hugetlb: revert use of page_cache_next_miss()

Ackerley Tng reported an issue with hugetlbfs fallocate as noted in the
Closes tag.  The issue showed up after the conversion of hugetlb page
cache lookup code to use page_cache_next_miss.  User visible effects are:

- hugetlbfs fallocate incorrectly returns -EEXIST if pages are presnet
  in the file.
- hugetlb pages will not be included in core dumps if they need to be
  brought in via GUP.
- userfaultfd UFFDIO_COPY will not notice pages already present in the
  cache.  It may try to allocate a new page and potentially return
  ENOMEM as opposed to EEXIST.

Revert the use page_cache_next_miss() in hugetlb code.

IMPORTANT NOTE FOR STABLE BACKPORTS:
This patch will apply cleanly to v6.3.  However, due to the change of
filemap_get_folio() return values, it will not function correctly.  This
patch must be modified for stable backports.

[dan.carpenter@linaro.org: fix hugetlbfs_pagecache_present()]
Link: https://lkml.kernel.org/r/efa86091-6a2c-4064-8f55-9b44e1313015@moroto.mountain
Link: https://lkml.kernel.org/r/20230621212403.174710-2-mike.kravetz@oracle.com
Fixes: d0ce0e47b323 ("mm/hugetlb: convert hugetlb fault paths to use alloc_hugetlb_folio()")
Signed-off-by: Mike Kravetz <mike.kravetz@oracle.com>
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Reported-by: Ackerley Tng <ackerleytng@google.com>
Closes: https://lore.kernel.org/linux-mm/cover.1683069252.git.ackerleytng@google.com
Reviewed-by: Sidhartha Kumar <sidhartha.kumar@oracle.com>
Cc: Erdem Aktas <erdemaktas@google.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Vishal Annapurve <vannapurve@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

Revert "page cache: fix page_cache_next/prev_miss off by one"

This reverts commit 9425c591e06a9ab27a145ba655fb50532cf0bcc9

The reverted commit fixed up routines primarily used by readahead code
such that they could also be used by hugetlb.  Unfortunately, this
caused a performance regression as pointed out by the Closes: tag.

The hugetlb code which uses page_cache_next_miss will be addressed in
a subsequent patch.

Link: https://lkml.kernel.org/r/20230621212403.174710-1-mike.kravetz@oracle.com
Fixes: 9425c591e06a ("page cache: fix page_cache_next/prev_miss off by one")
Signed-off-by: Mike Kravetz <mike.kravetz@oracle.com>
Reported-by: kernel test robot <oliver.sang@intel.com>
Closes: https://lore.kernel.org/oe-lkp/202306211346.1e9ff03e-oliver.sang@intel.com
Reviewed-by: Sidhartha Kumar <sidhartha.kumar@oracle.com>
Cc: Ackerley Tng <ackerleytng@google.com>
Cc: Erdem Aktas <erdemaktas@google.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Vishal Annapurve <vannapurve@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/vmscan: fix root proactive reclaim unthrottling unbalanced node

When memory.reclaim was introduced, it became the first case where
cgroup_reclaim() is true for the root cgroup.  Johannes concluded [1] that
for most cases this is okay, except for one case.  Historically, kswapd
would throttle reclaim on a node if a lot of pages marked for reclaim are
under writeback (aka the node is congested).  This occurred by setting
LRUVEC_CONGESTED bit in lruvec->flags.  The bit would be cleared when the
node is balanced.

Similarly, cgroup reclaim would set the same bit when an lruvec is
congested, and clear it on the way out of reclaim (to throttle local
reclaimers).

Before the introduction of memory.reclaim, the root memcg was the only
target of kswapd reclaim, and non-root memcgs were the only targets of
cgroup reclaim, so they would never interfere.  Using the same bit for
both was fine.  After memory.reclaim, it is possible for cgroup reclaim on
the root cgroup to clear the bit set by kswapd.  This would result in
reclaim on the node to be unthrottled before the node is balanced.

Fix this by introducing separate bits for cgroup-level and node-level
congestion.  kswapd can unthrottle an lruvec that is marked as congested
by cgroup reclaim (as the entire node should no longer be congested), but
not vice versa (to prevent premature unthrottling before the entire node
is balanced).

[1]https://lore.kernel.org/lkml/20230405200150.GA35884@cmpxchg.org/

Link: https://lkml.kernel.org/r/20230621023101.432780-1-yosryahmed@google.com
Signed-off-by: Yosry Ahmed <yosryahmed@google.com>
Reported-by: Johannes Weiner <hannes@cmpxchg.org>
Closes: https://lore.kernel.org/lkml/20230405200150.GA35884@cmpxchg.org/
Cc: Michal Hocko <mhocko@suse.com>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Yu Zhao <yuzhao@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: memcg: rename and document global_reclaim()

Evidently, global_reclaim() can be a confusing name.  Especially that it
used to exist before with a subtly different definition (removed by commit
b5ead35e7e1d ("mm: vmscan: naming fixes: global_reclaim() and
sane_reclaim()").  It can be interpreted as non-cgroup reclaim, even
though it returns true for cgroup reclaim on the root memcg (through
memory.reclaim).

Rename it to root_reclaim() in an attempt to make it less ambiguous, and
add documentation to it as well as cgroup_reclaim.

Link: https://lkml.kernel.org/r/20230621023053.432374-1-yosryahmed@google.com
Signed-off-by: Yosry Ahmed <yosryahmed@google.com>
Reported-by: Johannes Weiner <hannes@cmpxchg.org>
Closes: https://lore.kernel.org/lkml/20230405200150.GA35884@cmpxchg.org/
Acked-by: Yu Zhao <yuzhao@google.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Shakeel Butt <shakeelb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: kill [add|del]_page_to_lru_list()

Now no one call [add|del]_page_to_lru_list(), let's drop unused page
interfaces.

Link:https://lkml.kernel.org/r/20230619110718.65679-2-wangkefeng.wang@huawei.com
Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
Acked-by: Yu Zhao <yuzhao@google.com>
Reviewed-by: Baolin Wang <baolin.wang@linux.alibaba.com>
Cc: James Gowans <jgowans@amazon.com>
Cc: Matthew Wilcox <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: compaction: convert to use a folio in isolate_migratepages_block()

Directly use a folio instead of page_folio() when page successfully
isolated (hugepage and movable page) and after folio_get_nontail_page(),
which removes several calls to compound_head().

Link: https://lkml.kernel.org/r/20230619110718.65679-1-wangkefeng.wang@huawei.com
Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
Cc: Baolin Wang <baolin.wang@linux.alibaba.com>
Cc: James Gowans <jgowans@amazon.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Yu Zhao <yuzhao@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: zswap: fix double invalidate with exclusive loads

If exclusive loads are enabled for zswap, we invalidate the entry before
returning from zswap_frontswap_load(), after dropping the local reference.
However, the tree lock is dropped during decompression after the local
reference is acquired, so the entry could be invalidated before we drop
the local ref.  If this happens, the entry is freed once we drop the local
ref, and zswap_invalidate_entry() tries to invalidate an already freed
entry.

Fix this by:
(a) Making sure zswap_invalidate_entry() is always called with a local
    ref held, to avoid being called on a freed entry.
(b) Making sure zswap_invalidate_entry() only drops the ref if the entry
    was actually on the rbtree. Otherwise, another invalidation could
    have already happened, and the initial ref is already dropped.

With these changes, there is no need to check that there is no need to
make sure the entry still exists in the tree in zswap_reclaim_entry()
before invalidating it, as zswap_reclaim_entry() will make this check
internally.

Link: https://lkml.kernel.org/r/20230621093009.637544-1-yosryahmed@google.com
Fixes: b9c91c43412f ("mm: zswap: support exclusive loads")
Signed-off-by: Yosry Ahmed <yosryahmed@google.com>
Reported-by: Hyeonggon Yoo <42.hyeyoo@gmail.com>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: Nhat Pham <nphamcs@gmail.com>
Cc: Seth Jennings <sjenning@redhat.com>
Cc: Vitaly Wool <vitaly.wool@konsulko.com>
Cc: Yu Zhao <yuzhao@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: remove unnecessary pagevec includes

These files no longer need pagevec.h, mostly due to function declarations
being moved out of it.

Link: https://lkml.kernel.org/r/20230621164557.3510324-14-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: remove references to pagevec

Most of these should just refer to the LRU cache rather than the data
structure used to implement the LRU cache.

Link: https://lkml.kernel.org/r/20230621164557.3510324-13-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: rename invalidate_mapping_pagevec to mapping_try_invalidate

We don't use pagevecs for the LRU cache any more, and we don't know that
the failed invalidations were due to the folio being in an LRU cache.  So
rename it to be more accurate.

Link: https://lkml.kernel.org/r/20230621164557.3510324-12-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: remove struct pagevec

All users are now converted to use the folio_batch so we can get rid of
this data structure.

Link: https://lkml.kernel.org/r/20230621164557.3510324-11-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

net: convert sunrpc from pagevec to folio_batch

Remove the last usage of pagevecs.  There is a slight change here; we now
free the folio_batch as soon as it fills up instead of freeing the
folio_batch when we try to add a page to a full batch.  This should have
no effect in practice.

Link: https://lkml.kernel.org/r/20230621164557.3510324-10-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Acked-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

i915: convert i915_gpu_error to use a folio_batch

Remove one of the last remaining users of pagevec.

Link: https://lkml.kernel.org/r/20230621164557.3510324-9-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

pagevec: rename fbatch_count()

This should always have been called folio_batch_count().

Link: https://lkml.kernel.org/r/20230621164557.3510324-8-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: remove check_move_unevictable_pages()

All callers have now been converted to call
check_move_unevictable_folios().

Link: https://lkml.kernel.org/r/20230621164557.3510324-7-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

drm: convert drm_gem_put_pages() to use a folio_batch

Remove a few hidden compound_head() calls by converting the returned page
to a folio once and using the folio APIs.

