If the number of unusable blocks is not larger than
unusable capacity, we can skip GC when checkpoint
disabling.
Signed-off-by: Weichao Guo <guoweichao@oppo.com>
Signed-off-by: Chao Yu <chao@kernel.org>
[Jaegeuk Kim: Fix missing gc_mode assignment]
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

Previously, during foreground GC, if victims contain data of pinned file,
it will fail migration of the data, and meanwhile i_gc_failures of that
pinned file may increase, and when it exceeds threshold, GC will unpin
the file, result in breaking pinfile's semantics.

In order to mitigate such condition, let's record and skip section which
has pinned file's data and give priority to select unpinned one.
Signed-off-by: Chao Yu <chao.yu@oppo.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

As Yanming reported in bugzilla:

https://bugzilla.kernel.org/show_bug.cgi?id=215916

The kernel message is shown below:

kernel BUG at fs/f2fs/segment.c:2560!
Call Trace:
 allocate_segment_by_default+0x228/0x440
 f2fs_allocate_data_block+0x13d1/0x31f0
 do_write_page+0x18d/0x710
 f2fs_outplace_write_data+0x151/0x250
 f2fs_do_write_data_page+0xef9/0x1980
 move_data_page+0x6af/0xbc0
 do_garbage_collect+0x312f/0x46f0
 f2fs_gc+0x6b0/0x3bc0
 f2fs_balance_fs+0x921/0x2260
 f2fs_write_single_data_page+0x16be/0x2370
 f2fs_write_cache_pages+0x428/0xd00
 f2fs_write_data_pages+0x96e/0xd50
 do_writepages+0x168/0x550
 __writeback_single_inode+0x9f/0x870
 writeback_sb_inodes+0x47d/0xb20
 __writeback_inodes_wb+0xb2/0x200
 wb_writeback+0x4bd/0x660
 wb_workfn+0x5f3/0xab0
 process_one_work+0x79f/0x13e0
 worker_thread+0x89/0xf60
 kthread+0x26a/0x300
 ret_from_fork+0x22/0x30
RIP: 0010:new_curseg+0xe8d/0x15f0

The root cause is: ckpt.valid_block_count is inconsistent with SIT table,
stat info indicates filesystem has free blocks, but SIT table indicates
filesystem has no free segment.

So that during garbage colloection, it triggers panic when LFS allocator
fails to find free segment.

This patch tries to fix this issue by checking consistency in between
ckpt.valid_block_count and block accounted from SIT.

Cc: stable@vger.kernel.org
Reported-by: Ming Yan <yanming@tju.edu.cn>
Signed-off-by: Chao Yu <chao.yu@oppo.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

As Yanming reported in bugzilla:

https://bugzilla.kernel.org/show_bug.cgi?id=215914

The root cause is: in a very small sized image, it's very easy to
exceed threshold of foreground GC, if we calculate free space and
dirty data based on section granularity, in corner case,
has_not_enough_free_secs() will always return true, result in
deadloop in f2fs_gc().

So this patch refactors has_not_enough_free_secs() as below to fix
this issue:
1. calculate needed space based on block granularity, and separate
all blocks to two parts, section part, and block part, comparing
section part to free section, and comparing block part to free space
in openned log.
2. account F2FS_DIRTY_NODES, F2FS_DIRTY_IMETA and F2FS_DIRTY_DENTS
as node block consumer;
3. account F2FS_DIRTY_DENTS as data block consumer;

Cc: stable@vger.kernel.org
Reported-by: Ming Yan <yanming@tju.edu.cn>
Signed-off-by: Chao Yu <chao.yu@oppo.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

As Yanming reported in bugzilla:

https://bugzilla.kernel.org/show_bug.cgi?id=215894

I have encountered a bug in F2FS file system in kernel v5.17.

