core: Export perf_exclude_event()
uprobes: Reuse return_instances between multiple uretprobes within task
uprobes: Ensure return_instance is detached from the list before freeing
uprobes: Decouple return_instance list traversal and freeing
uprobes: Simplify session consumer tracking
uprobes: add speculative lockless VMA-to-inode-to-uprobe resolution
uprobes: simplify find_active_uprobe_rcu() VMA checks
mm: introduce mmap_lock_speculate_{try_begin|retry}
mm: convert mm_lock_seq to a proper seqcount
mm/gup: Use raw_seqcount_try_begin()
seqlock: add raw_seqcount_try_begin
perf/x86/rapl: Add core energy counter support for AMD CPUs
perf/x86/rapl: Move the cntr_mask to rapl_pmus struct
perf/x86/rapl: Remove the global variable rapl_msrs
...
Pull x86 cpuid updates from Borislav Petkov:
- Remove the less generic CPU matching infra around struct x86_cpu_desc
and use the generic struct x86_cpu_id thing
- Remove magic naked numbers for CPUID functions and use proper defines
of the prefix CPUID_LEAF_*. Consolidate some of the crazy use around
the tree
- Smaller cleanups and improvements
* tag 'x86_cpu_for_v6.14_rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/cpu: Make all all CPUID leaf names consistent
x86/fpu: Remove unnecessary CPUID level check
x86/fpu: Move CPUID leaf definitions to common code
x86/tsc: Remove CPUID "frequency" leaf magic numbers.
x86/tsc: Move away from TSC leaf magic numbers
x86/cpu: Move TSC CPUID leaf definition
x86/cpu: Refresh DCA leaf reading code
x86/cpu: Remove unnecessary MwAIT leaf checks
x86/cpu: Use MWAIT leaf definition
x86/cpu: Move MWAIT leaf definition to common header
x86/cpu: Remove 'x86_cpu_desc' infrastructure
x86/cpu: Move AMD erratum 1386 table over to 'x86_cpu_id'
x86/cpu: Replace PEBS use of 'x86_cpu_desc' use with 'x86_cpu_id'
x86/cpu: Expose only stepping min/max interface
x86/cpu: Introduce new microcode matching helper
x86/cpufeature: Document cpu_feature_enabled() as the default to use
x86/paravirt: Remove the WBINVD callback
x86/cpufeatures: Free up unused feature bits
The same CXL PMONs support is also avaiable on GNR. Apply
spr_uncore_cxlcm and spr_uncore_cxldp to GNR as well.
The other units were broken on early HW samples, so they were ignored in
the early enabling patch. The issue has been fixed and verified on the
later production HW. Add UPI, B2UPI, B2HOT, PCIEX16 and PCIEX8 for GNR.
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Eric Hu <eric.hu@intel.com>
Link: https://lkml.kernel.org/r/20250108143017.1793781-2-kan.liang@linux.intel.com
The below warning may be triggered on GNR when the PCIE uncore units are
exposed.
WARNING: CPU: 4 PID: 1 at arch/x86/events/intel/uncore.c:1169 uncore_pci_pmu_register+0x158/0x190
The current uncore driver assumes that all the devices in the same PMU
have the exact same devfn. It's true for the previous platforms. But it
doesn't work for the new PCIE uncore units on GNR.
The assumption doesn't make sense. There is no reason to limit the
devices from the same PMU to the same devfn. Also, the current code just
throws the warning, but still registers the device. The WARN_ON_ONCE()
should be removed.
The func_id is used by the later event_init() to check if a event->pmu
has valid devices. For cpu and mmio uncore PMUs, they are always valid.
For pci uncore PMUs, it's set when the PMU is registered. It can be
replaced by the pmu->registered. Clean up the func_id.
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Eric Hu <eric.hu@intel.com>
Link: https://lkml.kernel.org/r/20250108143017.1793781-1-kan.liang@linux.intel.com
The new RDPMC enhancement, metrics clear mode, is to clear the
PERF_METRICS-related resources as well as the fixed-function performance
monitoring counter 3 after the read is performed. It is available for
ring 3. The feature is enumerated by the
IA32_PERF_CAPABILITIES.RDPMC_CLEAR_METRICS[bit 19]. To enable the
feature, the IA32_FIXED_CTR_CTRL.METRICS_CLEAR_EN[bit 14] must be set.