Link: https://lkml.kernel.org/r/20230621164557.3510324-6-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

i915: convert shmem_sg_free_table() to use a folio_batch

Remove a few hidden compound_head() calls by converting the returned page
to a folio once and using the folio APIs.  We also only increment the
refcount on the folio once instead of once for each page.  Ideally, we
would have a for_each_sgt_folio macro, but until then this will do.

Link: https://lkml.kernel.org/r/20230621164557.3510324-5-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

scatterlist: add sg_set_folio()

This wrapper for sg_set_page() lets drivers add folios to a scatterlist
more easily.  We could, perhaps, do better by using a different page in
the folio if offset is larger than UINT_MAX, but let's hope we get a
better data structure than this before we need to care about such large
folios.

Link: https://lkml.kernel.org/r/20230621164557.3510324-4-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: add __folio_batch_release()

This performs the same role as __pagevec_release(), ie skipping the check
for batch length of 0.

Link: https://lkml.kernel.org/r/20230621164557.3510324-3-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

afs: convert pagevec to folio_batch in afs_extend_writeback()

Patch series "Remove pagevecs".

Removes a folio->page->folio conversion for each folio that's involved.
More importantly, removes one of the last few uses of a pagevec.

Link: https://lkml.kernel.org/r/20230621164557.3510324-1-willy@infradead.org
Link: https://lkml.kernel.org/r/20230621164557.3510324-2-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: page_alloc: use the correct type of list for free pages

Commit bf75f200569d ("mm/page_alloc: add page->buddy_list and
page->pcp_list") introduces page->buddy_list and page->pcp_list as a union
with page->lru, but missed to change get_page_from_free_area() to use
page->buddy_list to clarify the correct type of list for a free page.

Link: https://lkml.kernel.org/r/7e7ab533247d40c0ea0373c18a6a48e5667f9e10.1687333557.git.baolin.wang@linux.alibaba.com
Signed-off-by: Baolin Wang <baolin.wang@linux.alibaba.com>
Acked-by: Mel Gorman <mgorman@techsingularity.net>
Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: backing-dev: make bdi_class a static const structure

Now that the driver core allows for struct class to be in read-only
memory, move the bdi_class structure to be declared at build time placing
it into read-only memory, instead of having to be dynamically allocated at
load time.

Link: https://lkml.kernel.org/r/20230620183314.682822-2-gregkh@linuxfoundation.org
Signed-off-by: Ivan Orlov <ivan.orlov0322@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Suggested-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/swapfile: delete outdated pte_offset_map() comment

Delete a triply out-of-date comment from add_swap_count_continuation():
1. vmalloc_to_page() changed from pte_offset_map() to pte_offset_kernel()
2. pte_offset_map() changed from using kmap_atomic() to kmap_local_page()
3. kmap_atomic() changed from using fixed FIX_KMAP addresses in 2.6.37.

Link: https://lkml.kernel.org/r/9022632b-ba9d-8cb0-c25-4be9786481b5@google.com
Signed-off-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: pass nid to reserve_bootmem_region()

early_pfn_to_nid() is called frequently in init_reserved_page(), it
returns the node id of the PFN.  These PFN are probably from the same
memory region, they have the same node id.  It's not necessary to call
early_pfn_to_nid() for each PFN.

Pass nid to reserve_bootmem_region() and drop the call to
early_pfn_to_nid() in init_reserved_page().  Also, set nid on all reserved
pages before doing this, as some reserved memory regions may not be set
nid.

The most beneficial function is memmap_init_reserved_pages() if
CONFIG_DEFERRED_STRUCT_PAGE_INIT is enabled.

The following data was tested on an x86 machine with 190GB of RAM.

before:
memmap_init_reserved_pages()  67ms

after:
memmap_init_reserved_pages()  20ms

Link: https://lkml.kernel.org/r/20230619023406.424298-1-yajun.deng@linux.dev
Signed-off-by: Yajun Deng <yajun.deng@linux.dev>
Reviewed-by: Mike Rapoport (IBM) <rppt@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/gup: do not return 0 from pin_user_pages_fast() for bad args

These routines are not intended to return zero, the callers cannot do
anything sane with a 0 return.  They should return an error which means
future calls to GUP will not succeed, or they should return some non-zero
number of pinned pages which means GUP should be called again.

If start + nr_pages overflows it should return -EOVERFLOW to signal the
arguments are invalid.

Syzkaller keeps tripping on this when fuzzing GUP arguments.

Link: https://lkml.kernel.org/r/0-v1-3d5ed1f20d50+104-gup_overflow_jgg@nvidia.com
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Reported-by: syzbot+353c7be4964c6253f24a@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/all/000000000000094fdd05faa4d3a4@google.com
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Lorenzo Stoakes <lstoakes@gmail.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: fix shmem THP counters on migration

The per node numa_stat values for shmem don't change on page migration for
THP:

  grep shmem /sys/fs/cgroup/machine.slice/.../memory.numa_stat:

    shmem N0=1092616192 N1=10485760
    shmem_thp N0=1092616192 N1=10485760

  migratepages 9181 0 1:

    shmem N0=0 N1=1103101952
    shmem_thp N0=1092616192 N1=10485760

Fix that by updating shmem_thp counters likewise to shmem counters on page
migration.

[jglauber@digitalocean.com: use folio_test_pmd_mappable instead of folio_test_transhuge]
Link: https://lkml.kernel.org/r/20230622094720.510540-1-jglauber@digitalocean.com
Link: https://lkml.kernel.org/r/20230619103351.234837-1-jglauber@digitalocean.com
Signed-off-by: Jan Glauber <jglauber@digitalocean.com>
Reviewed-by: Baolin Wang <baolin.wang@linux.alibaba.com>
Cc: "Huang, Ying" <ying.huang@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

selftests: cgroup: fix unexpected failure on test_memcg_sock

Before server got a client connection, there were some memory allocations
in the test memcg, such as user stack.  So do not count those allocations
which are not related to socket when checking socket memory accounting.

Link: https://lkml.kernel.org/r/20230619124735.2124-1-haifeng.xu@shopee.com
Signed-off-by: Haifeng Xu <haifeng.xu@shopee.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Michal Koutný <mkoutny@suse.com>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Shakeel Butt <shakeelb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/memcontrol: do not tweak node in mem_cgroup_init()

mem_cgroup_init() request for allocations from each possible node, and
it's used to be a problem because NODE_DATA is not allocated for offline
node. Things have already changed since commit 09f49dca570a9 ("mm: handle
uninitialized numa nodes gracefully"), so it's unnecessary to check for
!node_online nodes here.

How to test?

qemu-system-x86_64 \
  -kernel vmlinux \
  -initrd full.rootfs.cpio.gz \
  -append "console=ttyS0,115200 root=/dev/ram0 nokaslr earlyprintk=serial oops=panic panic_on_warn" \
  -drive format=qcow2,file=vm_disk.qcow2,media=disk,if=ide \
  -enable-kvm \
  -cpu host \
  -m 8G,slots=2,maxmem=16G \
  -smp cores=4,threads=1,sockets=2  \
  -object memory-backend-ram,id=mem0,size=4G \
  -object memory-backend-ram,id=mem1,size=4G \
  -numa node,memdev=mem0,cpus=0-3,nodeid=0 \
  -numa node,memdev=mem1,cpus=4-7,nodeid=1 \
  -numa node,nodeid=2 \
  -net nic,model=virtio,macaddr=52:54:00:12:34:58 \
  -net user \
  -nographic \
  -rtc base=localtime \
  -gdb tcp::6000

Guest state when booting:

[    0.048881] NUMA: Node 0 [mem 0x00000000-0x0009ffff] + [mem 0x00100000-0xbfffffff] -> [mem 0x00000000-0xbfffffff]
[    0.050489] NUMA: Node 0 [mem 0x00000000-0xbfffffff] + [mem 0x100000000-0x13fffffff] -> [mem 0x00000000-0x13fffffff]
[    0.052173] NODE_DATA(0) allocated [mem 0x13fffc000-0x13fffffff]
[    0.053164] NODE_DATA(1) allocated [mem 0x23fffa000-0x23fffdfff]
[    0.054187] Zone ranges:
[    0.054587]   DMA      [mem 0x0000000000001000-0x0000000000ffffff]
[    0.055551]   DMA32    [mem 0x0000000001000000-0x00000000ffffffff]
[    0.056515]   Normal   [mem 0x0000000100000000-0x000000023fffffff]
[    0.057484] Movable zone start for each node
[    0.058149] Early memory node ranges
[    0.058705]   node   0: [mem 0x0000000000001000-0x000000000009efff]
[    0.059679]   node   0: [mem 0x0000000000100000-0x00000000bffdffff]
[    0.060659]   node   0: [mem 0x0000000100000000-0x000000013fffffff]
[    0.061649]   node   1: [mem 0x0000000140000000-0x000000023fffffff]
[    0.062638] Initmem setup node 0 [mem 0x0000000000001000-0x000000013fffffff]
[    0.063745] Initmem setup node 1 [mem 0x0000000140000000-0x000000023fffffff]
[    0.064855]   DMA zone: 158 reserved pages exceeds freesize 0
[    0.065746] Initializing node 2 as memoryless
[    0.066437] Initmem setup node 2 as memoryless
[    0.067132]   DMA zone: 158 reserved pages exceeds freesize 0
[    0.068037] On node 0, zone DMA: 1 pages in unavailable ranges
[    0.068265] On node 0, zone DMA: 97 pages in unavailable ranges
[    0.124755] On node 0, zone Normal: 32 pages in unavailable ranges

cat /sys/devices/system/node/online
0-1
cat /sys/devices/system/node/possible
0-2

Link: https://lkml.kernel.org/r/20230619130442.2487-1-haifeng.xu@shopee.com
Signed-off-by: Haifeng Xu <haifeng.xu@shopee.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Shakeel Butt <shakeelb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

kasan,kmsan: remove __GFP_KSWAPD_RECLAIM usage from kasan/kmsan

syzbot is reporting lockdep warning in __stack_depot_save(), for
the caller of __stack_depot_save() (i.e. __kasan_record_aux_stack() in
this report) is responsible for masking __GFP_KSWAPD_RECLAIM flag in
order not to wake kswapd which in turn wakes kcompactd.

