- 16 Feb, 2022 9 commits
-
-
Frederic Weisbecker authored
There can be two subsequent calls to housekeeping_setup() due to "nohz_full=" and "isolcpus=" that can mix up. The two passes each have their own way to deal with an empty housekeeping set of CPUs. Consolidate this part and remove the awful "tmp" based naming. Signed-off-by:
Frederic Weisbecker <frederic@kernel.org> Signed-off-by:
Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by:
Juri Lelli <juri.lelli@redhat.com> Reviewed-by:
Phil Auld <pauld@redhat.com> Link: https://lore.kernel.org/r/20220207155910.527133-6-frederic@kernel.org
-
Frederic Weisbecker authored
Refer to housekeeping APIs using single feature types instead of flags. This prevents from passing multiple isolation features at once to housekeeping interfaces, which soon won't be possible anymore as each isolation features will have their own cpumask. Signed-off-by:
-
Frederic Weisbecker authored
To prepare for supporting each feature of the housekeeping cpumask toward cpuset, prepare each of the HK_FLAG_* entries to move to their own cpumask with enforcing to fetch them individually. The new constraint is that multiple HK_FLAG_* entries can't be mixed together anymore in a single call to housekeeping cpumask(). This will later allow, for example, to runtime modify the cpulist passed through "isolcpus=", "nohz_full=" and "rcu_nocbs=" kernel boot parameters. Signed-off-by:
-
Frederic Weisbecker authored
To prepare for supporting each feature of the housekeeping cpumask toward cpuset, prepare each of the HK_FLAG_* entries to move to their own cpumask with enforcing to fetch them individually. The new constraint is that multiple HK_FLAG_* entries can't be mixed together anymore in a single call to housekeeping cpumask(). This will later allow, for example, to runtime modify the cpulist passed through "isolcpus=", "nohz_full=" and "rcu_nocbs=" kernel boot parameters. Signed-off-by:
Tejun Heo <tj@kernel.org> Link: https://lore.kernel.org/r/20220207155910.527133-3-frederic@kernel.org
-
Frederic Weisbecker authored
To prepare for supporting each feature of the housekeeping cpumask toward cpuset, prepare each of the HK_FLAG_* entries to move to their own cpumask with enforcing to fetch them individually. The new constraint is that multiple HK_FLAG_* entries can't be mixed together anymore in a single call to housekeeping cpumask(). This will later allow, for example, to runtime modify the cpulist passed through "isolcpus=", "nohz_full=" and "rcu_nocbs=" kernel boot parameters. Signed-off-by:
-
Zhaoyang Huang authored
When a new threshold breaching stall happens after a psi event was generated and within the window duration, the new event is not generated because the events are rate-limited to one per window. If after that no new stall is recorded then the event will not be generated even after rate-limiting duration has passed. This is happening because with no new stall, window_update will not be called even though threshold was previously breached. To fix this, record threshold breaching occurrence and generate the event once window duration is passed. Suggested-by:
Suren Baghdasaryan <surenb@google.com> Signed-off-by:
Zhaoyang Huang <zhaoyang.huang@unisoc.com> Signed-off-by:
Johannes Weiner <hannes@cmpxchg.org> Acked-by:
-
Huang Ying authored
In a typical memory tiering system, there's no CPU in slow (PMEM) NUMA nodes. But if the number of the hint page faults on a PMEM node is the max for a task, The current NUMA balancing policy may try to place the task on the PMEM node instead of DRAM node. This is unreasonable, because there's no CPU in PMEM NUMA nodes. To fix this, CPU-less nodes are ignored when searching the migration target node for a task in this patch. To test the patch, we run a workload that accesses more memory in PMEM node than memory in DRAM node. Without the patch, the PMEM node will be chosen as preferred node in task_numa_placement(). While the DRAM node will be chosen instead with the patch. Signed-off-by:
