Pull scheduler updates from Ingo Molnar: - Improve the scalability of the CFS bandwidth unthrottling logic with large number of CPUs. - Fix & rework various cpuidle routines, simplify interaction with the generic scheduler code. Add __cpuidle methods as noinstr to objtool's noinstr detection and fix boatloads of cpuidle bugs & quirks. - Add new ABI: introduce MEMBARRIER_CMD_GET_REGISTRATIONS, to query previously issued registrations. - Limit scheduler slice duration to the sysctl_sched_latency period, to improve scheduling granularity with a large number of SCHED_IDLE tasks. - Debuggability enhancement on sys_exit(): warn about disabled IRQs, but also enable them to prevent a cascade of followup problems and repeat warnings. - Fix the rescheduling logic in prio_changed_dl(). - Micro-optimize cpufreq and sched-util methods. - Micro-optimize ttwu_runnable() - Micro-optimize the idle-scanning in update_numa_stats(), select_idle_capacity() and steal_cookie_task(). - Update the RSEQ code & self-tests - Constify various scheduler methods - Remove unused methods - Refine __init tags - Documentation updates - Misc other cleanups, fixes * tag 'sched-core-2023-02-20' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (110 commits) sched/rt: pick_next_rt_entity(): check list_entry sched/deadline: Add more reschedule cases to prio_changed_dl() sched/fair: sanitize vruntime of entity being placed sched/fair: Remove capacity inversion detection sched/fair: unlink misfit task from cpu overutilized objtool: mem*() are not uaccess safe cpuidle: Fix poll_idle() noinstr annotation sched/clock: Make local_clock() noinstr sched/clock/x86: Mark sched_clock() noinstr x86/pvclock: Improve atomic update of last_value in pvclock_clocksource_read() x86/atomics: Always inline arch_atomic64*() cpuidle: tracing, preempt: Squash _rcuidle tracing cpuidle: tracing: Warn about !rcu_is_watching() cpuidle: lib/bug: Disable rcu_is_watching() during WARN/BUG cpuidle: drivers: firmware: psci: Dont instrument suspend code KVM: selftests: Fix build of rseq test exit: Detect and fix irq disabled state in oops cpuidle, arm64: Fix the ARM64 cpuidle logic cpuidle: mvebu: Fix duplicate flags assignment sched/fair: Limit sched slice duration ...
Switch from a request_queue pointer and reference to a gendisk once for the throttle information in struct task_struct. Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Andreas Herrmann <aherrmann@suse.de> Link: https://lore.kernel.org/r/20230203150400.3199230-8-hch@lst.de Signed-off-by: Jens Axboe <axboe@kernel.dk>
This feature allows the scheduler to expose a per-memory map concurrency ID to user-space. This concurrency ID is within the possible cpus range, and is temporarily (and uniquely) assigned while threads are actively running within a memory map. If a memory map has fewer threads than cores, or is limited to run on few cores concurrently through sched affinity or cgroup cpusets, the concurrency IDs will be values close to 0, thus allowing efficient use of user-space memory for per-cpu data structures. This feature is meant to be exposed by a new rseq thread area field. The primary purpose of this feature is to do the heavy-lifting needed by memory allocators to allow them to use per-cpu data structures efficiently in the following situations: - Single-threaded applications, - Multi-threaded applications on large systems (many cores) with limited cpu affinity mask, - Multi-threaded applications on large systems (many cores) with restricted cgroup cpuset per container. One of the key concern from scheduler maintainers is the overhead associated with additional spin locks or atomic operations in the scheduler fast-path. This is why the following optimization is implemented. On context switch between threads belonging to the same memory map, transfer the mm_cid from prev to next without any atomic ops. This takes care of use-cases involving frequent context switch between threads belonging to the same memory map. Additional optimizations can be done if the spin locks added when context switching between threads belonging to different memory maps end up being a performance bottleneck. Those are left out of this patch though. A performance impact would have to be clearly demonstrated to justify the added complexity. The credit goes to Paul Turner (Google) for the original virtual cpu id idea. This feature is implemented based on the discussions with Paul Turner and Peter Oskolkov (Google), but I took the liberty to implement scheduler fast-path optimizations and my own NUMA-awareness scheme. The rumor has it that Google have been running a rseq vcpu_id extension internally in production for a year. The tcmalloc source code indeed has comments hinting at a vcpu_id prototype extension to the rseq system call [1]. The following benchmarks do not show any significant overhead added to the scheduler context switch by this feature: * perf bench sched messaging (process) Baseline: 86.5±0.3 ms With mm_cid: 86.7±2.6 ms * perf bench sched messaging (threaded) Baseline: 84.3±3.0 ms With mm_cid: 84.7±2.6 ms * hackbench (process) Baseline: 82.9±2.7 ms With mm_cid: 82.9±2.9 ms * hackbench (threaded) Baseline: 85.2±2.6 ms With mm_cid: 84.4±2.9 ms [1] https://github.com/google/tcmalloc/blob/master/tcmalloc/internal/linux_syscall_support.h#L26 Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lore.kernel.org/r/20221122203932.231377-8-mathieu.desnoyers@efficios.com
