mm: move tlb_flush_pending inline helpers to mm_inline.h

linux/mm_types.h should only define structure definitions, to make it
cheap to include elsewhere.  The atomic_t helper function definitions
are particularly large, so it's better to move the helpers using those
into the existing linux/mm_inline.h and only include that where needed.

As a follow-up, we may want to go through all the indirect includes in
mm_types.h and reduce them as much as possible.

Link: https://lkml.kernel.org/r/20211207125710.2503446-2-arnd@kernel.org
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Colin Cross <ccross@google.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Peter Xu <peterx@redhat.com>
Cc: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Yu Zhao <yuzhao@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Eric Biederman <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm: move anon_vma declarations to linux/mm_inline.h

The patch to add anonymous vma names causes a build failure in some
configurations:

  include/linux/mm_types.h: In function 'is_same_vma_anon_name':
  include/linux/mm_types.h:924:37: error: implicit declaration of function 'strcmp' [-Werror=implicit-function-declaration]
    924 |         return name && vma_name && !strcmp(name, vma_name);
        |                                     ^~~~~~
  include/linux/mm_types.h:22:1: note: 'strcmp' is defined in header '<string.h>'; did you forget to '#include <string.h>'?

This should not really be part of linux/mm_types.h in the first place,
as that header is meant to only contain structure defintions and need a
minimum set of indirect includes itself.

While the header clearly includes more than it should at this point,
let's not make it worse by including string.h as well, which would pull
in the expensive (compile-speed wise) fortify-string logic.

Move the new functions into a separate header that only needs to be
included in a couple of locations.

Link: https://lkml.kernel.org/r/20211207125710.2503446-1-arnd@kernel.org
Fixes: "mm: add a field to store names for private anonymous memory"
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Colin Cross <ccross@google.com>
Cc: Eric Biederman <ebiederm@xmission.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Peter Xu <peterx@redhat.com>
Cc: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Yu Zhao <yuzhao@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm: add anonymous vma name refcounting

While forking a process with high number (64K) of named anonymous vmas
the overhead caused by strdup() is noticeable.  Experiments with ARM64
Android device show up to 40% performance regression when forking a
process with 64k unpopulated anonymous vmas using the max name lengths
vs the same process with the same number of anonymous vmas having no
name.

Introduce anon_vma_name refcounted structure to avoid the overhead of
copying vma names during fork() and when splitting named anonymous vmas.

When a vma is duplicated, instead of copying the name we increment the
refcount of this structure.  Multiple vmas can point to the same
anon_vma_name as long as they increment the refcount.  The name member
of anon_vma_name structure is assigned at structure allocation time and
is never changed.  If vma name changes then the refcount of the original
structure is dropped, a new anon_vma_name structure is allocated to hold
the new name and the vma pointer is updated to point to the new
structure.

With this approach the fork() performance regressions is reduced 3-4x
times and with usecases using more reasonable number of VMAs (a few
thousand) the regressions is not measurable.

Link: https://lkml.kernel.org/r/20211019215511.3771969-3-surenb@google.com
Signed-off-by: Suren Baghdasaryan <surenb@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Colin Cross <ccross@google.com>
Cc: Cyrill Gorcunov <gorcunov@openvz.org>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: David Rientjes <rientjes@google.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Jan Glauber <jan.glauber@gmail.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: John Stultz <john.stultz@linaro.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rob Landley <rob@landley.net>
Cc: "Serge E. Hallyn" <serge.hallyn@ubuntu.com>
Cc: Shaohua Li <shli@fusionio.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm: add a field to store names for private anonymous memory

In many userspace applications, and especially in VM based applications
like Android uses heavily, there are multiple different allocators in
use.  At a minimum there is libc malloc and the stack, and in many cases
there are libc malloc, the stack, direct syscalls to mmap anonymous
memory, and multiple VM heaps (one for small objects, one for big
objects, etc.).  Each of these layers usually has its own tools to
inspect its usage; malloc by compiling a debug version, the VM through
heap inspection tools, and for direct syscalls there is usually no way
to track them.

On Android we heavily use a set of tools that use an extended version of
the logic covered in Documentation/vm/pagemap.txt to walk all pages
mapped in userspace and slice their usage by process, shared (COW) vs.
unique mappings, backing, etc.  This can account for real physical
memory usage even in cases like fork without exec (which Android uses
heavily to share as many private COW pages as possible between
processes), Kernel SamePage Merging, and clean zero pages.  It produces
a measurement of the pages that only exist in that process (USS, for
unique), and a measurement of the physical memory usage of that process
with the cost of shared pages being evenly split between processes that
share them (PSS).

If all anonymous memory is indistinguishable then figuring out the real
physical memory usage (PSS) of each heap requires either a pagemap
walking tool that can understand the heap debugging of every layer, or
for every layer's heap debugging tools to implement the pagemap walking
logic, in which case it is hard to get a consistent view of memory
across the whole system.

Tracking the information in userspace leads to all sorts of problems.
It either needs to be stored inside the process, which means every
process has to have an API to export its current heap information upon
request, or it has to be stored externally in a filesystem that somebody
needs to clean up on crashes.  It needs to be readable while the process
is still running, so it has to have some sort of synchronization with
every layer of userspace.  Efficiently tracking the ranges requires
reimplementing something like the kernel vma trees, and linking to it
from every layer of userspace.  It requires more memory, more syscalls,
more runtime cost, and more complexity to separately track regions that
the kernel is already tracking.

This patch adds a field to /proc/pid/maps and /proc/pid/smaps to show a
userspace-provided name for anonymous vmas.  The names of named
anonymous vmas are shown in /proc/pid/maps and /proc/pid/smaps as
[anon:<name>].

Userspace can set the name for a region of memory by calling

   prctl(PR_SET_VMA, PR_SET_VMA_ANON_NAME, start, len, (unsigned long)name)

Setting the name to NULL clears it.  The name length limit is 80 bytes
including NUL-terminator and is checked to contain only printable ascii
characters (including space), except '[',']','\','$' and '`'.

Ascii strings are being used to have a descriptive identifiers for vmas,
which can be understood by the users reading /proc/pid/maps or
/proc/pid/smaps.  Names can be standardized for a given system and they
can include some variable parts such as the name of the allocator or a
library, tid of the thread using it, etc.

The name is stored in a pointer in the shared union in vm_area_struct
that points to a null terminated string.  Anonymous vmas with the same
name (equivalent strings) and are otherwise mergeable will be merged.
The name pointers are not shared between vmas even if they contain the
same name.  The name pointer is stored in a union with fields that are
only used on file-backed mappings, so it does not increase memory usage.

CONFIG_ANON_VMA_NAME kernel configuration is introduced to enable this
feature.  It keeps the feature disabled by default to prevent any
additional memory overhead and to avoid confusing procfs parsers on
systems which are not ready to support named anonymous vmas.

The patch is based on the original patch developed by Colin Cross, more
specifically on its latest version [1] posted upstream by Sumit Semwal.
It used a userspace pointer to store vma names.  In that design, name
pointers could be shared between vmas.  However during the last
upstreaming attempt, Kees Cook raised concerns [2] about this approach
and suggested to copy the name into kernel memory space, perform
validity checks [3] and store as a string referenced from
vm_area_struct.

One big concern is about fork() performance which would need to strdup
anonymous vma names.  Dave Hansen suggested experimenting with
worst-case scenario of forking a process with 64k vmas having longest
possible names [4].  I ran this experiment on an ARM64 Android device
and recorded a worst-case regression of almost 40% when forking such a
process.

This regression is addressed in the followup patch which replaces the
pointer to a name with a refcounted structure that allows sharing the
name pointer between vmas of the same name.  Instead of duplicating the
string during fork() or when splitting a vma it increments the refcount.

[1] https://lore.kernel.org/linux-mm/20200901161459.11772-4-sumit.semwal@linaro.org/
[2] https://lore.kernel.org/linux-mm/202009031031.D32EF57ED@keescook/
[3] https://lore.kernel.org/linux-mm/202009031022.3834F692@keescook/
[4] https://lore.kernel.org/linux-mm/5d0358ab-8c47-2f5f-8e43-23b89d6a8e95@intel.com/

Changes for prctl(2) manual page (in the options section):

PR_SET_VMA
Sets an attribute specified in arg2 for virtual memory areas
starting from the address specified in arg3 and spanning the
size specified in arg4. arg5 specifies the value of the attribute
to be set. Note that assigning an attribute to a virtual memory
area might prevent it from being merged with adjacent virtual
memory areas due to the difference in that attribute's value.