"Huang, Ying" <ying.huang@intel.com> Signed-off-by:
-
Huang Ying authored
The NUMA topology parameters (sched_numa_topology_type, sched_domains_numa_levels, and sched_max_numa_distance, etc.) identified by scheduler may be wrong for systems with CPU-less nodes. For example, the ACPI SLIT of a system with CPU-less persistent memory (Intel Optane DCPMM) nodes is as follows, [000h 0000 4] Signature : "SLIT" [System Locality Information Table] [004h 0004 4] Table Length : 0000042C [008h 0008 1] Revision : 01 [009h 0009 1] Checksum : 59 [00Ah 0010 6] Oem ID : "XXXX" [010h 0016 8] Oem Table ID : "XXXXXXX" [018h 0024 4] Oem Revision : 00000001 [01Ch 0028 4] Asl Compiler ID : "INTL" [020h 0032 4] Asl Compiler Revision : 20091013 [024h 0036 8] Localities : 0000000000000004 [02Ch 0044 4] Locality 0 : 0A 15 11 1C [030h 0048 4] Locality 1 : 15 0A 1C 11 [034h 0052 4] Locality 2 : 11 1C 0A 1C [038h 0056 4] Locality 3 : 1C 11 1C 0A While the `numactl -H` output is as follows, available: 4 nodes (0-3) node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 node 0 size: 64136 MB node 0 free: 5981 MB node 1 cpus: 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 node 1 size: 64466 MB node 1 free: 10415 MB node 2 cpus: node 2 size: 253952 MB node 2 free: 253920 MB node 3 cpus: node 3 size: 253952 MB node 3 free: 253951 MB node distances: node 0 1 2 3 0: 10 21 17 28 1: 21 10 28 17 2: 17 28 10 28 3: 28 17 28 10 In this system, there are only 2 sockets. In each memory controller, both DRAM and PMEM DIMMs are installed. Although the physical NUMA topology is simple, the logical NUMA topology becomes a little complex. Because both the distance(0, 1) and distance (1, 3) are less than the distance (0, 3), it appears that node 1 sits between node 0 and node 3. And the whole system appears to be a glueless mesh NUMA topology type. But it's definitely not, there is even no CPU in node 3. This isn't a practical problem now yet. Because the PMEM nodes (node 2 and node 3 in example system) are offlined by default during system boot. So init_numa_topology_type() called during system boot will ignore them and set sched_numa_topology_type to NUMA_DIRECT. And init_numa_topology_type() is only called at runtime when a CPU of a never-onlined-before node gets plugged in. And there's no CPU in the PMEM nodes. But it appears better to fix this to make the code more robust. To test the potential problem. We have used a debug patch to call init_numa_topology_type() when the PMEM node is onlined (in __set_migration_target_nodes()). With that, the NUMA parameters identified by scheduler is as follows, sched_numa_topology_type: NUMA_GLUELESS_MESH sched_domains_numa_levels: 4 sched_max_numa_distance: 28 To fix the issue, the CPU-less nodes are ignored when the NUMA topology parameters are identified. Because a node may become CPU-less or not at run time because of CPU hotplug, the NUMA topology parameters need to be re-initialized at runtime for CPU hotplug too. With the patch, the NUMA parameters identified for the example system above is as follows, sched_numa_topology_type: NUMA_DIRECT sched_domains_numa_levels: 2 sched_max_numa_distance: 21 Suggested-by:
-
Yury Norov authored
In some places, kernel/sched code calls cpumask_weight() to check if any bit of a given cpumask is set. We can do it more efficiently with cpumask_empty() because cpumask_empty() stops traversing the cpumask as soon as it finds first set bit, while cpumask_weight() counts all bits unconditionally. Signed-off-by:
Yury Norov <yury.norov@gmail.com> Signed-off-by:
-
- 11 Feb, 2022 4 commits
-
-
Huang Ying authored
After commit 8a99b683 ("sched: Move SCHED_DEBUG sysctl to debugfs"), some NUMA balancing sysctls enclosed with SCHED_DEBUG has been moved to debugfs. This patch move the document for these sysctls from Documentation/admin-guide/sysctl/kernel.rst to Documentation/scheduler/sched-debug.rst to make the document consistent with the code. Signed-off-by:
Valentin Schneider <valentin.schneider@arm.com> Acked-by:
Mel Gorman <mgorman@techsingularity.net> Link: https://lkml.kernel.org/r/20220210052514.3038279-1-ying.huang@intel.com
-
Mel Gorman authored
Commit 7d2b5dd0 ("sched/numa: Allow a floating imbalance between NUMA nodes") allowed an imbalance between NUMA nodes such that communicating tasks would not be pulled apart by the load balancer. This works fine when there is a 1:1 relationship between LLC and node but can be suboptimal for multiple LLCs if independent tasks prematurely use CPUs sharing cache. Zen* has multiple LLCs per node with local memory channels and due to the allowed imbalance, it's far harder to tune some workloads to run optimally than it is on hardware that has 1 LLC per node. This patch allows an imbalance to exist up to the point where LLCs should be balanced between nodes. On a Zen3 machine running STREAM parallelised with OMP to have on instance per LLC the results and without binding, the results are 5.17.0-rc0 5.17.0-rc0 vanilla sched-numaimb-v6 MB/sec copy-16 162596.94 ( 0.00%) 580559.74 ( 257.05%) MB/sec scale-16 136901.28 ( 0.00%) 374450.52 ( 173.52%) MB/sec add-16 157300.70 ( 0.00%) 564113.76 ( 258.62%) MB/sec triad-16 151446.88 ( 0.00%) 564304.24 ( 272.61%) STREAM can use directives to force the spread if the OpenMP is new enough but that doesn't help if an application uses threads and it's not known in advance how many threads will be created. Coremark is a CPU and cache intensive benchmark parallelised with threads. When running with 1 thread per core, the vanilla kernel allows threads to contend on cache. With the patch; 5.17.0-rc0 5.17.0-rc0 vanilla sched-numaimb-v5 Min Score-16 368239.36 ( 0.00%) 389816.06 ( 5.86%) Hmean Score-16 388607.33 ( 0.00%) 427877.08 * 10.11%* Max Score-16 408945.69 ( 0.00%) 481022.17 ( 17.62%) Stddev Score-16 15247.04 ( 0.00%) 24966.82 ( -63.75%) CoeffVar Score-16 3.92 ( 0.00%) 5.82 ( -48.48%) It can also make a big difference for semi-realistic workloads like specjbb which can execute arbitrary numbers of threads without advance knowledge of how they should be placed. Even in cases where the average performance is neutral, the results are more stable. 5.17.0-rc0 5.17.0-rc0 vanilla sched-numaimb-v6 Hmean tput-1 71631.55 ( 0.00%) 73065.57 ( 2.00%) Hmean tput-8 582758.78 ( 0.00%) 556777.23 ( -4.46%) Hmean tput-16 1020372.75 ( 0.00%) 1009995.26 ( -1.02%) Hmean tput-24 1416430.67 ( 0.00%) 1398700.11 ( -1.25%) Hmean tput-32 1687702.72 ( 0.00%) 1671357.04 ( -0.97%) Hmean tput-40 1798094.90 ( 0.00%) 2015616.46 * 12.10%* Hmean tput-48 1972731.77 ( 0.00%) 2333233.72 ( 18.27%) Hmean tput-56 2386872.38 ( 0.00%) 2759483.38 ( 15.61%) Hmean tput-64 2909475.33 ( 0.00%) 2925074.69 ( 0.54%) Hmean tput-72 2585071.36 ( 0.00%) 2962443.97 ( 14.60%) Hmean tput-80 2994387.24 ( 0.00%) 3015980.59 ( 0.72%) Hmean tput-88 3061408.57 ( 0.00%) 3010296.16 ( -1.67%) Hmean tput-96 3052394.82 ( 0.00%) 2784743.41 ( -8.77%) Hmean tput-104 2997814.76 ( 0.00%) 2758184.50 ( -7.99%) Hmean tput-112 2955353.29 ( 0.00%) 2859705.09 ( -3.24%) Hmean tput-120 2889770.71 ( 0.00%) 2764478.46 ( -4.34%) Hmean tput-128 2871713.84 ( 0.00%) 2750136.73 ( -4.23%) Stddev tput-1 5325.93 ( 0.00%) 2002.53 ( 62.40%) Stddev tput-8 6630.54 ( 0.00%) 10905.00 ( -64.47%) Stddev tput-16 25608.58 ( 0.00%) 6851.16 ( 73.25%) Stddev tput-24 12117.69 ( 0.00%) 4227.79 ( 65.11%) Stddev