Low priority tasks (e.g., SCHED_OTHER) can suffer starvation if tasks
with higher priority (e.g., SCHED_FIFO) monopolize CPU(s).
RT Throttling has been introduced a while ago as a (mostly debug)
countermeasure one can utilize to reserve some CPU time for low priority
tasks (usually background type of work, e.g. workqueues, timers, etc.).
It however has its own problems (see documentation) and the undesired
effect of unconditionally throttling FIFO tasks even when no lower
priority activity needs to run (there are mechanisms to fix this issue
as well, but, again, with their own problems).
Introduce deadline servers to service low priority tasks needs under
starvation conditions. Deadline servers are built extending SCHED_DEADLINE
implementation to allow 2-level scheduling (a sched_deadline entity
becomes a container for lower priority scheduling entities).
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Daniel Bristot de Oliveira <bristot@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/4968601859d920335cf85822eb573a5f179f04b8.1699095159.git.bristot@kernel.org
Now that trace_sched_stat_runtime() no longer takes a vruntime
argument, the task specific bits are identical between
update_curr_common() and update_curr().
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tracing the runtime delta makes sense, observer can sum over time.
Tracing the absolute vruntime makes less sense, inconsistent:
absolute-vs-delta, but also vruntime delta can be computed from
runtime delta.
Removing the vruntime thing also makes the two tracepoint sites
identical, allowing to unify the code in a later patch.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Since the RB-tree is now sorted by deadline, let's first try the
leftmost entity which has the earliest virtual deadline. I've done
some benchmarks to see its effectiveness.
All the benchmarks are done inside a normal cpu cgroup in a clean
environment with cpu turbo disabled, on a dual-CPU Intel Xeon(R)
Platinum 8260 with 2 NUMA nodes each of which has 24C/48T.
hackbench: process/thread + pipe/socket + 1/2/4/8 groups
netperf: TCP/UDP + STREAM/RR + 24/48/72/96/192 threads
tbench: loopback 24/48/72/96/192 threads
schbench: 1/2/4/8 mthreads
direct: cfs_rq has only one entity
parity: RUN_TO_PARITY
fast: O(1) fastpath
slow: heap search
(%) direct parity fast slow
hackbench 92.95 2.02 4.91 0.12
netperf 68.08 6.60 24.18 1.14
tbench 67.55 11.22 20.61 0.62
schbench 69.91 2.65 25.73 1.71
The above results indicate that this fastpath really makes task
selection more efficient.
Signed-off-by: Abel Wu <wuyun.abel@bytedance.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20231115033647.80785-4-wuyun.abel@bytedance.com
Since commit fc137c0dda ("sched/numa: enhance vma scanning logic")
NUMA Balancing allows updating PTEs to trap NUMA hinting faults if the
task had previously accessed VMA. However unconditional scan of VMAs are
allowed during initial phase of VMA creation until process's
mm numa_scan_seq reaches 2 even though current task had not accessed VMA.
Rationale:
- Without initial scan subsequent PTE update may never happen.
- Give fair opportunity to all the VMAs to be scanned and subsequently
understand the access pattern of all the VMAs.
But it has a corner case where, if a VMA is created after some time,
process's mm numa_scan_seq could be already greater than 2.
For e.g., values of mm numa_scan_seq when VMAs are created by running
mmtest autonuma benchmark briefly looks like:
start_seq=0 : 459
start_seq=2 : 138
start_seq=3 : 144
start_seq=4 : 8
start_seq=8 : 1
start_seq=9 : 1
This results in no unconditional PTE updates for those VMAs created after
some time.
Fix:
- Note down the initial value of mm numa_scan_seq in per VMA start_seq.
- Allow unconditional scan till start_seq + 2.
Result:
SUT: AMD EPYC Milan with 2 NUMA nodes 256 cpus.
base kernel: upstream 6.6-rc6 with Mels patches [1] applied.
kernbench
========== base patched %gain
Amean elsp-128 165.09 ( 0.00%) 164.78 * 0.19%*
Duration User 41404.28 41375.08
Duration System 9862.22 9768.48
Duration Elapsed 519.87 518.72
Ops NUMA PTE updates 1041416.00 831536.00
Ops NUMA hint faults 263296.00 220966.00
Ops NUMA pages migrated 258021.00 212769.00
Ops AutoNUMA cost 1328.67 1114.69
autonumabench
NUMA01_THREADLOCAL
==================
Amean elsp-NUMA01_THREADLOCAL 81.79 (0.00%) 67.74 * 17.18%*
Duration User 54832.73 47379.67
Duration System 75.00 185.75
Duration Elapsed 576.72 476.09
Ops NUMA PTE updates 394429.00 11121044.00
Ops NUMA hint faults 1001.00 8906404.00
Ops NUMA pages migrated 288.00 2998694.00
Ops AutoNUMA cost 7.77 44666.84
Signed-off-by: Raghavendra K T <raghavendra.kt@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Mel Gorman <mgorman@suse.de>
Link: https://lore.kernel.org/r/2ea7cbce80ac7c62e90cbfb9653a7972f902439f.1697816692.git.raghavendra.kt@amd.com
should_we_balance is called for the decision to do load-balancing.