I have uploaded the system call sequence as case.c, and a fuzzed image can
be found in google net disk

The kernel should enable CONFIG_KASAN=y and CONFIG_KASAN_INLINE=y. You can
reproduce the bug by running the following commands:

kernel BUG at fs/f2fs/segment.c:2291!
Call Trace:
 f2fs_invalidate_blocks+0x193/0x2d0
 f2fs_fallocate+0x2593/0x4a70
 vfs_fallocate+0x2a5/0xac0
 ksys_fallocate+0x35/0x70
 __x64_sys_fallocate+0x8e/0xf0
 do_syscall_64+0x3b/0x90
 entry_SYSCALL_64_after_hwframe+0x44/0xae

The root cause is, after image was fuzzed, block mapping info in inode
will be inconsistent with SIT table, so in f2fs_fallocate(), it will cause
panic when updating SIT with invalid blkaddr.

Let's fix the issue by adding sanity check on block address before updating
SIT table with it.

Cc: stable@vger.kernel.org
Reported-by: Ming Yan <yanming@tju.edu.cn>
Signed-off-by: Chao Yu <chao.yu@oppo.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

As Yanming reported in bugzilla:

https://bugzilla.kernel.org/show_bug.cgi?id=215897

I have encountered a bug in F2FS file system in kernel v5.17.

The kernel should enable CONFIG_KASAN=y and CONFIG_KASAN_INLINE=y. You can
reproduce the bug by running the following commands:

The kernel message is shown below:

kernel BUG at fs/f2fs/f2fs.h:2511!
Call Trace:
 f2fs_remove_inode_page+0x2a2/0x830
 f2fs_evict_inode+0x9b7/0x1510
 evict+0x282/0x4e0
 do_unlinkat+0x33a/0x540
 __x64_sys_unlinkat+0x8e/0xd0
 do_syscall_64+0x3b/0x90
 entry_SYSCALL_64_after_hwframe+0x44/0xae

The root cause is: .total_valid_block_count or .total_valid_node_count
could fuzzed to zero, then once dec_valid_node_count() was called, it
will cause BUG_ON(), this patch fixes to print warning info and set
SBI_NEED_FSCK into CP instead of panic.

Cc: stable@vger.kernel.org
Reported-by: Ming Yan <yanming@tju.edu.cn>
Signed-off-by: Chao Yu <chao.yu@oppo.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

If there's not enough free sections each of which consistis of large segments,
we can hit no free section for upcoming section allocation. Let's reclaim some
prefree segments by writing checkpoints.
Signed-off-by: Byungki Lee <dominicus79@gmail.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

As Yanming reported in bugzilla:

https://bugzilla.kernel.org/show_bug.cgi?id=215904

The kernel message is shown below:

kernel BUG at fs/f2fs/inode.c:825!
Call Trace:
 evict+0x282/0x4e0
 __dentry_kill+0x2b2/0x4d0
 shrink_dentry_list+0x17c/0x4f0
 shrink_dcache_parent+0x143/0x1e0
 do_one_tree+0x9/0x30
 shrink_dcache_for_umount+0x51/0x120
 generic_shutdown_super+0x5c/0x3a0
 kill_block_super+0x90/0xd0
 kill_f2fs_super+0x225/0x310
 deactivate_locked_super+0x78/0xc0
 cleanup_mnt+0x2b7/0x480
 task_work_run+0xc8/0x150
 exit_to_user_mode_prepare+0x14a/0x150
 syscall_exit_to_user_mode+0x1d/0x40
 do_syscall_64+0x48/0x90

The root cause is: inode node and dnode node share the same nid,
so during f2fs_evict_inode(), dnode node truncation will invalidate
its NAT entry, so when truncating inode node, it fails due to
invalid NAT entry, result in inode is still marked as dirty, fix
this issue by clearing dirty for inode and setting SBI_NEED_FSCK
flag in filesystem.

output from dump.f2fs:
[print_node_info: 354] Node ID [0xf:15] is inode
i_nid[0]                      		[0x       f : 15]

Cc: stable@vger.kernel.org
Reported-by: Ming Yan <yanming@tju.edu.cn>
Signed-off-by: Chao Yu <chao.yu@oppo.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

F2FS zoned support has power of 2 zone size assumption in many places
such as in __f2fs_issue_discard_zone, init_blkz_info. As the power of 2
requirement has been removed from the block layer, explicitly add a
condition in f2fs to allow only power of 2 zone size devices.

