Patch series "Multi-Gen LRU Framework", v14.
What's new
==========
1. OpenWrt, in addition to Android, Arch Linux Zen, Armbian, ChromeOS,
Liquorix, post-factum and XanMod, is now shipping MGLRU on 5.15.
2. Fixed long-tailed direct reclaim latency seen on high-memory (TBs)
machines. The old direct reclaim backoff, which tries to enforce a
minimum fairness among all eligible memcgs, over-swapped by about
(total_mem>>DEF_PRIORITY)-nr_to_reclaim. The new backoff, which
pulls the plug on swapping once the target is met, trades some
fairness for curtailed latency:
https://lore.kernel.org/r/20220918080010.2920238-10-yuzhao@google.com/
3. Fixed minior build warnings and conflicts. More comments and nits.
TLDR
====
The current page reclaim is too expensive in terms of CPU usage and it
often makes poor choices about what to evict. This patchset offers an
alternative solution that is performant, versatile and
straightforward.
Patchset overview
=================
The design and implementation overview is in patch 14:
https://lore.kernel.org/r/20220918080010.2920238-15-yuzhao@google.com/
01. mm: x86, arm64: add arch_has_hw_pte_young()
02. mm: x86: add CONFIG_ARCH_HAS_NONLEAF_PMD_YOUNG
Take advantage of hardware features when trying to clear the accessed
bit in many PTEs.
03. mm/vmscan.c: refactor shrink_node()
04. Revert "include/linux/mm_inline.h: fold __update_lru_size() into
its sole caller"
Minor refactors to improve readability for the following patches.
05. mm: multi-gen LRU: groundwork
Adds the basic data structure and the functions that insert pages to
and remove pages from the multi-gen LRU (MGLRU) lists.
06. mm: multi-gen LRU: minimal implementation
A minimal implementation without optimizations.
07. mm: multi-gen LRU: exploit locality in rmap
Exploits spatial locality to improve efficiency when using the rmap.
08. mm: multi-gen LRU: support page table walks
Further exploits spatial locality by optionally scanning page tables.
09. mm: multi-gen LRU: optimize multiple memcgs
Optimizes the overall performance for multiple memcgs running mixed
types of workloads.
10. mm: multi-gen LRU: kill switch
Adds a kill switch to enable or disable MGLRU at runtime.
11. mm: multi-gen LRU: thrashing prevention
12. mm: multi-gen LRU: debugfs interface
Provide userspace with features like thrashing prevention, working set
estimation and proactive reclaim.
13. mm: multi-gen LRU: admin guide
14. mm: multi-gen LRU: design doc
Add an admin guide and a design doc.
Benchmark results
=================
Independent lab results
-----------------------
Based on the popularity of searches [01] and the memory usage in
Google's public cloud, the most popular open-source memory-hungry
applications, in alphabetical order, are:
Apache Cassandra Memcached
Apache Hadoop MongoDB
Apache Spark PostgreSQL
MariaDB (MySQL) Redis
An independent lab evaluated MGLRU with the most widely used benchmark
suites for the above applications. They posted 960 data points along
with kernel metrics and perf profiles collected over more than 500
hours of total benchmark time. Their final reports show that, with 95%
confidence intervals (CIs), the above applications all performed
significantly better for at least part of their benchmark matrices.
On 5.14:
1. Apache Spark [02] took 95% CIs [9.28, 11.19]% and [12.20, 14.93]%
less wall time to sort three billion random integers, respectively,
under the medium- and the high-concurrency conditions, when
overcommitting memory. There were no statistically significant
changes in wall time for the rest of the benchmark matrix.
2. MariaDB [03] achieved 95% CIs [5.24, 10.71]% and [20.22, 25.97]%
more transactions per minute (TPM), respectively, under the medium-
and the high-concurrency conditions, when overcommitting memory.
There were no statistically significant changes in TPM for the rest
of the benchmark matrix.
3. Memcached [04] achieved 95% CIs [23.54, 32.25]%, [20.76, 41.61]%
and [21.59, 30.02]% more operations per second (OPS), respectively,
for sequential access, random access and Gaussian (distribution)
access, when THP=always; 95% CIs [13.85, 15.97]% and
[23.94, 29.92]% more OPS, respectively, for random access and
Gaussian access, when THP=never. There were no statistically
significant changes in OPS for the rest of the benchmark matrix.
4. MongoDB [05] achieved 95% CIs [2.23, 3.44]%, [6.97, 9.73]% and
[2.16, 3.55]% more operations per second (OPS), respectively, for
exponential (distribution) access, random access and Zipfian
(distribution) access, when underutilizing memory; 95% CIs
[8.83, 10.03]%, [21.12, 23.14]% and [5.53, 6.46]% more OPS,
respectively, for exponential access, random access and Zipfian
access, when overcommitting memory.
