| Age | Commit message (Collapse) | Author |
|---|
|
The mem_cgroup_size helper is used only in apply_proportional_protection
to read the current memory usage. Its semantics are unclear and
inconsistent with other sites, which directly call page_counter_read for
the same purpose.
Remove this helper and get its usage via mem_cgroup_protection for
clarity. Additionally, rename the local variable 'cgroup_size' to 'usage'
to better reflect its meaning.
No functional changes intended.
Link: https://lkml.kernel.org/r/20251211013019.2080004-3-chenridong@huaweicloud.com
Signed-off-by: Chen Ridong <chenridong@huawei.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Michal Hocko <mhocko@suse.com>
Acked-by: Shakeel Butt <shakeel.butt@linux.dev>
Cc: Michal Koutný <mkoutny@suse.com>
Cc: Axel Rasmussen <axelrasmussen@google.com>
Cc: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Cc: Lu Jialin <lujialin4@huawei.com>
Cc: Muchun Song <muchun.song@linux.dev>
Cc: Qi Zheng <zhengqi.arch@bytedance.com>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Wei Xu <weixugc@google.com>
Cc: Yuanchu Xie <yuanchu@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
Use batch clearing in clear_contig_highpages() instead of clearing a
single page at a time. Exposing larger ranges enables the processor to
optimize based on extent.
To do this we just switch to using clear_user_highpages() which would in
turn use clear_user_pages() or clear_pages().
Batched clearing, when running under non-preemptible models, however, has
latency considerations. In particular, we need periodic invocations of
cond_resched() to keep to reasonable preemption latencies. This is a
problem because the clearing primitives do not, or might not be able to,
call cond_resched() to check if preemption is needed.
So, limit the worst case preemption latency by doing the clearing in units
of no more than PROCESS_PAGES_NON_PREEMPT_BATCH pages. (Preemptible
models already define away most of cond_resched(), so the batch size is
ignored when running under those.)
PROCESS_PAGES_NON_PREEMPT_BATCH: for architectures with "fast" clear-pages
(ones that define clear_pages()), we define it as 32MB worth of pages.
This is meant to be large enough to allow the processor to optimize the
operation and yet small enough that we see reasonable preemption latency
for when this optimization is not possible (ex. slow microarchitectures,
memory bandwidth saturation.)
This specific value also allows for a cacheline allocation elision
optimization (which might help unrelated applications by not evicting
potentially useful cache lines) that kicks in recent generations of AMD
Zen processors at around LLC-size (32MB is a typical size).
At the same time 32MB is small enough that even with poor clearing
bandwidth (say ~10GBps), time to clear 32MB should be well below the
scheduler's default warning threshold
(sysctl_resched_latency_warn_ms=100).
"Slow" architectures (don't have clear_pages()) will continue to use the
base value (single page).
Performance
==
Testing a demand fault workload shows a decent improvement in bandwidth
with pg-sz=1GB. Bandwidth with pg-sz=2MB stays flat.
$ perf bench mem mmap -p $pg-sz -f demand -s 64GB -l 5
contiguous-pages batched-pages
(GBps +- %stdev) (GBps +- %stdev)
pg-sz=2MB 23.58 +- 1.95% 25.34 +- 1.18% + 7.50% preempt=*
pg-sz=1GB 25.09 +- 0.79% 39.22 +- 2.32% + 56.31% preempt=none|voluntary
pg-sz=1GB 25.71 +- 0.03% 52.73 +- 0.20% [#] +110.16% preempt=full|lazy
[#] We perform much better with preempt=full|lazy because, not
needing explicit invocations of cond_resched() we can clear the
full extent (pg-sz=1GB) as a single unit which the processor
can optimize for.
(Unless otherwise noted, all numbers are on AMD Genoa (EPYC 9J13);
region-size=64GB, local node; 2.56 GHz, boost=0.)
Analysis
==
pg-sz=1GB: the improvement we see falls in two buckets depending on the
batch size in use.
For batch-size=32MB the number of cachelines allocated (L1-dcache-loads)
-- which stay relatively flat for smaller batches, start to drop off
because cacheline allocation elision kicks in. And as can be seen below,
at batch-size=1GB, we stop allocating cachelines almost entirely. (Not
visible here but from testing with intermediate sizes, the allocation
change kicks in only at batch-size=32MB and ramps up from there.)
contigous-pages 6,949,417,798 L1-dcache-loads # 883.599 M/sec ( +- 0.01% ) (35.75%)
3,226,709,573 L1-dcache-load-misses # 46.43% of all L1-dcache accesses ( +- 0.05% ) (35.75%)
batched,32MB 2,290,365,772 L1-dcache-loads # 471.171 M/sec ( +- 0.36% ) (35.72%)
1,144,426,272 L1-dcache-load-misses # 49.97% of all L1-dcache accesses ( +- 0.58% ) (35.70%)
batched,1GB 63,914,157 L1-dcache-loads # 17.464 M/sec ( +- 8.08% ) (35.73%)
22,074,367 L1-dcache-load-misses # 34.54% of all L1-dcache accesses ( +- 16.70% ) (35.70%)
The dropoff is also visible in L2 prefetch hits (miss numbers are
on similar lines):
contiguous-pages 3,464,861,312 l2_pf_hit_l2.all # 437.722 M/sec ( +- 0.74% ) (15.69%)
batched,32MB 883,750,087 l2_pf_hit_l2.all # 181.223 M/sec ( +- 1.18% ) (15.71%)
batched,1GB 8,967,943 l2_pf_hit_l2.all # 2.450 M/sec ( +- 17.92% ) (15.77%)
This largely decouples the frontend from the backend since the clearing
operation does not need to wait on loads from memory (we still need
cacheline ownership but that's a shorter path). This is most visible if
we rerun the test above with (boost=1, 3.66 GHz).
$ perf bench mem mmap -p $pg-sz -f demand -s 64GB -l 5
contiguous-pages batched-pages
(GBps +- %stdev) (GBps +- %stdev)
pg-sz=2MB 26.08 +- 1.72% 26.13 +- 0.92% - preempt=*
pg-sz=1GB 26.99 +- 0.62% 48.85 +- 2.19% + 80.99% preempt=none|voluntary
pg-sz=1GB 27.69 +- 0.18% 75.18 +- 0.25% +171.50% preempt=full|lazy
Comparing the batched-pages numbers from the boost=0 ones and these: for a
clock-speed gain of 42% we gain 24.5% for batch-size=32MB and 42.5% for
batch-size=1GB. In comparison the baseline contiguous-pages case and both
the pg-sz=2MB ones are largely backend bound so gain no more than ~10%.
Other platforms tested, Intel Icelakex (Oracle X9) and ARM64 Neoverse-N1
(Ampere Altra) both show an improvement of ~35% for pg-sz=2MB|1GB. The
first goes from around 8GBps to 11GBps and the second from 32GBps to 44
GBPs.
[ankur.a.arora@oracle.com: move the unit computation and make it a const
Link: https://lkml.kernel.org/r/20260108060406.1693853-1-ankur.a.arora@oracle.com
Link: https://lkml.kernel.org/r/20260107072009.1615991-8-ankur.a.arora@oracle.com
Signed-off-by: Ankur Arora <ankur.a.arora@oracle.com>
Acked-by: David Hildenbrand (Red Hat) <david@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: "Borislav Petkov (AMD)" <bp@alien8.de>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Konrad Rzessutek Wilk <konrad.wilk@oracle.com>
Cc: Lance Yang <ioworker0@gmail.com>
Cc: "Liam R. Howlett" <Liam.Howlett@oracle.com>
Cc: Li Zhe <lizhe.67@bytedance.com>
Cc: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Cc: Mateusz Guzik <mjguzik@gmail.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Raghavendra K T <raghavendra.kt@amd.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
Define clear_user_highpages() which uses the range clearing primitive,
clear_user_pages(). We can safely use this when CONFIG_HIGHMEM is
disabled and if the architecture does not have clear_user_highpage.
The first is needed to ensure that contiguous page ranges stay contiguous
which precludes intermediate maps via HIGMEM. The second, because if the
architecture has clear_user_highpage(), it likely needs flushing magic
when clearing the page, magic that we aren't privy to.
For both of those cases, just fallback to a loop around
clear_user_highpage().
Link: https://lkml.kernel.org/r/20260107072009.1615991-4-ankur.a.arora@oracle.com
Signed-off-by: Ankur Arora <ankur.a.arora@oracle.com>
Acked-by: David Hildenbrand (Red Hat) <david@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: "Borislav Petkov (AMD)" <bp@alien8.de>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Konrad Rzessutek Wilk <konrad.wilk@oracle.com>
Cc: Lance Yang <ioworker0@gmail.com>
Cc: "Liam R. Howlett" <Liam.Howlett@oracle.com>
Cc: Li Zhe <lizhe.67@bytedance.com>
Cc: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Cc: Mateusz Guzik <mjguzik@gmail.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Raghavendra K T <raghavendra.kt@amd.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
Introduce clear_pages(), to be overridden by architectures that support
more efficient clearing of consecutive pages.
Also introduce clear_user_pages(), however, we will not expect this
function to be overridden anytime soon.
As we do for clear_user_page(), define clear_user_pages() only if the
architecture does not define clear_user_highpage().
That is because if the architecture does define clear_user_highpage(),
then it likely needs some flushing magic when clearing user pages or
highpages. This means we can get away without defining
clear_user_pages(), since, much like its single page sibling, its only
potential user is the generic clear_user_highpages() which should instead
be using clear_user_highpage().