Introduce the extensible rseq ABI, where the feature size supported by the kernel and the required alignment are communicated to user-space through ELF auxiliary vectors. This allows user-space to call rseq registration with a rseq_len of either 32 bytes for the original struct rseq size (which includes padding), or larger. If rseq_len is larger than 32 bytes, then it must be large enough to contain the feature size communicated to user-space through ELF auxiliary vectors. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lore.kernel.org/r/20221122203932.231377-4-mathieu.desnoyers@efficios.com
Pull MM updates from Andrew Morton: - More userfaultfs work from Peter Xu - Several convert-to-folios series from Sidhartha Kumar and Huang Ying - Some filemap cleanups from Vishal Moola - David Hildenbrand added the ability to selftest anon memory COW handling - Some cpuset simplifications from Liu Shixin - Addition of vmalloc tracing support by Uladzislau Rezki - Some pagecache folioifications and simplifications from Matthew Wilcox - A pagemap cleanup from Kefeng Wang: we have VM_ACCESS_FLAGS, so use it - Miguel Ojeda contributed some cleanups for our use of the __no_sanitize_thread__ gcc keyword. This series should have been in the non-MM tree, my bad - Naoya Horiguchi improved the interaction between memory poisoning and memory section removal for huge pages - DAMON cleanups and tuneups from SeongJae Park - Tony Luck fixed the handling of COW faults against poisoned pages - Peter Xu utilized the PTE marker code for handling swapin errors - Hugh Dickins reworked compound page mapcount handling, simplifying it and making it more efficient - Removal of the autonuma savedwrite infrastructure from Nadav Amit and David Hildenbrand - zram support for multiple compression streams from Sergey Senozhatsky - David Hildenbrand reworked the GUP code's R/O long-term pinning so that drivers no longer need to use the FOLL_FORCE workaround which didn't work very well anyway - Mel Gorman altered the page allocator so that local IRQs can remnain enabled during per-cpu page allocations - Vishal Moola removed the try_to_release_page() wrapper - Stefan Roesch added some per-BDI sysfs tunables which are used to prevent network block devices from dirtying excessive amounts of pagecache - David Hildenbrand did some cleanup and repair work on KSM COW breaking - Nhat Pham and Johannes Weiner have implemented writeback in zswap's zsmalloc backend - Brian Foster has fixed a longstanding corner-case oddity in file[map]_write_and_wait_range() - sparse-vmemmap changes for MIPS, LoongArch and NIOS2 from Feiyang Chen - Shiyang Ruan has done some work on fsdax, to make its reflink mode work better under xfstests. Better, but still not perfect - Christoph Hellwig has removed the .writepage() method from several filesystems. They only need .writepages() - Yosry Ahmed wrote a series which fixes the memcg reclaim target beancounting - David Hildenbrand has fixed some of our MM selftests for 32-bit machines - Many singleton patches, as usual * tag 'mm-stable-2022-12-13' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm: (313 commits) mm/hugetlb: set head flag before setting compound_order in __prep_compound_gigantic_folio mm: mmu_gather: allow more than one batch of delayed rmaps mm: fix typo in struct pglist_data code comment kmsan: fix memcpy tests mm: add cond_resched() in swapin_walk_pmd_entry() mm: do not show fs mm pc for VM_LOCKONFAULT pages selftests/vm: ksm_functional_tests: fixes for 32bit selftests/vm: cow: fix compile warning on 32bit selftests/vm: madv_populate: fix missing MADV_POPULATE_(READ|WRITE) definitions mm/gup_test: fix PIN_LONGTERM_TEST_READ with highmem mm,thp,rmap: fix races between updates of subpages_mapcount mm: memcg: fix swapcached stat accounting mm: add nodes= arg to memory.reclaim mm: disable top-tier fallback to reclaim on proactive reclaim selftests: cgroup: make sure reclaim target memcg is unprotected selftests: cgroup: refactor proactive reclaim code to reclaim_until() mm: memcg: fix stale protection of reclaim target memcg mm/mmap: properly unaccount memory on mas_preallocate() failure omfs: remove ->writepage jfs: remove ->writepage ...