On 5.15:
5. Apache Cassandra [06] achieved 95% CIs [1.06, 4.10]%, [1.94, 5.43]%
and [4.11, 7.50]% more operations per second (OPS), respectively,
for exponential (distribution) access, random access and Zipfian
(distribution) access, when swap was off; 95% CIs [0.50, 2.60]%,
[6.51, 8.77]% and [3.29, 6.75]% more OPS, respectively, for
exponential access, random access and Zipfian access, when swap was
on.
6. Apache Hadoop [07] took 95% CIs [5.31, 9.69]% and [2.02, 7.86]%
less average wall time to finish twelve parallel TeraSort jobs,
respectively, under the medium- and the high-concurrency
conditions, when swap was on. There were no statistically
significant changes in average wall time for the rest of the
benchmark matrix.
7. PostgreSQL [08] achieved 95% CI [1.75, 6.42]% more transactions per
minute (TPM) under the high-concurrency condition, when swap was
off; 95% CIs [12.82, 18.69]% and [22.70, 46.86]% more TPM,
respectively, under the medium- and the high-concurrency
conditions, when swap was on. There were no statistically
significant changes in TPM for the rest of the benchmark matrix.
8. Redis [09] achieved 95% CIs [0.58, 5.94]%, [6.55, 14.58]% and
[11.47, 19.36]% more total operations per second (OPS),
respectively, for sequential access, random access and Gaussian
(distribution) access, when THP=always; 95% CIs [1.27, 3.54]%,
[10.11, 14.81]% and [8.75, 13.64]% more total OPS, respectively,
for sequential access, random access and Gaussian access, when
THP=never.
Our lab results
---------------
To supplement the above results, we ran the following benchmark suites
on 5.16-rc7 and found no regressions [10].
fs_fio_bench_hdd_mq pft
fs_lmbench pgsql-hammerdb
fs_parallelio redis
fs_postmark stream
hackbench sysbenchthread
kernbench tpcc_spark
memcached unixbench
multichase vm-scalability
mutilate will-it-scale
nginx
[01] https://trends.google.com
[02] https://lore.kernel.org/r/20211102002002.92051-1-bot@edi.works/
[03] https://lore.kernel.org/r/20211009054315.47073-1-bot@edi.works/
[04] https://lore.kernel.org/r/20211021194103.65648-1-bot@edi.works/
[05] https://lore.kernel.org/r/20211109021346.50266-1-bot@edi.works/
[06] https://lore.kernel.org/r/20211202062806.80365-1-bot@edi.works/
[07] https://lore.kernel.org/r/20211209072416.33606-1-bot@edi.works/
[08] https://lore.kernel.org/r/20211218071041.24077-1-bot@edi.works/
[09] https://lore.kernel.org/r/20211122053248.57311-1-bot@edi.works/
[10] https://lore.kernel.org/r/20220104202247.2903702-1-yuzhao@google.com/
Read-world applications
=======================
Third-party testimonials
------------------------
Konstantin reported [11]:
I have Archlinux with 8G RAM + zswap + swap. While developing, I
have lots of apps opened such as multiple LSP-servers for different
langs, chats, two browsers, etc... Usually, my system gets quickly
to a point of SWAP-storms, where I have to kill LSP-servers,
restart browsers to free memory, etc, otherwise the system lags
heavily and is barely usable.
1.5 day ago I migrated from 5.11.15 kernel to 5.12 + the LRU
patchset, and I started up by opening lots of apps to create memory
pressure, and worked for a day like this. Till now I had not a
single SWAP-storm, and mind you I got 3.4G in SWAP. I was never
getting to the point of 3G in SWAP before without a single
SWAP-storm.
Vaibhav from IBM reported [12]:
In a synthetic MongoDB Benchmark, seeing an average of ~19%
throughput improvement on POWER10(Radix MMU + 64K Page Size) with
MGLRU patches on top of 5.16 kernel for MongoDB + YCSB across
three different request distributions, namely, Exponential, Uniform
and Zipfan.
Shuang from U of Rochester reported [13]:
With the MGLRU, fio achieved 95% CIs [38.95, 40.26]%, [4.12, 6.64]%
and [9.26, 10.36]% higher throughput, respectively, for random
access, Zipfian (distribution) access and Gaussian (distribution)
access, when the average number of jobs per CPU is 1; 95% CIs
[42.32, 49.15]%, [9.44, 9.89]% and [20.99, 22.86]% higher
throughput, respectively, for random access, Zipfian access and
Gaussian access, when the average number of jobs per CPU is 2.