Link: https://lkml.kernel.org/r/20260107072009.1615991-3-ankur.a.arora@oracle.com
Signed-off-by: Ankur Arora <ankur.a.arora@oracle.com>
Acked-by: David Hildenbrand (Red Hat) <david@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: "Borislav Petkov (AMD)" <bp@alien8.de>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Konrad Rzessutek Wilk <konrad.wilk@oracle.com>
Cc: Lance Yang <ioworker0@gmail.com>
Cc: "Liam R. Howlett" <Liam.Howlett@oracle.com>
Cc: Li Zhe <lizhe.67@bytedance.com>
Cc: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Cc: Mateusz Guzik <mjguzik@gmail.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Raghavendra K T <raghavendra.kt@amd.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
Patch series "mm: folio_zero_user: clear page ranges", v11.
This series adds clearing of contiguous page ranges for hugepages.
The series improves on the current discontiguous clearing approach in two
ways:
- clear pages in a contiguous fashion.
- use batched clearing via clear_pages() wherever exposed.
The first is useful because it allows us to make much better use of
hardware prefetchers.
The second, enables advertising the real extent to the processor. Where
specific instructions support it (ex. string instructions on x86; "mops"
on arm64 etc), a processor can optimize based on this because, instead of
seeing a sequence of 8-byte stores, or a sequence of 4KB pages, it sees a
larger unit being operated on.
For instance, AMD Zen uarchs (for extents larger than LLC-size) switch to
a mode where they start eliding cacheline allocation. This is helpful not
just because it results in higher bandwidth, but also because now the
cache is not evicting useful cachelines and replacing them with zeroes.
Demand faulting a 64GB region shows performance improvement:
$ perf bench mem mmap -p $pg-sz -f demand -s 64GB -l 5
baseline +series
(GBps +- %stdev) (GBps +- %stdev)
pg-sz=2MB 11.76 +- 1.10% 25.34 +- 1.18% [*] +115.47% preempt=*
pg-sz=1GB 24.85 +- 2.41% 39.22 +- 2.32% + 57.82% preempt=none|voluntary
pg-sz=1GB (similar) 52.73 +- 0.20% [#] +112.19% preempt=full|lazy
[*] This improvement is because switching to sequential clearing
allows the hardware prefetchers to do a much better job.
[#] For pg-sz=1GB a large part of the improvement is because of the
cacheline elision mentioned above. preempt=full|lazy improves upon
that because, not needing explicit invocations of cond_resched() to
ensure reasonable preemption latency, it can clear the full extent
as a single unit. In comparison the maximum extent used for
preempt=none|voluntary is PROCESS_PAGES_NON_PREEMPT_BATCH (32MB).
When provided the full extent the processor forgoes allocating
cachelines on this path almost entirely.
(The hope is that eventually, in the fullness of time, the lazy
preemption model will be able to do the same job that none or
voluntary models are used for, allowing us to do away with
cond_resched().)
Raghavendra also tested previous version of the series on AMD Genoa and
sees similar improvement [1] with preempt=lazy.
$ perf bench mem map -p $page-size -f populate -s 64GB -l 10
base patched change
pg-sz=2MB 12.731939 GB/sec 26.304263 GB/sec 106.6%
pg-sz=1GB 26.232423 GB/sec 61.174836 GB/sec 133.2%
This patch (of 8):
Let's drop all variants that effectively map to clear_page() and provide
it in a generic variant instead.
We'll use the macro clear_user_page to indicate whether an architecture
provides it's own variant.
Also, clear_user_page() is only called from the generic variant of
clear_user_highpage(), so define it only if the architecture does not
provide a clear_user_highpage(). And, for simplicity define it in
linux/highmem.h.
Note that for parisc, clear_page() and clear_user_page() map to
clear_page_asm(), so we can just get rid of the custom clear_user_page()
implementation. There is a clear_user_page_asm() function on parisc, that
seems to be unused. Not sure what's up with that.
Link: https://lkml.kernel.org/r/20260107072009.1615991-1-ankur.a.arora@oracle.com
Link: https://lkml.kernel.org/r/20260107072009.1615991-2-ankur.a.arora@oracle.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Co-developed-by: Ankur Arora <ankur.a.arora@oracle.com>
Signed-off-by: Ankur Arora <ankur.a.arora@oracle.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Ankur Arora <ankur.a.arora@oracle.com>
Cc: "Borislav Petkov (AMD)" <bp@alien8.de>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: David Hildenbrand <david@kernel.org>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Konrad Rzessutek Wilk <konrad.wilk@oracle.com>
Cc: Lance Yang <ioworker0@gmail.com>
Cc: "Liam R. Howlett" <Liam.Howlett@oracle.com>
Cc: Li Zhe <lizhe.67@bytedance.com>
Cc: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Cc: Mateusz Guzik <mjguzik@gmail.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Raghavendra K T <raghavendra.kt@amd.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
Despite recent efforts to prevent lazy_mmu sections from nesting, it
remains difficult to ensure that it never occurs - and in fact it does
occur on arm64 in certain situations (CONFIG_DEBUG_PAGEALLOC). Commit
1ef3095b1405 ("arm64/mm: Permit lazy_mmu_mode to be nested") made nesting
tolerable on arm64, but without truly supporting it: the inner call to
leave() disables the batching optimisation before the outer section ends.
This patch actually enables lazy_mmu sections to nest by tracking the
nesting level in task_struct, in a similar fashion to e.g.
pagefault_{enable,disable}(). This is fully handled by the generic
lazy_mmu helpers that were recently introduced.
lazy_mmu sections were not initially intended to nest, so we need to
clarify the semantics w.r.t. the arch_*_lazy_mmu_mode() callbacks. This
patch takes the following approach:
* The outermost calls to lazy_mmu_mode_{enable,disable}() trigger
calls to arch_{enter,leave}_lazy_mmu_mode() - this is unchanged.
* Nested calls to lazy_mmu_mode_{enable,disable}() are not forwarded
to the arch via arch_{enter,leave} - lazy MMU remains enabled so
the assumption is that these callbacks are not relevant. However,
existing code may rely on a call to disable() to flush any batched
state, regardless of nesting. arch_flush_lazy_mmu_mode() is
therefore called in that situation.
A separate interface was recently introduced to temporarily pause the lazy
MMU mode: lazy_mmu_mode_{pause,resume}(). pause() fully exits the mode
*regardless of the nesting level*, and resume() restores the mode at the
same nesting level.
pause()/resume() are themselves allowed to nest, so we actually store two
nesting levels in task_struct: enable_count and pause_count. A new helper
is_lazy_mmu_mode_active() is introduced to determine whether we are
currently in lazy MMU mode; this will be used in subsequent patches to
replace the various ways arch's currently track whether the mode is
enabled.
In summary (enable/pause represent the values *after* the call):
lazy_mmu_mode_enable() -> arch_enter() enable=1 pause=0
lazy_mmu_mode_enable() -> ø enable=2 pause=0
lazy_mmu_mode_pause() -> arch_leave() enable=2 pause=1
lazy_mmu_mode_resume() -> arch_enter() enable=2 pause=0
lazy_mmu_mode_disable() -> arch_flush() enable=1 pause=0
lazy_mmu_mode_disable() -> arch_leave() enable=0 pause=0
Note: is_lazy_mmu_mode_active() is added to <linux/sched.h> to allow
arch headers included by <linux/pgtable.h> to use it.
Link: https://lkml.kernel.org/r/20251215150323.2218608-10-kevin.brodsky@arm.com
Signed-off-by: Kevin Brodsky <kevin.brodsky@arm.com>
Acked-by: David Hildenbrand (Red Hat) <david@kernel.org>
Reviewed-by: Yeoreum Yun <yeoreum.yun@arm.com>
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Cc: Andreas Larsson <andreas@gaisler.com>
Cc: Anshuman Khandual <anshuman.khandual@arm.com>
Cc: Borislav Betkov <bp@alien8.de>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christophe Leroy <christophe.leroy@csgroup.eu>
Cc: David Hildenbrand <david@redhat.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: David Woodhouse <dwmw2@infradead.org>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jann Horn <jannh@google.com>
Cc: Juegren Gross <jgross@suse.com>
Cc: Liam Howlett <liam.howlett@oracle.com>
Cc: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Cc: Madhavan Srinivasan <maddy@linux.ibm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
Cc: Ryan Roberts <ryan.roberts@arm.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Thomas Gleinxer <tglx@linutronix.de>
Cc: Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Will Deacon <will@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
The lazy MMU mode cannot be used in interrupt context. This is documented
in <linux/pgtable.h>, but isn't consistently handled across architectures.
arm64 ensures that calls to lazy_mmu_mode_* have no effect in interrupt
context, because such calls do occur in certain configurations - see
commit b81c688426a9 ("arm64/mm: Disable barrier batching in interrupt
contexts"). Other architectures do not check this situation, most likely
because it hasn't occurred so far.
Let's handle this in the new generic lazy_mmu layer, in the same fashion
as arm64: bail out of lazy_mmu_mode_* if in_interrupt(). Also remove the
arm64 handling that is now redundant.
Both arm64 and x86/Xen also ensure that any lazy MMU optimisation is
disabled while in interrupt (see queue_pte_barriers() and
xen_get_lazy_mode() respectively). This will be handled in the generic
layer in a subsequent patch.