This condition will be relaxed once those calculation based on power of
2 is made generic.
Signed-off-by: Luis Chamberlain <mcgrof@kernel.org>
Signed-off-by: Pankaj Raghav <p.raghav@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

Instead of calling bdev_zone_sectors() multiple times, call
it once and cache the value locally. This will make the
subsequent change easier to read.
Signed-off-by: Luis Chamberlain <mcgrof@kernel.org>
Signed-off-by: Pankaj Raghav <p.raghav@samsung.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

There are multiple calculations and reads of fields of sbi that should
be protected by stat_lock. As stat_lock is not used to read these
values in statfs, this can lead to inconsistent results.
Extend the locking to prevent this issue.
Commit c9c8ed50 ("f2fs: fix to avoid potential race on
sbi->unusable_block_count access/update")
already added the use of sbi->stat_lock in statfs in
order to make the calculation of multiple, different fields atomic so
that results are consistent. This is similar to that patch regarding the
change in statfs.
Signed-off-by: Niels Dossche <dossche.niels@gmail.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

xfstests/generic/475 can give EIO all the time which give an infinite loop
to flush node page like below. Let's avoid it.

[16418.518551] Call Trace:
[16418.518553]  ? dm_submit_bio+0x48/0x400
[16418.518574]  ? submit_bio_checks+0x1ac/0x5a0
[16418.525207]  __submit_bio+0x1a9/0x230
[16418.525210]  ? kmem_cache_alloc+0x29e/0x3c0
[16418.525223]  submit_bio_noacct+0xa8/0x2b0
[16418.525226]  submit_bio+0x4d/0x130
[16418.525238]  __submit_bio+0x49/0x310 [f2fs]
[16418.525339]  ? bio_add_page+0x6a/0x90
[16418.525344]  f2fs_submit_page_bio+0x134/0x1f0 [f2fs]
[16418.525365]  read_node_page+0x125/0x1b0 [f2fs]
[16418.525388]  __get_node_page.part.0+0x58/0x3f0 [f2fs]
[16418.525409]  __get_node_page+0x2f/0x60 [f2fs]
[16418.525431]  f2fs_get_dnode_of_data+0x423/0x860 [f2fs]
[16418.525452]  ? asm_sysvec_apic_timer_interrupt+0x12/0x20
[16418.525458]  ? __mod_memcg_state.part.0+0x2a/0x30
[16418.525465]  ? __mod_memcg_lruvec_state+0x27/0x40
[16418.525467]  ? __xa_set_mark+0x57/0x70
[16418.525472]  f2fs_do_write_data_page+0x10e/0x7b0 [f2fs]
[16418.525493]  f2fs_write_single_data_page+0x555/0x830 [f2fs]
[16418.525514]  ? sysvec_apic_timer_interrupt+0x4e/0x90
[16418.525518]  ? asm_sysvec_apic_timer_interrupt+0x12/0x20
[16418.525523]  f2fs_write_cache_pages+0x303/0x880 [f2fs]
[16418.525545]  ? blk_flush_plug_list+0x47/0x100
[16418.525548]  f2fs_write_data_pages+0xfd/0x320 [f2fs]
[16418.525569]  do_writepages+0xd5/0x210
[16418.525648]  filemap_fdatawrite_wbc+0x7d/0xc0
[16418.525655]  filemap_fdatawrite+0x50/0x70
[16418.525658]  f2fs_sync_dirty_inodes+0xa4/0x230 [f2fs]
[16418.525679]  f2fs_write_checkpoint+0x16d/0x1720 [f2fs]
[16418.525699]  ? ttwu_do_wakeup+0x1c/0x160
[16418.525709]  ? ttwu_do_activate+0x6d/0xd0
[16418.525711]  ? __wait_for_common+0x11d/0x150
[16418.525715]  kill_f2fs_super+0xca/0x100 [f2fs]
[16418.525733]  deactivate_locked_super+0x3b/0xb0
[16418.525739]  deactivate_super+0x40/0x50
[16418.525741]  cleanup_mnt+0x139/0x190
[16418.525747]  __cleanup_mnt+0x12/0x20
[16418.525749]  task_work_run+0x6d/0xa0
[16418.525765]  exit_to_user_mode_prepare+0x1ad/0x1b0
[16418.525771]  syscall_exit_to_user_mode+0x27/0x50
[16418.525774]  do_syscall_64+0x48/0xc0
[16418.525776]  entry_SYSCALL_64_after_hwframe+0x44/0xae
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