Daniel from Michigan Tech reported [14]:
With Memcached allocating ~100GB of byte-addressable Optante,
performance improvement in terms of throughput (measured as queries
per second) was about 10% for a series of workloads.
Large-scale deployments
-----------------------
We've rolled out MGLRU to tens of millions of ChromeOS users and
about a million Android users. Google's fleetwide profiling [15] shows
an overall 40% decrease in kswapd CPU usage, in addition to
improvements in other UX metrics, e.g., an 85% decrease in the number
of low-memory kills at the 75th percentile and an 18% decrease in
app launch time at the 50th percentile.
The downstream kernels that have been using MGLRU include:
1. Android [16]
2. Arch Linux Zen [17]
3. Armbian [18]
4. ChromeOS [19]
5. Liquorix [20]
6. OpenWrt [21]
7. post-factum [22]
8. XanMod [23]
[11] https://lore.kernel.org/r/140226722f2032c86301fbd326d91baefe3d7d23.camel@yandex.ru/
[12] https://lore.kernel.org/r/87czj3mux0.fsf@vajain21.in.ibm.com/
[13] https://lore.kernel.org/r/20220105024423.26409-1-szhai2@cs.rochester.edu/
[14] https://lore.kernel.org/r/CA+4-3vksGvKd18FgRinxhqHetBS1hQekJE2gwco8Ja-bJWKtFw@mail.gmail.com/
[15] https://dl.acm.org/doi/10.1145/2749469.2750392
[16] https://android.com
[17] https://archlinux.org
[18] https://armbian.com
[19] https://chromium.org
[20] https://liquorix.net
[21] https://openwrt.org
[22] https://codeberg.org/pf-kernel
[23] https://xanmod.org
Summary
=======
The facts are:
1. The independent lab results and the real-world applications
indicate substantial improvements; there are no known regressions.
2. Thrashing prevention, working set estimation and proactive reclaim
work out of the box; there are no equivalent solutions.
3. There is a lot of new code; no smaller changes have been
demonstrated similar effects.
Our options, accordingly, are:
1. Given the amount of evidence, the reported improvements will likely
materialize for a wide range of workloads.
2. Gauging the interest from the past discussions, the new features
will likely be put to use for both personal computers and data
centers.
3. Based on Google's track record, the new code will likely be well
maintained in the long term. It'd be more difficult if not
impossible to achieve similar effects with other approaches.
This patch (of 14):
Some architectures automatically set the accessed bit in PTEs, e.g., x86
and arm64 v8.2. On architectures that do not have this capability,
clearing the accessed bit in a PTE usually triggers a page fault following
the TLB miss of this PTE (to emulate the accessed bit).
Being aware of this capability can help make better decisions, e.g.,
whether to spread the work out over a period of time to reduce bursty page
faults when trying to clear the accessed bit in many PTEs.
Note that theoretically this capability can be unreliable, e.g.,
hotplugged CPUs might be different from builtin ones. Therefore it should
not be used in architecture-independent code that involves correctness,
e.g., to determine whether TLB flushes are required (in combination with
the accessed bit).
Link: https://lkml.kernel.org/r/20220918080010.2920238-1-yuzhao@google.com
Link: https://lkml.kernel.org/r/20220918080010.2920238-2-yuzhao@google.com
Signed-off-by: Yu Zhao <yuzhao@google.com>
Reviewed-by: Barry Song <baohua@kernel.org>
Acked-by: Brian Geffon <bgeffon@google.com>
Acked-by: Jan Alexander Steffens (heftig) <heftig@archlinux.org>
Acked-by: Oleksandr Natalenko <oleksandr@natalenko.name>
Acked-by: Steven Barrett <steven@liquorix.net>
Acked-by: Suleiman Souhlal <suleiman@google.com>
Acked-by: Will Deacon <will@kernel.org>
Tested-by: Daniel Byrne <djbyrne@mtu.edu>
Tested-by: Donald Carr <d@chaos-reins.com>
Tested-by: Holger Hoffstätte <holger@applied-asynchrony.com>
Tested-by: Konstantin Kharlamov <Hi-Angel@yandex.ru>
Tested-by: Shuang Zhai <szhai2@cs.rochester.edu>
Tested-by: Sofia Trinh <sofia.trinh@edi.works>
Tested-by: Vaibhav Jain <vaibhav@linux.ibm.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Hillf Danton <hdanton@sina.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: linux-arm-kernel@lists.infradead.org
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michael Larabel <Michael@MichaelLarabel.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Tejun Heo <tj@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Cc: Qi Zheng <zhengqi.arch@bytedance.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Once upon a time, we only support accounting thrashing of page cache.