Since kasan/kmsan functions might be called with arbitrary locks held,
mask __GFP_KSWAPD_RECLAIM flag from all GFP_NOWAIT/GFP_ATOMIC allocations
in kasan/kmsan.

Note that kmsan_save_stack_with_flags() is changed to mask both
__GFP_DIRECT_RECLAIM flag and __GFP_KSWAPD_RECLAIM flag, for
wakeup_kswapd() from wake_all_kswapds() from __alloc_pages_slowpath()
calls wakeup_kcompactd() if __GFP_KSWAPD_RECLAIM flag is set and
__GFP_DIRECT_RECLAIM flag is not set.

Link: https://lkml.kernel.org/r/656cb4f5-998b-c8d7-3c61-c2d37aa90f9a@I-love.SAKURA.ne.jp
Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Reported-by: syzbot <syzbot+ece2915262061d6e0ac1@syzkaller.appspotmail.com>
Closes: https://syzkaller.appspot.com/bug?extid=ece2915262061d6e0ac1
Reviewed-by: "Huang, Ying" <ying.huang@intel.com>
Reviewed-by: Alexander Potapenko <glider@google.com>
Cc: Andrey Konovalov <andreyknvl@gmail.com>
Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Marco Elver <elver@google.com>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: Vincenzo Frascino <vincenzo.frascino@arm.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: compaction: skip memory hole rapidly when isolating migratable pages

On some machines, the normal zone can have a large memory hole like below
memory layout, and we can see the range from 0x100000000 to 0x1800000000
is a hole.  So when isolating some migratable pages, the scanner can meet
the hole and it will take more time to skip the large hole.  From my
measurement, I can see the isolation scanner will take 80us ~ 100us to
skip the large hole [0x100000000 - 0x1800000000].

So adding a new helper to fast search next online memory section to skip
the large hole can help to find next suitable pageblock efficiently.  With
this patch, I can see the large hole scanning only takes < 1us.

[    0.000000] Zone ranges:
[    0.000000]   DMA      [mem 0x0000000040000000-0x00000000ffffffff]
[    0.000000]   DMA32    empty
[    0.000000]   Normal   [mem 0x0000000100000000-0x0000001fa7ffffff]
[    0.000000] Movable zone start for each node
[    0.000000] Early memory node ranges
[    0.000000]   node   0: [mem 0x0000000040000000-0x0000000fffffffff]
[    0.000000]   node   0: [mem 0x0000001800000000-0x0000001fa3c7ffff]
[    0.000000]   node   0: [mem 0x0000001fa3c80000-0x0000001fa3ffffff]
[    0.000000]   node   0: [mem 0x0000001fa4000000-0x0000001fa402ffff]
[    0.000000]   node   0: [mem 0x0000001fa4030000-0x0000001fa40effff]
[    0.000000]   node   0: [mem 0x0000001fa40f0000-0x0000001fa73cffff]
[    0.000000]   node   0: [mem 0x0000001fa73d0000-0x0000001fa745ffff]
[    0.000000]   node   0: [mem 0x0000001fa7460000-0x0000001fa746ffff]
[    0.000000]   node   0: [mem 0x0000001fa7470000-0x0000001fa758ffff]
[    0.000000]   node   0: [mem 0x0000001fa7590000-0x0000001fa7ffffff]

[baolin.wang@linux.alibaba.com: limit next_ptn to not exceed cc->free_pfn]
Link: https://lkml.kernel.org/r/a1d859c28af0c7e85e91795e7473f553eb180a9d.1686813379.git.baolin.wang@linux.alibaba.com
Link: https://lkml.kernel.org/r/75b4c8ca36bf44ad8c42bf0685ac19d272e426ec.1686705221.git.baolin.wang@linux.alibaba.com
Signed-off-by: Baolin Wang <baolin.wang@linux.alibaba.com>
Suggested-by: David Hildenbrand <david@redhat.com>
Acked-by: David Hildenbrand <david@redhat.com>
Acked-by: "Huang, Ying" <ying.huang@intel.com>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

kasan, doc: note kasan.fault=panic_on_write behaviour for async modes

Note the behaviour of kasan.fault=panic_on_write for async modes, since
all asynchronous faults will result in panic (even if they are reads).

Link: https://lkml.kernel.org/r/ZJHfL6vavKUZ3Yd8@elver.google.com
Fixes: 452c03fdbed0 ("kasan: add support for kasan.fault=panic_on_write")
Signed-off-by: Marco Elver <elver@google.com>
Reviewed-by: Andrey Konovalov <andreyknvl@gmail.com>
Cc: Aleksandr Nogikh <nogikh@google.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Taras Madan <tarasmadan@google.com>
Cc: Vincenzo Frascino <vincenzo.frascino@arm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

Merge mm-hotfixes-stable into mm-stable to pick up depended-upon changes.

mm/mglru: make memcg_lru->lock irq safe

lru_gen_rotate_memcg() can happen in softirq if memory.soft_limit_in_bytes
is set.  This requires memcg_lru->lock to be irq safe.  Lockdep warns on
this.

This problem only affects memcg v1.

Link: https://lkml.kernel.org/r/20230619193821.2710944-1-yuzhao@google.com
Fixes: e4dde56cd208 ("mm: multi-gen LRU: per-node lru_gen_folio lists")
Signed-off-by: Yu Zhao <yuzhao@google.com>
Reported-by: syzbot+87c490fd2be656269b6a@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=87c490fd2be656269b6a
Reviewed-by: Yosry Ahmed <yosryahmed@google.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: page_alloc: make compound_page_dtors static

It's only used inside page_alloc.c now. So make it static and remove the
declaration in mm.h.

Link: https://lkml.kernel.org/r/20230617034622.1235913-1-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

Docs/admin-guide/mm/damon/usage: update the ways for getting monitoring results

The recommended ways for getting DAMON monitoring results are using
tried_regions sysfs directory for partial snapshot of the results, and
DAMON tracepoint for full record of the results.  However, the
tried_regions sysfs directory usage has not sufficiently updated on some
sections of the DAMON usage document.  Update those.

Link: https://lkml.kernel.org/r/20230616191742.87531-8-sj@kernel.org
Signed-off-by: SeongJae Park <sj@kernel.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

Docs/admin-guide/mm/damon/usage: clarify quotas and watermarks sysfs interface

Explanation of DAMOS quotas and watermarks are not clearly explaining the
meaning and expectation of each file.  Add more clarification for those.

Link: https://lkml.kernel.org/r/20230616191742.87531-7-sj@kernel.org
Signed-off-by: SeongJae Park <sj@kernel.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

Docs/admin-guide/mm/damon/usage: link design document for DAMOS

The background and concept of DAMOS is redundantly documented, in the
design document and the usage document.  Replace the duplicated ones in
usage document with links to the design document.

Link: https://lkml.kernel.org/r/20230616191742.87531-6-sj@kernel.org
Signed-off-by: SeongJae Park <sj@kernel.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

Docs/admin-guide/mm/damon/usage: remove unnecessary sentences about supported address spaces

Brief explanation of DAMON user space tool and sysfs interface are
unnecessarily and repeatedly mentioning the list of address spaces that
DAMON is supporting.  Remove those.

Link: https://lkml.kernel.org/r/20230616191742.87531-5-sj@kernel.org
Signed-off-by: SeongJae Park <sj@kernel.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

Docs/admin-guide/mm/damon/usage: fix typos in references and commas

Fix typos including a unnecessary comma and incomplete ':ref:' keywords.

Link: https://lkml.kernel.org/r/20230616191742.87531-4-sj@kernel.org
Signed-off-by: SeongJae Park <sj@kernel.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

Docs/admin-guide/mm/damon/start: update DAMOS example command

DAMON user-space tool, damo, has deprecated[1] its old DAMOS schemes
specification format.  However, an example of DAMON documentation is still
using it.  Update the example to use one of the alternative options.

[1] https://github.com/awslabs/damo/commit/e9950ae68f6c

Link: https://lkml.kernel.org/r/20230616191742.87531-3-sj@kernel.org
Signed-off-by: SeongJae Park <sj@kernel.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

Docs/mm/damon/design: document 'age' of region

Patch series "Docs/{mm,admin-guide}damon: update design and usage docs".

Update DAMON design and usage documents for outdated and unnecessarily
duplicated parts.

This patch (of 7):

The 'age' of each region in DAMON monitoring results is an important
concept for both monitoring part and DAMOS.  And DAMOS section of the
design document is mentioning it.  However, the age itself is not
explained in the document.  Add a section for that.

Link: https://lkml.kernel.org/r/20230616191742.87531-1-sj@kernel.org
Link: https://lkml.kernel.org/r/20230616191742.87531-2-sj@kernel.org
Signed-off-by: SeongJae Park <sj@kernel.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: move mm_count into its own cache line

The mm_struct mm_count field is frequently updated by mmgrab/mmdrop
performed by context switch.  This causes false-sharing for surrounding
mm_struct fields which are read-mostly.

This has been observed on a 2sockets/112core/224cpu Intel Sapphire Rapids
server running hackbench, and by the kernel test robot will-it-scale
testcase.

Move the mm_count field into its own cache line to prevent false-sharing
with other mm_struct fields.

Move mm_count to the first field of mm_struct to minimize the amount of
padding required: rather than adding padding before and after the mm_count
field, padding is only added after mm_count.