Currently, arg2 must be one of:

PR_SET_VMA_ANON_NAME
Set a name for anonymous virtual memory areas. arg5 should
be a pointer to a null-terminated string containing the
name. The name length including null byte cannot exceed
80 bytes. If arg5 is NULL, the name of the appropriate
anonymous virtual memory areas will be reset. The name
can contain only printable ascii characters (including
                space), except '[',']','\','$' and '`'.

                This feature is available only if the kernel is built with
                the CONFIG_ANON_VMA_NAME option enabled.

[surenb@google.com: docs: proc.rst: /proc/PID/maps: fix malformed table]
Link: https://lkml.kernel.org/r/20211123185928.2513763-1-surenb@google.com
[surenb: rebased over v5.15-rc6, replaced userpointer with a kernel copy,
 added input sanitization and CONFIG_ANON_VMA_NAME config. The bulk of the
 work here was done by Colin Cross, therefore, with his permission, keeping
 him as the author]

Link: https://lkml.kernel.org/r/20211019215511.3771969-2-surenb@google.com
Signed-off-by: Colin Cross <ccross@google.com>
Signed-off-by: Suren Baghdasaryan <surenb@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Cyrill Gorcunov <gorcunov@openvz.org>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: David Rientjes <rientjes@google.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Jan Glauber <jan.glauber@gmail.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: John Stultz <john.stultz@linaro.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rob Landley <rob@landley.net>
Cc: "Serge E. Hallyn" <serge.hallyn@ubuntu.com>
Cc: Shaohua Li <shli@fusionio.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm: rearrange madvise code to allow for reuse

Patch series "mm: rearrange madvise code to allow for reuse", v11.

Avoid performance regression of the new anon vma name field refcounting it.

I checked the image sizes with allnoconfig builds:

  unpatched Linus' ToT
     text    data     bss     dec     hex filename
  1324759      32   73928 1398719 1557bf vmlinux

  After the first patch is applied (madvise refactoring)
     text    data     bss     dec     hex filename
  1322346      32   73928 1396306 154e52 vmlinux
  >>> 2413 bytes decrease vs ToT <<<

  After all patches applied with CONFIG_ANON_VMA_NAME=n
     text    data     bss     dec     hex filename
  1322337      32   73928 1396297 154e49 vmlinux
  >>> 2422 bytes decrease vs ToT <<<

  After all patches applied with CONFIG_ANON_VMA_NAME=y
     text    data     bss     dec     hex filename
  1325228      32   73928 1399188 155994 vmlinux
  >>> 469 bytes increase vs ToT <<<

This patch (of 3):

Refactor the madvise syscall to allow for parts of it to be reused by a
prctl syscall that affects vmas.

Move the code that walks vmas in a virtual address range into a function
that takes a function pointer as a parameter.  The only caller for now
is sys_madvise, which uses it to call madvise_vma_behavior on each vma,
but the next patch will add an additional caller.

Move handling all vma behaviors inside madvise_behavior, and rename it
to madvise_vma_behavior.

Move the code that updates the flags on a vma, including splitting or
merging the vma as necessary, into a new function called
madvise_update_vma.  The next patch will add support for updating a new
anon_name field as well.

Link: https://lkml.kernel.org/r/20211019215511.3771969-1-surenb@google.com
Signed-off-by: Colin Cross <ccross@google.com>
Signed-off-by: Suren Baghdasaryan <surenb@google.com>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Jan Glauber <jan.glauber@gmail.com>
Cc: John Stultz <john.stultz@linaro.org>
Cc: Rob Landley <rob@landley.net>
Cc: Cyrill Gorcunov <gorcunov@openvz.org>
Cc: Kees Cook <keescook@chromium.org>
Cc: "Serge E. Hallyn" <serge.hallyn@ubuntu.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Hugh Dickins <hughd@google.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Shaohua Li <shli@fusionio.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Minchan Kim <minchan@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm: remove redundant check about FAULT_FLAG_ALLOW_RETRY bit

Since commit 4064b9827063 ("mm: allow VM_FAULT_RETRY for multiple
times") allowed VM_FAULT_RETRY for multiple times, the
FAULT_FLAG_ALLOW_RETRY bit of fault_flag will not be changed in the page
fault path, so the following check is no longer needed:

flags & FAULT_FLAG_ALLOW_RETRY

So just remove it.

[akpm@linux-foundation.org: coding style fixes]

Link: https://lkml.kernel.org/r/20211110123358.36511-1-zhengqi.arch@bytedance.com
Signed-off-by: Qi Zheng <zhengqi.arch@bytedance.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Kirill Shutemov <kirill@shutemov.name>
Cc: Peter Xu <peterx@redhat.com>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Chengming Zhou <zhouchengming@bytedance.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

tools/testing/selftests/vm/userfaultfd.c: use swap() to make code cleaner

Fix the following coccicheck REVIEW:

 tools/testing/selftests/vm/userfaultfd.c:1531:21-22:use swap() to make code cleaner

Link: https://lkml.kernel.org/r/20211124031632.35317-1-chi.minghao@zte.com.cn
Signed-off-by: chiminghao <chi.minghao@zte.com.cn>
Reported-by: Zeal Robot <zealci@zte.com.cn>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

memcg: add per-memcg vmalloc stat

The kvmalloc* allocation functions can fallback to vmalloc allocations
and more often on long running machines.  In addition the kernel does
have __GFP_ACCOUNT kvmalloc* calls.  So, often on long running machines,
the memory.stat does not tell the complete picture which type of memory
is charged to the memcg.  So add a per-memcg vmalloc stat.

[shakeelb@google.com: page_memcg() within rcu lock, per Muchun]
Link: https://lkml.kernel.org/r/20211222052457.1960701-1-shakeelb@google.com
[akpm@linux-foundation.org: remove cast, per Muchun]
[shakeelb@google.com: remove area->page[0] checks and move to page by page accounting per Michal]
Link: https://lkml.kernel.org/r/20220104222341.3972772-1-shakeelb@google.com
Link: https://lkml.kernel.org/r/20211221215336.1922823-1-shakeelb@google.com
Signed-off-by: Shakeel Butt <shakeelb@google.com>
Acked-by: Roman Gushchin <guro@fb.com>
Reviewed-by: Muchun Song <songmuchun@bytedance.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/memcg: use struct_size() helper in kzalloc()

Make use of the struct_size() helper instead of an open-coded version,
in order to avoid any potential type mistakes or integer overflows that,
in the worst scenario, could lead to heap overflows.

Link: https://github.com/KSPP/linux/issues/160
Link: https://lkml.kernel.org/r/20211216022024.127375-1-wangweiyang2@huawei.com
Signed-off-by: Wang Weiyang <wangweiyang2@huawei.com>
Reviewed-by: Muchun Song <songmuchun@bytedance.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Vladimir Davydov <vdavydov.dev@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

memcg: better bounds on the memcg stats updates

Commit 11192d9c124d ("memcg: flush stats only if updated") added
tracking of memcg stats updates which is used by the readers to flush
only if the updates are over a certain threshold.  However each
individual update can correspond to a large value change for a given
stat.  For example adding or removing a hugepage to an LRU changes the
stat by thp_nr_pages (512 on x86_64).

Treating the update related to THP as one can keep the stat off, in
theory, by (thp_nr_pages * nr_cpus * CHARGE_BATCH) before flush.

To handle such scenarios, this patch adds consideration of the stat
update value as well instead of just the update event.  In addition let
the asyn flusher unconditionally flush the stats to put time limit on
the stats skew and hopefully a lot less readers would need to flush.

Link: https://lkml.kernel.org/r/20211118065350.697046-1-shakeelb@google.com
Signed-off-by: Shakeel Butt <shakeelb@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: "Michal Koutný" <mkoutny@suse.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/memcg: add oom_group_kill memory event

Our container agent wants to know when a container exits if it was OOM
killed or not to report to the user.  We use memory.oom.group = 1 to
ensure that OOM kills within the container's cgroup kill everything.
Existing memory.events are insufficient for knowing if this triggered:

1) Our current approach reads memory.events oom_kill and reports the
   container was killed if the value is non-zero. This is erroneous in
   some cases where containers create their children cgroups with
   memory.oom.group=1 as such OOM kills will get counted against the
   container cgroup's oom_kill counter despite not actually OOM killing
   the entire container.

2) Reading memory.events.local will fail to identify OOM kills in leaf
   cgroups (that don't set memory.oom.group) within the container
   cgroup.