Two ways were considered to enable the feature.
- Expose a knob in the sysfs globally. One user may affect the
measurement of other users when changing the knob. The solution is
dropped.
- Introduce a new event format, metrics_clear, for the slots event to
disable/enable the feature only for the current process. Users can
utilize the feature as needed.
The latter solution is implemented in the patch.
The current KVM doesn't support the perf metrics yet. For
virtualization, the feature can be enabled later separately.
Suggested-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Andi Kleen <ak@linux.intel.com>
Reviewed-by: Ian Rogers <irogers@google.com>
Link: https://lkml.kernel.org/r/20241211160318.235056-1-kan.liang@linux.intel.com
The 'x86_cpu_desc' and 'x86_cpu_id' structures are very similar.
Reduce duplicate infrastructure by moving the few users of
'x86_cpu_desc' to the much more common variant.
The existing X86_MATCH_VFM_STEPS() helper matches ranges of
steppings. Instead of introducing a single-stepping match function
which could get confusing when paired with the range, just use
the stepping min/max match helper and use min==max.
Note that this makes the table more vertically compact because
multiple entries like this:
INTEL_CPU_DESC(INTEL_SKYLAKE_X, 4, 0x00000000),
INTEL_CPU_DESC(INTEL_SKYLAKE_X, 5, 0x00000000),
INTEL_CPU_DESC(INTEL_SKYLAKE_X, 6, 0x00000000),
INTEL_CPU_DESC(INTEL_SKYLAKE_X, 7, 0x00000000),
can be consolidated down to a single stepping range.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Link: https://lore.kernel.org/all/20241213185131.8B610039%40davehans-spike.ostc.intel.com
The only difference between 5 and 6 is the new counters snapshotting
group, without the following counters snapshotting enabling patches,
it's impossible to utilize the feature in a PEBS record. It's safe to
share the same code path with format 5.
Add format 6, so the end user can at least utilize the legacy PEBS
features.
Fixes: a932aa0e86 ("perf/x86: Add Lunar Lake and Arrow Lake support")
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20241216204505.748363-1-kan.liang@linux.intel.com
While IBS is available for per-thread profiling, still regular users
cannot open an event due to the default paranoid setting (2) which
doesn't allow unprivileged users to get kernel samples. That means
it needs to set exclude_kernel bit in the attribute but IBS driver
would reject it since it has PERF_PMU_CAP_NO_EXCLUDE. This is not what
we want and I've been getting requests to fix this issue.
This should be done in the hardware, but until we get the HW fix we may
allow exclude_{kernel,user,hv} in the attribute and silently drop the
samples in the PMU IRQ handler. It won't guarantee the sampling
frequency or even it'd miss some with fixed period too. Not ideal,
but that'd still be helpful to regular users.
To minimize the confusion, let's add 'swfilt' bit to attr.config2 which
is exposed in the sysfs format directory so that users can figure out
if the kernel support the privilege filters by software.
$ perf record -e ibs_op/swfilt=1/u true
This uses perf_exclude_event() which checks regs->cs. But it should be
fine because set_linear_ip() also updates the CS according to the RIP
provided by IBS.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Tested-by: Ravi Bangoria <ravi.bangoria@amd.com>
Reviewed-by: Ravi Bangoria <ravi.bangoria@amd.com>
Link: https://lore.kernel.org/r/20241203180441.1634709-3-namhyung@kernel.org
Add a new "power_core" PMU and "energy-core" event for monitoring
energy consumption by each individual core. The existing energy-cores
event aggregates the energy consumption of CPU cores at the package level.
This new event aligns with the AMD's per-core energy counters.
Tested the package level and core level PMU counters with workloads
pinned to different CPUs.