Note that I noticed this odd comment in mm_struct:

commit 2e3025434a6b ("mm: relocate 'write_protect_seq' in struct mm_struct")

                /*
                 * With some kernel config, the current mmap_lock's offset
                 * inside 'mm_struct' is at 0x120, which is very optimal, as
                 * its two hot fields 'count' and 'owner' sit in 2 different
                 * cachelines,  and when mmap_lock is highly contended, both
                 * of the 2 fields will be accessed frequently, current layout
                 * will help to reduce cache bouncing.
                 *
                 * So please be careful with adding new fields before
                 * mmap_lock, which can easily push the 2 fields into one
                 * cacheline.
                 */
                struct rw_semaphore mmap_lock;

This comment is rather odd for a few reasons:

- It requires addition/removal of mm_struct fields to carefully consider
  field alignment of _other_ fields,
- It expresses the wish to keep an "optimal" alignment for a specific
  kernel config.

I suspect that the author of this comment may want to revisit this topic
and perhaps introduce a split-struct approach for struct rw_semaphore,
if the need is to place various fields of this structure in different
cache lines.

Link: https://lkml.kernel.org/r/20230515143536.114960-1-mathieu.desnoyers@efficios.com
Fixes: 223baf9d17f2 ("sched: Fix performance regression introduced by mm_cid")
Fixes: af7f588d8f73 ("sched: Introduce per-memory-map concurrency ID")
Link: https://lore.kernel.org/lkml/7a0c1db1-103d-d518-ed96-1584a28fbf32@efficios.com
Reported-by: kernel test robot <yujie.liu@intel.com>
Link: https://lore.kernel.org/oe-lkp/202305151017.27581d75-yujie.liu@intel.com
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Reviewed-by: Aaron Lu <aaron.lu@intel.com>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Olivier Dion <odion@efficios.com>
Cc: <michael.christie@oracle.com>
Cc: Feng Tang <feng.tang@intel.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: Peter Xu <peterx@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/memcg: remove return value of mem_cgroup_scan_tasks()

No user checks the return value of mem_cgroup_scan_tasks(). Make the
return value void.

Link: https://lkml.kernel.org/r/20230616063030.977586-1-zhangpeng362@huawei.com
Signed-off-by: ZhangPeng <zhangpeng362@huawei.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Kefeng Wang <wangkefeng.wang@huawei.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Muchun Song <muchun.song@linux.dev>
Cc: Nanyong Sun <sunnanyong@huawei.com>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Shakeel Butt <shakeelb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/damon/core-test: add a test for damon_set_attrs()

Commit 5ff6e2fff88e ("mm/damon/core: fix divide error in
damon_nr_accesses_to_accesses_bp()") fixed a bug by adding arguments
validation in damon_set_attrs().  Add a unit test for the added validation
to ensure the bug cannot occur again.

Link: https://lkml.kernel.org/r/20230615183323.87561-1-sj@kernel.org
Signed-off-by: SeongJae Park <sj@kernel.org>
Reviewed-by: Kefeng Wang <wangkefeng.wang@huawei.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: remove is_longterm_pinnable_page() and reimplement folio_is_longterm_pinnable()

folio_is_longterm_pinnable() already exists as a wrapper function.  Now
that the whole implementation of is_longterm_pinnable_page() can be
implemented using folios, folio_is_longterm_pinnable() can be made its own
standalone function - and we can remove is_longterm_pinnable_page().

Link: https://lkml.kernel.org/r/20230614021312.34085-6-vishal.moola@gmail.com
Signed-off-by: Vishal Moola (Oracle) <vishal.moola@gmail.com>
Reviewed-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Lorenzo Stoakes <lstoakes@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/gup.c: reorganize try_get_folio()

try_get_folio() takes in a page, then chooses to do some folio operations
based on the flags (either FOLL_GET or FOLL_PIN).  We can rewrite this
function to be more purpose oriented.

After calling try_get_folio(), if neither FOLL_GET nor FOLL_PIN are set,
warn and fail.  If FOLL_GET is set we can return the result.  If FOLL_GET
is not set then FOLL_PIN is set, so we pin the folio.

This change assists with folio conversions, and makes the function more
readable.

Link: https://lkml.kernel.org/r/20230614021312.34085-5-vishal.moola@gmail.com
Signed-off-by: Vishal Moola (Oracle) <vishal.moola@gmail.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/gup_test.c: convert verify_dma_pinned() to us folios

verify_dma_pinned() checks that pages are dma-pinned.  We can convert this
to use folios.

Link: https://lkml.kernel.org/r/20230614021312.34085-4-vishal.moola@gmail.com
Signed-off-by: Vishal Moola (Oracle) <vishal.moola@gmail.com>
Reviewed-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Lorenzo Stoakes <lstoakes@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mmzone: introduce folio_migratetype()

Introduce folio_migratetype() as a folio equivalent for
get_pageblock_migratetype().  This function intends to return the
migratetype the folio is located in, hence the name choice.

Link: https://lkml.kernel.org/r/20230614021312.34085-3-vishal.moola@gmail.com
Signed-off-by: Vishal Moola (Oracle) <vishal.moola@gmail.com>
Reviewed-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mmzone: introduce folio_is_zone_movable()

Patch series "Replace is_longterm_pinnable_page()", v2.

This patchset introduces some more helper functions for the folio
conversions, and converts all callers of is_longterm_pinnable_page() to
use folios.

This patch (of 5):

Introduce folio_is_zone_movable() to act as a folio equivalent for
is_zone_movable_page().  This is to assist in later folio conversions.

Link: https://lkml.kernel.org/r/20230614021312.34085-1-vishal.moola@gmail.com
Link: https://lkml.kernel.org/r/20230614021312.34085-2-vishal.moola@gmail.com
Signed-off-by: Vishal Moola (Oracle) <vishal.moola@gmail.com>
Reviewed-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

kasan: add support for kasan.fault=panic_on_write

KASAN's boot time kernel parameter 'kasan.fault=' currently supports
'report' and 'panic', which results in either only reporting bugs or also
panicking on reports.

However, some users may wish to have more control over when KASAN reports
result in a kernel panic: in particular, KASAN reported invalid _writes_
are of special interest, because they have greater potential to corrupt
random kernel memory or be more easily exploited.

To panic on invalid writes only, introduce 'kasan.fault=panic_on_write',
which allows users to choose to continue running on invalid reads, but
panic only on invalid writes.

Link: https://lkml.kernel.org/r/20230614095158.1133673-1-elver@google.com
Signed-off-by: Marco Elver <elver@google.com>
Reviewed-by: Alexander Potapenko <glider@google.com>
Cc: Aleksandr Nogikh <nogikh@google.com>
Cc: Andrey Konovalov <andreyknvl@gmail.com>
Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Taras Madan <tarasmadan@google.com>
Cc: Vincenzo Frascino <vincenzo.frascino@arm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

zram: further limit recompression threshold

Recompression threshold should be below huge-size-class watermark.  Any
object larger than huge-size-class is a "huge object" and occupies a
whole physical page on the zsmalloc side, in other words it's
incompressible, as far as zsmalloc is concerned.

Link: https://lkml.kernel.org/r/20230614141338.3480029-1-senozhatsky@chromium.org
Signed-off-by: Sergey Senozhatsky <senozhatsky@chromium.org>
Suggested-by: Brian Geffon <bgeffon@google.com>
Acked-by: Brian Geffon <bgeffon@google.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Sergey Senozhatsky <senozhatsky@chromium.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: zswap: invaldiate entry after writeback

When an entry started writeback, it used to be invalidated with ref count
logic alone, meaning that it would stay on the tree until all references
were put.  The problem with this behavior is that as soon as the writeback
started, the ownership of the data held by the entry is passed to the
swapcache and should not be left in zswap too.  Currently there are no
known issues because of this, but this change explicitly invalidates an
entry that started writeback to reduce opportunities for future bugs.

This patch is a follow up on the series titled "mm: zswap: move writeback
LRU from zpool to zswap" + commit f090b7949768("mm: zswap: support
exclusive loads").

Link: https://lkml.kernel.org/r/20230614143122.74471-1-cerasuolodomenico@gmail.com
Signed-off-by: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
Suggested-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Nhat Pham <nphamcs@gmail.com>
Cc: Seth Jennings <sjenning@redhat.com>
Cc: Vitaly Wool <vitaly.wool@konsulko.com>
Cc: Yosry Ahmed <yosryahmed@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: kill lock|unlock_page_memcg()

Since commit c7c3dec1c9db ("mm: rmap: remove lock_page_memcg()"),
no more user, kill lock_page_memcg() and unlock_page_memcg().

Link: https://lkml.kernel.org/r/20230614143612.62575-1-wangkefeng.wang@huawei.com
Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/page_owner/cma: show pfn in cma/page_owner with hex format

cma: display pfn as well as pfn_to_page(pfn)

page_owner: display pfn in hex rather than decimal

Link: https://lkml.kernel.org/r/20230613092533.15449-1-quic_yingangl@quicinc.com
Signed-off-by: Kassey Li <quic_yingangl@quicinc.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

buffer: convert block_truncate_page() to use a folio

Support large folios in block_truncate_page() and avoid three hidden calls
to compound_head().

[willy@infradead.org: fix check of filemap_grab_folio() return value]
Link: https://lkml.kernel.org/r/ZItZOt+XxV12HtzL@casper.infradead.org
Link: https://lkml.kernel.org/r/20230612210141.730128-15-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Andreas Gruenbacher <agruenba@redhat.com>
Cc: Bob Peterson <rpeterso@redhat.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Luis Chamberlain <mcgrof@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

buffer: use a folio in __find_get_block_slow()

Saves a call to compound_head() and may be needed to support block size >
PAGE_SIZE.

Link: https://lkml.kernel.org/r/20230612210141.730128-14-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Andreas Gruenbacher <agruenba@redhat.com>
Cc: Bob Peterson <rpeterso@redhat.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Luis Chamberlain <mcgrof@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

buffer: convert link_dev_buffers to take a folio

Its one caller already has a folio, so switch it to use the folio API.
Removes a hidden call to compound_head().