Link: https://lkml.kernel.org/r/20251215150323.2218608-9-kevin.brodsky@arm.com
Signed-off-by: Kevin Brodsky <kevin.brodsky@arm.com>
Acked-by: David Hildenbrand (Red Hat) <david@kernel.org>
Reviewed-by: Anshuman Khandual <anshuman.khandual@arm.com>
Reviewed-by: Yeoreum Yun <yeoreum.yun@arm.com>
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Cc: Andreas Larsson <andreas@gaisler.com>
Cc: Borislav Betkov <bp@alien8.de>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christophe Leroy <christophe.leroy@csgroup.eu>
Cc: David Hildenbrand <david@redhat.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: David Woodhouse <dwmw2@infradead.org>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jann Horn <jannh@google.com>
Cc: Juegren Gross <jgross@suse.com>
Cc: Liam Howlett <liam.howlett@oracle.com>
Cc: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Cc: Madhavan Srinivasan <maddy@linux.ibm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
Cc: Ryan Roberts <ryan.roberts@arm.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Thomas Gleinxer <tglx@linutronix.de>
Cc: Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Will Deacon <will@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
The implementation of the lazy MMU mode is currently entirely
arch-specific; core code directly calls arch helpers:
arch_{enter,leave}_lazy_mmu_mode().
We are about to introduce support for nested lazy MMU sections. As things
stand we'd have to duplicate that logic in every arch implementing
lazy_mmu - adding to a fair amount of logic already duplicated across
lazy_mmu implementations.
This patch therefore introduces a new generic layer that calls the
existing arch_* helpers. Two pair of calls are introduced:
* lazy_mmu_mode_enable() ... lazy_mmu_mode_disable()
This is the standard case where the mode is enabled for a given
block of code by surrounding it with enable() and disable()
calls.
* lazy_mmu_mode_pause() ... lazy_mmu_mode_resume()
This is for situations where the mode is temporarily disabled
by first calling pause() and then resume() (e.g. to prevent any
batching from occurring in a critical section).
The documentation in <linux/pgtable.h> will be updated in a subsequent
patch.
No functional change should be introduced at this stage. The
implementation of enable()/resume() and disable()/pause() is currently
identical, but nesting support will change that.
Most of the call sites have been updated using the following Coccinelle
script:
@@
@@
{
...
- arch_enter_lazy_mmu_mode();
+ lazy_mmu_mode_enable();
...
- arch_leave_lazy_mmu_mode();
+ lazy_mmu_mode_disable();
...
}
@@
@@
{
...
- arch_leave_lazy_mmu_mode();
+ lazy_mmu_mode_pause();
...
- arch_enter_lazy_mmu_mode();
+ lazy_mmu_mode_resume();
...
}
A couple of notes regarding x86:
* Xen is currently the only case where explicit handling is required
for lazy MMU when context-switching. This is purely an
implementation detail and using the generic lazy_mmu_mode_*
functions would cause trouble when nesting support is introduced,
because the generic functions must be called from the current task.
For that reason we still use arch_leave() and arch_enter() there.
* x86 calls arch_flush_lazy_mmu_mode() unconditionally in a few
places, but only defines it if PARAVIRT_XXL is selected, and we
are removing the fallback in <linux/pgtable.h>. Add a new fallback
definition to <asm/pgtable.h> to keep things building.
Link: https://lkml.kernel.org/r/20251215150323.2218608-8-kevin.brodsky@arm.com
Signed-off-by: Kevin Brodsky <kevin.brodsky@arm.com>
Acked-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Anshuman Khandual <anshuman.khandual@arm.com>
Reviewed-by: Yeoreum Yun <yeoreum.yun@arm.com>
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Cc: Andreas Larsson <andreas@gaisler.com>
Cc: Borislav Betkov <bp@alien8.de>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christophe Leroy <christophe.leroy@csgroup.eu>
Cc: David Hildenbrand (Red Hat) <david@kernel.org>
Cc: David S. Miller <davem@davemloft.net>
Cc: David Woodhouse <dwmw2@infradead.org>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jann Horn <jannh@google.com>
Cc: Juegren Gross <jgross@suse.com>
Cc: Liam Howlett <liam.howlett@oracle.com>
Cc: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Cc: Madhavan Srinivasan <maddy@linux.ibm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
Cc: Ryan Roberts <ryan.roberts@arm.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Thomas Gleinxer <tglx@linutronix.de>
Cc: Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Will Deacon <will@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
Architectures currently opt in for implementing lazy_mmu helpers by
defining __HAVE_ARCH_ENTER_LAZY_MMU_MODE.
In preparation for introducing a generic lazy_mmu layer that will require
storage in task_struct, let's switch to a cleaner approach: instead of
defining a macro, select a CONFIG option.
This patch introduces CONFIG_ARCH_HAS_LAZY_MMU_MODE and has each arch
select it when it implements lazy_mmu helpers.
__HAVE_ARCH_ENTER_LAZY_MMU_MODE is removed and <linux/pgtable.h> relies on
the new CONFIG instead.
On x86, lazy_mmu helpers are only implemented if PARAVIRT_XXL is selected.
This creates some complications in arch/x86/boot/, because a few files
manually undefine PARAVIRT* options. As a result <asm/paravirt.h> does
not define the lazy_mmu helpers, but this breaks the build as
<linux/pgtable.h> only defines them if !CONFIG_ARCH_HAS_LAZY_MMU_MODE.
There does not seem to be a clean way out of this - let's just undefine
that new CONFIG too.
Link: https://lkml.kernel.org/r/20251215150323.2218608-7-kevin.brodsky@arm.com
Signed-off-by: Kevin Brodsky <kevin.brodsky@arm.com>
Acked-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
Reviewed-by: Ryan Roberts <ryan.roberts@arm.com>
Reviewed-by: Yeoreum Yun <yeoreum.yun@arm.com>
Acked-by: Andreas Larsson <andreas@gaisler.com> [sparc]
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Cc: Anshuman Khandual <anshuman.khandual@arm.com>
Cc: Borislav Betkov <bp@alien8.de>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christophe Leroy <christophe.leroy@csgroup.eu>
Cc: David Hildenbrand (Red Hat) <david@kernel.org>
Cc: David S. Miller <davem@davemloft.net>
Cc: David Woodhouse <dwmw2@infradead.org>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jann Horn <jannh@google.com>
Cc: Juegren Gross <jgross@suse.com>
Cc: Liam Howlett <liam.howlett@oracle.com>
Cc: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Cc: Madhavan Srinivasan <maddy@linux.ibm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Thomas Gleinxer <tglx@linutronix.de>
Cc: Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Will Deacon <will@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
The lazy MMU mode documentation makes clear that an implementation should
not assume that preemption is disabled or any lock is held upon entry to
the mode; however it says nothing about what code using the lazy MMU
interface should expect.
In practice sleeping is forbidden (for generic code) while the lazy MMU
mode is active: say it explicitly.
Link: https://lkml.kernel.org/r/20251215150323.2218608-6-kevin.brodsky@arm.com
Signed-off-by: Kevin Brodsky <kevin.brodsky@arm.com>
Reviewed-by: Yeoreum Yun <yeoreum.yun@arm.com>
Acked-by: David Hildenbrand (Red Hat) <david@kernel.org>
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Cc: Andreas Larsson <andreas@gaisler.com>
Cc: Anshuman Khandual <anshuman.khandual@arm.com>
Cc: Borislav Betkov <bp@alien8.de>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christophe Leroy <christophe.leroy@csgroup.eu>
Cc: David S. Miller <davem@davemloft.net>
Cc: David Woodhouse <dwmw2@infradead.org>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jann Horn <jannh@google.com>
Cc: Juegren Gross <jgross@suse.com>
Cc: Liam Howlett <liam.howlett@oracle.com>
Cc: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Cc: Madhavan Srinivasan <maddy@linux.ibm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
Cc: Ryan Roberts <ryan.roberts@arm.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Thomas Gleinxer <tglx@linutronix.de>
Cc: Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Will Deacon <will@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
HIPPI has not been relevant for over two decades. It was rapidly
eclipsed by Fibre Channel, and even when it was new, it was
confined to very high-end hardware. The HIPPI code has only
received tree-wide changes and fixes by inspection in the entire
Git history. Remove HIPPI support and the rrunner HIPPI driver,
and move the former maintainer to the CREDITS file. Keep the
include/uapi/linux/if_hippi.h header because it is used by the TUN
code, and to avoid breaking userspace, however unlikely that may be.
Signed-off-by: Ethan Nelson-Moore <enelsonmoore@gmail.com>
Link: https://patch.msgid.link/20260119022451.22344-1-enelsonmoore@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
tcp_rate_skb_delivered() is only called from tcp_input.c.
Move it there and make it static.
Both gcc and clang are (auto)inlining it, TCP performance
is increased at a small space cost.
$ scripts/bloat-o-meter -t vmlinux.old vmlinux.new
add/remove: 0/2 grow/shrink: 3/0 up/down: 509/-187 (322)
Function old new delta
tcp_sacktag_walk 1682 1867 +185
tcp_ack 5230 5405 +175
tcp_shifted_skb 437 586 +149
__pfx_tcp_rate_skb_delivered 16 - -16
tcp_rate_skb_delivered 171 - -171
Total: Before=22566192, After=22566514, chg +0.00%
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Neal Cardwell <ncardwell@google.com>
Link: https://patch.msgid.link/20260118123204.2315993-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Pavel Begunkov says:
====================
Add support for providers with large rx buffer
Many modern NICs support configurable receive buffer lengths, and zcrx and
memory providers can use buffers larger than 4K to improve performance.
When paired with hw-gro larger rx buffer sizes can drastically reduce
the number of buffers traversing the stack and save a lot of processing
time. It also allows to give to users larger contiguous chunks of data.
Single stream benchmarks showed up to ~30% CPU util improvement.