This patch adds a new oom_group_kill event when memory.oom.group
triggers to allow userspace to cleanly identify when an entire cgroup is
oom killed.

[schatzberg.dan@gmail.com: changes from Johannes and Chris]
Link: https://lkml.kernel.org/r/20211213162511.2492267-1-schatzberg.dan@gmail.com
Link: https://lkml.kernel.org/r/20211203162426.3375036-1-schatzberg.dan@gmail.com
Signed-off-by: Dan Schatzberg <schatzberg.dan@gmail.com>
Reviewed-by: Roman Gushchin <guro@fb.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Chris Down <chris@chrisdown.name>
Reviewed-by: Shakeel Butt <shakeelb@google.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Zefan Li <lizefan.x@bytedance.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Vladimir Davydov <vdavydov.dev@gmail.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Alex Shi <alexs@kernel.org>
Cc: Wei Yang <richard.weiyang@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/page_counter: remove an incorrect call to propagate_protected_usage()

propagate_protected_usage() is called to propagate the usage change in
the page_counter structure.  But there is a call to this function from
page_counter_try_charge() when there is actually no usage change.  Hence
this call should be removed.

Link: https://lkml.kernel.org/r/20211118181125.3918222-1-dqiao@redhat.com
Signed-off-by: Donghai Qiao <dqiao@redhat.com>
Reviewed-by: Roman Gushchin <guro@fb.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm: memcontrol: make cgroup_memory_nokmem static

Commit 494c1dfe855e ("mm: memcg/slab: create a new set of kmalloc-cg-<n>
caches") makes cgroup_memory_nokmem global, however, it is unnecessary
because there is already a function mem_cgroup_kmem_disabled() which
exports it.

Just make it static and replace it with mem_cgroup_kmem_disabled() in
mm/slab_common.c.

Link: https://lkml.kernel.org/r/20211109065418.21693-1-songmuchun@bytedance.com
Signed-off-by: Muchun Song <songmuchun@bytedance.com>
Acked-by: Chris Down <chris@chrisdown.name>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Vladimir Davydov <vdavydov.dev@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/frontswap.c: use non-atomic '__set_bit()' when possible

The 'a' and 'b' bitmaps are local to this function, so no concurrent
access can occur.  So the non-atomic '__set_bit()' can be used to save a
few cycles.

Link: https://lkml.kernel.org/r/e52476da5cee57151745c5c3c934a69798dc6fa4.1638132190.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

shmem: fix a race between shmem_unused_huge_shrink and shmem_evict_inode

Fix a data race in commit 779750d20b93 ("shmem: split huge pages beyond
i_size under memory pressure").

Here are call traces causing race:

   Call Trace 1:
     shmem_unused_huge_shrink+0x3ae/0x410
     ? __list_lru_walk_one.isra.5+0x33/0x160
     super_cache_scan+0x17c/0x190
     shrink_slab.part.55+0x1ef/0x3f0
     shrink_node+0x10e/0x330
     kswapd+0x380/0x740
     kthread+0xfc/0x130
     ? mem_cgroup_shrink_node+0x170/0x170
     ? kthread_create_on_node+0x70/0x70
     ret_from_fork+0x1f/0x30

   Call Trace 2:
     shmem_evict_inode+0xd8/0x190
     evict+0xbe/0x1c0
     do_unlinkat+0x137/0x330
     do_syscall_64+0x76/0x120
     entry_SYSCALL_64_after_hwframe+0x3d/0xa2

A simple explanation:

Image there are 3 items in the local list (@list).  In the first
traversal, A is not deleted from @list.

  1)    A->B->C
        ^
        |
        pos (leave)

In the second traversal, B is deleted from @list.  Concurrently, A is
deleted from @list through shmem_evict_inode() since last reference
counter of inode is dropped by other thread.  Then the @list is corrupted.

  2)    A->B->C
        ^  ^
        |  |
     evict pos (drop)

We should make sure the inode is either on the global list or deleted from
any local list before iput().

Fixed by moving inodes back to global list before we put them.

[akpm@linux-foundation.org: coding style fixes]

Link: https://lkml.kernel.org/r/20211125064502.99983-1-ligang.bdlg@bytedance.com
Fixes: 779750d20b93 ("shmem: split huge pages beyond i_size under memory pressure")
Signed-off-by: Gang Li <ligang.bdlg@bytedance.com>
Reviewed-by: Muchun Song <songmuchun@bytedance.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm: shmem: don't truncate page if memory failure happens

The current behavior of memory failure is to truncate the page cache
regardless of dirty or clean.  If the page is dirty the later access
will get the obsolete data from disk without any notification to the
users.  This may cause silent data loss.  It is even worse for shmem
since shmem is in-memory filesystem, truncating page cache means
discarding data blocks.  The later read would return all zero.

The right approach is to keep the corrupted page in page cache, any
later access would return error for syscalls or SIGBUS for page fault,
until the file is truncated, hole punched or removed.  The regular
storage backed filesystems would be more complicated so this patch is
focused on shmem.  This also unblock the support for soft offlining
shmem THP.

[akpm@linux-foundation.org: coding style fixes]
[arnd@arndb.de: fix uninitialized variable use in me_pagecache_clean()]
Link: https://lkml.kernel.org/r/20211022064748.4173718-1-arnd@kernel.org
[Fix invalid pointer dereference in shmem_read_mapping_page_gfp() with a
 slight different implementation from what Ajay Garg <ajaygargnsit@gmail.com>
 and Muchun Song <songmuchun@bytedance.com> proposed and reworked the
 error handling of shmem_write_begin() suggested by Linus]
Link: https://lore.kernel.org/linux-mm/20211111084617.6746-1-ajaygargnsit@gmail.com/
Link: https://lkml.kernel.org/r/20211020210755.23964-6-shy828301@gmail.com
Link: https://lkml.kernel.org/r/20211116193247.21102-1-shy828301@gmail.com
Signed-off-by: Yang Shi <shy828301@gmail.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Cc: Hugh Dickins <hughd@google.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Ajay Garg <ajaygargnsit@gmail.com>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Andy Lavr <andy.lavr@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/gup.c: stricter check on THP migration entry during follow_pmd_mask

When BUG_ON check for THP migration entry, the existing code only check
thp_migration_supported case, but not for !thp_migration_supported case.
If !thp_migration_supported() and !pmd_present(), the original code may
dead loop in theory.  To make the BUG_ON check consistent, we need catch
both cases.

Move the BUG_ON check one step earlier, because if the bug happen we
should know it instead of depend on FOLL_MIGRATION been used by caller.

Because pmdval instead of *pmd is read by the is_pmd_migration_entry()
check, the existing code don't help to avoid useless locking within
pmd_migration_entry_wait(), so remove that check.

Link: https://lkml.kernel.org/r/20211217062559.737063-1-lixinhai.lxh@gmail.com
Signed-off-by: Li Xinhai <lixinhai.lxh@gmail.com>
Reviewed-by: "Huang, Ying" <ying.huang@intel.com>
Reviewed-by: Miaohe Lin <linmiaohe@huawei.com>
Cc: Zi Yan <ziy@nvidia.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

gup: avoid multiple user access locking/unlocking in fault_in_{read/write}able

fault_in_readable() and fault_in_writeable() perform __get_user() and
__put_user() in a loop, implying multiple user access locking/unlocking.

To avoid that, use user access blocks.

Link: https://lkml.kernel.org/r/720dcf79314acca1a78fae56d478cc851952149d.1637084492.git.christophe.leroy@csgroup.eu
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/truncate.c: remove unneeded variable

Return value directly instead of taking this in another redundant
variable.

Link: https://lkml.kernel.org/r/20211207083222.401594-1-chi.minghao@zte.com.cn
Signed-off-by: chiminghao <chi.minghao@zte.com.cn>
Reported-by: Zeal Robot <zealci@zte.com.cm>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Pankaj Gupta <pankaj.gupta@ionos.com>
Reviewed-by: Muchun Song <songmuchun@bytedance.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/debug_vm_pgtable: update comments regarding migration swap entries

Commit 4dd845b5a3e5 ("mm/swapops: rework swap entry manipulation code")
had changed migtation entry related helpers.  Just update
debug_vm_pgatble() synced documentation to reflect those changes.

Link: https://lkml.kernel.org/r/1641880417-24848-1-git-send-email-anshuman.khandual@arm.com
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm,fs: split dump_mapping() out from dump_page()

dump_mapping() is a big chunk of dump_page(), and it'd be handy to be
able to call it when we don't have a struct page.  Split it out and move
it to fs/inode.c.  Take the opportunity to simplify some of the debug
messages a little.