The block layer for zoned disk can reorder the FUA'ed IOs. Let's use flush
command to keep the write order.
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

Syzbot triggers two WARNs in f2fs_is_valid_blkaddr and
__is_bitmap_valid. For example, in f2fs_is_valid_blkaddr,
if type is DATA_GENERIC_ENHANCE or DATA_GENERIC_ENHANCE_READ,
it invokes WARN_ON if blkaddr is not in the right range.
The call trace is as follows:

 f2fs_get_node_info+0x45f/0x1070
 read_node_page+0x577/0x1190
 __get_node_page.part.0+0x9e/0x10e0
 __get_node_page
 f2fs_get_node_page+0x109/0x180
 do_read_inode
 f2fs_iget+0x2a5/0x58b0
 f2fs_fill_super+0x3b39/0x7ca0

Fix these two WARNs by replacing WARN_ON with dump_stack.

Reported-by: syzbot+763ae12a2ede1d99d4dc@syzkaller.appspotmail.com
Signed-off-by: Dongliang Mu <mudongliangabcd@gmail.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

To move the list iterator variable into the list_for_each_entry_*()
macro in the future it should be avoided to use the list iterator
variable after the loop body.

To *never* use the list iterator variable after the loop it was
concluded to use a separate iterator variable instead of a
found boolean [1].

This removes the need to use a found variable and simply checking if
the variable was set, can determine if the break/goto was hit.

Link: https://lore.kernel.org/all/CAHk-=wgRr_D8CB-D9Kg-c=EHreAsk5SqXPwr9Y7k9sA6cWXJ6w@mail.gmail.com/Signed-off-by: Jakob Koschel <jakobkoschel@gmail.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

In preparation to limit the scope of a list iterator to the list
traversal loop, the usage of the list iterator variable 'next' should
be avoided past the loop body [1].

Instead of calling list_move_tail() on 'next' after the loop, it is
called within the loop if the correct location was found.
After the loop it covers the case if no location was found and it
should be inserted based on the 'head' of the list.

Link: https://lore.kernel.org/all/CAHk-=wgRr_D8CB-D9Kg-c=EHreAsk5SqXPwr9Y7k9sA6cWXJ6w@mail.gmail.com/Signed-off-by: Jakob Koschel <jakobkoschel@gmail.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

The list iterator variable will be a bogus pointer if no break was hit.
Dereferencing it (cur->page in this case) could load an out-of-bounds/undefined
value making it unsafe to use that in the comparision to determine if the
specific element was found.

Since 'cur->page' *can* be out-ouf-bounds it cannot be guaranteed that
by chance (or intention of an attacker) it matches the value of 'page'
even though the correct element was not found.

This is fixed by using a separate list iterator variable for the loop
and only setting the original variable if a suitable element was found.
Then determing if the element was found is simply checking if the
variable is set.

Fixes: 8c242db9 ("f2fs: fix stale ATOMIC_WRITTEN_PAGE private pointer")
Signed-off-by: Jakob Koschel <jakobkoschel@gmail.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

As Wenqing reported in bugzilla:

https://bugzilla.kernel.org/show_bug.cgi?id=215765

It will cause a kernel panic with steps:
- mkdir mnt
- mount tmp40.img mnt
- ls mnt

folio_mark_dirty+0x33/0x50
f2fs_add_regular_entry+0x541/0xad0 [f2fs]
f2fs_add_dentry+0x6c/0xb0 [f2fs]
f2fs_do_add_link+0x182/0x230 [f2fs]
__recover_dot_dentries+0x2d6/0x470 [f2fs]
f2fs_lookup+0x5af/0x6a0 [f2fs]
__lookup_slow+0xac/0x200
lookup_slow+0x45/0x70
walk_component+0x16c/0x250
path_lookupat+0x8b/0x1f0
filename_lookup+0xef/0x250
user_path_at_empty+0x46/0x70
vfs_statx+0x98/0x190
__do_sys_newlstat+0x41/0x90
__x64_sys_newlstat+0x1a/0x30
do_syscall_64+0x37/0xb0
entry_SYSCALL_64_after_hwframe+0x44/0xae

The root cause is for special file: e.g. character, block, fifo or
socket file, f2fs doesn't assign address space operations pointer array
for mapping->a_ops field, so, in a fuzzed image, if inline_dots flag was
tagged in special file, during lookup(), when f2fs runs into
__recover_dot_dentries(), it will cause NULL pointer access once
f2fs_add_regular_entry() calls a_ops->set_dirty_page().