Link: https://lkml.kernel.org/r/20230612210141.730128-13-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Andreas Gruenbacher <agruenba@redhat.com>
Cc: Bob Peterson <rpeterso@redhat.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Luis Chamberlain <mcgrof@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

buffer: convert init_page_buffers() to folio_init_buffers()

Use the folio API and pass the folio from both callers.  Saves a hidden
call to compound_head().

Link: https://lkml.kernel.org/r/20230612210141.730128-12-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Andreas Gruenbacher <agruenba@redhat.com>
Cc: Bob Peterson <rpeterso@redhat.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Luis Chamberlain <mcgrof@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

buffer: convert grow_dev_page() to use a folio

Get a folio from the page cache instead of a page, then use the folio API
throughout.  Removes a few calls to compound_head() and may be needed to
support block size > PAGE_SIZE.

Link: https://lkml.kernel.org/r/20230612210141.730128-11-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Andreas Gruenbacher <agruenba@redhat.com>
Cc: Bob Peterson <rpeterso@redhat.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Luis Chamberlain <mcgrof@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

buffer: convert page_zero_new_buffers() to folio_zero_new_buffers()

Most of the callers already have a folio; convert reiserfs_write_end() to
have a folio.  Removes a couple of hidden calls to compound_head().

Link: https://lkml.kernel.org/r/20230612210141.730128-10-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Andreas Gruenbacher <agruenba@redhat.com>
Cc: Bob Peterson <rpeterso@redhat.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Luis Chamberlain <mcgrof@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

buffer: convert __block_commit_write() to take a folio

This removes a hidden call to compound_head() inside
__block_commit_write() and moves it to those callers which are still page
based.  Also make block_write_end() safe for large folios.

Link: https://lkml.kernel.org/r/20230612210141.730128-9-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Andreas Gruenbacher <agruenba@redhat.com>
Cc: Bob Peterson <rpeterso@redhat.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Luis Chamberlain <mcgrof@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

buffer: convert block_page_mkwrite() to use a folio

If any page in a folio is dirtied, dirty the entire folio.  Removes a
number of hidden calls to compound_head() and references to page->mapping
and page->index.  Fixes a pre-existing bug where we could mark a folio as
dirty if the file is truncated to a multiple of the page size just as we
take the page fault.  I don't believe this bug has any bad effect, it's
just inefficient.

Link: https://lkml.kernel.org/r/20230612210141.730128-8-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Andreas Gruenbacher <agruenba@redhat.com>
Cc: Bob Peterson <rpeterso@redhat.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Luis Chamberlain <mcgrof@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

buffer: make block_write_full_page() handle large folios correctly

Keep the interface as struct page, but work entirely on the folio
internally.  Removes several PAGE_SIZE assumptions and removes some
references to page->index and page->mapping.

Link: https://lkml.kernel.org/r/20230612210141.730128-7-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Tested-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Bob Peterson <rpeterso@redhat.com>
Cc: Andreas Gruenbacher <agruenba@redhat.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Luis Chamberlain <mcgrof@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

gfs2: support ludicrously large folios in gfs2_trans_add_databufs()

We may someday support folios larger than 4GB, so use a size_t for the
byte count within a folio to prevent unpleasant truncations.

Link: https://lkml.kernel.org/r/20230612210141.730128-6-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Tested-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Luis Chamberlain <mcgrof@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

buffer: convert __block_write_full_page() to __block_write_full_folio()

Remove nine hidden calls to compound_head() by using a folio instead of a
page.

Link: https://lkml.kernel.org/r/20230612210141.730128-5-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Tested-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Bob Peterson <rpeterso@redhat.com>
Cc: Andreas Gruenbacher <agruenba@redhat.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Luis Chamberlain <mcgrof@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

gfs2: convert gfs2_write_jdata_page() to gfs2_write_jdata_folio()

Add support for large folios and remove some accesses to page->mapping and
page->index.

Link: https://lkml.kernel.org/r/20230612210141.730128-4-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Tested-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Luis Chamberlain <mcgrof@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

gfs2: pass a folio to __gfs2_jdata_write_folio()

Remove a couple of folio->page conversions in the callers, and two calls
to compound_head() in the function itself.  Rename it from
__gfs2_jdata_writepage() to __gfs2_jdata_write_folio().

Link: https://lkml.kernel.org/r/20230612210141.730128-3-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Tested-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Luis Chamberlain <mcgrof@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

gfs2: use a folio inside gfs2_jdata_writepage()

Patch series "gfs2/buffer folio changes for 6.5", v3.

This kind of started off as a gfs2 patch series, then became entwined with
buffer heads once I realised that gfs2 was the only remaining caller of
__block_write_full_page().  For those not in the gfs2 world, the big point
of this series is that block_write_full_page() should now handle large
folios correctly.

This patch (of 14):

Replace a few implicit calls to compound_head() with one explicit one.

Link: https://lkml.kernel.org/r/20230612210141.730128-1-willy@infradead.org
Link: https://lkml.kernel.org/r/20230612210141.730128-2-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Tested-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
Cc: Andreas Gruenbacher <agruenba@redhat.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Luis Chamberlain <mcgrof@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/khugepaged: use DEFINE_READ_MOSTLY_HASHTABLE macro

These are equivalent, but DEFINE_READ_MOSTLY_HASHTABLE exists to define
a hashtable in the .data..read_mostly section.

Link: https://lkml.kernel.org/r/20230609-khugepage-v1-1-dad4e8382298@google.com
Signed-off-by: Nick Desaulniers <ndesaulniers@google.com>
Reviewed-by: Yang Shi <shy828301@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

percpu-internal/pcpu_chunk: re-layout pcpu_chunk structure to reduce false sharing

When running UnixBench/Execl throughput case, false sharing is observed
due to frequent read on base_addr and write on free_bytes, chunk_md.

UnixBench/Execl represents a class of workload where bash scripts are
spawned frequently to do some short jobs.  It will do system call on execl
frequently, and execl will call mm_init to initialize mm_struct of the
process.  mm_init will call __percpu_counter_init for percpu_counters
initialization.  Then pcpu_alloc is called to read the base_addr of
pcpu_chunk for memory allocation.  Inside pcpu_alloc, it will call
pcpu_alloc_area to allocate memory from a specified chunk.  This function
will update "free_bytes" and "chunk_md" to record the rest free bytes and
other meta data for this chunk.  Correspondingly, pcpu_free_area will also
update these 2 members when free memory.

Call trace from perf is as below:
+   57.15%  0.01%  execl   [kernel.kallsyms] [k] __percpu_counter_init
+   57.13%  0.91%  execl   [kernel.kallsyms] [k] pcpu_alloc
-   55.27% 54.51%  execl   [kernel.kallsyms] [k] osq_lock
   - 53.54% 0x654278696e552f34
        main
        __execve
        entry_SYSCALL_64_after_hwframe
        do_syscall_64
        __x64_sys_execve
        do_execveat_common.isra.47
        alloc_bprm
        mm_init
        __percpu_counter_init
        pcpu_alloc
      - __mutex_lock.isra.17

In current pcpu_chunk layout, `base_addr' is in the same cache line with
`free_bytes' and `chunk_md', and `base_addr' is at the last 8 bytes.  This
patch moves `bound_map' up to `base_addr', to let `base_addr' locate in a
new cacheline.

With this change, on Intel Sapphire Rapids 112C/224T platform, based on
v6.4-rc4, the 160 parallel score improves by 24%.

The pcpu_chunk struct is a backing data structure per chunk, so the
additional memory should not be dramatic.  A chunk covers ballpark
between 64kb and 512kb memory depending on some config and boot time
stuff, so I believe the additional memory used here is nominal at best.

Working the #s on my desktop:
Percpu:            58624 kB
28 cores -> ~2.1MB of percpu memory.
At say ~128KB per chunk -> 33 chunks, generously 40 chunks.
Adding alignment might bump the chunk size ~64 bytes, so in total ~2KB
of overhead?

I believe we can do a little better to avoid eating that full padding,
so likely less than that.

[dennis@kernel.org: changelog details]
Link: https://lkml.kernel.org/r/20230610030730.110074-1-yu.ma@intel.com
Signed-off-by: Yu Ma <yu.ma@intel.com>
Reviewed-by: Tim Chen <tim.c.chen@linux.intel.com>
Acked-by: Dennis Zhou <dennis@kernel.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Liam R. Howlett <Liam.Howlett@oracle.com>
Cc: Shakeel Butt <shakeelb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

memory tier: remove unneeded !IS_ENABLED(CONFIG_MIGRATION) check

establish_demotion_targets() is defined while CONFIG_MIGRATION is
enabled. There's no need to check it again.

Link: https://lkml.kernel.org/r/20230610034114.981861-1-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Yang Shi <shy828301@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: compaction: mark kcompactd_run() and kcompactd_stop() __meminit

Add __meminit to kcompactd_run() and kcompactd_stop() to ensure they're
default to __init when memory hotplug is not enabled.

Link: https://lkml.kernel.org/r/20230610034615.997813-1-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: Baolin Wang <baolin.wang@linux.alibaba.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: remove unused vma_init_lock()

commit c7f8f31c00d1 ("mm: separate vma->lock from vm_area_struct")
left this behind.

Link: https://lkml.kernel.org/r/20230610101956.20592-1-yuehaibing@huawei.com
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

kernel: pid_namespace: remove unused set_memfd_noexec_scope()

This inline function is unused, remove it.

Link: https://lkml.kernel.org/r/20230610102858.31488-1-yuehaibing@huawei.com
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Cc: Jeff Xu <jeffxu@google.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Luis Chamberlain <mcgrof@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

userfaultfd: fix regression in userfaultfd_unmap_prep()

Android reported a performance regression in the userfaultfd unmap path.
A closer inspection on the userfaultfd_unmap_prep() change showed that a
second tree walk would be necessary in the reworked code.