Pull scheduler updates from Ingo Molnar: - Implement persistent user-requested affinity: introduce affinity_context::user_mask and unconditionally preserve the user-requested CPU affinity masks, for long-lived tasks to better interact with cpusets & CPU hotplug events over longer timespans, without destroying the original affinity intent if the underlying topology changes. - Uclamp updates: fix relationship between uclamp and fits_capacity() - PSI fixes - Misc fixes & updates * tag 'sched-core-2022-12-12' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: sched: Clear ttwu_pending after enqueue_task() sched/psi: Use task->psi_flags to clear in CPU migration sched/psi: Stop relying on timer_pending() for poll_work rescheduling sched/psi: Fix avgs_work re-arm in psi_avgs_work() sched/psi: Fix possible missing or delayed pending event sched: Always clear user_cpus_ptr in do_set_cpus_allowed() sched: Enforce user requested affinity sched: Always preserve the user requested cpumask sched: Introduce affinity_context sched: Add __releases annotations to affine_move_task() sched/fair: Check if prev_cpu has highest spare cap in feec() sched/fair: Consider capacity inversion in util_fits_cpu() sched/fair: Detect capacity inversion sched/uclamp: Cater for uclamp in find_energy_efficient_cpu()'s early exit condition sched/uclamp: Make cpu_overutilized() use util_fits_cpu() sched/uclamp: Make asym_fits_capacity() use util_fits_cpu() sched/uclamp: Make select_idle_capacity() use util_fits_cpu() sched/uclamp: Fix fits_capacity() check in feec() sched/uclamp: Make task_fits_capacity() use util_fits_cpu() sched/uclamp: Fix relationship between uclamp and migration margin
Currently mm_struct maintains rss_stats which are updated on page fault and the unmapping codepaths. For page fault codepath the updates are cached per thread with the batch of TASK_RSS_EVENTS_THRESH which is 64. The reason for caching is performance for multithreaded applications otherwise the rss_stats updates may become hotspot for such applications. However this optimization comes with the cost of error margin in the rss stats. The rss_stats for applications with large number of threads can be very skewed. At worst the error margin is (nr_threads * 64) and we have a lot of applications with 100s of threads, so the error margin can be very high. Internally we had to reduce TASK_RSS_EVENTS_THRESH to 32. Recently we started seeing the unbounded errors for rss_stats for specific applications which use TCP rx0cp. It seems like vm_insert_pages() codepath does not sync rss_stats at all. This patch converts the rss_stats into percpu_counter to convert the error margin from (nr_threads * 64) to approximately (nr_cpus ^ 2). However this conversion enable us to get the accurate stats for situations where accuracy is more important than the cpu cost. This patch does not make such tradeoffs - we can just use percpu_counter_add_local() for the updates and percpu_counter_sum() (or percpu_counter_sync() + percpu_counter_read) for the readers. At the moment the readers are either procfs interface, oom_killer and memory reclaim which I think are not performance critical and should be ok with slow read. However I think we can make that change in a separate patch. Link: https://lkml.kernel.org/r/20221024052841.3291983-1-shakeelb@google.com Signed-off-by: Shakeel Butt <shakeelb@google.com> Cc: Marek Szyprowski <m.szyprowski@samsung.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The commit d583d360a620 ("psi: Fix psi state corruption when schedule()
races with cgroup move") fixed a race problem by making cgroup_move_task()
use task->psi_flags instead of looking at the scheduler state.
We can extend task->psi_flags usage to CPU migration, which should be
a minor optimization for performance and code simplicity.