Results with workload pinned to CPU 4 in core 4 on an AMD Zen4 Genoa
machine:
$ sudo perf stat --per-core -e power_core/energy-core/ -- taskset -c 4 stress-ng --matrix 1 --timeout 5s
stress-ng: info: [21250] setting to a 5 second run per stressor
stress-ng: info: [21250] dispatching hogs: 1 matrix
stress-ng: info: [21250] successful run completed in 5.00s
Performance counter stats for 'system wide':
S0-D0-C0 1 0.00 Joules power_core/energy-core/
S0-D0-C1 1 0.00 Joules power_core/energy-core/
S0-D0-C2 1 0.00 Joules power_core/energy-core/
S0-D0-C3 1 0.00 Joules power_core/energy-core/
S0-D0-C4 1 8.43 Joules power_core/energy-core/
S0-D0-C5 1 0.00 Joules power_core/energy-core/
S0-D0-C6 1 0.00 Joules power_core/energy-core/
S0-D0-C7 1 0.00 Joules power_core/energy-core/
S0-D1-C8 1 0.00 Joules power_core/energy-core/
S0-D1-C9 1 0.00 Joules power_core/energy-core/
S0-D1-C10 1 0.00 Joules power_core/energy-core/
Signed-off-by: Dhananjay Ugwekar <Dhananjay.Ugwekar@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: "Gautham R. Shenoy" <gautham.shenoy@amd.com>
Link: https://lore.kernel.org/r/20241115060805.447565-11-Dhananjay.Ugwekar@amd.com
Rename struct rapl_pmu variables from "pmu" to "rapl_pmu", to
avoid any confusion between the variables of two different
structs pmu and rapl_pmu. As rapl_pmu also contains a pointer to
struct pmu, which leads to situations in code like pmu->pmu,
which is needlessly confusing. Above scenario is replaced with
much more readable rapl_pmu->pmu with this change.
Also rename "pmus" member in rapl_pmus struct, for same reason.
No functional change.
Signed-off-by: Dhananjay Ugwekar <Dhananjay.Ugwekar@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: "Gautham R. Shenoy" <gautham.shenoy@amd.com>
Reviewed-by: Zhang Rui <rui.zhang@intel.com>
Tested-by: Zhang Rui <rui.zhang@intel.com>
Link: https://lore.kernel.org/r/20241115060805.447565-5-Dhananjay.Ugwekar@amd.com
The current code may iterate all the PEBS records in the DS area several
times. The first loop is to find all active events and calculate the
available records for each event. Then iterate the whole buffer again
and again to process available records until all active events are
processed.
The algorithm is inherited from the old generations. The old PEBS
hardware does not deal well with the situation when events happen near
each other. SW has to drop the error records. Multiple iterations are
required.
The hardware limit has been addressed on newer platforms with adaptive
PEBS. A simple one-iteration algorithm is introduced.
The samples are output by record order with the patch, rather than the
event order. It doesn't impact the post-processing. The perf tool always
sorts the records by time before presenting them to the end user.
In an NMI, the last record has to be specially handled. Add a last[]
variable to track the last unprocessed record of each event.
Test:
11 PEBS events are used in the perf test. Only the basic information is
collected.
perf record -e instructions:up,...,instructions:up -c 2000003 benchmark
The ftrace is used to record the duration of the
intel_pmu_drain_pebs_icl().
The average duration reduced from 62.04us to 57.94us.
A small improvement can be observed with the new algorithm.
Also, the implementation becomes simpler and more straightforward.
Suggested-by: Stephane Eranian <eranian@google.com>
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Link: https://lore.kernel.org/r/20241119135504.1463839-5-kan.liang@linux.intel.com
Factor out functions to process normal and the last PEBS records, which
can be shared with the later patch.
Move the event updating related codes (intel_pmu_save_and_restart())
to the end, where all samples have been processed.
For the current usage, it doesn't matter when perf updates event counts
and reset the counter. Because all counters are stopped when the PEBS
buffer is drained.
Drop the return of the !intel_pmu_save_and_restart(event) check. Because
it never happen. The intel_pmu_save_and_restart(event) only returns 0,
when !hwc->event_base or the period_left > 0.
- The !hwc->event_base is impossible for the PEBS event, since the PEBS
event is only available on GP and fixed counters, which always have
a valid hwc->event_base.
- The check only happens for the case of non-AUTO_RELOAD and single
PEBS, which implies that the event must be overflowed. The period_left
must be always <= 0 for an overflowed event after the
x86_pmu_update().