Fix the regression by passing each VMA that will be unmapped through to
the userfaultfd_unmap_prep() function as they are added to the unmap list,
instead of re-walking the tree for the VMA.

Link: https://lkml.kernel.org/r/20230601015402.2819343-1-Liam.Howlett@oracle.com
Fixes: 69dbe6daf104 ("userfaultfd: use maple tree iterator to iterate VMAs")
Signed-off-by: Liam R. Howlett <Liam.Howlett@oracle.com>
Reported-by: Suren Baghdasaryan <surenb@google.com>
Suggested-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/folio: replace set_compound_order with folio_set_order

The patch ("mm/folio: Avoid special handling for order value 0 in
folio_set_order") [1] removed the need for special handling of order = 0
in folio_set_order.  Now, folio_set_order and set_compound_order becomes
similar function.  This patch removes the set_compound_order and uses
folio_set_order instead.

[1] https://lore.kernel.org/all/20230609183032.13E08C433D2@smtp.kernel.org/

Link: https://lkml.kernel.org/r/20230612093514.689846-1-tsahu@linux.ibm.com
Signed-off-by: Tarun Sahu <tsahu@linux.ibm.com>
Reviewed-by Sidhartha Kumar <sidhartha.kumar@oracle.com>
Reviewed-by: Muchun Song <songmuchun@bytedance.com>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: zswap: remove zswap_header

Previously, zswap_header served the purpose of storing the swpentry within
zpool pages.  This allowed zpool implementations to pass relevant
information to the writeback function.  However, with the current
implementation, writeback is directly handled within zswap.  Consequently,
there is no longer a necessity for zswap_header, as the swp_entry_t can be
stored directly in zswap_entry.

Link: https://lkml.kernel.org/r/20230612093815.133504-8-cerasuolodomenico@gmail.com
Signed-off-by: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
Tested-by: Yosry Ahmed <yosryahmed@google.com>
Suggested-by: Yosry Ahmed <yosryahmed@google.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Nhat Pham <nphamcs@gmail.com>
Cc: Sergey Senozhatsky <senozhatsky@chromium.org>
Cc: Seth Jennings <sjenning@redhat.com>
Cc: Vitaly Wool <vitaly.wool@konsulko.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: zswap: simplify writeback function

zswap_writeback_entry() used to be a callback for the backends, which
don't know about struct zswap_entry.

Now that the only user is the generic zswap LRU reclaimer, it can be
simplified: pass the pinned zswap_entry directly, and consolidate the
refcount management in the shrink function.

Link: https://lkml.kernel.org/r/20230612093815.133504-7-cerasuolodomenico@gmail.com
Signed-off-by: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
Tested-by: Yosry Ahmed <yosryahmed@google.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Nhat Pham <nphamcs@gmail.com>
Cc: Sergey Senozhatsky <senozhatsky@chromium.org>
Cc: Seth Jennings <sjenning@redhat.com>
Cc: Vitaly Wool <vitaly.wool@konsulko.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: zswap: remove shrink from zpool interface

Now that all three zswap backends have removed their shrink code, it is
no longer necessary for the zpool interface to include shrink/writeback
endpoints.

Link: https://lkml.kernel.org/r/20230612093815.133504-6-cerasuolodomenico@gmail.com
Signed-off-by: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
Reviewed-by: Yosry Ahmed <yosryahmed@google.com>
Acked-by: Nhat Pham <nphamcs@gmail.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: Sergey Senozhatsky <senozhatsky@chromium.org>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Seth Jennings <sjenning@redhat.com>
Cc: Vitaly Wool <vitaly.wool@konsulko.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: zswap: remove page reclaim logic from zsmalloc

Switch zsmalloc to the new generic zswap LRU and remove its custom
implementation.

Link: https://lkml.kernel.org/r/20230612093815.133504-5-cerasuolodomenico@gmail.com
Signed-off-by: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Nhat Pham <nphamcs@gmail.com>
Acked-by: Minchan Kim <minchan@kernel.org>
Tested-by: Yosry Ahmed <yosryahmed@google.com>
Acked-by: Sergey Senozhatsky <senozhatsky@chromium.org>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Seth Jennings <sjenning@redhat.com>
Cc: Vitaly Wool <vitaly.wool@konsulko.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: zswap: remove page reclaim logic from z3fold

Switch z3fold to the new generic zswap LRU and remove its custom
implementation.

Link: https://lkml.kernel.org/r/20230612093815.133504-4-cerasuolodomenico@gmail.com
Signed-off-by: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Nhat Pham <nphamcs@gmail.com>
Cc: Sergey Senozhatsky <senozhatsky@chromium.org>
Cc: Seth Jennings <sjenning@redhat.com>
Cc: Vitaly Wool <vitaly.wool@konsulko.com>
Cc: Yosry Ahmed <yosryahmed@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: zswap: remove page reclaim logic from zbud

Switch zbud to the new generic zswap LRU and remove its custom
implementation.

Link: https://lkml.kernel.org/r/20230612093815.133504-3-cerasuolodomenico@gmail.com
Signed-off-by: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Nhat Pham <nphamcs@gmail.com>
Cc: Sergey Senozhatsky <senozhatsky@chromium.org>
Cc: Seth Jennings <sjenning@redhat.com>
Cc: Vitaly Wool <vitaly.wool@konsulko.com>
Cc: Yosry Ahmed <yosryahmed@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: zswap: add pool shrinking mechanism

Patch series "mm: zswap: move writeback LRU from zpool to zswap", v3.

This series aims to improve the zswap reclaim mechanism by reorganizing
the LRU management. In the current implementation, the LRU is maintained
within each zpool driver, resulting in duplicated code across the three
drivers. The proposed change consists in moving the LRU management from
the individual implementations up to the zswap layer.

The primary objective of this refactoring effort is to simplify the
codebase. By unifying the reclaim loop and consolidating LRU handling
within zswap, we can eliminate redundant code and improve
maintainability. Additionally, this change enables the reclamation of
stored pages in their actual LRU order. Presently, the zpool drivers
link backing pages in an LRU, causing compressed pages with different
LRU positions to be written back simultaneously.

The series consists of several patches. The first patch implements the
LRU and the reclaim loop in zswap, but it is not used yet because all
three driver implementations are marked as zpool_evictable.
The following three commits modify each zpool driver to be not
zpool_evictable, allowing the use of the reclaim loop in zswap.
As the drivers removed their shrink functions, the zpool interface is
then trimmed by removing zpool_evictable, zpool_ops, and zpool_shrink.
Finally, the code in zswap is further cleaned up by simplifying the
writeback function and removing the now unnecessary zswap_header.

This patch (of 7):

Each zpool driver (zbud, z3fold and zsmalloc) implements its own shrink
function, which is called from zpool_shrink.  However, with this commit, a
unified shrink function is added to zswap.  The ultimate goal is to
eliminate the need for zpool_shrink once all zpool implementations have
dropped their shrink code.

To ensure the functionality of each commit, this change focuses solely on
adding the mechanism itself.  No modifications are made to the backends,
meaning that functionally, there are no immediate changes.  The zswap
mechanism will only come into effect once the backends have removed their
shrink code.  The subsequent commits will address the modifications needed
in the backends.

Link: https://lkml.kernel.org/r/20230612093815.133504-1-cerasuolodomenico@gmail.com
Link: https://lkml.kernel.org/r/20230612093815.133504-2-cerasuolodomenico@gmail.com
Signed-off-by: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
Acked-by: Nhat Pham <nphamcs@gmail.com>
Tested-by: Yosry Ahmed <yosryahmed@google.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: Yosry Ahmed <yosryahmed@google.com>
Reviewed-by: Sergey Senozhatsky <senozhatsky@chromium.org>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Seth Jennings <sjenning@redhat.com>
Cc: Vitaly Wool <vitaly.wool@konsulko.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

selftests: mm: remove duplicate unneeded defines

Remove all defines which aren't needed after correctly including the
kernel header files.

Link: https://lkml.kernel.org/r/20230612095347.996335-2-usama.anjum@collabora.com
Signed-off-by: Muhammad Usama Anjum <usama.anjum@collabora.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Stefan Roesch <shr@devkernel.io>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

selftests: mm: remove wrong kernel header inclusion

It is wrong to include unprocessed user header files directly.  They are
processed to "<source_tree>/usr/include" by running "make headers" and
they are included in selftests by kselftest makefiles automatically with
help of KHDR_INCLUDES variable.  These headers should always bulilt first
before building kselftests.

Link: https://lkml.kernel.org/r/20230612095347.996335-1-usama.anjum@collabora.com
Fixes: 07115fcc15b4 ("selftests/mm: add new selftests for KSM")
Signed-off-by: Muhammad Usama Anjum <usama.anjum@collabora.com>
Acked-by: David Hildenbrand <david@redhat.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Stefan Roesch <shr@devkernel.io>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: ptep_get() conversion

Convert all instances of direct pte_t* dereferencing to instead use
ptep_get() helper.  This means that by default, the accesses change from a
C dereference to a READ_ONCE().  This is technically the correct thing to
do since where pgtables are modified by HW (for access/dirty) they are
volatile and therefore we should always ensure READ_ONCE() semantics.

But more importantly, by always using the helper, it can be overridden by
the architecture to fully encapsulate the contents of the pte.  Arch code
is deliberately not converted, as the arch code knows best.  It is
intended that arch code (arm64) will override the default with its own
implementation that can (e.g.) hide certain bits from the core code, or
determine young/dirty status by mixing in state from another source.

Conversion was done using Coccinelle:

----

// $ make coccicheck \
//          COCCI=ptepget.cocci \
//          SPFLAGS="--include-headers" \
//          MODE=patch

virtual patch

@ depends on patch @
pte_t *v;
@@

- *v
+ ptep_get(v)

----

Then reviewed and hand-edited to avoid multiple unnecessary calls to
ptep_get(), instead opting to store the result of a single call in a
variable, where it is correct to do so.  This aims to negate any cost of
READ_ONCE() and will benefit arch-overrides that may be more complex.