Link: https://lkml.kernel.org/r/20211121121056.2870061-1-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: William Kucharski <william.kucharski@oracle.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

kasan: fix quarantine conflicting with init_on_free

KASAN's quarantine might save its metadata inside freed objects.  As
this happens after the memory is zeroed by the slab allocator when
init_on_free is enabled, the memory coming out of quarantine is not
properly zeroed.

This causes lib/test_meminit.c tests to fail with Generic KASAN.

Zero the metadata when the object is removed from quarantine.

Link: https://lkml.kernel.org/r/2805da5df4b57138fdacd671f5d227d58950ba54.1640037083.git.andreyknvl@google.com
Fixes: 6471384af2a6 ("mm: security: introduce init_on_alloc=1 and init_on_free=1 boot options")
Signed-off-by: Andrey Konovalov <andreyknvl@google.com>
Reviewed-by: Marco Elver <elver@google.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Andrey Konovalov <andreyknvl@gmail.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

kasan: test: add test case for double-kmem_cache_destroy()

Add a test case for double-kmem_cache_destroy() detection.

Link: https://lkml.kernel.org/r/20211119142219.1519617-2-elver@google.com
Signed-off-by: Marco Elver <elver@google.com>
Reviewed-by: Andrey Konovalov <andreyknvl@gmail.com>
Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Christoph Lameter <cl@linux.com>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: David Rientjes <rientjes@google.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

kasan: add ability to detect double-kmem_cache_destroy()

Because mm/slab_common.c is not instrumented with software KASAN modes,
it is not possible to detect use-after-free of the kmem_cache passed
into kmem_cache_destroy().  In particular, because of the s->refcount--
and subsequent early return if non-zero, KASAN would never be able to
see the double-free via kmem_cache_free(kmem_cache, s).  To be able to
detect a double-kmem_cache_destroy(), check accessibility of the
kmem_cache, and in case of failure return early.

While KASAN_HW_TAGS is able to detect such bugs, by checking
accessibility and returning early we fail more gracefully and also avoid
corrupting reused objects (where tags mismatch).

A recent case of a double-kmem_cache_destroy() was detected by KFENCE:
https://lkml.kernel.org/r/0000000000003f654905c168b09d@google.com, which
was not detectable by software KASAN modes.

Link: https://lkml.kernel.org/r/20211119142219.1519617-1-elver@google.com
Signed-off-by: Marco Elver <elver@google.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Reviewed-by: Andrey Konovalov <andreyknvl@gmail.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Cc: Christoph Lameter <cl@linux.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Pekka Enberg <penberg@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

kasan: test: add globals left-out-of-bounds test

Add a test checking that KASAN generic can also detect out-of-bounds
accesses to the left of globals.

Unfortunately it seems that GCC doesn't catch this (tested GCC 10, 11).
The main difference between GCC's globals redzoning and Clang's is that
GCC relies on using increased alignment to producing padding, where
Clang's redzoning implementation actually adds real data after the
global and doesn't rely on alignment to produce padding.  I believe this
is the main reason why GCC can't reliably catch globals out-of-bounds in
this case.

Given this is now a known issue, to avoid failing the whole test suite,
skip this test case with GCC.

Link: https://lkml.kernel.org/r/20211117130714.135656-1-elver@google.com
Signed-off-by: Marco Elver <elver@google.com>
Reported-by: Kaiwan N Billimoria <kaiwan.billimoria@gmail.com>
Reviewed-by: Andrey Konovalov <andreyknvl@gmail.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Kaiwan N Billimoria <kaiwan.billimoria@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

device-dax: compound devmap support

Use the newly added compound devmap facility which maps the assigned dax
ranges as compound pages at a page size of @align.

dax devices are created with a fixed @align (huge page size) which is
enforced through as well at mmap() of the device.  Faults, consequently
happen too at the specified @align specified at the creation, and those
don't change throughout dax device lifetime.  MCEs unmap a whole dax
huge page, as well as splits occurring at the configured page size.

Performance measured by gup_test improves considerably for
unpin_user_pages() and altmap with NVDIMMs:

  $ gup_test -f /dev/dax1.0 -m 16384 -r 10 -S -a -n 512 -w
  (pin_user_pages_fast 2M pages) put:~71 ms -> put:~22 ms
  [altmap]
  (pin_user_pages_fast 2M pages) get:~524ms put:~525 ms -> get: ~127ms put:~71ms

   $ gup_test -f /dev/dax1.0 -m 129022 -r 10 -S -a -n 512 -w
  (pin_user_pages_fast 2M pages) put:~513 ms -> put:~188 ms
  [altmap with -m 127004]
  (pin_user_pages_fast 2M pages) get:~4.1 secs put:~4.12 secs -> get:~1sec put:~563ms

.. as well as unpin_user_page_range_dirty_lock() being just as effective
as THP/hugetlb[0] pages.

[0] https://lore.kernel.org/linux-mm/20210212130843.13865-5-joao.m.martins@oracle.com/

Link: https://lkml.kernel.org/r/20211202204422.26777-12-joao.m.martins@oracle.com
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Reviewed-by: Dan Williams <dan.j.williams@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Dave Jiang <dave.jiang@intel.com>
Cc: Jane Chu <jane.chu@oracle.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Vishal Verma <vishal.l.verma@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

device-dax: remove pfn from __dev_dax_{pte,pmd,pud}_fault()

After moving the page mapping to be set prior to pte insertion, the pfn
in dev_dax_huge_fault() no longer is necessary.  Remove it, as well as
the @pfn argument passed to the internal fault handler helpers.

[akpm@linux-foundation.org: fix CONFIG_HAVE_ARCH_TRANSPARENT_HUGEPAGE_PUD=n build]

Link: https://lkml.kernel.org/r/20211202204422.26777-11-joao.m.martins@oracle.com
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Suggested-by: Christoph Hellwig <hch@lst.de>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Jiang <dave.jiang@intel.com>
Cc: Jane Chu <jane.chu@oracle.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Vishal Verma <vishal.l.verma@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

device-dax: set mapping prior to vmf_insert_pfn{,_pmd,pud}()

Normally, the @page mapping is set prior to inserting the page into a
page table entry.  Make device-dax adhere to the same ordering, rather
than setting mapping after the PTE is inserted.

The address_space never changes and it is always associated with the
same inode and underlying pages.  So, the page mapping is set once but
cleared when the struct pages are removed/freed (i.e.  after
{devm_}memunmap_pages()).

Link: https://lkml.kernel.org/r/20211202204422.26777-10-joao.m.martins@oracle.com
Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Jiang <dave.jiang@intel.com>
Cc: Jane Chu <jane.chu@oracle.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Vishal Verma <vishal.l.verma@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

device-dax: factor out page mapping initialization

Move initialization of page->mapping into a separate helper.

This is in preparation to move the mapping set to be prior to inserting
the page table entry and also for tidying up compound page handling into
one helper.

Link: https://lkml.kernel.org/r/20211202204422.26777-9-joao.m.martins@oracle.com
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Jiang <dave.jiang@intel.com>
Cc: Jane Chu <jane.chu@oracle.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Vishal Verma <vishal.l.verma@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

device-dax: ensure dev_dax->pgmap is valid for dynamic devices

Right now, only static dax regions have a valid @pgmap pointer in its
struct dev_dax.  Dynamic dax case however, do not.

In preparation for device-dax compound devmap support, make sure that
dev_dax pgmap field is set after it has been allocated and initialized.

dynamic dax device have the @pgmap is allocated at probe() and it's
managed by devm (contrast to static dax region which a pgmap is provided
and dax core kfrees it).  So in addition to ensure a valid @pgmap, clear
the pgmap when the dynamic dax device is released to avoid the same
pgmap ranges to be re-requested across multiple region device reconfigs.

Add a static_dev_dax() and use that helper in dev_dax_probe() to ensure
the initialization differences between dynamic and static regions are
more explicit.  While at it, consolidate the ranges initialization when
we allocate the @pgmap for the dynamic dax region case.  Also take the
opportunity to document the differences between static and dynamic da
regions.

Link: https://lkml.kernel.org/r/20211202204422.26777-8-joao.m.martins@oracle.com
Suggested-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Dave Jiang <dave.jiang@intel.com>
Cc: Jane Chu <jane.chu@oracle.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Vishal Verma <vishal.l.verma@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

device-dax: use struct_size()

Use the struct_size() helper for the size of a struct with variable
array member at the end, rather than manually calculating it.