Co-developed-by: "Peter Zijlstra (Intel)" <peterz@infradead.org>
Signed-off-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://lore.kernel.org/r/20241119135504.1463839-4-kan.liang@linux.intel.com
The PEBS kernel warnings can still be observed with the below case.
when the below commands are running in parallel for a while.
while true;
do
perf record --no-buildid -a --intr-regs=AX \
-e cpu/event=0xd0,umask=0x81/pp \
-c 10003 -o /dev/null ./triad;
done &
while true;
do
perf record -e 'cpu/mem-loads,ldlat=3/uP' -W -d -- ./dtlb
done
The commit b752ea0c28 ("perf/x86/intel/ds: Flush PEBS DS when changing
PEBS_DATA_CFG") intends to flush the entire PEBS buffer before the
hardware is reprogrammed. However, it fails in the above case.
The first perf command utilizes the large PEBS, while the second perf
command only utilizes a single PEBS. When the second perf event is
added, only the n_pebs++. The intel_pmu_pebs_enable() is invoked after
intel_pmu_pebs_add(). So the cpuc->n_pebs == cpuc->n_large_pebs check in
the intel_pmu_drain_large_pebs() fails. The PEBS DS is not flushed.
The new PEBS event should not be taken into account when flushing the
existing PEBS DS.
The check is unnecessary here. Before the hardware is reprogrammed, all
the stale records must be drained unconditionally.
For single PEBS or PEBS-vi-pt, the DS must be empty. The drain_pebs()
can handle the empty case. There is no harm to unconditionally drain the
PEBS DS.
Fixes: b752ea0c28 ("perf/x86/intel/ds: Flush PEBS DS when changing PEBS_DATA_CFG")
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20241119135504.1463839-2-kan.liang@linux.intel.com
Currently, space for raw sample data is always allocated within sample
records for both BPF output and tracepoint events. This leads to unused
space in sample records when raw sample data is not requested.
This patch enforces checking sample type of an event in
perf_sample_save_raw_data(). So raw sample data will only be saved if
explicitly requested, reducing overhead when it is not needed.
Fixes: 0a9081cf0a ("perf/core: Add perf_sample_save_raw_data() helper")
Signed-off-by: Yabin Cui <yabinc@google.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Ian Rogers <irogers@google.com>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Link: https://lore.kernel.org/r/20240515193610.2350456-2-yabinc@google.com
Previously any PMU overflow interrupt that fired while a VCPU was
loaded was recorded as a guest event whether it truly was or not. This
resulted in nonsense perf recordings that did not honor
perf_event_attr.exclude_guest and recorded guest IPs where it should
have recorded host IPs.
Rework the sampling logic to only record guest samples for events with
exclude_guest = 0. This way any host-only events with exclude_guest
set will never see unexpected guest samples. The behaviour of events
with exclude_guest = 0 is unchanged.
Note that events configured to sample both host and guest may still
misattribute a PMI that arrived in the host as a guest event depending
on KVM arch and vendor behavior.
Signed-off-by: Colton Lewis <coltonlewis@google.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Acked-by: Mark Rutland <mark.rutland@arm.com>
Acked-by: Kan Liang <kan.liang@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lore.kernel.org/r/20241113190156.2145593-6-coltonlewis@google.com
Fix the following warning:
CC [M] arch/x86/events/amd/uncore.o
arch/x86/events/amd/uncore.c: In function ‘amd_uncore_umc_ctx_init’:
arch/x86/events/amd/uncore.c:951:52: warning: ‘%d’ directive output may be truncated writing between 1 and 10 bytes into a region of size 8 [-Wformat-truncation=]
snprintf(pmu->name, sizeof(pmu->name), "amd_umc_%d", index);
^~
arch/x86/events/amd/uncore.c:951:43: note: directive argument in the range [0, 2147483647]
snprintf(pmu->name, sizeof(pmu->name), "amd_umc_%d", index);
^~~~~~~~~~~~
arch/x86/events/amd/uncore.c:951:4: note: ‘snprintf’ output between 10 and 19 bytes into a destination of size 16
snprintf(pmu->name, sizeof(pmu->name), "amd_umc_%d", index);
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
As far as I can see, there can't be more than UNCORE_GROUP_MAX (256)
groups and each group can't have more than 255 PMU, so the number
printed by this %d can't exceed 65279, that's only 5 digits and would
fit into the buffer. So it's a false positive warning. But we can
make the compiler happy by declaring index as a 16-bit number.