Signed-off-by: Chengming Zhou <zhouchengming@bytedance.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Link: https://lore.kernel.org/r/20220926081931.45420-1-zhouchengming@bytedance.com
There have been various issues and limitations with the way perf uses
(task) contexts to track events. Most notable is the single hardware
PMU task context, which has resulted in a number of yucky things (both
proposed and merged).
Notably:
- HW breakpoint PMU
- ARM big.little PMU / Intel ADL PMU
- Intel Branch Monitoring PMU
- AMD IBS PMU
- S390 cpum_cf PMU
- PowerPC trace_imc PMU
*Current design:*
Currently we have a per task and per cpu perf_event_contexts:
task_struct::perf_events_ctxp[] <-> perf_event_context <-> perf_cpu_context
^ | ^ | ^
`---------------------------------' | `--> pmu ---'
v ^
perf_event ------'
Each task has an array of pointers to a perf_event_context. Each
perf_event_context has a direct relation to a PMU and a group of
events for that PMU. The task related perf_event_context's have a
pointer back to that task.
Each PMU has a per-cpu pointer to a per-cpu perf_cpu_context, which
includes a perf_event_context, which again has a direct relation to
that PMU, and a group of events for that PMU.
The perf_cpu_context also tracks which task context is currently
associated with that CPU and includes a few other things like the
hrtimer for rotation etc.
Each perf_event is then associated with its PMU and one
perf_event_context.
*Proposed design:*
New design proposed by this patch reduce to a single task context and
a single CPU context but adds some intermediate data-structures:
task_struct::perf_event_ctxp -> perf_event_context <- perf_cpu_context
^ | ^ ^
`---------------------------' | |
| | perf_cpu_pmu_context <--.
| `----. ^ |
| | | |
| v v |
| ,--> perf_event_pmu_context |
| | |
| | |
v v |
perf_event ---> pmu ----------------'
With the new design, perf_event_context will hold all events for all
pmus in the (respective pinned/flexible) rbtrees. This can be achieved
by adding pmu to rbtree key:
{cpu, pmu, cgroup, group_index}
Each perf_event_context carries a list of perf_event_pmu_context which
is used to hold per-pmu-per-context state. For example, it keeps track
of currently active events for that pmu, a pmu specific task_ctx_data,
a flag to tell whether rotation is required or not etc.
Additionally, perf_cpu_pmu_context is used to hold per-pmu-per-cpu
state like hrtimer details to drive the event rotation, a pointer to
perf_event_pmu_context of currently running task and some other
ancillary information.
Each perf_event is associated to it's pmu, perf_event_context and
perf_event_pmu_context.
Further optimizations to current implementation are possible. For
example, ctx_resched() can be optimized to reschedule only single pmu
events.
Much thanks to Ravi for picking this up and pushing it towards
completion.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Co-developed-by: Ravi Bangoria <ravi.bangoria@amd.com>
Signed-off-by: Ravi Bangoria <ravi.bangoria@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20221008062424.313-1-ravi.bangoria@amd.com
Pull more MM updates from Andrew Morton: - fix a race which causes page refcounting errors in ZONE_DEVICE pages (Alistair Popple) - fix userfaultfd test harness instability (Peter Xu) - various other patches in MM, mainly fixes * tag 'mm-stable-2022-10-13' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm: (29 commits) highmem: fix kmap_to_page() for kmap_local_page() addresses mm/page_alloc: fix incorrect PGFREE and PGALLOC for high-order page mm/selftest: uffd: explain the write missing fault check mm/hugetlb: use hugetlb_pte_stable in migration race check mm/hugetlb: fix race condition of uffd missing/minor handling zram: always expose rw_page LoongArch: update local TLB if PTE entry exists mm: use update_mmu_tlb() on the second thread kasan: fix array-bounds warnings in tests hmm-tests: add test for migrate_device_range() nouveau/dmem: evict device private memory during release nouveau/dmem: refactor nouveau_dmem_fault_copy_one() mm/migrate_device.c: add migrate_device_range() mm/migrate_device.c: refactor migrate_vma and migrate_deivce_coherent_page() mm/memremap.c: take a pgmap reference on page allocation mm: free device private pages have zero refcount mm/memory.c: fix race when faulting a device private page mm/damon: use damon_sz_region() in appropriate place mm/damon: move sz_damon_region to damon_sz_region lib/test_meminit: add checks for the allocation functions ...