E.g. comparison for 4K vs 32K buffers using a 200Gbit NIC:
packets=23987040 (MB=2745098), rps=199559 (MB/s=22837)
CPU %usr %nice %sys %iowait %irq %soft %idle
0 1.53 0.00 27.78 2.72 1.31 66.45 0.22
packets=24078368 (MB=2755550), rps=200319 (MB/s=22924)
CPU %usr %nice %sys %iowait %irq %soft %idle
0 0.69 0.00 8.26 31.65 1.83 57.00 0.57
This series adds net infrastructure for memory providers configuring
the size and implements it for bnxt. It's an opt-in feature for drivers,
they should advertise support for the parameter in the qops and must check
if the hardware supports the given size. It's limited to memory providers
as it drastically simplifies implementation. It doesn't affect the fast
path zcrx uAPI, and the user exposed parameter is defined in zcrx terms,
which allows it to be flexible and adjusted in the future.
A liburing example can be found at [2]
full branch:
[1] https://github.com/isilence/linux.git zcrx/large-buffers-v8
Liburing example:
[2] https://github.com/isilence/liburing.git zcrx/rx-buf-len
* tag 'net-queue-rx-buf-len-v9' of https://github.com/isilence/linux:
io_uring/zcrx: document area chunking parameter
selftests: iou-zcrx: test large chunk sizes
eth: bnxt: support qcfg provided rx page size
eth: bnxt: adjust the fill level of agg queues with larger buffers
eth: bnxt: store rx buffer size per queue
net: pass queue rx page size from memory provider
net: add bare bone queue configs
net: reduce indent of struct netdev_queue_mgmt_ops members
net: memzero mp params when closing a queue
====================
Link: https://patch.msgid.link/
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
This reverts commit 77b9c4a438fc66e2ab004c411056b3fb71a54f2c, reversing
changes made to 4515ec4ad58a37e70a9e1256c0b993958c9b7497:
931420a2fc36 ("selftests/net: Add netkit container tests")
ab771c938d9a ("selftests/net: Make NetDrvContEnv support queue leasing")
6be87fbb2776 ("selftests/net: Add env for container based tests")
61d99ce3dfc2 ("selftests/net: Add bpf skb forwarding program")
920da3634194 ("netkit: Add xsk support for af_xdp applications")
eef51113f8af ("netkit: Add netkit notifier to check for unregistering devices")
b5ef109d22d4 ("netkit: Implement rtnl_link_ops->alloc and ndo_queue_create")
b5c3fa4a0b16 ("netkit: Add single device mode for netkit")
0073d2fd679d ("xsk: Proxy pool management for leased queues")
1ecea95dd3b5 ("xsk: Extend xsk_rcv_check validation")
804bf334d08a ("net: Proxy netdev_queue_get_dma_dev for leased queues")
0caa9a8ddec3 ("net: Proxy net_mp_{open,close}_rxq for leased queues")
ff8889ff9107 ("net, ethtool: Disallow leased real rxqs to be resized")
9e2103f36110 ("net: Add lease info to queue-get response")
31127deddef4 ("net: Implement netdev_nl_queue_create_doit")
a5546e18f77c ("net: Add queue-create operation")
The series will conflict with io_uring work, and the code needs more
polish.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
A kernel function bpf_foo marked with KF_IMPLICIT_ARGS flag is
expected to have two associated types in BTF:
* `bpf_foo` with a function prototype that omits implicit arguments
* `bpf_foo_impl` with a function prototype that matches the kernel
declaration of `bpf_foo`, but doesn't have a ksym associated with
its name
In order to support kfuncs with implicit arguments, the verifier has
to know how to resolve a call of `bpf_foo` to the correct BTF function
prototype and address.
To implement this, in add_kfunc_call() kfunc flags are checked for
KF_IMPLICIT_ARGS. For such kfuncs a BTF func prototype is adjusted to
the one found for `bpf_foo_impl` (func_name + "_impl" suffix, by
convention) function in BTF.
This effectively changes the signature of the `bpf_foo` kfunc in the
context of verification: from one without implicit args to the one
with full argument list.
The values of implicit arguments by design are provided by the
verifier, and so they can only be of particular types. In this patch
the only allowed implicit arg type is a pointer to struct
bpf_prog_aux.
In order for the verifier to correctly set an implicit bpf_prog_aux
arg value at runtime, is_kfunc_arg_prog() is extended to check for the
arg type. At a point when prog arg is determined in check_kfunc_args()
the kfunc with implicit args already has a prototype with full
argument list, so the existing value patch mechanism just works.
If a new kfunc with KF_IMPLICIT_ARG is declared for an existing kfunc
that uses a __prog argument (a legacy case), the prototype
substitution works in exactly the same way, assuming the kfunc follows
the _impl naming convention. The difference is only in how _impl
prototype is added to the BTF, which is not the verifier's
concern. See a subsequent resolve_btfids patch for details.
__prog suffix is still supported at this point, but will be removed in
a subsequent patch, after current users are moved to KF_IMPLICIT_ARGS.
Introduction of KF_IMPLICIT_ARGS revealed an issue with zero-extension
tracking, because an explicit rX = 0 in place of the verifier-supplied
argument is now absent if the arg is implicit (the BPF prog doesn't
pass a dummy NULL anymore). To mitigate this, reset the subreg_def of
all caller saved registers in check_kfunc_call() [1].
[1] https://lore.kernel.org/bpf/b4a760ef828d40dac7ea6074d39452bb0dc82caa.camel@gmail.com/
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Ihor Solodrai <ihor.solodrai@linux.dev>
Link: https://lore.kernel.org/r/20260120222638.3976562-4-ihor.solodrai@linux.dev
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
btf_kfunc_id_set_contains() is called by fetch_kfunc_meta() in the BPF
verifier to get the kfunc flags stored in the .BTF_ids ELF section.
If it returns NULL instead of a valid pointer, it's interpreted as an
illegal kfunc usage failing the verification.
There are two potential reasons for btf_kfunc_id_set_contains() to
return NULL:
1. Provided kfunc BTF id is not present in relevant kfunc id sets.
2. The kfunc is not allowed, as determined by the program type
specific filter [1].
The filter functions accept a pointer to `struct bpf_prog`, so they
might implicitly depend on earlier stages of verification, when
bpf_prog members are set.
For example, bpf_qdisc_kfunc_filter() in linux/net/sched/bpf_qdisc.c
inspects prog->aux->st_ops [2], which is initialized in:
check_attach_btf_id() -> check_struct_ops_btf_id()
So far this hasn't been an issue, because fetch_kfunc_meta() is the
only caller of btf_kfunc_id_set_contains().
However in subsequent patches of this series it is necessary to
inspect kfunc flags earlier in BPF verifier, in the add_kfunc_call().
To resolve this, refactor btf_kfunc_id_set_contains() into two
interface functions:
* btf_kfunc_flags() that simply returns pointer to kfunc_flags
without applying the filters
* btf_kfunc_is_allowed() that both checks for kfunc_flags existence
(which is a requirement for a kfunc to be allowed) and applies the
prog filters
See [3] for the previous version of this patch.
[1] https://lore.kernel.org/all/20230519225157.760788-7-aditi.ghag@isovalent.com/
[2] https://lore.kernel.org/all/20250409214606.2000194-4-ameryhung@gmail.com/
[3] https://lore.kernel.org/bpf/20251029190113.3323406-3-ihor.solodrai@linux.dev/
Reviewed-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Ihor Solodrai <ihor.solodrai@linux.dev>
Link: https://lore.kernel.org/r/20260120222638.3976562-2-ihor.solodrai@linux.dev
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
Pull misc fixes from Andrew Morton:
- A patch series from David Hildenbrand which fixes a few things
related to hugetlb PMD sharing
- The remainder are singletons, please see their changelogs for details
* tag 'mm-hotfixes-stable-2026-01-20-13-09' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm:
mm: restore per-memcg proactive reclaim with !CONFIG_NUMA
mm/kfence: fix potential deadlock in reboot notifier
Docs/mm/allocation-profiling: describe sysctrl limitations in debug mode
mm: do not copy page tables unnecessarily for VM_UFFD_WP
mm/hugetlb: fix excessive IPI broadcasts when unsharing PMD tables using mmu_gather
mm/rmap: fix two comments related to huge_pmd_unshare()
mm/hugetlb: fix two comments related to huge_pmd_unshare()
mm/hugetlb: fix hugetlb_pmd_shared()
mm: remove unnecessary and incorrect mmap lock assert
x86/kfence: avoid writing L1TF-vulnerable PTEs
mm/vma: do not leak memory when .mmap_prepare swaps the file
migrate: correct lock ordering for hugetlb file folios
panic: only warn about deprecated panic_print on write access
fs/writeback: skip AS_NO_DATA_INTEGRITY mappings in wait_sb_inodes()
mm: take into account mm_cid size for mm_struct static definitions
mm: rename cpu_bitmap field to flexible_array
mm: add missing static initializer for init_mm::mm_cid.lock
|
|
Add Nova Lake (NVL) audio Device ID
The ID will be used by HDA legacy, SOF audio stack and the driver
to determine which audio stack should be used (intel-dsp-config).
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Reviewed-by: Liam Girdwood <liam.r.girdwood@intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Takashi Iwai <tiwai@suse.de>
Link: https://patch.msgid.link/20260120193507.14019-2-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
MT7987
Add thermal controller definition for MT7987.
Signed-off-by: Frank Wunderlich <frank-w@public-files.de>
Acked-by: Rob Herring (Arm) <robh@kernel.org>
Link: https://patch.msgid.link/20251223175710.25850-2-linux@fw-web.de
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
|
|
Directly returning delegations on close when over the watermark is
rather suboptimal as these delegations are much more likely to be reused
than those that have been unused for a long time. Switch to returning
unused delegations from a new LRU list when we are above the threshold and
there are reclaimable delegations instead.