Then Joonsoo introduced workingset detection for anonymous pages and we
gained the ability to account thrashing of them[1].
For page cache thrashing accounting, there is no suitable place to do it
in fs level likes swap_readpage(). So we have to do it in
folio_wait_bit_common().
Then for anonymous pages thrashing accounting, we have to do it in both
swap_readpage() and folio_wait_bit_common(). This likes PSI, so we should
let thrashing accounting supports re-entrance detection.
This patch is to prepare complete thrashing accounting, and is based on
patch "filemap: make the accounting of thrashing more consistent".
[1] commit aae466b005 ("mm/swap: implement workingset detection for anonymous LRU")
Link: https://lkml.kernel.org/r/20220815071134.74551-1-yang.yang29@zte.com.cn
Signed-off-by: Yang Yang <yang.yang29@zte.com.cn>
Signed-off-by: CGEL ZTE <cgel.zte@gmail.com>
Reviewed-by: Ran Xiaokai <ran.xiaokai@zte.com.cn>
Reviewed-by: wangyong <wang.yong12@zte.com.cn>
Acked-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Introduce a variable swap_migration_ad_supported to cache whether the arch
supports swap migration A/D bits.
Here one thing to mention is that SWP_MIG_TOTAL_BITS will internally
reference the other macro MAX_PHYSMEM_BITS, which is a function call on
x86 (constant on all the rest of archs).
It's safe to reference it in swapfile_init() because when reaching here
we're already during initcalls level 4 so we must have initialized 5-level
pgtable for x86_64 (right after early_identify_cpu() finishes).
- start_kernel
- setup_arch
- early_cpu_init
- get_cpu_cap --> fetch from CPUID (including X86_FEATURE_LA57)
- early_identify_cpu --> clear X86_FEATURE_LA57 (if early lvl5 not enabled (USE_EARLY_PGTABLE_L5))
- arch_call_rest_init
- rest_init
- kernel_init
- kernel_init_freeable
- do_basic_setup
- do_initcalls --> calls swapfile_init() (initcall level 4)
This should slightly speed up the migration swap entry handlings.
Link: https://lkml.kernel.org/r/20220811161331.37055-8-peterx@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Andi Kleen <andi.kleen@intel.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Huang Ying <ying.huang@intel.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: "Kirill A . Shutemov" <kirill@shutemov.name>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Nadav Amit <nadav.amit@gmail.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Dave Hansen <dave.hansen@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
We used to have swapfile_maximum_size() fetching a maximum value of
swapfile size per-arch.
As the caller of max_swapfile_size() grows, this patch introduce a
variable "swapfile_maximum_size" and cache the value of old
max_swapfile_size(), so that we don't need to calculate the value every
time.
Caching the value in swapfile_init() is safe because when reaching the
phase we should have initialized all the relevant information. Here the
major arch to take care of is x86, which defines the max swapfile size
based on L1TF mitigation.
Here both X86_BUG_L1TF or l1tf_mitigation should have been setup properly
when reaching swapfile_init(). As a reference, the code path looks like
this for x86:
- start_kernel
- setup_arch
- early_cpu_init
- early_identify_cpu --> setup X86_BUG_L1TF
- parse_early_param
- l1tf_cmdline --> set l1tf_mitigation
- check_bugs
- l1tf_select_mitigation --> set l1tf_mitigation
- arch_call_rest_init
- rest_init
- kernel_init
- kernel_init_freeable
- do_basic_setup
- do_initcalls --> calls swapfile_init() (initcall level 4)
The swapfile size only depends on swp pte format on non-x86 archs, so
caching it is safe too.
Since at it, rename max_swapfile_size() to arch_max_swapfile_size()
because arch can define its own function, so it's more straightforward to
have "arch_" as its prefix. At the meantime, export swapfile_maximum_size
to replace the old usages of max_swapfile_size().
[peterx@redhat.com: declare arch_max_swapfile_size) in swapfile.h]
Link: https://lkml.kernel.org/r/YxTh1GuC6ro5fKL5@xz-m1.local
Link: https://lkml.kernel.org/r/20220811161331.37055-7-peterx@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: "Huang, Ying" <ying.huang@intel.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Andi Kleen <andi.kleen@intel.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: "Kirill A . Shutemov" <kirill@shutemov.name>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Nadav Amit <nadav.amit@gmail.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Dave Hansen <dave.hansen@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
When page migration happens, we always ignore the young/dirty bit settings
in the old pgtable, and marking the page as old in the new page table
using either pte_mkold() or pmd_mkold(), and keeping the pte clean.