Fixes: 510022a8 ("f2fs: add F2FS_INLINE_DOTS to recover missing dot dentries")
Reported-by: Wenqing Liu <wenqingliu0120@gmail.com>
Signed-off-by: Chao Yu <chao.yu@oppo.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

This was used in Android for a long time. Let's upstream it.
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

This can be removed, since f2fs_alloc_nid() actually doesn't require to block
checkpoint and __f2fs_build_free_nids() is covered by nm_i->nat_tree_lock.
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

So that it can reduce the possibility that file be unpinned forcely by
foreground GC due to .i_gc_failures[GC_FAILURE_PIN] exceeds threshold.
Signed-off-by: Chao Yu <chao.yu@oppo.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

In order to skip migrating section which contains data of pinned
file in advance.
Signed-off-by: Chao Yu <chao.yu@oppo.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

Pull f2fs fixes from Jaegeuk Kim:
 "This includes major bug fixes introduced in 5.18-rc1 and 5.17+:

   - Remove obsolete whint_mode (5.18-rc1)

   - Fix IO split issue caused by op_flags change in f2fs (5.18-rc1)

   - Fix a wrong condition check to detect IO failure loop (5.18-rc1)

   - Fix wrong data truncation during roll-forward (5.17+)"

* tag 'f2fs-fix-5.18' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs:
  f2fs: should not truncate blocks during roll-forward recovery
  f2fs: fix wrong condition check when failing metapage read
  f2fs: keep io_flags to avoid IO split due to different op_flags in two fio holders
  f2fs: remove obsolete whint_mode

It turns out that for the CONFIG_MMU=n builds, vmalloc_huge() was never
defined, since it's defined in mm/vmalloc.c, which doesn't get built for
the no-MMU configurations.

Just implement the trivial wrapper for the no-MMU case too.  In fact,
just make it an alias to the existing __vmalloc() function that has the
same signature.

Link: https://lore.kernel.org/all/CAMuHMdVdx2V1uhv_152Sw3_z2xE0spiaWp1d6Ko8-rYmAxUBAg@mail.gmail.com/
Link: https://lore.kernel.org/all/CA+G9fYscb1y4a17Sf5G_Aibt+WuSf-ks_Qjw9tYFy=A4sjCEug@mail.gmail.com/
Link: https://lore.kernel.org/all/20220425150356.GA4138752@roeck-us.net/Reported-and-tested-by: Linux Kernel Functional Testing <lkft@linaro.org>
Reported-and-tested-by: Geert Uytterhoeven <geert@linux-m68k.org>
Reported-by: Sudip Mukherjee <sudipm.mukherjee@gmail.com>
Reported-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Pull scheduler fix from Borislav Petkov:

 - Fix a corner case when calculating sched runqueue variables

That fix also removes a check for a zero divisor in the code, without
mentioning it.  Vincent clarified that it's ok after I whined about it:

  https://lore.kernel.org/all/CAKfTPtD2QEyZ6ADd5WrwETMOX0XOwJGnVddt7VHgfURdqgOS-Q@mail.gmail.com/

* tag 'sched_urgent_for_v5.18_rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  sched/pelt: Fix attach_entity_load_avg() corner case

Pull powerpc fixes from Michael Ellerman:

 - Partly revert a change to our timer_interrupt() that caused lockups
   with high res timers disabled.

 - Fix a bug in KVM TCE handling that could corrupt kernel memory.

 - Two commits fixing Power9/Power10 perf alternative event selection.

Thanks to Alexey Kardashevskiy, Athira Rajeev, David Gibson, Frederic
Barrat, Madhavan Srinivasan, Miguel Ojeda, and Nicholas Piggin.