When sched ticks invoke this function, only one CPU should return
true. However, in the current code, two CPUs can return true. The
following situation, where b means busy and i means idle, is an
example, because CPU 0 and CPU 2 return true.
[0, 1] [2, 3]
b b i b
This fix checks if there exists an idle CPU with busy sibling(s)
after looking for a CPU on an idle core. If some idle CPUs with busy
siblings are found, just the first one should do load-balancing.
Fixes: b1bfeab9b0 ("sched/fair: Consider the idle state of the whole core for load balance")
Signed-off-by: Keisuke Nishimura <keisuke.nishimura@inria.fr>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Chen Yu <yu.c.chen@intel.com>
Reviewed-by: Shrikanth Hegde <sshegde@linux.vnet.ibm.com>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/20231031133821.1570861-1-keisuke.nishimura@inria.fr
519fabc7aa ("psi: remove 500ms min window size limitation for
triggers") breaks unprivileged psi polling on cgroups.
Historically, we had a privilege check for polling in the open() of a
pressure file in /proc, but were erroneously missing it for the open()
of cgroup pressure files.
When unprivileged polling was introduced in d82caa2735 ("sched/psi:
Allow unprivileged polling of N*2s period"), it needed to filter
privileges depending on the exact polling parameters, and as such
moved the CAP_SYS_RESOURCE check from the proc open() callback to
psi_trigger_create(). Both the proc files as well as cgroup files go
through this during write(). This implicitly added the missing check
for privileges required for HT polling for cgroups.
When 519fabc7aa ("psi: remove 500ms min window size limitation for
triggers") followed right after to remove further restrictions on the
RT polling window, it incorrectly assumed the cgroup privilege check
was still missing and added it to the cgroup open(), mirroring what we
used to do for proc files in the past.
As a result, unprivileged poll requests that would be supported now
get rejected when opening the cgroup pressure file for writing.
Remove the cgroup open() check. psi_trigger_create() handles it.
Fixes: 519fabc7aa ("psi: remove 500ms min window size limitation for triggers")
Reported-by: Luca Boccassi <bluca@debian.org>
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Luca Boccassi <bluca@debian.org>
Acked-by: Suren Baghdasaryan <surenb@google.com>
Cc: stable@vger.kernel.org # 6.5+
Link: https://lore.kernel.org/r/20231026164114.2488682-1-hannes@cmpxchg.org
vruntime of the (on_rq && !0-lag) entity needs to be adjusted when
it gets re-weighted, and the calculations can be simplified based
on the fact that re-weight won't change the w-average of all the
entities. Please check the proofs in comments.
But adjusting vruntime can also cause position change in RB-tree
hence require re-queue to fix up which might be costly. This might
be avoided by deferring adjustment to the time the entity actually
leaves tree (dequeue/pick), but that will negatively affect task
selection and probably not good enough either.
Fixes: 147f3efaa2 ("sched/fair: Implement an EEVDF-like scheduling policy")
Signed-off-by: Abel Wu <wuyun.abel@bytedance.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20231107090510.71322-2-wuyun.abel@bytedance.com
Pull parisc architecture fixes from Helge Deller:
- Include the upper 5 address bits when inserting TLB entries on a
64-bit kernel.
On physical machines those are ignored, but in qemu it's nice to have
them included and to be correct.