Link: https://lkml.kernel.org/r/20211202204422.26777-7-joao.m.martins@oracle.com
Suggested-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Dave Jiang <dave.jiang@intel.com>
Cc: Jane Chu <jane.chu@oracle.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Vishal Verma <vishal.l.verma@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

device-dax: use ALIGN() for determining pgoff

Rather than calculating @pgoff manually, switch to ALIGN() instead.

Link: https://lkml.kernel.org/r/20211202204422.26777-6-joao.m.martins@oracle.com
Suggested-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Reviewed-by: Dan Williams <dan.j.williams@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Dave Jiang <dave.jiang@intel.com>
Cc: Jane Chu <jane.chu@oracle.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Vishal Verma <vishal.l.verma@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/memremap: add ZONE_DEVICE support for compound pages

Add a new @vmemmap_shift property for struct dev_pagemap which specifies
that a devmap is composed of a set of compound pages of order
@vmemmap_shift, instead of base pages.  When a compound page devmap is
requested, all but the first page are initialised as tail pages instead
of order-0 pages.

For certain ZONE_DEVICE users like device-dax which have a fixed page
size, this creates an opportunity to optimize GUP and GUP-fast walkers,
treating it the same way as THP or hugetlb pages.

Additionally, commit 7118fc2906e2 ("hugetlb: address ref count racing in
prep_compound_gigantic_page") removed set_page_count() because the
setting of page ref count to zero was redundant.  devmap pages don't
come from page allocator though and only head page refcount is used for
compound pages, hence initialize tail page count to zero.

Link: https://lkml.kernel.org/r/20211202204422.26777-5-joao.m.martins@oracle.com
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Reviewed-by: Dan Williams <dan.j.williams@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Dave Jiang <dave.jiang@intel.com>
Cc: Jane Chu <jane.chu@oracle.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Vishal Verma <vishal.l.verma@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/page_alloc: refactor memmap_init_zone_device() page init

Move struct page init to an helper function __init_zone_device_page().

This is in preparation for sharing the storage for compound page
metadata.

Link: https://lkml.kernel.org/r/20211202204422.26777-4-joao.m.martins@oracle.com
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Reviewed-by: Dan Williams <dan.j.williams@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Dave Jiang <dave.jiang@intel.com>
Cc: Jane Chu <jane.chu@oracle.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Vishal Verma <vishal.l.verma@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/page_alloc: split prep_compound_page into head and tail subparts

Patch series "mm, device-dax: Introduce compound pages in devmap", v7.

This series converts device-dax to use compound pages, and moves away
from the 'struct page per basepage on PMD/PUD' that is done today.

Doing so
 1) unlocks a few noticeable improvements on unpin_user_pages() and
    makes device-dax+altmap case 4x times faster in pinning (numbers
    below and in last patch)
 2) as mentioned in various other threads it's one important step
    towards cleaning up ZONE_DEVICE refcounting.

I've split the compound pages on devmap part from the rest based on
recent discussions on devmap pending and future work planned[5][6].
There is consensus that device-dax should be using compound pages to
represent its PMD/PUDs just like HugeTLB and THP, and that leads to less
specialization of the dax parts.  I will pursue the rest of the work in
parallel once this part is merged, particular the GUP-{slow,fast}
improvements [7] and the tail struct page deduplication memory savings
part[8].

To summarize what the series does:

Patch 1: Prepare hwpoisoning to work with dax compound pages.

Patches 2-3: Split the current utility function of prep_compound_page()
into head and tail and use those two helpers where appropriate to take
advantage of caches being warm after __init_single_page().  This is used
when initializing zone device when we bring up device-dax namespaces.

Patches 4-10: Add devmap support for compound pages in device-dax.
memmap_init_zone_device() initialize its metadata as compound pages, and
it introduces a new devmap property known as vmemmap_shift which
outlines how the vmemmap is structured (defaults to base pages as done
today).  The property describe the page order of the metadata
essentially.  While at it do a few cleanups in device-dax in patches
5-9.  Finally enable device-dax usage of devmap @vmemmap_shift to a
value based on its own @align property.  @vmemmap_shift returns 0 by
default (which is today's case of base pages in devmap, like fsdax or
the others) and the usage of compound devmap is optional.  Starting with
device-dax (*not* fsdax) we enable it by default.  There are a few
pinning improvements particular on the unpinning case and altmap, as
well as unpin_user_page_range_dirty_lock() being just as effective as
THP/hugetlb[0] pages.

    $ gup_test -f /dev/dax1.0 -m 16384 -r 10 -S -a -n 512 -w
    (pin_user_pages_fast 2M pages) put:~71 ms -> put:~22 ms
    [altmap]
    (pin_user_pages_fast 2M pages) get:~524ms put:~525 ms -> get: ~127ms put:~71ms

     $ gup_test -f /dev/dax1.0 -m 129022 -r 10 -S -a -n 512 -w
    (pin_user_pages_fast 2M pages) put:~513 ms -> put:~188 ms
    [altmap with -m 127004]
    (pin_user_pages_fast 2M pages) get:~4.1 secs put:~4.12 secs -> get:~1sec put:~563ms

Tested on x86 with 1Tb+ of pmem (alongside registering it with RDMA with
and without altmap), alongside gup_test selftests with dynamic dax
regions and static dax regions.  Coupled with ndctl unit tests for
dynamic dax devices that exercise all of this.  Note, for dynamic dax
regions I had to revert commit 8aa83e6395 ("x86/setup: Call
early_reserve_memory() earlier"), it is a known issue that this commit
broke efi_fake_mem=.

This patch (of 11):

Split the utility function prep_compound_page() into head and tail
counterparts, and use them accordingly.

This is in preparation for sharing the storage for compound page
metadata.

Link: https://lkml.kernel.org/r/20211202204422.26777-1-joao.m.martins@oracle.com
Link: https://lkml.kernel.org/r/20211202204422.26777-3-joao.m.martins@oracle.com
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Acked-by: Mike Kravetz <mike.kravetz@oracle.com>
Reviewed-by: Dan Williams <dan.j.williams@intel.com>
Reviewed-by: Muchun Song <songmuchun@bytedance.com>
Cc: Vishal Verma <vishal.l.verma@intel.com>
Cc: Dave Jiang <dave.jiang@intel.com>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Jane Chu <jane.chu@oracle.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm: defer kmemleak object creation of module_alloc()

Yongqiang reports a kmemleak panic when module insmod/rmmod with KASAN
enabled(without KASAN_VMALLOC) on x86[1].

When the module area allocates memory, it's kmemleak_object is created
successfully, but the KASAN shadow memory of module allocation is not
ready, so when kmemleak scan the module's pointer, it will panic due to
no shadow memory with KASAN check.

  module_alloc
    __vmalloc_node_range
      kmemleak_vmalloc
kmemleak_scan
  update_checksum
    kasan_module_alloc
      kmemleak_ignore

Note, there is no problem if KASAN_VMALLOC enabled, the modules area
entire shadow memory is preallocated.  Thus, the bug only exits on ARCH
which supports dynamic allocation of module area per module load, for
now, only x86/arm64/s390 are involved.

Add a VM_DEFER_KMEMLEAK flags, defer vmalloc'ed object register of
kmemleak in module_alloc() to fix this issue.

[1] https://lore.kernel.org/all/6d41e2b9-4692-5ec4-b1cd-cbe29ae89739@huawei.com/

[wangkefeng.wang@huawei.com: fix build]
Link: https://lkml.kernel.org/r/20211125080307.27225-1-wangkefeng.wang@huawei.com
[akpm@linux-foundation.org: simplify ifdefs, per Andrey]
Link: https://lkml.kernel.org/r/CA+fCnZcnwJHUQq34VuRxpdoY6_XbJCDJ-jopksS5Eia4PijPzw@mail.gmail.com
Link: https://lkml.kernel.org/r/20211124142034.192078-1-wangkefeng.wang@huawei.com
Fixes: 793213a82de4 ("s390/kasan: dynamic shadow mem allocation for modules")
Fixes: 39d114ddc682 ("arm64: add KASAN support")
Fixes: bebf56a1b176 ("kasan: enable instrumentation of global variables")
Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
Reported-by: Yongqiang Liu <liuyongqiang13@huawei.com>
Cc: Andrey Konovalov <andreyknvl@gmail.com>
Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Christian Borntraeger <borntraeger@linux.ibm.com>
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Kefeng Wang <wangkefeng.wang@huawei.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm: kmemleak: alloc gray object for reserved region with direct map

Reserved regions with direct mapping may contain references to other
regions.  CMA region with fixed location is reserved without creating
kmemleak_object for it.