Signed-off-by: Jean Delvare <jdelvare@suse.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Sandipan Das <sandipan.das@amd.com>
Link: https://lore.kernel.org/r/20241105095253.18f34b4d@endymion.delvare
If the trace data buffer becomes full, a truncated flag [T] is reported
in PERF_RECORD_AUX. In some cases, the size reported is 0, even though
data must have been added to make the buffer full.
That happens when the buffer fills up from empty to full before the
Intel PT driver has updated the buffer position. Then the driver
calculates the new buffer position before calculating the data size.
If the old and new positions are the same, the data size is reported
as 0, even though it is really the whole buffer size.
Fix by detecting when the buffer position is wrapped, and adjust the
data size calculation accordingly.
Example
Use a very small buffer size (8K) and observe the size of truncated [T]
data. Before the fix, it is possible to see records of 0 size.
Before:
$ perf record -m,8K -e intel_pt// uname
Linux
[ perf record: Woken up 2 times to write data ]
[ perf record: Captured and wrote 0.105 MB perf.data ]
$ perf script -D --no-itrace | grep AUX | grep -F '[T]'
Warning:
AUX data lost 2 times out of 3!
5 19462712368111 0x19710 [0x40]: PERF_RECORD_AUX offset: 0 size: 0 flags: 0x1 [T]
5 19462712700046 0x19ba8 [0x40]: PERF_RECORD_AUX offset: 0x170 size: 0xe90 flags: 0x1 [T]
After:
$ perf record -m,8K -e intel_pt// uname
Linux
[ perf record: Woken up 3 times to write data ]
[ perf record: Captured and wrote 0.040 MB perf.data ]
$ perf script -D --no-itrace | grep AUX | grep -F '[T]'
Warning:
AUX data lost 2 times out of 3!
1 113720802995 0x4948 [0x40]: PERF_RECORD_AUX offset: 0 size: 0x2000 flags: 0x1 [T]
1 113720979812 0x6b10 [0x40]: PERF_RECORD_AUX offset: 0x2000 size: 0x2000 flags: 0x1 [T]
Fixes: 52ca9ced3f ("perf/x86/intel/pt: Add Intel PT PMU driver")
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20241022155920.17511-2-adrian.hunter@intel.com
There are extra codes in the CPU hotplug function to allocate rapl pmus.
The generic PMU hotplug support is hard to be applied.
As long as the rapl pmus can be allocated upfront for each die/socket,
the code doesn't need to be implemented in the CPU hotplug function.
Move the code to the init_rapl_pmus(), and allocate a PMU for each
possible die/socket.
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Oliver Sang <oliver.sang@intel.com>
Link: https://lore.kernel.org/r/20241010142604.770192-1-kan.liang@linux.intel.com
Warning at every leaking bits can cause a flood of message, triggering
various stall-warning mechanisms to fire, including CSD locks, which
makes the machine to be unusable.
Track the bits that are being leaked, and only warn when a new bit is
set.
That said, this patch will help with the following issues:
1) It will tell us which bits are being set, so, it is easy to
communicate it back to vendor, and to do a root-cause analyzes.
2) It avoid the machine to be unusable, because, worst case
scenario, the user gets less than 60 WARNs (one per unhandled bit).
Suggested-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Breno Leitao <leitao@debian.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Sandipan Das <sandipan.das@amd.com>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Link: https://lkml.kernel.org/r/20241001141020.2620361-1-leitao@debian.org
ArrowLake-H contains 3 different uarchs, LionCove, Skymont and Crestmont.
It is different with previous hybrid processors which only contains two
kinds of uarchs.
This patch adds PMU support for ArrowLake-H processor, adds ARL-H
specific events which supports the 3 kinds of uarchs, such as
td_retiring_arl_h, and extends some existed format attributes like
offcore_rsp to make them be available to support ARL-H as well. Althrough
these format attributes like offcore_rsp have been extended to support
ARL-H, they can still support the regular hybrid platforms with 2 kinds
of uarchs since the helper hybrid_format_is_visible() would filter PMU
types and only show the format attribute for available PMUs.
Signed-off-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Kan Liang <kan.liang@linux.intel.com>
Tested-by: Yongwei Ma <yongwei.ma@intel.com>
Link: https://lkml.kernel.org/r/20240820073853.1974746-5-dapeng1.mi@linux.intel.com
The upcoming ARL-H hybrid processor contains 2 different atom uarchs
which have different PMU capabilities. To distinguish these atom uarchs,
CPUID.1AH.EAX[23:0] defines a native model ID which can be used to
uniquely identify the uarch of the core by combining with core type.