- Stop the 64-bit kernel and show a warning if someone tries to boot on
a machine with a 32-bit CPU
- Fix a "no previous prototype" warning in parport-gsc
* tag 'parisc-for-6.7-rc1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux:
parisc: Prevent booting 64-bit kernels on PA1.x machines
parport: gsc: mark init function static
parisc/pgtable: Do not drop upper 5 address bits of physical address
Pull LoongArch updates from Huacai Chen:
- support PREEMPT_DYNAMIC with static keys
- relax memory ordering for atomic operations
- support BPF CPU v4 instructions for LoongArch
- some build and runtime warning fixes
* tag 'loongarch-6.7' of git://git.kernel.org/pub/scm/linux/kernel/git/chenhuacai/linux-loongson:
selftests/bpf: Enable cpu v4 tests for LoongArch
LoongArch: BPF: Support signed mod instructions
LoongArch: BPF: Support signed div instructions
LoongArch: BPF: Support 32-bit offset jmp instructions
LoongArch: BPF: Support unconditional bswap instructions
LoongArch: BPF: Support sign-extension mov instructions
LoongArch: BPF: Support sign-extension load instructions
LoongArch: Add more instruction opcodes and emit_* helpers
LoongArch/smp: Call rcutree_report_cpu_starting() earlier
LoongArch: Relax memory ordering for atomic operations
LoongArch: Mark __percpu functions as always inline
LoongArch: Disable module from accessing external data directly
LoongArch: Support PREEMPT_DYNAMIC with static keys
Pull powerpc fixes from Michael Ellerman:
- Finish a refactor of pgprot_framebuffer() which dependend
on some changes that were merged via the drm tree
- Fix some kernel-doc warnings to quieten the bots
Thanks to Nathan Lynch and Thomas Zimmermann.
* tag 'powerpc-6.7-2' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
powerpc/rtas: Fix ppc_rtas_rmo_buf_show() kernel-doc
powerpc/pseries/rtas-work-area: Fix rtas_work_area_reserve_arena() kernel-doc
powerpc/fb: Call internal __phys_mem_access_prot() in fbdev code
powerpc: Remove file parameter from phys_mem_access_prot()
powerpc/machdep: Remove trailing whitespaces
Pull smb client fixes from Steve French:
- ctime caching fix (for setxattr)
- encryption fix
- DNS resolver mount fix
- debugging improvements
- multichannel fixes including cases where server stops or starts
supporting multichannel after mount
- reconnect fix
- minor cleanups
* tag '6.7-rc-smb3-client-fixes-part2' of git://git.samba.org/sfrench/cifs-2.6:
cifs: update internal module version number for cifs.ko
cifs: handle when server stops supporting multichannel
cifs: handle when server starts supporting multichannel
Missing field not being returned in ioctl CIFS_IOC_GET_MNT_INFO
smb3: allow dumping session and tcon id to improve stats analysis and debugging
smb: client: fix mount when dns_resolver key is not available
smb3: fix caching of ctime on setxattr
smb3: minor cleanup of session handling code
cifs: reconnect work should have reference on server struct
cifs: do not pass cifs_sb when trying to add channels
cifs: account for primary channel in the interface list
cifs: distribute channels across interfaces based on speed
cifs: handle cases where a channel is closed
smb3: more minor cleanups for session handling routines
smb3: minor RDMA cleanup
cifs: Fix encryption of cleared, but unset rq_iter data buffers
Pull probes fixes from Masami Hiramatsu:
- Documentation update: Add a note about argument and return value
fetching is the best effort because it depends on the type.
- objpool: Fix to make internal global variables static in
test_objpool.c.
- kprobes: Unify kprobes_exceptions_nofify() prototypes. There are the
same prototypes in asm/kprobes.h for some architectures, but some of
them are missing the prototype and it causes a warning. So move the
prototype into linux/kprobes.h.
- tracing: Fix to check the tracepoint event and return event at
parsing stage. The tracepoint event doesn't support %return but if
$retval exists, it will be converted to %return silently. This finds
that case and rejects it.
- tracing: Fix the order of the descriptions about the parameters of
__kprobe_event_gen_cmd_start() to be consistent with the argument
list of the function.