tput-32 27577.16 ( 0.00%) 8761.05 ( 68.23%) Stddev tput-40 59505.86 ( 0.00%) 2048.49 ( 96.56%) Stddev tput-48 168330.30 ( 0.00%) 93058.08 ( 44.72%) Stddev tput-56 219540.39 ( 0.00%) 30687.02 ( 86.02%) Stddev tput-64 121750.35 ( 0.00%) 9617.36 ( 92.10%) Stddev tput-72 223387.05 ( 0.00%) 34081.13 ( 84.74%) Stddev tput-80 128198.46 ( 0.00%) 22565.19 ( 82.40%) Stddev tput-88 136665.36 ( 0.00%) 27905.97 ( 79.58%) Stddev tput-96 111925.81 ( 0.00%) 99615.79 ( 11.00%) Stddev tput-104 146455.96 ( 0.00%) 28861.98 ( 80.29%) Stddev tput-112 88740.49 ( 0.00%) 58288.23 ( 34.32%) Stddev tput-120 186384.86 ( 0.00%) 45812.03 ( 75.42%) Stddev tput-128 78761.09 ( 0.00%) 57418.48 ( 27.10%) Similarly, for embarassingly parallel problems like NPB-ep, there are improvements due to better spreading across LLC when the machine is not fully utilised. vanilla sched-numaimb-v6 Min ep.D 31.79 ( 0.00%) 26.11 ( 17.87%) Amean ep.D 31.86 ( 0.00%) 26.17 * 17.86%* Stddev ep.D 0.07 ( 0.00%) 0.05 ( 24.41%) CoeffVar ep.D 0.22 ( 0.00%) 0.20 ( 7.97%) Max ep.D 31.93 ( 0.00%) 26.21 ( 17.91%) Signed-off-by:
Gautham R. Shenoy <gautham.shenoy@amd.com> Tested-by:
K Prateek Nayak <kprateek.nayak@amd.com> Link: https://lore.kernel.org/r/20220208094334.16379-3-mgorman@techsingularity.net
-
Mel Gorman authored
There are inconsistencies when determining if a NUMA imbalance is allowed that should be corrected. o allow_numa_imbalance changes types and is not always examining the destination group so both the type should be corrected as well as the naming. o find_idlest_group uses the sched_domain's weight instead of the group weight which is different to find_busiest_group o find_busiest_group uses the source group instead of the destination which is different to task_numa_find_cpu o Both find_idlest_group and find_busiest_group should account for the number of running tasks if a move was allowed to be consistent with task_numa_find_cpu Fixes: 7d2b5dd0 ("sched/numa: Allow a floating imbalance between NUMA nodes") Signed-off-by:
-
Mathieu Desnoyers authored
Just before the 2.35 release of glibc, the __rseq_offset userspace ABI was changed from int to ptrdiff_t. Adapt to this change in the kernel selftests. Signed-off-by:
Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by:
-
- 02 Feb, 2022 16 commits
-
-
Zhen Ni authored
move autogroup sysctls to autogroup.c and use the new register_sysctl_init() to register the sysctl interface. Signed-off-by:
Zhen Ni <nizhen@uniontech.com> Signed-off-by:
-
Mathieu Desnoyers authored
Rather than use rseq_get_abi() and pass its result through a register to the inline assembler, directly access the per-thread rseq area through a memory reference combining the %gs segment selector, the constant offset of the field in struct rseq, and the rseq_offset value (in a register). Signed-off-by:
-
Mathieu Desnoyers authored
Rather than use rseq_get_abi() and pass its result through a register to the inline assembler, directly access the per-thread rseq area through a memory reference combining the %fs segment selector, the constant offset of the field in struct rseq, and the rseq_offset value (in a register). Signed-off-by:
-
Mathieu Desnoyers authored
gcc and clang each have their own compiler bugs with respect to asm goto. Implement a work-around for compiler versions known to have those bugs. gcc prior to 4.8.2 miscompiles asm goto. https://gcc.gnu.org/bugzilla/show_bug.cgi?id=58670 gcc prior to 8.1.0 miscompiles asm goto at O1. https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103908 clang prior to version 13.0.1 miscompiles asm goto at O2. https://github.com/llvm/llvm-project/issues/52735 Work around these issues by adding a volatile inline asm with memory clobber in the fallthrough after the asm goto and at each label target. Emit this for all compilers in case other similar issues are found in the future. Signed-off-by:
-
Mathieu Desnoyers authored
The arm and mips work-around for asm goto size guess issues are not properly documented, and lack reference to specific compiler versions, upstream compiler bug tracker entry, and reproducer. I can only find a loosely documented patch in my original LKML rseq post refering to gcc < 7 on ARM, but it does not appear to be sufficient to track the exact issue. Also, I am not sure MIPS really has the same limitation. Therefore, remove the work-around until we can properly document this. Signed-off-by:
-
Mathieu Desnoyers authored
Signed-off-by:
-
Mathieu Desnoyers authored
The semantic of off_t is for file offsets. We mean to use it as an offset from a pointer. We really expect it to fit in a single register, and not use a 64-bit type on 32-bit architectures. Fix runtime issues on ppc32 where the offset is always 0 due to inconsistency between the argument type (off_t -> 64-bit) and type expected by the inline assembler (32-bit). Signed-off-by:
-
Mathieu Desnoyers authored
Building the rseq basic test with gcc version 5.4.0 20160609 (Ubuntu 5.4.0-6ubuntu1~16.04.12) Target: powerpc-linux-gnu leads to these errors: /tmp/ccieEWxU.s: Assembler messages: /tmp/ccieEWxU.s:118: Error: syntax error; found `,', expected `(' /tmp/ccieEWxU.s:118: Error: junk at end of line: `,8' /tmp/ccieEWxU.s:121: Error: syntax error; found `,', expected `(' /tmp/ccieEWxU.s:121: Error: junk at end of line: `,8' /tmp/ccieEWxU.s:626: Error: syntax error; found `,', expected `(' /tmp/ccieEWxU.s:626: Error: junk at end of line: `,8' /tmp/ccieEWxU.s:629: Error: syntax error; found `,', expected `(' /tmp/ccieEWxU.s:629: Error: junk at end of line: `,8' /tmp/ccieEWxU.s:735: Error: syntax error; found `,', expected `(' /tmp/ccieEWxU.s:735: Error: junk at end of line: `,8' /tmp/ccieEWxU.s:738: Error: syntax error; found `,', expected `(' /tmp/ccieEWxU.s:738: Error: junk at end of line: `,8' /tmp/ccieEWxU.s:741: Error: syntax error; found `,', expected `(' /tmp/ccieEWxU.s:741: Error: junk at end of line: `,8' Makefile:581: recipe for target 'basic_percpu_ops_test.o' failed Based on discussion with Linux powerpc maintainers and review of the use of the "m" operand in powerpc kernel code, add the missing %Un%Xn (where n is operand number) to the lwz, stw, ld, and std instructions when used with "m" operands. Using "WORD" to mean either a 32-bit or 64-bit type depending on the architecture is misleading. The term "WORD" really means a 32-bit type in both 32-bit and 64-bit powerpc assembler. The intent here is to wrap load/store to intptr_t into common macros for both 32-bit and 64-bit. Rename the macros with a RSEQ_ prefix, and use the terms "INT" for always 32-bit type, and "LONG" for architecture bitness-sized type. Signed-off-by: -
Mathieu Desnoyers authored
ppc32 incorrectly uses padding as rseq_cs pointer field. Fix this by using the rseq_cs.arch.ptr field. Use this field across all architectures. Signed-off-by:
-
Mathieu Desnoyers authored
glibc-2.35 (upcoming release date 2022-02-01) exposes the rseq per-thread data in the TCB, accessible at an offset from the thread pointer, rather than through an actual Thread-Local Storage (TLS) variable, as the Linux kernel selftests initially expected. The __rseq_abi TLS and glibc-2.35's ABI for per-thread data cannot actively coexist in a process, because the kernel supports only a single rseq registration per thread. Here is the scheme introduced to ensure selftests can work both with an older glibc and with glibc-2.35+: - librseq exposes its own "rseq_offset, rseq_size, rseq_flags" ABI. - librseq queries for glibc rseq ABI (__rseq_offset, __rseq_size, __rseq_flags) using dlsym() in a librseq library constructor. If those are found, copy their values into rseq_offset, rseq_size, and rseq_flags. - Else, if those glibc symbols are not found, handle rseq registration from librseq and use its own IE-model TLS to implement the rseq ABI per-thread storage. Signed-off-by:
-
Mathieu Desnoyers authored
This is done in preparation for the selftest uplift to become compatible with glibc-2.35. glibc-2.35 exposes the rseq per-thread data in the TCB, accessible at an offset from the thread pointer. The toolchains do not implement accessing the thread pointer on all architectures. Provide thread pointer getters for ppc and x86 which lack (or lacked until recently) toolchain support. Signed-off-by:
-
Mathieu Desnoyers authored
This is done in preparation for the selftest uplift to become compatible with glibc-2.35. glibc-2.35 exposes the rseq per-thread data in the TCB, accessible at an offset from the thread pointer, rather than through an actual Thread-Local Storage (TLS) variable, as the kernel selftests initially expected. Introduce a rseq_get_abi() helper, initially using the __rseq_abi TLS variable, in preparation for changing this userspace ABI for one which is compatible with glibc-2.35. Note that the __rseq_abi TLS and glibc-2.35's ABI for per-thread data cannot actively coexist in a process, because the kernel supports only a single rseq registration per thread. Signed-off-by:
-
Mathieu Desnoyers authored
This is done in preparation for the selftest uplift to become compatible with glibc-2.35. All accesses to the __rseq_abi fields are volatile, but remove the volatile from the TLS variable declaration, otherwise we are stuck with volatile for the upcoming rseq_get_abi() helper. Signed-off-by:
-
Mathieu Desnoyers authored
Signed-off-by:
-
Mathieu Desnoyers authored
The rseq rseq_cs.ptr.{ptr32,padding} uapi endianness handling is entirely wrong on 32-bit little endian: a preprocessor logic mistake wrongly uses the big endian field layout on 32-bit little endian architectures. Fortunately, those ptr32 accessors were never used within the kernel, and only meant as a convenience for user-space. Remove those and replace the whole rseq_cs union by a __u64 type, as this is the only thing really needed to express the ABI. Document how 32-bit architectures are meant to interact with this field. Fixes: ec9c82e0 ("rseq: uapi: Declare rseq_cs field as union, update includes") Signed-off-by: -
Mathieu Desnoyers authored
The Linux kernel rseq uapi header has a broken layout for the rseq_cs.ptr field on 32-bit little endian architectures. The entire rseq_cs.ptr field is planned for removal, leaving only the 64-bit rseq_cs.ptr64 field available. Both glibc and librseq use their own copy of the Linux kernel uapi header, where they introduce proper union fields to access to the 32-bit low order bits of the rseq_cs pointer on 32-bit architectures. Introduce a copy of the Linux kernel uapi headers in the Linux kernel selftests. Signed-off-by:
-
- 27 Jan, 2022 8 commits
-
-
Suren Baghdasaryan authored
When CONFIG_CGROUPS is disabled psi code generates the following warnings: kernel/sched/psi.c:1112:21: warning: no previous prototype for 'psi_trigger_create' [-Wmissing-prototypes] 1112 | struct psi_trigger *psi_trigger_create(struct psi_group *group, | ^~~~~~~~~~~~~~~~~~ kernel/sched/psi.c:1182:6: warning: no previous prototype for 'psi_trigger_destroy' [-Wmissing-prototypes] 1182 | void psi_trigger_destroy(struct psi_trigger *t) | ^~~~~~~~~~~~~~~~~~~ kernel/sched/psi.c:1249:10: warning: no previous prototype for 'psi_trigger_poll' [-Wmissing-prototypes] 1249 | __poll_t psi_trigger_poll(void **trigger_ptr, | ^~~~~~~~~~~~~~~~ Change declarations of these functions in the header to provide the prototypes even when they are unused. Fixes: 0e94682b ("psi: introduce psi monitor") Reported-by:kernel test robot <lkp@intel.com> Signed-off-by:
-
Suren Baghdasaryan authored
When CONFIG_PROC_FS is disabled psi code generates the following warnings: kernel/sched/psi.c:1364:30: warning: 'psi_cpu_proc_ops' defined but not used [-Wunused-const-variable=] 1364 | static const struct proc_ops psi_cpu_proc_ops = { | ^~~~~~~~~~~~~~~~ kernel/sched/psi.c:1355:30: warning: 'psi_memory_proc_ops' defined but not used [-Wunused-const-variable=] 1355 | static const struct proc_ops psi_memory_proc_ops = { | ^~~~~~~~~~~~~~~~~~~ kernel/sched/psi.c:1346:30: warning: 'psi_io_proc_ops' defined but not used [-Wunused-const-variable=] 1346 | static const struct proc_ops psi_io_proc_ops = { | ^~~~~~~~~~~~~~~ Make definitions of these structures and related functions conditional on CONFIG_PROC_FS config. Fixes: 0e94682b ("psi: introduce psi monitor") Reported-by: -
Qais Yousef authored
iowait_boost signal is applied independently of util and doesn't take into account uclamp settings of the rq. An io heavy task that is capped by uclamp_max could still request higher frequency because sugov_iowait_apply() doesn't clamp the boost via uclamp_rq_util_with() like effective_cpu_util() does. Make sure that iowait_boost honours uclamp requests by calling uclamp_rq_util_with() when applying the boost. Fixes: 982d9cdc ("sched/cpufreq, sched/uclamp: Add clamps for FAIR and RT tasks") Signed-off-by:
Qais Yousef <qais.yousef@arm.com> Signed-off-by:
Rafael J. Wysocki <rafael.j.wysocki@intel.com> Link: https://lore.kernel.org/r/20211216225320.2957053-3-qais.yousef@arm.com
-
Qais Yousef authored
sugov_update_single_{freq, perf}() contains a 'busy' filter that ensures we don't bring the frqeuency down if there's no idle time (CPU is busy). The problem is that with uclamp_max we will have scenarios where a busy task is capped to run at a lower frequency and this filter prevents applying the capping when this task starts running. We handle this by skipping the filter when uclamp is enabled and the rq is being capped by uclamp_max. We introduce a new function uclamp_rq_is_capped() to help detecting when this capping is taking effect. Some code shuffling was required to allow using cpu_util_{cfs, rt}() in this new function. On 2 Core SMT2 Intel laptop I see: Without this patch: uclampset -M 0 sysbench --test=cpu --threads = 4 run produces a score of ~3200 consistently. Which is the highest possible. Compiling the kernel also results in frequency running at max 3.1GHz all the time - running uclampset -M 400 to cap it has no effect without this patch. With this patch: uclampset -M 0 sysbench --test=cpu --threads = 4 run produces a score of ~1100 with some outliers in ~1700. Uclamp max aggregates the performance requirements, so having high values sometimes is expected if some other task happens to require that frequency starts running at the same time. When compiling the kernel with uclampset -M 400 I can see the frequencies mostly in the ~2GHz region. Helpful to conserve power and prevent heating when not plugged in. Fixes: 982d9cdc ("sched/cpufreq, sched/uclamp: Add clamps for FAIR and RT tasks") Signed-off-by: -
Qais Yousef authored
We can't use this tracepoint in modules without having the symbol exported first, fix that. Fixes: 76504793 ("sched/pelt: Add support to track thermal pressure") Signed-off-by:
-
Johannes Weiner authored
Suren wrote the poll() interface, which is a significant part of the psi code and represents a large user of psi itself (Android). It's a good idea to have him look at psi patches as well, and it's good to have two people following things in case one of us is traveling. Signed-off-by:
-
Honglei Wang authored