Pass over referenced delegations during the first pass to give delegations
that aren't in active used by frequently used for stat() or similar another
chance to not be instantly reclaimed. This scheme works the same as the
referenced flags in the VFS inode and dentry caches.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Anna Schumaker <anna.schumaker@oracle.com>
|
|
Searching for returnable delegations in the per-server delegations list
can be very expensive. While commit e04bbf6b1bbe ("NFS: Avoid quadratic
search when freeing delegations.") reduced the overhead a bit, the
fact that all the non-returnable delegations have to be searched limits
the amount of optimizations that can be done.
Fix this by introducing a separate list that only contains delegations
scheduled for return.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Anna Schumaker <anna.schumaker@oracle.com>
|
|
Add LVTS thermal controller binding for MediaTek MT8196.
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Tested-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Tested-by: Frank Wunderlich <frank-w@public-files.de>
Signed-off-by: Laura Nao <laura.nao@collabora.com>
Link: https://patch.msgid.link/20251125-mt8196-lvts-v4-v5-1-6db7eb903fb7@collabora.com
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
|
|
Currently, reset_idmap_scratch() performs a 4.7KB memset() in every
states_equal() call. Optimize this by using a counter to track used
ID mappings, replacing the O(N) memset() with an O(1) reset and
bounding the search loop in check_ids().
Signed-off-by: Qiliang Yuan <realwujing@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/bpf/20260120023234.77673-1-realwujing@gmail.com
|
|
Merge series from Claudiu <claudiu.beznea@tuxon.dev>:
This series adds cleanups for the Renesas RZ SSI driver.
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/mszyprowski/linux
Pull dma-mapping fixes from Marek Szyprowski:
- minor fixes for the corner cases of the SWIOTLB pool management
(Robin Murphy)
* tag 'dma-mapping-6.19-2026-01-20' of git://git.kernel.org/pub/scm/linux/kernel/git/mszyprowski/linux:
dma/pool: Avoid allocating redundant pools
mm_zone: Generalise has_managed_dma()
dma/pool: Improve pool lookup
|
|
Merge series from Abdurrahman Hussain <abdurrahman@nexthop.ai>:
Additionally, make interrupts optional to allow the driver to fall back
to its existing polling mode on systems where interrupts are either missing
or broken.
|
|
Commit ab04b530e7e8 ("mm: introduce copy-on-fork VMAs and make
VM_MAYBE_GUARD one") aggregates flags checks in vma_needs_copy(),
including VM_UFFD_WP.
However in doing so, it incorrectly performed this check against src_vma.
This check was done on the assumption that all relevant flags are copied
upon fork.
However the userfaultfd logic is very innovative in that it implements
custom logic on fork in dup_userfaultfd(), including a rather well hidden
case where lacking UFFD_FEATURE_EVENT_FORK causes VM_UFFD_WP to not be
propagated to the destination VMA.
And indeed, vma_needs_copy(), prior to this patch, did check this property
on dst_vma, not src_vma.
Since all the other relevant flags are copied on fork, we can simply fix
this by checking against dst_vma.
While we're here, we fix a comment against VM_COPY_ON_FORK (noting that it
did indeed already reference dst_vma) to make it abundantly clear that we
must check against the destination VMA.
Link: https://lkml.kernel.org/r/20260114110006.1047071-1-lorenzo.stoakes@oracle.com
Fixes: ab04b530e7e8 ("mm: introduce copy-on-fork VMAs and make VM_MAYBE_GUARD one")
Signed-off-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Reported-by: Chris Mason <clm@meta.com>
Closes: https://lore.kernel.org/all/20260113231257.3002271-1-clm@meta.com/
Acked-by: David Hildenbrand (Red Hat) <david@kernel.org>
Acked-by: Pedro Falcato <pfalcato@suse.de>
Cc: Liam Howlett <liam.howlett@oracle.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
mmu_gather
As reported, ever since commit 1013af4f585f ("mm/hugetlb: fix
huge_pmd_unshare() vs GUP-fast race") we can end up in some situations
where we perform so many IPI broadcasts when unsharing hugetlb PMD page
tables that it severely regresses some workloads.
In particular, when we fork()+exit(), or when we munmap() a large
area backed by many shared PMD tables, we perform one IPI broadcast per
unshared PMD table.
There are two optimizations to be had:
(1) When we process (unshare) multiple such PMD tables, such as during
exit(), it is sufficient to send a single IPI broadcast (as long as
we respect locking rules) instead of one per PMD table.
Locking prevents that any of these PMD tables could get reused before
we drop the lock.
(2) When we are not the last sharer (> 2 users including us), there is
no need to send the IPI broadcast. The shared PMD tables cannot
become exclusive (fully unshared) before an IPI will be broadcasted
by the last sharer.
Concurrent GUP-fast could walk into a PMD table just before we
unshared it. It could then succeed in grabbing a page from the
shared page table even after munmap() etc succeeded (and supressed
an IPI). But there is not difference compared to GUP-fast just
sleeping for a while after grabbing the page and re-enabling IRQs.
Most importantly, GUP-fast will never walk into page tables that are
no-longer shared, because the last sharer will issue an IPI
broadcast.
(if ever required, checking whether the PUD changed in GUP-fast
after grabbing the page like we do in the PTE case could handle
this)
So let's rework PMD sharing TLB flushing + IPI sync to use the mmu_gather
infrastructure so we can implement these optimizations and demystify the
code at least a bit. Extend the mmu_gather infrastructure to be able to
deal with our special hugetlb PMD table sharing implementation.
To make initialization of the mmu_gather easier when working on a single
VMA (in particular, when dealing with hugetlb), provide
tlb_gather_mmu_vma().
We'll consolidate the handling for (full) unsharing of PMD tables in
tlb_unshare_pmd_ptdesc() and tlb_flush_unshared_tables(), and track
in "struct mmu_gather" whether we had (full) unsharing of PMD tables.
Because locking is very special (concurrent unsharing+reuse must be
prevented), we disallow deferring flushing to tlb_finish_mmu() and instead
require an explicit earlier call to tlb_flush_unshared_tables().
From hugetlb code, we call huge_pmd_unshare_flush() where we make sure
that the expected lock protecting us from concurrent unsharing+reuse is
still held.
Check with a VM_WARN_ON_ONCE() in tlb_finish_mmu() that
tlb_flush_unshared_tables() was properly called earlier.
Document it all properly.
Notes about tlb_remove_table_sync_one() interaction with unsharing:
There are two fairly tricky things:
(1) tlb_remove_table_sync_one() is a NOP on architectures without
CONFIG_MMU_GATHER_RCU_TABLE_FREE.
Here, the assumption is that the previous TLB flush would send an
IPI to all relevant CPUs. Careful: some architectures like x86 only
send IPIs to all relevant CPUs when tlb->freed_tables is set.
The relevant architectures should be selecting
MMU_GATHER_RCU_TABLE_FREE, but x86 might not do that in stable
kernels and it might have been problematic before this patch.
Also, the arch flushing behavior (independent of IPIs) is different
when tlb->freed_tables is set. Do we have to enlighten them to also
take care of tlb->unshared_tables? So far we didn't care, so
hopefully we are fine. Of course, we could be setting
tlb->freed_tables as well, but that might then unnecessarily flush
too much, because the semantics of tlb->freed_tables are a bit
fuzzy.
This patch changes nothing in this regard.
(2) tlb_remove_table_sync_one() is not a NOP on architectures with
CONFIG_MMU_GATHER_RCU_TABLE_FREE that actually don't need a sync.
Take x86 as an example: in the common case (!pv, !X86_FEATURE_INVLPGB)
we still issue IPIs during TLB flushes and don't actually need the
second tlb_remove_table_sync_one().
This optimized can be implemented on top of this, by checking e.g., in
tlb_remove_table_sync_one() whether we really need IPIs. But as
described in (1), it really must honor tlb->freed_tables then to
send IPIs to all relevant CPUs.
Notes on TLB flushing changes:
(1) Flushing for non-shared PMD tables
We're converting from flush_hugetlb_tlb_range() to
tlb_remove_huge_tlb_entry(). Given that we properly initialize the
MMU gather in tlb_gather_mmu_vma() to be hugetlb aware, similar to
__unmap_hugepage_range(), that should be fine.
(2) Flushing for shared PMD tables
We're converting from various things (flush_hugetlb_tlb_range(),
tlb_flush_pmd_range(), flush_tlb_range()) to tlb_flush_pmd_range().
tlb_flush_pmd_range() achieves the same that
tlb_remove_huge_tlb_entry() would achieve in these scenarios.
Note that tlb_remove_huge_tlb_entry() also calls
__tlb_remove_tlb_entry(), however that is only implemented on
powerpc, which does not support PMD table sharing.
Similar to (1), tlb_gather_mmu_vma() should make sure that TLB
flushing keeps on working as expected.
Further, note that the ptdesc_pmd_pts_dec() in huge_pmd_share() is not a
concern, as we are holding the i_mmap_lock the whole time, preventing
concurrent unsharing. That ptdesc_pmd_pts_dec() usage will be removed
separately as a cleanup later.
There are plenty more cleanups to be had, but they have to wait until
this is fixed.