* tag 'powerpc-5.18-3' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
  powerpc/perf: Fix 32bit compile
  powerpc/perf: Fix power10 event alternatives
  powerpc/perf: Fix power9 event alternatives
  KVM: PPC: Fix TCE handling for VFIO
  powerpc/time: Always set decrementer in timer_interrupt()

Pull perf fixes from Borislav Petkov:

 - Add Sapphire Rapids CPU support

 - Fix a perf vmalloc-ed buffer mapping error (PERF_USE_VMALLOC in use)

* tag 'perf_urgent_for_v5.18_rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  perf/x86/cstate: Add SAPPHIRERAPIDS_X CPU support
  perf/core: Fix perf_mmap fail when CONFIG_PERF_USE_VMALLOC enabled

Pull EDAC fix from Borislav Petkov:

 - Read the reported error count from the proper register on
   synopsys_edac

* tag 'edac_urgent_for_v5.18_rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/ras/ras:
  EDAC/synopsys: Read the error count from the correct register

Since commit 559089e0 ("vmalloc: replace VM_NO_HUGE_VMAP with
VM_ALLOW_HUGE_VMAP"), the use of hugepage mappings for vmalloc is an
opt-in strategy, because it caused a number of problems that weren't
noticed until x86 enabled it too.

One of the issues was fixed by Nick Piggin in commit 3b8000ae
("mm/vmalloc: huge vmalloc backing pages should be split rather than
compound"), but I'm still worried about page protection issues, and
VM_FLUSH_RESET_PERMS in particular.

However, like the hash table allocation case (commit f2edd118:
"page_alloc: use vmalloc_huge for large system hash"), the use of
kvmalloc() should be safe from any such games, since the returned
pointer might be a SLUB allocation, and as such no user should
reasonably be using it in any odd ways.

We also know that the allocations are fairly large, since it falls back
to the vmalloc case only when a kmalloc() fails.  So using a hugepage
mapping seems both safe and relevant.

This patch does show a weakness in the opt-in strategy: since the opt-in
flag is in the 'vm_flags', not the usual gfp_t allocation flags, very
few of the usual interfaces actually expose it.

That's not much of an issue in this case that already used one of the
fairly specialized low-level vmalloc interfaces for the allocation, but
for a lot of other vmalloc() users that might want to opt in, it's going
to be very inconvenient.

We'll either have to fix any compatibility problems, or expose it in the
gfp flags (__GFP_COMP would have made a lot of sense) to allow normal
vmalloc() users to use hugepage mappings.  That said, the cases that
really matter were probably already taken care of by the hash tabel
allocation.

Link: https://lore.kernel.org/all/20220415164413.2727220-1-song@kernel.org/
Link: https://lore.kernel.org/all/CAHk-=whao=iosX1s5Z4SF-ZGa-ebAukJoAdUJFk5SPwnofV+Vg@mail.gmail.com/
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Paul Menzel <pmenzel@molgen.mpg.de>
Cc: Song Liu <songliubraving@fb.com>
Cc: Rick Edgecombe <rick.p.edgecombe@intel.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Use vmalloc_huge() in alloc_large_system_hash() so that large system
hash (>= PMD_SIZE) could benefit from huge pages.

Note that vmalloc_huge only allocates huge pages for systems with
HAVE_ARCH_HUGE_VMALLOC.
Signed-off-by: Song Liu <song@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Rik van Riel <riel@surriel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Pull ksmbd server fixes from Steve French:

 - cap maximum sector size reported to avoid mount problems

 - reference count fix

 - fix filename rename race

* tag '5.18-rc3-ksmbd-fixes' of git://git.samba.org/ksmbd:
  ksmbd: set fixed sector size to FS_SECTOR_SIZE_INFORMATION
  ksmbd: increment reference count of parent fp
  ksmbd: remove filename in ksmbd_file

Pull ARC fixes from Vineet Gupta:

 - Assorted fixes

* tag 'arc-5.18-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/vgupta/arc:
  ARC: remove redundant READ_ONCE() in cmpxchg loop
  ARC: atomic: cleanup atomic-llsc definitions
  arc: drop definitions of pgd_index() and pgd_offset{, _k}() entirely
  ARC: dts: align SPI NOR node name with dtschema
  ARC: Remove a redundant memset()
  ARC: fix typos in comments
  ARC: entry: fix syscall_trace_exit argument