Thus a 3rd hybrid pmu type "hybrid_tiny" is defined to mark the 2nd
atom uarch. The helper find_hybrid_pmu_for_cpu() would compare the
hybrid pmu type and dynamically read core native id from cpu to identify
the corresponding hybrid pmu structure.
Signed-off-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Kan Liang <kan.liang@linux.intel.com>
Tested-by: Yongwei Ma <yongwei.ma@intel.com>
Link: https://lkml.kernel.org/r/20240820073853.1974746-4-dapeng1.mi@linux.intel.com
Pull perf events updates from Ingo Molnar:
- Implement per-PMU context rescheduling to significantly improve
single-PMU performance, and related cleanups/fixes (Peter Zijlstra
and Namhyung Kim)
- Fix ancient bug resulting in a lot of events being dropped
erroneously at higher sampling frequencies (Luo Gengkun)
- uprobes enhancements:
- Implement RCU-protected hot path optimizations for better
performance:
"For baseline vs SRCU, peak througput increased from 3.7 M/s
(million uprobe triggerings per second) up to about 8 M/s. For
uretprobes it's a bit more modest with bump from 2.4 M/s to
5 M/s.
For SRCU vs RCU Tasks Trace, peak throughput for uprobes
increases further from 8 M/s to 10.3 M/s (+28%!), and for
uretprobes from 5.3 M/s to 5.8 M/s (+11%), as we have more
work to do on uretprobes side.
Even single-thread (no contention) performance is slightly
better: 3.276 M/s to 3.396 M/s (+3.5%) for uprobes, and 2.055
M/s to 2.174 M/s (+5.8%) for uretprobes."
(Andrii Nakryiko et al)
- Document mmap_lock, don't abuse get_user_pages_remote() (Oleg
Nesterov)
- Cleanups & fixes to prepare for future work:
- Remove uprobe_register_refctr()
- Simplify error handling for alloc_uprobe()
- Make uprobe_register() return struct uprobe *
- Fold __uprobe_unregister() into uprobe_unregister()
- Shift put_uprobe() from delete_uprobe() to uprobe_unregister()
- BPF: Fix use-after-free in bpf_uprobe_multi_link_attach()
(Oleg Nesterov)
- New feature & ABI extension: allow events to use PERF_SAMPLE READ
with inheritance, enabling sample based profiling of a group of
counters over a hierarchy of processes or threads (Ben Gainey)
- Intel uncore & power events updates:
- Add Arrow Lake and Lunar Lake support
- Add PERF_EV_CAP_READ_SCOPE
- Clean up and enhance cpumask and hotplug support
(Kan Liang)
- Add LNL uncore iMC freerunning support
- Use D0:F0 as a default device
(Zhenyu Wang)
- Intel PT: fix AUX snapshot handling race (Adrian Hunter)
- Misc fixes and cleanups (James Clark, Jiri Olsa, Oleg Nesterov and
Peter Zijlstra)
* tag 'perf-core-2024-09-18' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (40 commits)
dmaengine: idxd: Clean up cpumask and hotplug for perfmon
iommu/vt-d: Clean up cpumask and hotplug for perfmon
perf/x86/intel/cstate: Clean up cpumask and hotplug
perf: Add PERF_EV_CAP_READ_SCOPE
perf: Generic hotplug support for a PMU with a scope
uprobes: perform lockless SRCU-protected uprobes_tree lookup
rbtree: provide rb_find_rcu() / rb_find_add_rcu()
perf/uprobe: split uprobe_unregister()
uprobes: travers uprobe's consumer list locklessly under SRCU protection
uprobes: get rid of enum uprobe_filter_ctx in uprobe filter callbacks
uprobes: protected uprobe lifetime with SRCU
uprobes: revamp uprobe refcounting and lifetime management
bpf: Fix use-after-free in bpf_uprobe_multi_link_attach()
perf/core: Fix small negative period being ignored
perf: Really fix event_function_call() locking
perf: Optimize __pmu_ctx_sched_out()
perf: Add context time freeze
perf: Fix event_function_call() locking
perf: Extract a few helpers
perf: Optimize context reschedule for single PMU cases
...