That's fine from functional-wise, but that's not friendly to page reclaim
because the moving page can be actively accessed within the procedure.
Not to mention hardware setting the young bit can bring quite some
overhead on some systems, e.g. x86_64 needs a few hundreds nanoseconds to
set the bit. The same slowdown problem to dirty bits when the memory is
first written after page migration happened.
Actually we can easily remember the A/D bit configuration and recover the
information after the page is migrated. To achieve it, define a new set
of bits in the migration swap offset field to cache the A/D bits for old
pte. Then when removing/recovering the migration entry, we can recover
the A/D bits even if the page changed.
One thing to mention is that here we used max_swapfile_size() to detect
how many swp offset bits we have, and we'll only enable this feature if we
know the swp offset is big enough to store both the PFN value and the A/D
bits. Otherwise the A/D bits are dropped like before.
Link: https://lkml.kernel.org/r/20220811161331.37055-6-peterx@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: "Huang, Ying" <ying.huang@intel.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Andi Kleen <andi.kleen@intel.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: "Kirill A . Shutemov" <kirill@shutemov.name>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Nadav Amit <nadav.amit@gmail.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Dave Hansen <dave.hansen@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
We've got a bunch of special swap entries that stores PFN inside the swap
offset fields. To fetch the PFN, normally the user just calls
swp_offset() assuming that'll be the PFN.
Add a helper swp_offset_pfn() to fetch the PFN instead, fetching only the
max possible length of a PFN on the host, meanwhile doing proper check
with MAX_PHYSMEM_BITS to make sure the swap offsets can actually store the
PFNs properly always using the BUILD_BUG_ON() in is_pfn_swap_entry().
One reason to do so is we never tried to sanitize whether swap offset can
really fit for storing PFN. At the meantime, this patch also prepares us
with the future possibility to store more information inside the swp
offset field, so assuming "swp_offset(entry)" to be the PFN will not stand
any more very soon.
Replace many of the swp_offset() callers to use swp_offset_pfn() where
proper. Note that many of the existing users are not candidates for the
replacement, e.g.:
(1) When the swap entry is not a pfn swap entry at all, or,
(2) when we wanna keep the whole swp_offset but only change the swp type.
For the latter, it can happen when fork() triggered on a write-migration
swap entry pte, we may want to only change the migration type from
write->read but keep the rest, so it's not "fetching PFN" but "changing
swap type only". They're left aside so that when there're more
information within the swp offset they'll be carried over naturally in
those cases.
Since at it, dropping hwpoison_entry_to_pfn() because that's exactly what
the new swp_offset_pfn() is about.
Link: https://lkml.kernel.org/r/20220811161331.37055-4-peterx@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: "Huang, Ying" <ying.huang@intel.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Andi Kleen <andi.kleen@intel.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: "Kirill A . Shutemov" <kirill@shutemov.name>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Nadav Amit <nadav.amit@gmail.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Dave Hansen <dave.hansen@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Currently, struct efi_boot_memmap is a struct that is passed around
between callers of efi_get_memory_map() and the users of the resulting
data, and which carries pointers to various variables whose values are
provided by the EFI GetMemoryMap() boot service.
This is overly complex, and it is much easier to carry these values in
the struct itself. So turn the struct into one that carries these data
items directly, including a flex array for the variable number of EFI
memory descriptors that the boot service may return.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Practical experience (and advice from Alexei) tell us that bitfields in
structs lead to un-optimized assembly code. I've verified this change
does lead to better x86_64 assembly, both via objdump and playing with
code snippets in godbolt.org.
Using scripts/bloat-o-meter shows the code size is reduced with 24
bytes for xdp_convert_buff_to_frame() that gets inlined e.g. in
i40e_xmit_xdp_tx_ring() which were used for microbenchmarking.
Microbenchmarking results do show improvements, but very small and
varying between 0.5 to 2 nanosec improvement per packet.
The member @metasize is changed from u8 to u32. Future users of this
area could split this into two u16 fields. I've also benchmarked with
two u16 fields showing equal performance gains and code size reduction.
The moved member @frame_sz doesn't change sizeof struct due to existing
padding. Like xdp_buff member @frame_sz is placed next to @flags, which
allows compiler to optimize assignment of these.
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
Link: https://lore.kernel.org/r/166393728005.2213882.4162674859542409548.stgit@firesoul
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Pull last (?) hotfixes from Andrew Morton:
"26 hotfixes.