So add them as gray kmemleak objects.

Link: https://lkml.kernel.org/r/20211123090641.3654006-1-calvinzhang.cool@gmail.com
Signed-off-by: Calvin Zhang <calvinzhang.cool@gmail.com>
Cc: Rob Herring <robh+dt@kernel.org>
Cc: Frank Rowand <frowand.list@gmail.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

kmemleak: fix kmemleak false positive report with HW tag-based kasan enable

With HW tag-based kasan enable, We will get the warning when we free
object whose address starts with 0xFF.

It is because kmemleak rbtree stores tagged object and this freeing
object's tag does not match with rbtree object.

In the example below, kmemleak rbtree stores the tagged object in the
kmalloc(), and kfree() gets the pointer with 0xFF tag.

Call sequence:
    ptr = kmalloc(size, GFP_KERNEL);
    page = virt_to_page(ptr);
    offset = offset_in_page(ptr);
    kfree(page_address(page) + offset);
    ptr = kmalloc(size, GFP_KERNEL);

A sequence like that may cause the warning as following:

 1) Freeing unknown object:

    In kfree(), we will get free unknown object warning in
    kmemleak_free(). Because object(0xFx) in kmemleak rbtree and
    pointer(0xFF) in kfree() have different tag.

 2) Overlap existing:

    When we allocate that object with the same hw-tag again, we will
    find the overlap in the kmemleak rbtree and kmemleak thread will be
    killed.

kmemleak: Freeing unknown object at 0xffff000003f88000
CPU: 5 PID: 177 Comm: cat Not tainted 5.16.0-rc1-dirty #21
Hardware name: linux,dummy-virt (DT)
Call trace:
 dump_backtrace+0x0/0x1ac
 show_stack+0x1c/0x30
 dump_stack_lvl+0x68/0x84
 dump_stack+0x1c/0x38
 kmemleak_free+0x6c/0x70
 slab_free_freelist_hook+0x104/0x200
 kmem_cache_free+0xa8/0x3d4
 test_version_show+0x270/0x3a0
 module_attr_show+0x28/0x40
 sysfs_kf_seq_show+0xb0/0x130
 kernfs_seq_show+0x30/0x40
 seq_read_iter+0x1bc/0x4b0
 seq_read_iter+0x1bc/0x4b0
 kernfs_fop_read_iter+0x144/0x1c0
 generic_file_splice_read+0xd0/0x184
 do_splice_to+0x90/0xe0
 splice_direct_to_actor+0xb8/0x250
 do_splice_direct+0x88/0xd4
 do_sendfile+0x2b0/0x344
 __arm64_sys_sendfile64+0x164/0x16c
 invoke_syscall+0x48/0x114
 el0_svc_common.constprop.0+0x44/0xec
 do_el0_svc+0x74/0x90
 el0_svc+0x20/0x80
 el0t_64_sync_handler+0x1a8/0x1b0
 el0t_64_sync+0x1ac/0x1b0
...
kmemleak: Cannot insert 0xf2ff000003f88000 into the object search tree (overlaps existing)
CPU: 5 PID: 178 Comm: cat Not tainted 5.16.0-rc1-dirty #21
Hardware name: linux,dummy-virt (DT)
Call trace:
 dump_backtrace+0x0/0x1ac
 show_stack+0x1c/0x30
 dump_stack_lvl+0x68/0x84
 dump_stack+0x1c/0x38
 create_object.isra.0+0x2d8/0x2fc
 kmemleak_alloc+0x34/0x40
 kmem_cache_alloc+0x23c/0x2f0
 test_version_show+0x1fc/0x3a0
 module_attr_show+0x28/0x40
 sysfs_kf_seq_show+0xb0/0x130
 kernfs_seq_show+0x30/0x40
 seq_read_iter+0x1bc/0x4b0
 kernfs_fop_read_iter+0x144/0x1c0
 generic_file_splice_read+0xd0/0x184
 do_splice_to+0x90/0xe0
 splice_direct_to_actor+0xb8/0x250
 do_splice_direct+0x88/0xd4
 do_sendfile+0x2b0/0x344
 __arm64_sys_sendfile64+0x164/0x16c
 invoke_syscall+0x48/0x114
 el0_svc_common.constprop.0+0x44/0xec
 do_el0_svc+0x74/0x90
 el0_svc+0x20/0x80
 el0t_64_sync_handler+0x1a8/0x1b0
 el0t_64_sync+0x1ac/0x1b0
kmemleak: Kernel memory leak detector disabled
kmemleak: Object 0xf2ff000003f88000 (size 128):
kmemleak:   comm "cat", pid 177, jiffies 4294921177
kmemleak:   min_count = 1
kmemleak:   count = 0
kmemleak:   flags = 0x1
kmemleak:   checksum = 0
kmemleak:   backtrace:
     kmem_cache_alloc+0x23c/0x2f0
     test_version_show+0x1fc/0x3a0
     module_attr_show+0x28/0x40
     sysfs_kf_seq_show+0xb0/0x130
     kernfs_seq_show+0x30/0x40
     seq_read_iter+0x1bc/0x4b0
     kernfs_fop_read_iter+0x144/0x1c0
     generic_file_splice_read+0xd0/0x184
     do_splice_to+0x90/0xe0
     splice_direct_to_actor+0xb8/0x250
     do_splice_direct+0x88/0xd4
     do_sendfile+0x2b0/0x344
     __arm64_sys_sendfile64+0x164/0x16c
     invoke_syscall+0x48/0x114
     el0_svc_common.constprop.0+0x44/0xec
     do_el0_svc+0x74/0x90
kmemleak: Automatic memory scanning thread ended

[akpm@linux-foundation.org: whitespace tweak]

Link: https://lkml.kernel.org/r/20211118054426.4123-1-Kuan-Ying.Lee@mediatek.com
Signed-off-by: Kuan-Ying Lee <Kuan-Ying.Lee@mediatek.com>
Reviewed-by: Catalin Marinas <catalin.marinas@arm.com>
Cc: Doug Berger <opendmb@gmail.com>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm: slab: make slab iterator functions static

There is no external users of slab_start/next/stop(), so make them
static.  And the memory.kmem.slabinfo is deprecated, which outputs
nothing now, so move memcg_slab_show() into mm/memcontrol.c and rename
it to mem_cgroup_slab_show to be consistent with other function names.

Link: https://lkml.kernel.org/r/20211109133359.32881-1-songmuchun@bytedance.com
Signed-off-by: Muchun Song <songmuchun@bytedance.com>
Reviewed-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Christoph Lameter <cl@linux.com>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: David Rientjes <rientjes@google.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/slab_common: use WARN() if cache still has objects on destroy

Calling kmem_cache_destroy() while the cache still has objects allocated
is a kernel bug, and will usually result in the entire cache being
leaked.  While the message in kmem_cache_destroy() resembles a warning,
it is currently not implemented using a real WARN().

This is problematic for infrastructure testing the kernel, all of which
rely on the specific format of WARN()s to pick up on bugs.

Some 13 years ago this used to be a simple WARN_ON() in slub, but commit
d629d8195793 ("slub: improve kmem_cache_destroy() error message")
changed it into an open-coded warning to avoid confusion with a bug in
slub itself.

Instead, turn the open-coded warning into a real WARN() with the message
preserved, so that test systems can actually identify these issues, and
we get all the other benefits of using a normal WARN().  The warning
message is extended with "when called from <caller-ip>" to make it even
clearer where the fault lies.

For most configurations this is only a cosmetic change, however, note
that WARN() here will now also respect panic_on_warn.

Link: https://lkml.kernel.org/r/20211102170733.648216-1-elver@google.com
Signed-off-by: Marco Elver <elver@google.com>
Reviewed-by: Vlastimil Babka <vbabka@suse.cz>
Acked-by: David Rientjes <rientjes@google.com>
Cc: Christoph Lameter <cl@linux.com>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Ingo Molnar <mingo@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

fs/ioctl: remove unnecessary __user annotation

__user annotations are used by the checker (e.g sparse) to mark user
pointers.  However here __user is applied to a struct directly, without a
pointer being directly involved.

Although the presence of __user does not cause sparse to emit a warning,
__user should be removed for consistency with other uses of offsetof().

Note: No functional changes intended.