[david@kernel.org: fix kerneldoc]
Link: https://lkml.kernel.org/r/f223dd74-331c-412d-93fc-69e360a5006c@kernel.org
Link: https://lkml.kernel.org/r/20251223214037.580860-5-david@kernel.org
Fixes: 1013af4f585f ("mm/hugetlb: fix huge_pmd_unshare() vs GUP-fast race")
Signed-off-by: David Hildenbrand (Red Hat) <david@kernel.org>
Reported-by: Uschakow, Stanislav" <suschako@amazon.de>
Closes: https://lore.kernel.org/all/4d3878531c76479d9f8ca9789dc6485d@amazon.de/
Tested-by: Laurence Oberman <loberman@redhat.com>
Acked-by: Harry Yoo <harry.yoo@oracle.com>
Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Cc: Lance Yang <lance.yang@linux.dev>
Cc: Liu Shixin <liushixin2@huawei.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Rik van Riel <riel@surriel.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
Patch series "mm/hugetlb: fixes for PMD table sharing (incl. using
mmu_gather)", v3.
One functional fix, one performance regression fix, and two related
comment fixes.
I cleaned up my prototype I recently shared [1] for the performance fix,
deferring most of the cleanups I had in the prototype to a later point.
While doing that I identified the other things.
The goal of this patch set is to be backported to stable trees "fairly"
easily. At least patch #1 and #4.
Patch #1 fixes hugetlb_pmd_shared() not detecting any sharing
Patch #2 + #3 are simple comment fixes that patch #4 interacts with.
Patch #4 is a fix for the reported performance regression due to excessive
IPI broadcasts during fork()+exit().
The last patch is all about TLB flushes, IPIs and mmu_gather.
Read: complicated
There are plenty of cleanups in the future to be had + one reasonable
optimization on x86. But that's all out of scope for this series.
Runtime tested, with a focus on fixing the performance regression using
the original reproducer [2] on x86.
This patch (of 4):
We switched from (wrongly) using the page count to an independent shared
count. Now, shared page tables have a refcount of 1 (excluding
speculative references) and instead use ptdesc->pt_share_count to identify
sharing.
We didn't convert hugetlb_pmd_shared(), so right now, we would never
detect a shared PMD table as such, because sharing/unsharing no longer
touches the refcount of a PMD table.
Page migration, like mbind() or migrate_pages() would allow for migrating
folios mapped into such shared PMD tables, even though the folios are not
exclusive. In smaps we would account them as "private" although they are
"shared", and we would be wrongly setting the PM_MMAP_EXCLUSIVE in the
pagemap interface.
Fix it by properly using ptdesc_pmd_is_shared() in hugetlb_pmd_shared().
Link: https://lkml.kernel.org/r/20251223214037.580860-1-david@kernel.org
Link: https://lkml.kernel.org/r/20251223214037.580860-2-david@kernel.org
Link: https://lore.kernel.org/all/8cab934d-4a56-44aa-b641-bfd7e23bd673@kernel.org/ [1]
Link: https://lore.kernel.org/all/8cab934d-4a56-44aa-b641-bfd7e23bd673@kernel.org/ [2]
Fixes: 59d9094df3d7 ("mm: hugetlb: independent PMD page table shared count")
Signed-off-by: David Hildenbrand (Red Hat) <david@kernel.org>
Reviewed-by: Rik van Riel <riel@surriel.com>
Reviewed-by: Lance Yang <lance.yang@linux.dev>
Tested-by: Lance Yang <lance.yang@linux.dev>
Reviewed-by: Harry Yoo <harry.yoo@oracle.com>
Tested-by: Laurence Oberman <loberman@redhat.com>
Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Acked-by: Oscar Salvador <osalvador@suse.de>
Cc: Liu Shixin <liushixin2@huawei.com>
Cc: Uschakow, Stanislav" <suschako@amazon.de>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
When multiple io_uring rings poll on the same NVMe queue, one ring can
find completions belonging to another ring. The current code always
uses task_work to handle this, but this adds overhead for the common
single-ring case.
This patch passes the polling io_ring_ctx through io_comp_batch's new
poll_ctx field. In io_do_iopoll(), the polling ring's context is stored
in iob.poll_ctx before calling the iopoll callbacks.
In nvme_uring_cmd_end_io(), we now compare iob->poll_ctx with the
request's owning io_ring_ctx (via io_uring_cmd_ctx_handle()). If they
match (local context), we complete inline with io_uring_cmd_done32().
If they differ (remote context) or iob is NULL (non-iopoll path), we
use task_work as before.
This optimization eliminates task_work scheduling overhead for the
common case where a ring polls and finds its own completions.
~10% IOPS improvement is observed in the following benchmark:
fio/t/io_uring -b512 -d128 -c32 -s32 -p1 -F1 -O0 -P1 -u1 -n1 /dev/ng0n1
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Kanchan Joshi <joshi.k@samsung.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Add a third parameter 'const struct io_comp_batch *' to the rq_end_io_fn
callback signature. This allows end_io handlers to access the completion
batch context when requests are completed via blk_mq_end_request_batch().
The io_comp_batch is passed from blk_mq_end_request_batch(), while NULL
is passed from __blk_mq_end_request() and blk_mq_put_rq_ref() which don't
have batch context.
This infrastructure change enables drivers to detect whether they're
being called from a batched completion path (like iopoll) and access
additional context stored in the io_comp_batch.
Update all rq_end_io_fn implementations:
- block/blk-mq.c: blk_end_sync_rq
- block/blk-flush.c: flush_end_io, mq_flush_data_end_io
- drivers/nvme/host/ioctl.c: nvme_uring_cmd_end_io
- drivers/nvme/host/core.c: nvme_keep_alive_end_io
- drivers/nvme/host/pci.c: abort_endio, nvme_del_queue_end, nvme_del_cq_end
- drivers/nvme/target/passthru.c: nvmet_passthru_req_done
- drivers/scsi/scsi_error.c: eh_lock_door_done
- drivers/scsi/sg.c: sg_rq_end_io
- drivers/scsi/st.c: st_scsi_execute_end
- drivers/target/target_core_pscsi.c: pscsi_req_done
- drivers/md/dm-rq.c: end_clone_request
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Kanchan Joshi <joshi.k@samsung.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
This was used only to allow the s5m RTC driver to deal with the alarm
IRQ. That driver now uses a different approach to acquire that IRQ, and
::irq_data doesn't need to be kept around anymore.
Signed-off-by: André Draszik <andre.draszik@linaro.org>
Link: https://patch.msgid.link/20260113-s5m-alarm-v3-3-855a19db1277@linaro.org
Signed-off-by: Lee Jones <lee@kernel.org>
|
|
several netfilter compilation units rely on implicit includes
coming from nf_conntrack_proto_gre.h.
Clean this up and add the required dependencies where needed.
nf_conntrack.h requires net_generic() helper.
Place various gre/ppp/vlan includes to where they are needed.
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
conntrack, xtables and nftables are distinct subsystems, don't use them
in other subystems.
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
After the optimization to only perform one GC per jiffy, a new problem
was introduced. If more than 8 new connections are tracked per jiffy the
list won't be cleaned up fast enough possibly reaching the limit
wrongly.
In order to prevent this issue, only skip the GC if it was already
triggered during the same jiffy and the increment is lower than the
clean up limit. In addition, increase the clean up limit to 64
connections to avoid triggering GC too often and do more effective GCs.
This has been tested using a HTTP server and several
performance tools while having nft_connlimit/xt_connlimit or OVS limit
configured.
Output of slowhttptest + OVS limit at 52000 connections:
slow HTTP test status on 340th second:
initializing: 0
pending: 432
connected: 51998
error: 0
closed: 0
service available: YES
Fixes: d265929930e2 ("netfilter: nf_conncount: reduce unnecessary GC")
Reported-by: Aleksandra Rukomoinikova <ARukomoinikova@k2.cloud>
Closes: https://lore.kernel.org/netfilter/b2064e7b-0776-4e14-adb6-c68080987471@k2.cloud/
Signed-off-by: Fernando Fernandez Mancera <fmancera@suse.de>
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
of_drm_find_and_get_bridge()
of_drm_find_bridge() is deprecated. Move to its replacement
of_drm_find_and_get_bridge() which gets a bridge reference, and ensure it
is put when done. Also switch to the drm_bridge::next_bridge pointer.
This needs to handle both cases: when of_drm_find_panel() succeeds and when
it fails.
In the 'else' case (i.e. when of_drm_find_panel() fails), just switch to
of_drm_find_and_get_bridge() to ensure the bridge is not freed while in use
in the function tail, when it is stored in dsi->bridge.next_bridge.
In the 'then' case (i.e. when of_drm_find_panel() succeeds),
devm_drm_panel_bridge_add() already increments the refcount using devres
which ties the bridge allocation lifetime to the device lifetime, so we
would not need to do anything. However to have the same behaviour in both
branches take an additional reference here, so that the bridge needs to be
put whichever branch is taken without more complicated logic. Ensure to
clear the bridge pointer however, to avoid calling drm_bridge_put() on an
ERR_PTR.
Acked-by: Maxime Ripard <mripard@kernel.org>
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Link: https://patch.msgid.link/20260109-drm-bridge-alloc-getput-drm_of_find_bridge-2-v2-12-8bad3ef90b9f@bootlin.com
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
|
|
of_drm_find_bridge() is deprecated. Move to its replacement
of_drm_find_and_get_bridge() which gets a bridge reference, and ensure it
is put when done.
Acked-by: Maxime Ripard <mripard@kernel.org>
Link: https://patch.msgid.link/20260109-drm-bridge-alloc-getput-drm_of_find_bridge-2-v2-2-8bad3ef90b9f@bootlin.com
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
|
|
drm_of_panel_bridge_remove() uses of_graph_get_remote_node() to get a
device_node but does not put the node reference.