* tag 'probes-fixes-v6.7-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:
tracing/kprobes: Fix the order of argument descriptions
tracing: fprobe-event: Fix to check tracepoint event and return
kprobes: unify kprobes_exceptions_nofify() prototypes
lib: test_objpool: make global variables static
Documentation: tracing: Add a note about argument and retval access
Pull fbdev fixes and cleanups from Helge Deller:
- fix double free and resource leaks in imsttfb
- lots of remove callback cleanups and section mismatch fixes in
omapfb, amifb and atmel_lcdfb
- error code fix and memparse simplification in omapfb
* tag 'fbdev-for-6.7-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/deller/linux-fbdev: (31 commits)
fbdev: fsl-diu-fb: mark wr_reg_wa() static
fbdev: amifb: Convert to platform remove callback returning void
fbdev: amifb: Mark driver struct with __refdata to prevent section mismatch warning
fbdev: hyperv_fb: fix uninitialized local variable use
fbdev: omapfb/tpd12s015: Convert to platform remove callback returning void
fbdev: omapfb/tfp410: Convert to platform remove callback returning void
fbdev: omapfb/sharp-ls037v7dw01: Convert to platform remove callback returning void
fbdev: omapfb/opa362: Convert to platform remove callback returning void
fbdev: omapfb/hdmi: Convert to platform remove callback returning void
fbdev: omapfb/dvi: Convert to platform remove callback returning void
fbdev: omapfb/dsi-cm: Convert to platform remove callback returning void
fbdev: omapfb/dpi: Convert to platform remove callback returning void
fbdev: omapfb/analog-tv: Convert to platform remove callback returning void
fbdev: atmel_lcdfb: Convert to platform remove callback returning void
fbdev: omapfb/tpd12s015: Don't put .remove() in .exit.text and drop suppress_bind_attrs
fbdev: omapfb/tfp410: Don't put .remove() in .exit.text and drop suppress_bind_attrs
fbdev: omapfb/sharp-ls037v7dw01: Don't put .remove() in .exit.text and drop suppress_bind_attrs
fbdev: omapfb/opa362: Don't put .remove() in .exit.text and drop suppress_bind_attrs
fbdev: omapfb/hdmi: Don't put .remove() in .exit.text and drop suppress_bind_attrs
fbdev: omapfb/dvi: Don't put .remove() in .exit.text and drop suppress_bind_attrs
...
Pull drm fixes from Daniel Vetter:
"Dave's VPN to the big machine died, so it's on me to do fixes pr this
and next week while everyone else is at plumbers.
- big pile of amd fixes, but mostly for hw support newly added in 6.7
- i915 fixes, mostly minor things
- qxl memory leak fix
- vc4 uaf fix in mock helpers
- syncobj fix for DRM_SYNCOBJ_WAIT_FLAGS_WAIT_AVAILABLE"
* tag 'drm-next-2023-11-10' of git://anongit.freedesktop.org/drm/drm: (78 commits)
drm/amdgpu: fix error handling in amdgpu_vm_init
drm/amdgpu: Fix possible null pointer dereference
drm/amdgpu: move UVD and VCE sched entity init after sched init
drm/amdgpu: move kfd_resume before the ip late init
drm/amd: Explicitly check for GFXOFF to be enabled for s0ix
drm/amdgpu: Change WREG32_RLC to WREG32_SOC15_RLC where inst != 0 (v2)
drm/amdgpu: Use correct KIQ MEC engine for gfx9.4.3 (v5)
drm/amdgpu: add smu v13.0.6 pcs xgmi ras error query support
drm/amdgpu: fix software pci_unplug on some chips
drm/amd/display: remove duplicated argument
drm/amdgpu: correct mca debugfs dump reg list
drm/amdgpu: correct acclerator check architecutre dump
drm/amdgpu: add pcs xgmi v6.4.0 ras support
drm/amdgpu: Change extended-scope MTYPE on GC 9.4.3
drm/amdgpu: disable smu v13.0.6 mca debug mode by default
drm/amdgpu: Support multiple error query modes
drm/amdgpu: refine smu v13.0.6 mca dump driver
drm/amdgpu: Do not program PF-only regs in hdp_v4_0.c under SRIOV (v2)
drm/amdgpu: Skip PCTL0_MMHUB_DEEPSLEEP_IB write in jpegv4.0.3 under SRIOV
drm: amd: Resolve Sphinx unexpected indentation warning
...
Pull arm64 fixes from Catalin Marinas:
"Mostly PMU fixes and a reworking of the pseudo-NMI disabling on broken
MediaTek firmware:
- Move the MediaTek GIC quirk handling from irqchip to core. Before
the merging window commit 44bd78dd2b ("irqchip/gic-v3: Disable
pseudo NMIs on MediaTek devices w/ firmware issues") temporarily
addressed this issue. Fixed now at a deeper level in the arch code
- Reject events meant for other PMUs in the CoreSight PMU driver,
otherwise some of the core PMU events would disappear
- Fix the Armv8 PMUv3 driver driver to not truncate 64-bit registers,
causing some events to be invisible
- Remove duplicate declaration of __arm64_sys##name following the
patch to avoid prototype warning for syscalls
- Typos in the elf_hwcap documentation"
* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
arm64/syscall: Remove duplicate declaration
Revert "arm64: smp: avoid NMI IPIs with broken MediaTek FW"
arm64: Move MediaTek GIC quirk handling from irqchip to core
arm64/arm: arm_pmuv3: perf: Don't truncate 64-bit registers
perf: arm_cspmu: Reject events meant for other PMUs
Documentation/arm64: Fix typos in elf_hwcaps