8 are for issues which were introduced during this -rc cycle, 18 are
for earlier issues, and are cc:stable"
* tag 'mm-hotfixes-stable-2022-09-26' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm: (26 commits)
x86/uaccess: avoid check_object_size() in copy_from_user_nmi()
mm/page_isolation: fix isolate_single_pageblock() isolation behavior
mm,hwpoison: check mm when killing accessing process
mm/hugetlb: correct demote page offset logic
mm: prevent page_frag_alloc() from corrupting the memory
mm: bring back update_mmu_cache() to finish_fault()
frontswap: don't call ->init if no ops are registered
mm/huge_memory: use pfn_to_online_page() in split_huge_pages_all()
mm: fix madivse_pageout mishandling on non-LRU page
powerpc/64s/radix: don't need to broadcast IPI for radix pmd collapse flush
mm: gup: fix the fast GUP race against THP collapse
mm: fix dereferencing possible ERR_PTR
vmscan: check folio_test_private(), not folio_get_private()
mm: fix VM_BUG_ON in __delete_from_swap_cache()
tools: fix compilation after gfp_types.h split
mm/damon/dbgfs: fix memory leak when using debugfs_lookup()
mm/migrate_device.c: copy pte dirty bit to page
mm/migrate_device.c: add missing flush_cache_page()
mm/migrate_device.c: flush TLB while holding PTL
x86/mm: disable instrumentations of mm/pgprot.c
...
Add cpumask_nth_{,and,andnot} as wrappers around corresponding
find functions, and use it in cpumask_local_spread().
Signed-off-by: Yury Norov <yury.norov@gmail.com>
Kernel lacks for a function that searches for Nth bit in a bitmap.
Usually people do it like this:
for_each_set_bit(bit, mask, size)
if (n-- == 0)
return bit;
We can do it more efficiently, if we:
1. find a word containing Nth bit, using hweight(); and
2. find the bit, using a helper fns(), that works similarly to
__ffs() and ffz().
fns() is implemented as a simple loop. For x86_64, there's PDEP instruction
to do that: ret = clz(pdep(1 << idx, num)). However, for large bitmaps the
most of improvement comes from using hweight(), so I kept fns() simple.
New find_nth_bit() is ~70 times faster on x86_64/kvm in find_bit benchmark:
find_nth_bit: 7154190 ns, 16411 iterations
for_each_bit: 505493126 ns, 16315 iterations
With all that, a family of 3 new functions is added, and used where
appropriate in the following patches.
Signed-off-by: Yury Norov <yury.norov@gmail.com>
The function calculates Hamming weight of (bitmap1 & bitmap2). Now we
have to do like this:
tmp = bitmap_alloc(nbits);
bitmap_and(tmp, map1, map2, nbits);
weight = bitmap_weight(tmp, nbits);
bitmap_free(tmp);
This requires additional memory, adds pressure on alloc subsystem, and
way less cache-friendly than just:
weight = bitmap_weight_and(map1, map2, nbits);
The following patches apply it for cpumask functions.
Signed-off-by: Yury Norov <yury.norov@gmail.com>
All uses of dst_hold_and_use() have
been removed since commit 1202cdd665 ("Remove DECnet support
from kernel"), so remove it.
Signed-off-by: Gaosheng Cui <cuigaosheng1@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
All uses of sk_nulls_node_init() have
been removed since commit dbca1596bb ("ping: convert to RCU
lookups, get rid of rwlock"), so remove it.
Signed-off-by: Gaosheng Cui <cuigaosheng1@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
All uses of neigh_key_eq16() have
been removed since commit 1202cdd665 ("Remove DECnet support
from kernel"), so remove it.
Signed-off-by: Gaosheng Cui <cuigaosheng1@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The dust has settled a bit and it's become obvious what code is
totally common between nfsd_init_dirlist_pages() and
nfsd3_init_dirlist_pages(). Move that common code to SUNRPC.
The new helper brackets the existing xdr_init_decode_pages() API.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Currently, SUNRPC clears the whole of .pc_argsize before processing
each incoming RPC transaction. Add an extra parameter to struct
svc_procedure to enable upper layers to reduce the amount of each
operation's argument structure that is zeroed by SUNRPC.
The size of struct nfsd4_compoundargs, in particular, is a lot to
clear on each incoming RPC Call. A subsequent patch will cut this
down to something closer to what NFSv2 and NFSv3 uses.
This patch should cause no behavior changes.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Commit 2825a7f907 ("nfsd4: allow encoding across page boundaries")
added an explicit computation of the remaining length in the rq_res
XDR buffer.