Included is a fix for an issue in an earlier version of this patch that
was pointed out by kernel test robot.  The issue arose because config
MMU=n elides definition of the ptep helper functions, including
ptep_get().  HUGETLB_PAGE=n configs still define a simple
huge_ptep_clear_flush() for linking purposes, which dereferences the ptep.
So when both configs are disabled, this caused a build error because
ptep_get() is not defined.  Fix by continuing to do a direct dereference
when MMU=n.  This is safe because for this config the arch code cannot be
trying to virtualize the ptes because none of the ptep helpers are
defined.

Link: https://lkml.kernel.org/r/20230612151545.3317766-4-ryan.roberts@arm.com
Reported-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/oe-kbuild-all/202305120142.yXsNEo6H-lkp@intel.com/
Signed-off-by: Ryan Roberts <ryan.roberts@arm.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Alex Williamson <alex.williamson@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Andrey Konovalov <andreyknvl@gmail.com>
Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Cc: Christian Brauner <brauner@kernel.org>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: Dave Airlie <airlied@gmail.com>
Cc: Dimitri Sivanich <dimitri.sivanich@hpe.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jérôme Glisse <jglisse@redhat.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Lorenzo Stoakes <lstoakes@gmail.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Mike Rapoport (IBM) <rppt@kernel.org>
Cc: Muchun Song <muchun.song@linux.dev>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Oleksandr Tyshchenko <oleksandr_tyshchenko@epam.com>
Cc: Pavel Tatashin <pasha.tatashin@soleen.com>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: SeongJae Park <sj@kernel.org>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Uladzislau Rezki (Sony) <urezki@gmail.com>
Cc: Vincenzo Frascino <vincenzo.frascino@arm.com>
Cc: Yu Zhao <yuzhao@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: move ptep_get() and pmdp_get() helpers

There are many call sites that directly dereference a pte_t pointer.  This
makes it very difficult to properly encapsulate a page table in the arch
code without having to allocate shadow page tables.

We will shortly solve this by replacing all the call sites with ptep_get()
calls.  But there are call sites above the function definition in the
header file, so let's move ptep_get() to an earlier location to solve that
problem.  And move pmdp_get() at the same time to keep it close to
ptep_get().

Link: https://lkml.kernel.org/r/20230612151545.3317766-3-ryan.roberts@arm.com
Signed-off-by: Ryan Roberts <ryan.roberts@arm.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Alex Williamson <alex.williamson@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Andrey Konovalov <andreyknvl@gmail.com>
Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Cc: Christian Brauner <brauner@kernel.org>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: Dave Airlie <airlied@gmail.com>
Cc: Dimitri Sivanich <dimitri.sivanich@hpe.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jérôme Glisse <jglisse@redhat.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: kernel test robot <lkp@intel.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Lorenzo Stoakes <lstoakes@gmail.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Mike Rapoport (IBM) <rppt@kernel.org>
Cc: Muchun Song <muchun.song@linux.dev>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Oleksandr Tyshchenko <oleksandr_tyshchenko@epam.com>
Cc: Pavel Tatashin <pasha.tatashin@soleen.com>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: SeongJae Park <sj@kernel.org>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Uladzislau Rezki (Sony) <urezki@gmail.com>
Cc: Vincenzo Frascino <vincenzo.frascino@arm.com>
Cc: Yu Zhao <yuzhao@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: ptdump should use ptep_get_lockless()

Patch series "Encapsulate PTE contents from non-arch code", v3.

A series to improve the encapsulation of pte entries by disallowing
non-arch code from directly dereferencing pte_t pointers.

This means that by default, the accesses change from a C dereference to a
READ_ONCE().  This is technically the correct thing to do since where
pgtables are modified by HW (for access/dirty) they are volatile and
therefore we should always ensure READ_ONCE() semantics.

But more importantly, by always using the helper, it can be overridden by
the architecture to fully encapsulate the contents of the pte.  Arch code
is deliberately not converted, as the arch code knows best.  It is
intended that arch code (arm64) will override the default with its own
implementation that can (e.g.) hide certain bits from the core code, or
determine young/dirty status by mixing in state from another source.

This patch (of 3):

The page table dumper uses walk_page_range_novma() to walk the page
tables, which does not lock the PTL before calling the pte_entry()
callback.  Therefore, the page table dumper's callback must use
ptep_get_lockless() rather than ptep_get() to ensure that the pte it reads
is not torn or otherwise corrupt when racing with writers.

Link: https://lkml.kernel.org/r/20230612151545.3317766-1-ryan.roberts@arm.com
Link: https://lkml.kernel.org/r/20230612151545.3317766-2-ryan.roberts@arm.com
Signed-off-by: Ryan Roberts <ryan.roberts@arm.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Alex Williamson <alex.williamson@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Andrey Konovalov <andreyknvl@gmail.com>
Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Cc: Christian Brauner <brauner@kernel.org>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: Dave Airlie <airlied@gmail.com>
Cc: Dimitri Sivanich <dimitri.sivanich@hpe.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jérôme Glisse <jglisse@redhat.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Lorenzo Stoakes <lstoakes@gmail.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Mike Rapoport (IBM) <rppt@kernel.org>
Cc: Muchun Song <muchun.song@linux.dev>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Oleksandr Tyshchenko <oleksandr_tyshchenko@epam.com>
Cc: Pavel Tatashin <pasha.tatashin@soleen.com>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: SeongJae Park <sj@kernel.org>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Uladzislau Rezki (Sony) <urezki@gmail.com>
Cc: Vincenzo Frascino <vincenzo.frascino@arm.com>
Cc: Yu Zhao <yuzhao@google.com>
Cc: kernel test robot <lkp@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

sh: move the ARCH_DMA_MINALIGN definition to asm/cache.h

The sh architecture defines ARCH_DMA_MINALIGN in asm/page.h.  Move it to
asm/cache.h to allow a generic ARCH_DMA_MINALIGN definition in
linux/cache.h without redefine errors/warnings.

Link: https://lkml.kernel.org/r/20230613155245.1228274-4-catalin.marinas@arm.com
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: Rich Felker <dalias@libc.org>
Cc: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
Cc: Christophe Leroy <christophe.leroy@csgroup.eu>
Cc: kernel test robot <lkp@intel.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

microblaze: move the ARCH_{DMA,SLAB}_MINALIGN definitions to asm/cache.h

The microblaze architecture defines ARCH_DMA_MINALIGN in asm/page.h.  Move
it to asm/cache.h to allow a generic ARCH_DMA_MINALIGN definition in
linux/cache.h without redefine errors/warnings.

While at it, also move ARCH_SLAB_MINALIGN to asm/cache.h for
consistency.

Link: https://lkml.kernel.org/r/20230613155245.1228274-3-catalin.marinas@arm.com
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Christophe Leroy <christophe.leroy@csgroup.eu>
Cc: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
Cc: kernel test robot <lkp@intel.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Rich Felker <dalias@libc.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

powerpc: move the ARCH_DMA_MINALIGN definition to asm/cache.h

Patch series "Move the ARCH_DMA_MINALIGN definition to asm/cache.h".

The ARCH_KMALLOC_MINALIGN reduction series defines a generic
ARCH_DMA_MINALIGN in linux/cache.h:

https://lore.kernel.org/r/20230612153201.554742-2-catalin.marinas@arm.com/

Unfortunately, this causes a duplicate definition warning for
microblaze, powerpc (32-bit only) and sh as these architectures define
ARCH_DMA_MINALIGN in a different file than asm/cache.h. Move the macro
to asm/cache.h to avoid this issue and also bring them in line with the
other architectures.

This patch (of 3):

The powerpc architecture defines ARCH_DMA_MINALIGN in asm/page_32.h and
only if CONFIG_NOT_COHERENT_CACHE is enabled (32-bit platforms only).
Move this macro to asm/cache.h to allow a generic ARCH_DMA_MINALIGN
definition in linux/cache.h without redefine errors/warnings.

Link: https://lkml.kernel.org/r/20230613155245.1228274-1-catalin.marinas@arm.com
Link: https://lkml.kernel.org/r/20230613155245.1228274-2-catalin.marinas@arm.com
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202306131053.1ybvRRhO-lkp@intel.com/
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Christophe Leroy <christophe.leroy@csgroup.eu>
Cc: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Rich Felker <dalias@libc.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

arm64: enable ARCH_WANT_KMALLOC_DMA_BOUNCE for arm64

With the DMA bouncing of unaligned kmalloc() buffers now in place, enable
it for arm64 to allow the kmalloc-{8,16,32,48,96} caches.  In addition,
always create the swiotlb buffer even when the end of RAM is within the
32-bit physical address range (the swiotlb buffer can still be disabled on
the kernel command line).

Link: https://lkml.kernel.org/r/20230612153201.554742-18-catalin.marinas@arm.com
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Tested-by: Isaac J. Manjarres <isaacmanjarres@google.com>
Cc: Will Deacon <will@kernel.org>
Cc: Alasdair Kergon <agk@redhat.com>
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: Jerry Snitselaar <jsnitsel@redhat.com>
Cc: Joerg Roedel <joro@8bytes.org>
Cc: Jonathan Cameron <jic23@kernel.org>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Lars-Peter Clausen <lars@metafoo.de>
Cc: Logan Gunthorpe <logang@deltatee.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Mark Brown <broonie@kernel.org>
Cc: Mike Snitzer <snitzer@kernel.org>
Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Cc: Robin Murphy <robin.murphy@arm.com>
Cc: Saravana Kannan <saravanak@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: slab: reduce the kmalloc() minimum alignment if DMA bouncing possible

If an architecture opted in to DMA bouncing of unaligned kmalloc() buffers
(ARCH_WANT_KMALLOC_DMA_BOUNCE), reduce the minimum kmalloc() cache
alignment below cache-line size to ARCH_KMALLOC_MINALIGN.