The child processes will inherit numa_pages_migrated and total_numa_faults from the parent. It means even if there is no numa fault happen on the child, the statistics in /proc/$pid of the child process might show huge amount. This is a bit weird. Let's initialize them when do fork. Signed-off-by:
Honglei Wang <wanghonglei@didichuxing.com> Signed-off-by:
Mel Gorman <mgorman@suse.de> Link: https://lore.kernel.org/r/20220113133920.49900-1-wanghonglei@didichuxing.com
-
Bharata B Rao authored
The older format of /proc/pid/sched printed home node info which required the mempolicy and task lock around mpol_get(). However the format has changed since then and there is no need for sched_show_numa() any more to have mempolicy argument, asssociated mpol_get/put and task_lock/unlock. Remove them. Fixes: 397f2378 ("sched/numa: Fix numa balancing stats in /proc/pid/sched") Signed-off-by:
Bharata B Rao <bharata@amd.com> Signed-off-by:
Srikar Dronamraju <srikar@linux.vnet.ibm.com> Acked-by:
-
- 23 Jan, 2022 3 commits
-
-
Linus Torvalds authored
-
Linus Torvalds authored
Merge tag 'perf-tools-for-v5.17-2022-01-22' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux Pull more perf tools updates from Arnaldo Carvalho de Melo: - Fix printing 'phys_addr' in 'perf script'. - Fix failure to add events with 'perf probe' in ppc64 due to not removing leading dot (ppc64 ABIv1). - Fix cpu_map__item() python binding building. - Support event alias in form foo-bar-baz, add pmu-events and parse-event tests for it. - No need to setup affinities when starting a workload or attaching to a pid. - Use path__join() to compose a path instead of ad-hoc snprintf() equivalent. - Override attr->sample_period for non-libpfm4 events. - Use libperf cpumap APIs instead of accessing the internal state directly. - Sync x86 arch prctl headers and files changed by the new set_mempolicy_home_node syscall with the kernel sources. - Remove duplicate include in cpumap.h. - Remove redundant err variable. * tag 'perf-tools-for-v5.17-2022-01-22' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux: perf tools: Remove redundant err variable perf test: Add parse-events test for aliases with hyphens perf test: Add pmu-events test for aliases with hyphens perf parse-events: Support event alias in form foo-bar-baz perf evsel: Override attr->sample_period for non-libpfm4 events perf cpumap: Remove duplicate include in cpumap.h perf cpumap: Migrate to libperf cpumap api perf python: Fix cpu_map__item() building perf script: Fix printing 'phys_addr' failure issue tools headers UAPI: Sync files changed by new set_mempolicy_home_node syscall tools headers UAPI: Sync x86 arch prctl headers with the kernel sources perf machine: Use path__join() to compose a path instead of snprintf(dir, '/', filename) perf evlist: No need to setup affinities when disabling events for pid targets perf evlist: No need to setup affinities when enabling events for pid targets perf stat: No need to setup affinities when starting a workload perf affinity: Allow passing a NULL arg to affinity__cleanup() perf probe: Fix ppc64 'perf probe add events failed' case
-
git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-traceLinus Torvalds authored
Pull ftrace fix from Steven Rostedt: "Fix s390 breakage from sorting mcount tables. The latest merge of the tracing tree sorts the mcount table at build time. But s390 appears to do things differently (like always) and replaces the sorted table back to the original unsorted one. As the ftrace algorithm depends on it being sorted, bad things happen when it is not, and s390 experienced those bad things. Add a new config to tell the boot if the mcount table is sorted or not, and allow s390 to opt out of it" * tag 'trace-v5.17-3' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace: ftrace: Fix assuming build time sort works for s390
-