The computation appears to suffer from an "off-by-one" bug. Because
buflen is too large by one page, XDR encoding can run off the end of
the send buffer by eventually trying to use the struct page address
in rq_page_end, which always contains NULL.
Fixes: bddfdbcddb ("NFSD: Extract the svcxdr_init_encode() helper")
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Ensure that stream-based argument decoding can't go past the actual
end of the receive buffer. xdr_init_decode's calculation of the
value of xdr->end over-estimates the end of the buffer because the
Linux kernel RPC server code does not remove the size of the RPC
header from rqstp->rq_arg before calling the upper layer's
dispatcher.
The server-side still uses the svc_getnl() macros to decode the
RPC call header. These macros reduce the length of the head iov
but do not update the total length of the message in the buffer
(buf->len).
A proper fix for this would be to replace the use of svc_getnl() and
friends in the RPC header decoder, but that would be a large and
invasive change that would be difficult to backport.
Fixes: 5191955d6f ("SUNRPC: Prepare for xdr_stream-style decoding on the server-side")
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
In preparation for replacing __builtin_object_size() with
__builtin_dynamic_object_size(), all the compile-time size checks
need to check that the bounds comparisons are, in fact, known at
compile-time. Enforce what was guaranteed with __bos(). In other words,
since all uses of __bos() were constant expressions, it was not required
to test for this. When these change to __bdos(), they _may_ be constant
expressions, and the checks are only valid when the prior condition
holds. This results in no binary differences.
Cc: linux-hardening@vger.kernel.org
Link: https://lore.kernel.org/lkml/20220920192202.190793-3-keescook@chromium.org
Signed-off-by: Kees Cook <keescook@chromium.org>
Grorgi writes:
"interconnect changes for 6.1
These are the interconnect changes for the 6.1-rc1 merge window, which
this time are tiny. One is a series to convert the remove() callback of
platform devices to return void instead of int. The other change is
enabling modular support for a driver."
Signed-off-by: Georgi Djakov <djakov@kernel.org>
* tag 'icc-6.1-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/djakov/icc:
interconnect: qcom: Kconfig: Make INTERCONNECT_QCOM tristate
interconnect: imx: Make imx_icc_unregister() return void
interconnect: Make icc_provider_del() return void
interconnect: sm8450: Ignore return value of icc_provider_del() in .remove()
interconnect: osm-l3: Ignore return value of icc_provider_del() in .remove()
interconnect: msm8974: Ignore return value of icc_provider_del() in .remove()
interconnect: icc-rpmh: Ignore return value of icc_provider_del() in .remove()
interconnect: icc-rpm: Ignore return value of icc_provider_del() in .remove()
interconnect: imx: Ignore return value of icc_provider_del() in .remove()
With CONFIG_CFI_CLANG, assembly functions called indirectly
from C code must be annotated with type identifiers to pass CFI
checking. In order to make this easier, the compiler emits a
__kcfi_typeid_<function> symbol for each address-taken function
declaration in C, which contains the expected type identifier that
we can refer to in assembly code.
Add a typed version of SYM_FUNC_START, which emits the type
identifier before the function. Architectures that support KCFI can
define their own __CFI_TYPE macro to override the default preamble
format.
As an example, for the x86_64 blowfish_dec_blk function, the
compiler emits the following type symbol:
$ readelf -sW vmlinux | grep __kcfi_typeid_blowfish_dec_blk
120204: 00000000ef478db5 0 NOTYPE WEAK DEFAULT ABS
__kcfi_typeid_blowfish_dec_blk
And SYM_TYPED_FUNC_START will generate the following preamble based
on the __CFI_TYPE definition for the architecture:
$ objdump -dr arch/x86/crypto/blowfish-x86_64-asm_64.o
...
0000000000000400 <__cfi_blowfish_dec_blk>:
...
40b: b8 00 00 00 00 mov $0x0,%eax
40c: R_X86_64_32 __kcfi_typeid_blowfish_dec_blk
0000000000000410 <blowfish_dec_blk>:
...
Note that the address of all assembly functions annotated with
SYM_TYPED_FUNC_START must be taken in C code that's linked into the
binary or the missing __kcfi_typeid_ symbol will result in a linker
error with CONFIG_CFI_CLANG. If the code that contains the indirect
call is not always compiled in, __ADDRESSABLE(functionname) can be
used to ensure that the __kcfi_typeid_ symbol is emitted.