Link: https://lkml.kernel.org/r/20230612153201.554742-17-catalin.marinas@arm.com
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Reviewed-by: Vlastimil Babka <vbabka@suse.cz>
Tested-by: Isaac J. Manjarres <isaacmanjarres@google.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Robin Murphy <robin.murphy@arm.com>
Cc: Alasdair Kergon <agk@redhat.com>
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: Jerry Snitselaar <jsnitsel@redhat.com>
Cc: Joerg Roedel <joro@8bytes.org>
Cc: Jonathan Cameron <jic23@kernel.org>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Lars-Peter Clausen <lars@metafoo.de>
Cc: Logan Gunthorpe <logang@deltatee.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Mark Brown <broonie@kernel.org>
Cc: Mike Snitzer <snitzer@kernel.org>
Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Cc: Saravana Kannan <saravanak@google.com>
Cc: Will Deacon <will@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

iommu/dma: force bouncing if the size is not cacheline-aligned

Similarly to the direct DMA, bounce small allocations as they may have
originated from a kmalloc() cache not safe for DMA. Unlike the direct
DMA, iommu_dma_map_sg() cannot call iommu_dma_map_sg_swiotlb() for all
non-coherent devices as this would break some cases where the iova is
expected to be contiguous (dmabuf). Instead, scan the scatterlist for
any small sizes and only go the swiotlb path if any element of the list
needs bouncing (note that iommu_dma_map_page() would still only bounce
those buffers which are not DMA-aligned).

To avoid scanning the scatterlist on the 'sync' operations, introduce an
SG_DMA_SWIOTLB flag set by iommu_dma_map_sg_swiotlb(). The
dev_use_swiotlb() function together with the newly added
dev_use_sg_swiotlb() now check for both untrusted devices and unaligned
kmalloc() buffers (suggested by Robin Murphy).

Link: https://lkml.kernel.org/r/20230612153201.554742-16-catalin.marinas@arm.com
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Reviewed-by: Robin Murphy <robin.murphy@arm.com>
Tested-by: Isaac J. Manjarres <isaacmanjarres@google.com>
Cc: Joerg Roedel <joro@8bytes.org>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Alasdair Kergon <agk@redhat.com>
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: Jerry Snitselaar <jsnitsel@redhat.com>
Cc: Jonathan Cameron <jic23@kernel.org>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Lars-Peter Clausen <lars@metafoo.de>
Cc: Logan Gunthorpe <logang@deltatee.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Mark Brown <broonie@kernel.org>
Cc: Mike Snitzer <snitzer@kernel.org>
Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Cc: Saravana Kannan <saravanak@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Will Deacon <will@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

dma-mapping: force bouncing if the kmalloc() size is not cache-line-aligned

For direct DMA, if the size is small enough to have originated from a
kmalloc() cache below ARCH_DMA_MINALIGN, check its alignment against
dma_get_cache_alignment() and bounce if necessary.  For larger sizes, it
is the responsibility of the DMA API caller to ensure proper alignment.

At this point, the kmalloc() caches are properly aligned but this will
change in a subsequent patch.

Architectures can opt in by selecting DMA_BOUNCE_UNALIGNED_KMALLOC.

Link: https://lkml.kernel.org/r/20230612153201.554742-15-catalin.marinas@arm.com
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Robin Murphy <robin.murphy@arm.com>
Tested-by: Isaac J. Manjarres <isaacmanjarres@google.com>
Cc: Alasdair Kergon <agk@redhat.com>
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: Jerry Snitselaar <jsnitsel@redhat.com>
Cc: Joerg Roedel <joro@8bytes.org>
Cc: Jonathan Cameron <jic23@kernel.org>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Lars-Peter Clausen <lars@metafoo.de>
Cc: Logan Gunthorpe <logang@deltatee.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Mark Brown <broonie@kernel.org>
Cc: Mike Snitzer <snitzer@kernel.org>
Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Cc: Saravana Kannan <saravanak@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Will Deacon <will@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

dma-mapping: name SG DMA flag helpers consistently

sg_is_dma_bus_address() is inconsistent with the naming pattern of its
corresponding setters and its own kerneldoc, so take the majority vote and
rename it sg_dma_is_bus_address() (and fix up the missing underscores in
the kerneldoc too).  This gives us a nice clear pattern where SG DMA flags
are SG_DMA_<NAME>, and the helpers for acting on them are
sg_dma_<action>_<name>().

Link: https://lkml.kernel.org/r/20230612153201.554742-14-catalin.marinas@arm.com
Signed-off-by: Robin Murphy <robin.murphy@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jerry Snitselaar <jsnitsel@redhat.com>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Link: https://lore.kernel.org/r/fa2eca2862c7ffc41b50337abffb2dfd2864d3ea.1685036694.git.robin.murphy@arm.com
Tested-by: Isaac J. Manjarres <isaacmanjarres@google.com>
Cc: Alasdair Kergon <agk@redhat.com>
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: Joerg Roedel <joro@8bytes.org>
Cc: Jonathan Cameron <jic23@kernel.org>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Lars-Peter Clausen <lars@metafoo.de>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Mark Brown <broonie@kernel.org>
Cc: Mike Snitzer <snitzer@kernel.org>
Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Cc: Saravana Kannan <saravanak@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Will Deacon <will@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

scatterlist: add dedicated config for DMA flags

The DMA flags field will be useful for users beyond PCI P2P, so upgrade to
its own dedicated config option.

[catalin.marinas@arm.com: use #ifdef CONFIG_NEED_SG_DMA_FLAGS in scatterlist.h]
[catalin.marinas@arm.com: update PCI_P2PDMA dma_flags comment in scatterlist.h]
Link: https://lkml.kernel.org/r/20230612153201.554742-13-catalin.marinas@arm.com
Signed-off-by: Robin Murphy <robin.murphy@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Tested-by: Isaac J. Manjarres <isaacmanjarres@google.com>
Cc: Alasdair Kergon <agk@redhat.com>
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: Jerry Snitselaar <jsnitsel@redhat.com>
Cc: Joerg Roedel <joro@8bytes.org>
Cc: Jonathan Cameron <jic23@kernel.org>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Lars-Peter Clausen <lars@metafoo.de>
Cc: Logan Gunthorpe <logang@deltatee.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Mark Brown <broonie@kernel.org>
Cc: Mike Snitzer <snitzer@kernel.org>
Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Cc: Saravana Kannan <saravanak@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Will Deacon <will@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

arm64: allow kmalloc() caches aligned to the smaller cache_line_size()

On arm64, ARCH_DMA_MINALIGN is 128, larger than the cache line size on
most of the current platforms (typically 64).  Define
ARCH_KMALLOC_MINALIGN to 8 (the default for architectures without their
own ARCH_DMA_MINALIGN) and override dma_get_cache_alignment() to return
cache_line_size(), probed at run-time.  The kmalloc() caches will be
limited to the cache line size.  This will allow the additional
kmalloc-{64,192} caches on most arm64 platforms.

Link: https://lkml.kernel.org/r/20230612153201.554742-12-catalin.marinas@arm.com
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Tested-by: Isaac J. Manjarres <isaacmanjarres@google.com>
Cc: Will Deacon <will@kernel.org>
Cc: Alasdair Kergon <agk@redhat.com>
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: Jerry Snitselaar <jsnitsel@redhat.com>
Cc: Joerg Roedel <joro@8bytes.org>
Cc: Jonathan Cameron <jic23@kernel.org>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Lars-Peter Clausen <lars@metafoo.de>
Cc: Logan Gunthorpe <logang@deltatee.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Mark Brown <broonie@kernel.org>
Cc: Mike Snitzer <snitzer@kernel.org>
Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Cc: Robin Murphy <robin.murphy@arm.com>
Cc: Saravana Kannan <saravanak@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

iio: core: use ARCH_DMA_MINALIGN instead of ARCH_KMALLOC_MINALIGN

ARCH_DMA_MINALIGN represents the minimum (static) alignment for safe DMA
operations while ARCH_KMALLOC_MINALIGN is the minimum kmalloc() objects
alignment.

Link: https://lkml.kernel.org/r/20230612153201.554742-11-catalin.marinas@arm.com
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Tested-by: Isaac J. Manjarres <isaacmanjarres@google.com>
Cc: Lars-Peter Clausen <lars@metafoo.de>
Cc: Alasdair Kergon <agk@redhat.com>
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: Jerry Snitselaar <jsnitsel@redhat.com>
Cc: Joerg Roedel <joro@8bytes.org>
Cc: Jonathan Cameron <jic23@kernel.org>
Cc: Logan Gunthorpe <logang@deltatee.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Mark Brown <broonie@kernel.org>
Cc: Mike Snitzer <snitzer@kernel.org>
Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Cc: Robin Murphy <robin.murphy@arm.com>
Cc: Saravana Kannan <saravanak@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Will Deacon <will@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

dm-crypt: use ARCH_DMA_MINALIGN instead of ARCH_KMALLOC_MINALIGN

ARCH_DMA_MINALIGN represents the minimum (static) alignment for safe DMA
operations while ARCH_KMALLOC_MINALIGN is the minimum kmalloc() objects
alignment.

Link: https://lkml.kernel.org/r/20230612153201.554742-10-catalin.marinas@arm.com
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Tested-by: Isaac J. Manjarres <isaacmanjarres@google.com>
Cc: Alasdair Kergon <agk@redhat.com>
Cc: Mike Snitzer <snitzer@kernel.org>
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: Jerry Snitselaar <jsnitsel@redhat.com>
Cc: Joerg Roedel <joro@8bytes.org>
Cc: Jonathan Cameron <jic23@kernel.org>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Lars-Peter Clausen <lars@metafoo.de>
Cc: Logan Gunthorpe <logang@deltatee.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Mark Brown <broonie@kernel.org>
Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Cc: Robin Murphy <robin.murphy@arm.com>
Cc: Saravana Kannan <saravanak@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Will Deacon <will@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>