Fixes: c70087e8f16f ("drm/drm_of: add drm_of_panel_bridge_remove function")
Cc: stable@vger.kernel.org # v4.15
Acked-by: Maxime Ripard <mripard@kernel.org>
Link: https://patch.msgid.link/20260109-drm-bridge-alloc-getput-drm_of_find_bridge-2-v2-1-8bad3ef90b9f@bootlin.com
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
|
|
Include linux/types.h for u8.
Reviewed-by: Andy Yan <andy.yan@rock-chips.com>
Link: https://patch.msgid.link/20260107152704.2290146-1-jani.nikula@intel.com
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
Add a netdevice notifier in netkit to watch for NETDEV_UNREGISTER events.
If the target device is indeed NETREG_UNREGISTERING and previously leased
a queue to a netkit device, then collect the related netkit devices and
batch-unregister_netdevice_many() them.
If this would not be done, then the netkit device would hold a reference
on the physical device preventing it from going away. However, in case of
both io_uring zero-copy as well as AF_XDP this situation is handled
gracefully and the allocated resources are torn down.
In the case where mentioned infra is used through netkit, the applications
have a reference on netkit, and netkit in turn holds a reference on the
physical device. In order to have netkit release the reference on the
physical device, we need such watcher to then unregister the netkit ones.
This is generally quite similar to the dependency handling in case of
tunnels (e.g. vxlan bound to a underlying netdev) where the tunnel device
gets removed along with the physical device.
# ip a
[...]
4: enp10s0f0np0: <BROADCAST,MULTICAST> mtu 1500 qdisc mq state DOWN group default qlen 1000
link/ether e8:eb:d3:a3:43:f6 brd ff:ff:ff:ff:ff:ff
inet 10.0.0.2/24 scope global enp10s0f0np0
valid_lft forever preferred_lft forever
[...]
8: nk@NONE: <BROADCAST,MULTICAST,NOARP> mtu 1500 qdisc noop state DOWN group default qlen 1000
link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff
[...]
# rmmod mlx5_ib
# rmmod mlx5_core
[ 309.261822] mlx5_core 0000:0a:00.0 mlx5_0: Port: 1 Link DOWN
[ 344.235236] mlx5_core 0000:0a:00.1: E-Switch: Unload vfs: mode(LEGACY), nvfs(0), necvfs(0), active vports(0)
[ 344.246948] mlx5_core 0000:0a:00.1: E-Switch: Disable: mode(LEGACY), nvfs(0), necvfs(0), active vports(0)
[ 344.463754] mlx5_core 0000:0a:00.1: E-Switch: Disable: mode(LEGACY), nvfs(0), necvfs(0), active vports(0)
[ 344.770155] mlx5_core 0000:0a:00.1: E-Switch: cleanup
[ 345.345709] mlx5_core 0000:0a:00.0: E-Switch: Unload vfs: mode(LEGACY), nvfs(0), necvfs(0), active vports(0)
[ 345.357524] mlx5_core 0000:0a:00.0: E-Switch: Disable: mode(LEGACY), nvfs(0), necvfs(0), active vports(0)
[ 350.995989] mlx5_core 0000:0a:00.0: E-Switch: Disable: mode(LEGACY), nvfs(0), necvfs(0), active vports(0)
[ 351.574396] mlx5_core 0000:0a:00.0: E-Switch: cleanup
# ip a
[...]
[ both enp10s0f0np0 and nk gone ]
[...]
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Co-developed-by: David Wei <dw@davidwei.uk>
Signed-off-by: David Wei <dw@davidwei.uk>
Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org>
Link: https://patch.msgid.link/20260115082603.219152-12-daniel@iogearbox.net
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
|
|
Add a single device mode for netkit instead of netkit pairs. The primary
target for the paired devices is to connect network namespaces, of course,
and support has been implemented in projects like Cilium [0]. For the rxq
leasing the plan is to support two main scenarios related to single device
mode:
* For the use-case of io_uring zero-copy, the control plane can either
set up a netkit pair where the peer device can perform rxq leasing which
is then tied to the lifetime of the peer device, or the control plane
can use a regular netkit pair to connect the hostns to a Pod/container
and dynamically add/remove rxq leasing through a single device without
having to interrupt the device pair. In the case of io_uring, the memory
pool is used as skb non-linear pages, and thus the skb will go its way
through the regular stack into netkit. Things like the netkit policy when
no BPF is attached or skb scrubbing etc apply as-is in case the paired
devices are used, or if the backend memory is tied to the single device
and traffic goes through a paired device.
* For the use-case of AF_XDP, the control plane needs to use netkit in the
single device mode. The single device mode currently enforces only a
pass policy when no BPF is attached, and does not yet support BPF link
attachments for AF_XDP. skbs sent to that device get dropped at the
moment. Given AF_XDP operates at a lower layer of the stack tying this
to the netkit pair did not make sense. In future, the plan is to allow
BPF at the XDP layer which can: i) process traffic coming from the AF_XDP
application (e.g. QEMU with AF_XDP backend) to filter egress traffic or
to push selected egress traffic up to the single netkit device to the
local stack (e.g. DHCP requests), and ii) vice-versa skbs sent to the
single netkit into the AF_XDP application (e.g. DHCP replies). Also,
the control-plane can dynamically manage rxq leasing for the single
netkit device without having to interrupt (e.g. down/up cycle) the main
netkit pair for the Pod which has traffic going in and out.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Co-developed-by: David Wei <dw@davidwei.uk>
Signed-off-by: David Wei <dw@davidwei.uk>
Reviewed-by: Jordan Rife <jordan@jrife.io>
Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org>
Link: https://docs.cilium.io/en/stable/operations/performance/tuning/#netkit-device-mode [0]
Link: https://patch.msgid.link/20260115082603.219152-10-daniel@iogearbox.net
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
|
|
When a process in a container wants to setup a memory provider, it will
use the virtual netdev and a leased rxq, and call net_mp_{open,close}_rxq
to try and restart the queue. At this point, proxy the queue restart on
the real rxq in the physical netdev.
For memory providers (io_uring zero-copy rx and devmem), it causes the
real rxq in the physical netdev to be filled from a memory provider that
has DMA mapped memory from a process within a container.
Signed-off-by: David Wei <dw@davidwei.uk>
Co-developed-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org>
Acked-by: Stanislav Fomichev <sdf@fomichev.me>
Link: https://patch.msgid.link/20260115082603.219152-6-daniel@iogearbox.net
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
|
|
Populate nested lease info to the queue-get response that returns the
ifindex, queue id with type and optionally netns id if the device
resides in a different netns.
Example with ynl client:
# ip a
[...]
4: enp10s0f0np0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 xdp/id:24 qdisc mq state UP group default qlen 1000
link/ether e8:eb:d3:a3:43:f6 brd ff:ff:ff:ff:ff:ff
inet 10.0.0.2/24 scope global enp10s0f0np0
valid_lft forever preferred_lft forever
inet6 fe80::eaeb:d3ff:fea3:43f6/64 scope link proto kernel_ll
valid_lft forever preferred_lft forever
[...]
# ethtool -i enp10s0f0np0
driver: mlx5_core
[...]
# ./pyynl/cli.py \
--spec ~/netlink/specs/netdev.yaml \
--do queue-get \
--json '{"ifindex": 4, "id": 15, "type": "rx"}'
{'id': 15,
'ifindex': 4,
'lease': {'ifindex': 8, 'netns-id': 0, 'queue': {'id': 1, 'type': 'rx'}},
'napi-id': 8227,
'type': 'rx',
'xsk': {}}
# ip netns list
foo (id: 0)
# ip netns exec foo ip a
[...]
8: nk@NONE: <BROADCAST,MULTICAST,NOARP,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff
inet6 fe80::200:ff:fe00:0/64 scope link proto kernel_ll
valid_lft forever preferred_lft forever
[...]
# ip netns exec foo ethtool -i nk
driver: netkit
[...]
# ip netns exec foo ls /sys/class/net/nk/queues/
rx-0 rx-1 tx-0
# ip netns exec foo ./pyynl/cli.py \
--spec ~/netlink/specs/netdev.yaml \
--do queue-get \
--json '{"ifindex": 8, "id": 1, "type": "rx"}'
{'id': 1, 'ifindex': 8, 'type': 'rx'}
Note that the caller of netdev_nl_queue_fill_one() holds the netdevice
lock. For the queue-get we do not lock both devices. When queues get
{un,}leased, both devices are locked, thus if __netif_get_rx_queue_peer()
returns true, the peer pointer points to a valid device. The netns-id
is fetched via peernet2id_alloc() similarly as done in OVS.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Co-developed-by: David Wei <dw@davidwei.uk>
Signed-off-by: David Wei <dw@davidwei.uk>
Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org>
Acked-by: Stanislav Fomichev <sdf@fomichev.me>
Link: https://patch.msgid.link/20260115082603.219152-4-daniel@iogearbox.net
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
|
|
Implement netdev_nl_queue_create_doit which creates a new rx queue in a
virtual netdev and then leases it to a rx queue in a physical netdev.
Example with ynl client:
# ./pyynl/cli.py \
--spec ~/netlink/specs/netdev.yaml \
--do queue-create \
--json '{"ifindex": 8, "type": "rx", "lease": {"ifindex": 4, "queue": {"type": "rx", "id": 15}}}'
{'id': 1}
Note that the netdevice locking order is always from the virtual to
the physical device.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Co-developed-by: David Wei <dw@davidwei.uk>
Signed-off-by: David Wei <dw@davidwei.uk>
Acked-by: Stanislav Fomichev <sdf@fomichev.me>
Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org>
Link: https://patch.msgid.link/20260115082603.219152-3-daniel@iogearbox.net
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
|
|
Add a ynl netdev family operation called queue-create that creates a
new queue on a netdevice:
name: queue-create
attribute-set: queue
flags: [admin-perm]
do:
request:
attributes:
- ifindex
- type
- lease
reply: &queue-create-op
attributes:
- id
This is a generic operation such that it can be extended for various
use cases in future. Right now it is mandatory to specify ifindex,
the queue type which is enforced to rx and a lease. The newly created
queue id is returned to the caller.