Signed-off-by: Sami Tolvanen <samitolvanen@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Tested-by: Kees Cook <keescook@chromium.org>
Tested-by: Nathan Chancellor <nathan@kernel.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20220908215504.3686827-7-samitolvanen@google.com
32-bit ABIs support passing 64-bit integers by registers via argument
translation. Commit 59c10c52f5 ("riscv: compat: syscall: Add
compat_sys_call_table implementation") implements the compat_arg_u64
macro for efficiently defining little endian compatibility syscalls.
Architectures supporting big endianness may benefit from reciprocal
argument translation, but are welcome also to implement their own.
Signed-off-by: Rohan McLure <rmclure@linux.ibm.com>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
Reviewed-by: Arnd Bergmann <arnd@anrdb.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20220921065605.1051927-10-rmclure@linux.ibm.com
The DARTs present in the M1 Pro/Max/Ultra SoC use a diffent PTE format.
They support a 42bit physical address space by shifting the paddr and
extending its mask inside the PTE.
They also come with mandatory sub-page protection now which we just
configure to always allow access to the entire page. This feature is
already present but optional on the previous DARTs which allows to
unconditionally configure it.
Signed-off-by: Sven Peter <sven@svenpeter.dev>
Co-developed-by: Janne Grunau <j@jannau.net>
Signed-off-by: Janne Grunau <j@jannau.net>
Reviewed-by: Rob Herring <robh@kernel.org>
Acked-by: Hector Martin <marcan@marcan.st>
Link: https://lore.kernel.org/r/20220916094152.87137-5-j@jannau.net
Signed-off-by: Joerg Roedel <jroedel@suse.de>
CONFIG_ARCH_HAS_HUGEPD is used to tell core mm when huge page
directories are used.
When they are not used, no need to provide hugepd_t or is_hugepd(),
just rely on the core mm fallback definition.
For that, change core mm behaviour so that CONFIG_ARCH_HAS_HUGEPD
is used instead of indirect is_hugepd macro existence.
powerpc being the only user of huge page directories, there is no
impact on other architectures.
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/da81462d93069bb90fe5e762dd3283a644318937.1662543243.git.christophe.leroy@csgroup.eu
There is a separate I/O failure tree to track the fail reads, so remove
the extra EXTENT_DAMAGED bit in the I/O tree as it's set but never used.
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We still have this oddity of stashing the io_failure_record in the
extent state for the io_failure_tree, which is leftover from when we
used to stuff private pointers in extent_io_trees.
However this doesn't make a lot of sense for the io failure records, we
can simply use a normal rb_tree for this. This will allow us to further
simplify the extent_io_tree code by removing the io_failure_rec pointer
from the extent state.
Convert the io_failure_tree to an rb tree + spinlock in the inode, and
then use our rb tree simple helpers to insert and find failed records.
This greatly cleans up this code and makes it easier to separate out the
extent_io_tree code.
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Currently we only have 3 qgroup flags:
- BTRFS_QGROUP_STATUS_FLAG_ON
- BTRFS_QGROUP_STATUS_FLAG_RESCAN
- BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT
These flags match the on-disk flags used in btrfs_qgroup_status.
But we're going to introduce extra runtime flags which will not reach
disks.
So here we introduce a new mask, BTRFS_QGROUP_STATUS_FLAGS_MASK, to
make sure only those flags can reach disks.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The problem of long mount time caused by block group item search is
already known for some time, and the solution of block group tree has
been proposed.
There is really no need to bound this feature into extent tree v2, just
introduce compat RO flag, BLOCK_GROUP_TREE, to correctly solve the
problem.
All the code handling block group root is already in the upstream
kernel, thus this patch really only needs to introduce the new compat RO
flag.
This patch introduces one extra artificial limitation on block group
tree feature, that free space cache v2 and no-holes feature must be
enabled to use this new compat RO feature.
This artificial requirement is mostly to reduce the test combinations,
and can be a guideline for future features, to mostly rely on the latest
default features.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Preserve the fs-verity status of a btrfs file across send/recv.
There is no facility for installing the Merkle tree contents directly on
the receiving filesystem, so we package up the parameters used to enable
verity found in the verity descriptor. This gives the receive side
enough information to properly enable verity again. Note that this means
that receive will have to re-compute the whole Merkle tree, similar to
how compression worked before encoded_write.
Since the file becomes read-only after verity is enabled, it is
important that verity is added to the send stream after any file writes.
Therefore, when we process a verity item, merely note that it happened,
then actually create the command in the send stream during
'finish_inode_if_needed'.
This also creates V3 of the send stream format, without any format
changes besides adding the new commands and attributes.
Signed-off-by: Boris Burkov <boris@bur.io>
Signed-off-by: David Sterba <dsterba@suse.com>