Link: https://lkml.kernel.org/r/20211122101256.7875-1-amit.kachhap@arm.com
Signed-off-by: Amit Daniel Kachhap <amit.kachhap@arm.com>
Cc: Vincenzo Frascino <Vincenzo.Frascino@arm.com>
Cc: Kevin Brodsky <Kevin.Brodsky@arm.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

ocfs2: remove redundant assignment to variable free_space

The variable 'free_space' is being initialized with a value that is not
read, it is being re-assigned later in the two paths of an if statement.
The early initialization is redundant and can be removed.

Link: https://lkml.kernel.org/r/20220112230411.1090761-1-colin.i.king@gmail.com
Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Acked-by: Joseph Qi <joseph.qi@linux.alibaba.com>
Cc: Mark Fasheh <mark@fasheh.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Junxiao Bi <junxiao.bi@oracle.com>
Cc: Changwei Ge <gechangwei@live.cn>
Cc: Gang He <ghe@suse.com>
Cc: Jun Piao <piaojun@huawei.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

ocfs2: cluster: use default_groups in kobj_type

There are currently two ways to create a set of sysfs files for a
kobj_type, through the default_attrs field, and the default_groups
field.

Move the ocfs2 cluster sysfs code to use default_groups field which has
been the preferred way since aa30f47cf666 ("kobject: Add support for
default attribute groups to kobj_type") so that we can soon get rid of
the obsolete default_attrs field.

Link: https://lkml.kernel.org/r/20220106102028.3345634-1-gregkh@linuxfoundation.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Joseph Qi <joseph.qi@linux.alibaba.com>
Tested-by: Joseph Qi <joseph.qi@linux.alibaba.com>
Cc: Mark Fasheh <mark@fasheh.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Junxiao Bi <junxiao.bi@oracle.com>
Cc: Changwei Ge <gechangwei@live.cn>
Cc: Gang He <ghe@suse.com>
Cc: Jun Piao <piaojun@huawei.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

ocfs2: remove redundant assignment to pointer root_bh

The variable 'root_bh' is being initialized with a value that is not
read, it is being re-assigned later on closer to its use.  The early
initialization is redundant and can be removed.

Link: https://lkml.kernel.org/r/20211228013719.620923-1-colin.i.king@gmail.com
Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Acked-by: Joseph Qi <joseph.qi@linux.alibaba.com>
Cc: Mark Fasheh <mark@fasheh.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Junxiao Bi <junxiao.bi@oracle.com>
Cc: Changwei Ge <gechangwei@live.cn>
Cc: Gang He <ghe@suse.com>
Cc: Jun Piao <piaojun@huawei.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

ocfs2: use default_groups in kobj_type

There are currently two ways to create a set of sysfs files for a
kobj_type, through the default_attrs field, and the default_groups
field.

Move the ocfs2 code to use default_groups field which has been the
preferred way since aa30f47cf666 ("kobject: Add support for default
attribute groups to kobj_type") so that we can soon get rid of the
obsolete default_attrs field.

Link: https://lkml.kernel.org/r/20211228144517.391660-1-gregkh@linuxfoundation.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Acked-by: Joseph Qi <joseph.qi@linux.alibaba.com>
Cc: Mark Fasheh <mark@fasheh.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Junxiao Bi <junxiao.bi@oracle.com>
Cc: Changwei Ge <gechangwei@live.cn>
Cc: Gang He <ghe@suse.com>
Cc: Jun Piao <piaojun@huawei.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

ocfs2: clearly handle ocfs2_grab_pages_for_write() return value

ocfs2_grab_pages_for_write() may return -EAGAIN if write context type is
mmap and it could not lock the target page.  In this case, we exit with
no error and no target page.  And then trigger the caller page_mkwrite()
to retry.

Since there are other caller types, e.g.  buffer and direct io, make the
return value handling more clear.

Link: https://lkml.kernel.org/r/20211206065051.103353-1-joseph.qi@linux.alibaba.com
Signed-off-by: Joseph Qi <joseph.qi@linux.alibaba.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Cc: Mark Fasheh <mark@fasheh.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Junxiao Bi <junxiao.bi@oracle.com>
Cc: Changwei Ge <gechangwei@live.cn>
Cc: Gang He <ghe@suse.com>
Cc: Jun Piao <piaojun@huawei.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

ocfs2: use BUG_ON instead of if condition followed by BUG.

This issue was detected with the help of Coccinelle.

Link: https://lkml.kernel.org/r/20211105014424.75372-1-zhang.mingyu@zte.com.cn
Signed-off-by: Zhang Mingyu <zhang.mingyu@zte.com.cn>
Reported-by: Zeal Robot <zealci@zte.com.cn>
Acked-by: Joseph Qi <joseph.qi@linux.alibaba.com>
Cc: Mark Fasheh <mark@fasheh.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Junxiao Bi <junxiao.bi@oracle.com>
Cc: Changwei Ge <gechangwei@live.cn>
Cc: Gang He <ghe@suse.com>
Cc: Jun Piao <piaojun@huawei.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

squashfs: provide backing_dev_info in order to disable read-ahead

Commit c1f6925e1091 ("mm: put readahead pages in cache earlier") causes
the read performance of squashfs to deteriorate.Through testing, we find
that the performance will be back by closing the readahead of squashfs.

So we want to learn the way of ubifs, provides backing_dev_info and
disable read-ahead

We tested the following data by fio.
squashfs image blocksize=128K
test command:

  fio --name basic --bs=? --filename="/mnt/test_file" --rw=? --iodepth=1 --ioengine=psync --runtime=200 --time_based

  turn on squashfs readahead in 5.10 kernel
  bs(k)      read/randread           MB/s
  4            randread              271
  128          randread              231
  1024         randread              246
  4            read                  310
  128          read                  245
  1024         read                  247

  turn off squashfs readahead in 5.10 kernel
  bs(k)      read/randread           MB/s
  4            randread              293
  128          randread              330
  1024         randread              363
  4            read                  338
  128          read                  360
  1024         read                  365

  turn on squashfs readahead and revert the
  commit c1f6925e1091("mm: put readahead
  pages in cache earlier") in 5.10 kernel
  bs(k)      read/randread           MB/s
  4           randread               289
  128         randread               306
  1024        randread               335
  4           read                   337
  128         read                   336
  1024        read                   338

Link: https://lkml.kernel.org/r/20211116113141.1391026-1-zhengliang6@huawei.com
Signed-off-by: Zheng Liang <zhengliang6@huawei.com>
Reviewed-by: Phillip Lougher <phillip@squashfs.org.uk>
Cc: Zhang Yi <yi.zhang@huawei.com>
Cc: Hou Tao <houtao1@huawei.com>
Cc: Miao Xie <miaoxie@huawei.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

fs/ntfs/attrib.c: fix one kernel-doc comment

The comments for the file should not be in kernel-doc format:

/**
 * attrib.c - NTFS attribute operations.  Part of the Linux-NTFS

as it causes it to be incorrectly identified for function
ntfs_map_runlist_nolock(), causing some warnings found by running
scripts/kernel-doc.:

  fs/ntfs/attrib.c:25: warning: Incorrect use of kernel-doc format:  * ntfs_map_runlist_nolock - map (a part of) a runlist of an ntfs inode
  fs/ntfs/attrib.c:71: warning: Function parameter or member 'ni' not described in 'ntfs_map_runlist_nolock'
  fs/ntfs/attrib.c:71: warning: Function parameter or member 'vcn' not described in 'ntfs_map_runlist_nolock'
  fs/ntfs/attrib.c:71: warning: Function parameter or member 'ctx' not described in 'ntfs_map_runlist_nolock'
  fs/ntfs/attrib.c:71: warning: expecting prototype for attrib.c - NTFS attribute operations.  Part of the Linux(). Prototype was for ntfs_map_runlist_nolock() instead

Link: https://lkml.kernel.org/r/20220106015145.67067-1-yang.lee@linux.alibaba.com
Signed-off-by: Yang Li <yang.lee@linux.alibaba.com>
Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Cc: Anton Altaparmakov <anton@tuxera.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

scripts/spelling.txt: add "oveflow"

Add typo "oveflow" for "overflow".  This typo was found and fixed in
tools/testing/selftests/bpf/prog_tests/btf_dump.c

Link: https://lore.kernel.org/all/20211122070528.837806-1-dfustini@baylibre.com/
Link: https://lkml.kernel.org/r/20211122072302.839102-1-dfustini@baylibre.com
Signed-off-by: Drew Fustini <dfustini@baylibre.com>
Suggested-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Cc: Colin Ian King <colin.king@intel.com>
Cc: Drew Fustini <dfustini@baylibre.com>
Cc: zuoqilin <zuoqilin@yulong.com>
Cc: Tom Saeger <tom.saeger@oracle.com>
Cc: Sven Eckelmann <sven@narfation.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

ia64: topology: use default_groups in kobj_type

There are currently two ways to create a set of sysfs files for a kobj_type,
through the default_attrs field, and the default_groups field.