A queue from a virtual device can have a lease which refers to another
queue from a physical device. This is useful for memory providers
and AF_XDP operations which take an ifindex and queue id to allow
applications to bind against virtual devices in containers. The lease
couples both queues together and allows to proxy the operations from
a virtual device in a container to the physical device.
In future, the nested lease attribute can be lifted and made optional
for other use-cases such as dynamic queue creation for physical
netdevs. The lack of lease and the specification of the physical
device as an ifindex will imply that we need a real queue to be
allocated. Similarly, the queue type enforcement to rx can then be
lifted as well to support tx.
An early implementation had only driver-specific integration [0], but
in order for other virtual devices to reuse, it makes sense to have
this as a generic API in core net.
For leasing queues, the virtual netdev must have real_num_rx_queue
less than num_rx_queues at the time of calling queue-create. The
queue-type must be rx as only rx queues are supported for leasing
for now. We also enforce that the queue-create ifindex must point
to a virtual device, and that the nested lease attribute's ifindex
must point to a physical device. The nested lease attribute set
contains a netns-id attribute which is currently only intended for
dumping as part of the queue-get operation. Also, it is modeled as
an s32 type similarly as done elsewhere in the stack.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Co-developed-by: David Wei <dw@davidwei.uk>
Signed-off-by: David Wei <dw@davidwei.uk>
Link: https://bpfconf.ebpf.io/bpfconf2025/bpfconf2025_material/lsfmmbpf_2025_netkit_borkmann.pdf [0]
Acked-by: Stanislav Fomichev <sdf@fomichev.me>
Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org>
Link: https://patch.msgid.link/20260115082603.219152-2-daniel@iogearbox.net
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
|
|
When the AP has an advertised TID to Link Mapping (TTLM) it shall
include the element in the association response. As such, when this
element is present it needs to be used for the currently dormant links.
See Draft P802.11REVmf_D1.0 section 35.3.7.2.3 ("Negotiation of TTLM")
for the details. The flag is also not usable in case userspace wants to
specify a negotiated TTLM during association.
Note that for the link reconfiguration case, mac80211 did not use the
information. Draft P802.11REVmf_D1.0 states in section 35.3.6.4 ("Link
reconfiguration to the setup links) that we "shall operate with all the
TIDs mapped to the newly added links ..."
All this means that the flag is not needed. The implementation should
parse the information from the association response.
Signed-off-by: Benjamin Berg <benjamin.berg@intel.com>
Reviewed-by: Johannes Berg <johannes.berg@intel.com>
Reviewed-by: Ilan Peer <ilan.peer@intel.com>
Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com>
Link: https://patch.msgid.link/20260118093904.754e057896a5.Ifd06f5ef839a93bfd54d0593dc932870f95f3242@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
|
|
Fast commits only log operations that have dedicated replay support.
EXT4_IOC_MOVE_EXT swaps extents between regular files and may copy
data, rewriting the affected inodes' block mapping layout without
going through the fast commit tracking paths.
In practice these operations are rare and usually followed by further
updates, but mixing them into a fast commit makes the overall
semantics harder to reason about and risks replay gaps if new call
sites appear.
Teach ext4 to mark the filesystem fast-commit ineligible for the
journal transactions used by move_extent_per_page() when
EXT4_IOC_MOVE_EXT runs.
This forces those transactions to fall back to a full commit,
ensuring that these multi-inode extent swaps are captured by the
normal journal rather than partially encoded in fast commit TLVs.
This change should not affect common workloads but makes online
defragmentation safer and easier to reason about under fast commit.
Testing:
1. prepare:
dd if=/dev/zero of=/root/fc_move.img bs=1M count=0 seek=256
mkfs.ext4 -O fast_commit -F /root/fc_move.img
mkdir -p /mnt/fc_move && mount -t ext4 -o loop \
/root/fc_move.img /mnt/fc_move
2. Created two files, ran EXT4_IOC_MOVE_EXT via e4defrag, and checked
the ineligible reason statistics:
fallocate -l 64M /mnt/fc_move/file1
cp /mnt/fc_move/file1 /mnt/fc_move/file2
e4defrag /mnt/fc_move/file1
cat /proc/fs/ext4/loop0/fc_info
shows "Move extents": > 0 and fc stats ineligible > 0.
Signed-off-by: Li Chen <me@linux.beauty>
Link: https://patch.msgid.link/20251211115146.897420-4-me@linux.beauty
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
|
|
Fast commits only log operations that have dedicated replay support.
Enabling fs-verity builds a Merkle tree and updates inode and orphan
state in ways that are not described by the fast commit replay tags.
In practice these operations are rare and usually followed by further
updates, but mixing them into a fast commit makes the overall
semantics harder to reason about and risks replay gaps if new call
sites appear.
Teach ext4 to mark the filesystem fast-commit ineligible when
ext4_end_enable_verity() starts its journal transaction.
This forces that transaction to fall back to a full commit, ensuring
that the fs-verity enable changes are captured by the normal journal
rather than partially encoded in fast commit TLVs.
This change should not affect common workloads but makes fs-verity
enable safer and easier to reason about under fast commit.
Testing:
1. prepare:
dd if=/dev/zero of=/root/fc_verity.img bs=1M count=0 seek=128
mkfs.ext4 -O fast_commit,verity -F /root/fc_verity.img
mkdir -p /mnt/fc_verity && mount -t ext4 -o loop /root/fc_verity.img /mnt/fc_verity
2. Enabled fs-verity on a file and verified reason accounting:
echo "data" > /mnt/fc_verity/verityfile
/root/enable_verity /mnt/fc_verity/verityfile
sync
tail -n 1 /proc/fs/ext4/loop0/fc_info
"fs-verity enable": 1
3. Enabled fs-verity on a second file, fsynced it, and checked that the
ineligible commit counter is updated too:
echo "data2" > /mnt/fc_verity/verityfile2
/root/enable_verity /mnt/fc_verity/verityfile2
/root/fsync_file /mnt/fc_verity/verityfile2
sync
/proc/fs/ext4/loop0/fc_info shows "fs-verity enable" incremented and
fc stats ineligible increased accordingly.
Signed-off-by: Li Chen <me@linux.beauty>
Link: https://patch.msgid.link/20251211115146.897420-3-me@linux.beauty
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
|
|
Fast commits only log operations that have dedicated replay support.
Inode format migration (indirect<->extent layout changes via
EXT4_IOC_MIGRATE or toggling EXT4_EXTENTS_FL) rewrites the block mapping
representation without going through the fast commit tracking paths.
In practice these migrations are rare and usually followed by further
updates, but mixing them into a fast commit makes the overall semantics
harder to reason about and risks replay gaps if new call sites appear.
Teach ext4 to mark the filesystem fast-commit ineligible when
ext4_ext_migrate() or ext4_ind_migrate() start their journal transactions.
This forces those transactions to fall back to a full commit, ensuring
that the entire inode layout change is captured by the normal journal
rather than partially encoded in fast commit TLVs. This change should
not affect common workloads but makes format migrations safer and easier
to reason about under fast commit.
Testing:
1. prepare:
dd if=/dev/zero of=/root/fc.img bs=1M count=0 seek=128
mkfs.ext4 -O fast_commit -F /root/fc.img
mkdir -p /mnt/fc && mount -t ext4 -o loop /root/fc.img /mnt/fc
2. Created a test file and toggled the extents flag to exercise both
ext4_ind_migrate() and ext4_ext_migrate():
touch /mnt/fc/migtest
chattr -e /mnt/fc/migtest
chattr +e /mnt/fc/migtest
3. Verified fast-commit ineligible statistics:
tail -n 1 /proc/fs/ext4/loop0/fc_info
"Inode format migration": 2
Signed-off-by: Li Chen <me@linux.beauty>
Link: https://patch.msgid.link/20251211115146.897420-2-me@linux.beauty
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
|
|
Above the while() loop in wait_sb_inodes(), we document that we must wait
for all pages under writeback for data integrity. Consequently, if a
mapping, like fuse, traditionally does not have data integrity semantics,
there is no need to wait at all; we can simply skip these inodes.
This restores fuse back to prior behavior where syncs are no-ops. This
fixes a user regression where if a system is running a faulty fuse server
that does not reply to issued write requests, this causes wait_sb_inodes()
to wait forever.
Link: https://lkml.kernel.org/r/20260105211737.4105620-2-joannelkoong@gmail.com
Fixes: 0c58a97f919c ("fuse: remove tmp folio for writebacks and internal rb tree")
Signed-off-by: Joanne Koong <joannelkoong@gmail.com>
Reported-by: Athul Krishna <athul.krishna.kr@protonmail.com>
Reported-by: J. Neuschäfer <j.neuschaefer@gmx.net>
Reviewed-by: Bernd Schubert <bschubert@ddn.com>
Tested-by: J. Neuschäfer <j.neuschaefer@gmx.net>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Bernd Schubert <bschubert@ddn.com>
Cc: Bonaccorso Salvatore <carnil@debian.org>
Cc: Christian Brauner <brauner@kernel.org>
Cc: David Hildenbrand <david@kernel.org>
Cc: Jan Kara <jack@suse.cz>
Cc: "Liam R. Howlett" <Liam.Howlett@oracle.com>
Cc: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Miklos Szeredi <miklos@szeredi.hu>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