Pull SCSI fix from James Bottomley:
 "One fix for an information leak caused by copying a buffer to
  userspace without checking for error first in the sr driver"

* tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:
  scsi: sr: Do not leak information in ioctl

Pull xen fixes from Juergen Gross:
 "A simple cleanup patch and a refcount fix for Xen on Arm"

* tag 'for-linus-5.18-rc4-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
  arm/xen: Fix some refcount leaks
  xen: Convert kmap() to kmap_local_page()

Pull more drm fixes from Dave Airlie:
 "Maarten was away, so Maxine stepped up and sent me the drm-fixes
  merge, so no point leaving it for another week.

  The big change is an OF revert around bridge/panels, it may have some
  driver fallout, but hopefully this revert gets them shook out in the
  next week easier.

  Otherwise it's a bunch of locking/refcounts across drivers, a radeon
  dma_resv logic fix and some raspberry pi panel fixes.

  panel:
   - revert of patch that broke panel/bridge issues

  dma-buf:
   - remove unused header file.

  amdgpu:
   - partial revert of locking change

  radeon:
   - fix dma_resv logic inversion

  panel:
   - pi touchscreen panel init fixes

  vc4:
   - build fix
   - runtime pm refcount fix

  vmwgfx:
   - refcounting fix"

* tag 'drm-fixes-2022-04-23' of git://anongit.freedesktop.org/drm/drm:
  drm/amdgpu: partial revert "remove ctx->lock" v2
  Revert "drm: of: Lookup if child node has panel or bridge"
  Revert "drm: of: Properly try all possible cases for bridge/panel detection"
  drm/vc4: Use pm_runtime_resume_and_get to fix pm_runtime_get_sync() usage
  drm/vmwgfx: Fix gem refcounting and memory evictions
  drm/vc4: Fix build error when CONFIG_DRM_VC4=y && CONFIG_RASPBERRYPI_FIRMWARE=m
  drm/panel/raspberrypi-touchscreen: Initialise the bridge in prepare
  drm/panel/raspberrypi-touchscreen: Avoid NULL deref if not initialised
  dma-buf-map: remove renamed header file
  drm/radeon: fix logic inversion in radeon_sync_resv

Pull input fixes from Dmitry Torokhov:

 - a new set of keycodes to be used by marine navigation systems

 - minor fixes to omap4-keypad and cypress-sf drivers

* tag 'input-for-v5.18-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:
  Input: add Marine Navigation Keycodes
  Input: omap4-keypad - fix pm_runtime_get_sync() error checking
  Input: cypress-sf - register a callback to disable the regulators

Pull block fixes from Jens Axboe:
 "Just two small regression fixes for bcache"

* tag 'block-5.18-2022-04-22' of git://git.kernel.dk/linux-block:
  bcache: fix wrong bdev parameter when calling bio_alloc_clone() in do_bio_hook()
  bcache: put bch_bio_map() back to correct location in journal_write_unlocked()

Pull io_uring fixes from Jens Axboe:
 "Just two small fixes - one fixing a potential leak for the iovec for
  larger requests added in this cycle, and one fixing a theoretical leak
  with CQE_SKIP and IOPOLL"

* tag 'io_uring-5.18-2022-04-22' of git://git.kernel.dk/linux-block:
  io_uring: fix leaks on IOPOLL and CQE_SKIP
  io_uring: free iovec if file assignment fails

Merge tag 'perf-tools-fixes-for-v5.18-2022-04-22' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux

Pull perf tools fixes from Arnaldo Carvalho de Melo:

 - Fix header include for LLVM >= 14 when building with libclang.

 - Allow access to 'data_src' for auxtrace in 'perf script' with ARM SPE
   perf.data files, fixing processing data with such attributes.

 - Fix error message for test case 71 ("Convert perf time to TSC") on
   s390, where it is not supported.

* tag 'perf-tools-fixes-for-v5.18-2022-04-22' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux:
  perf test: Fix error message for test case 71 on s390, where it is not supported
  perf report: Set PERF_SAMPLE_DATA_SRC bit for Arm SPE event
  perf script: Always allow field 'data_src' for auxtrace
  perf clang: Fix header include for LLVM >= 14