Move the ia64 topology sysfs code to use default_groups field which has
been the preferred way since aa30f47cf666 ("kobject: Add support for
default attribute groups to kobj_type") so that we can soon get rid of
the obsolete default_attrs field.

Link: https://lkml.kernel.org/r/20220104154800.1287947-1-gregkh@linuxfoundation.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: David Hildenbrand <david@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

ia64: fix typo in a comment

The double `the' in a comment is repeated, thus it should be removed.

Link: https://lkml.kernel.org/r/20211113030316.22650-1-wangborong@cdjrlc.com
Signed-off-by: Jason Wang <wangborong@cdjrlc.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

arch/ia64/kernel/setup.c: use swap() to make code cleaner

Use the macro 'swap()' defined in 'include/linux/minmax.h' to avoid
opencoding it.

Link: https://lkml.kernel.org/r/20211104001908.695110-1-yang.guang5@zte.com.cn
Reported-by: Zeal Robot <zealci@zte.com.cn>
Signed-off-by: Yang Guang <yang.guang5@zte.com.cn>
Cc: David Yang <davidcomponentone@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

ia64: module: use swap() to make code cleaner

Use the macro 'swap()' defined in 'include/linux/minmax.h' to avoid
opencoding it.

Link: https://lkml.kernel.org/r/20211104062642.1506539-1-yang.guang5@zte.com.cn
Signed-off-by: Yang Guang <yang.guang5@zte.com.cn>
Reported-by: Zeal Robot <zealci@zte.com.cn>
Cc: David Yang <davidcomponentone@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

trace/hwlat: make use of the helper function kthread_run_on_cpu()

Replace kthread_create_on_cpu/wake_up_process() with kthread_run_on_cpu()
to simplify the code.

Link: https://lkml.kernel.org/r/20211022025711.3673-7-caihuoqing@baidu.com
Signed-off-by: Cai Huoqing <caihuoqing@baidu.com>
Cc: Bernard Metzler <bmt@zurich.ibm.com>
Cc: Daniel Bristot de Oliveira <bristot@kernel.org>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Doug Ledford <dledford@redhat.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Joel Fernandes (Google) <joel@joelfernandes.org>
Cc: Josh Triplett <josh@joshtriplett.org>
Cc: Lai Jiangshan <jiangshanlai@gmail.com>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: "Paul E . McKenney" <paulmck@kernel.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

trace/osnoise: make use of the helper function kthread_run_on_cpu()

Replace kthread_create_on_cpu/wake_up_process() with kthread_run_on_cpu()
to simplify the code.

Link: https://lkml.kernel.org/r/20211022025711.3673-6-caihuoqing@baidu.com
Signed-off-by: Cai Huoqing <caihuoqing@baidu.com>
Cc: Bernard Metzler <bmt@zurich.ibm.com>
Cc: Daniel Bristot de Oliveira <bristot@kernel.org>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Doug Ledford <dledford@redhat.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Joel Fernandes (Google) <joel@joelfernandes.org>
Cc: Josh Triplett <josh@joshtriplett.org>
Cc: Lai Jiangshan <jiangshanlai@gmail.com>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: "Paul E . McKenney" <paulmck@kernel.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

rcutorture: make use of the helper function kthread_run_on_cpu()

Replace kthread_create_on_node/kthread_bind/wake_up_process() with
kthread_run_on_cpu() to simplify the code.

Link: https://lkml.kernel.org/r/20211022025711.3673-5-caihuoqing@baidu.com
Signed-off-by: Cai Huoqing <caihuoqing@baidu.com>
Cc: Bernard Metzler <bmt@zurich.ibm.com>
Cc: Daniel Bristot de Oliveira <bristot@kernel.org>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Doug Ledford <dledford@redhat.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Joel Fernandes (Google) <joel@joelfernandes.org>
Cc: Josh Triplett <josh@joshtriplett.org>
Cc: Lai Jiangshan <jiangshanlai@gmail.com>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: "Paul E . McKenney" <paulmck@kernel.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

ring-buffer: make use of the helper function kthread_run_on_cpu()

Replace kthread_create/kthread_bind/wake_up_process() with
kthread_run_on_cpu() to simplify the code.

Link: https://lkml.kernel.org/r/20211022025711.3673-4-caihuoqing@baidu.com
Signed-off-by: Cai Huoqing <caihuoqing@baidu.com>
Cc: Bernard Metzler <bmt@zurich.ibm.com>
Cc: Daniel Bristot de Oliveira <bristot@kernel.org>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Doug Ledford <dledford@redhat.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Joel Fernandes (Google) <joel@joelfernandes.org>
Cc: Josh Triplett <josh@joshtriplett.org>
Cc: Lai Jiangshan <jiangshanlai@gmail.com>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: "Paul E . McKenney" <paulmck@kernel.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

RDMA/siw: make use of the helper function kthread_run_on_cpu()

Replace kthread_create/kthread_bind/wake_up_process() with
kthread_run_on_cpu() to simplify the code.

Link: https://lkml.kernel.org/r/20211022025711.3673-3-caihuoqing@baidu.com
Signed-off-by: Cai Huoqing <caihuoqing@baidu.com>
Cc: Bernard Metzler <bmt@zurich.ibm.com>
Cc: Daniel Bristot de Oliveira <bristot@kernel.org>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Doug Ledford <dledford@redhat.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Joel Fernandes (Google) <joel@joelfernandes.org>
Cc: Josh Triplett <josh@joshtriplett.org>
Cc: Lai Jiangshan <jiangshanlai@gmail.com>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: "Paul E . McKenney" <paulmck@kernel.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

kthread: add the helper function kthread_run_on_cpu()

Add a new helper function kthread_run_on_cpu(), which includes
kthread_create_on_cpu/wake_up_process().

In some cases, use kthread_run_on_cpu() directly instead of
kthread_create_on_node/kthread_bind/wake_up_process() or
kthread_create_on_cpu/wake_up_process() or
kthreadd_create/kthread_bind/wake_up_process() to simplify the code.

[akpm@linux-foundation.org: export kthread_create_on_cpu to modules]

Link: https://lkml.kernel.org/r/20211022025711.3673-2-caihuoqing@baidu.com
Signed-off-by: Cai Huoqing <caihuoqing@baidu.com>
Cc: Bernard Metzler <bmt@zurich.ibm.com>
Cc: Cai Huoqing <caihuoqing@baidu.com>
Cc: Daniel Bristot de Oliveira <bristot@kernel.org>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Doug Ledford <dledford@redhat.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Joel Fernandes (Google) <joel@joelfernandes.org>
Cc: Josh Triplett <josh@joshtriplett.org>
Cc: Lai Jiangshan <jiangshanlai@gmail.com>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: "Paul E . McKenney" <paulmck@kernel.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Linux 5.16

Merge branch 'for-linus' of git://git./linux/kernel/git/dtor/input

Pull input fix from Dmitry Torokhov:
 "A small fixup to the Zinitix touchscreen driver to avoid enabling the
  IRQ line before we successfully requested it"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:
  Input: zinitix - make sure the IRQ is allocated before it gets enabled

Merge tag 'soc-fixes-5.16-5' of git://git./linux/kernel/git/soc/soc

Pull ARM SoC fix from Olof Johansson:
 "One more fix for 5.16

  I had missed one patch when I sent up what I thought was the last
  batch of fixes for this release. This one fixes issues on the
  Raspberry Pi platforms due to gpio init changes this release, so
  hopefully we can get it merged before final release is cut"

* tag 'soc-fixes-5.16-5' of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc:
  ARM: dts: gpio-ranges property is now required

Merge tag 'perf-tools-fixes-for-v5.16-2022-01-09' of git://git./linux/kernel/git/acme/linux

Pull perf tools fixes from Arnaldo Carvalho de Melo:

 - Revert "libtraceevent: Increase libtraceevent logging when verbose",
   breaks the build with libtraceevent-1.3.0, i.e. when building with
   'LIBTRACEEVENT_DYNAMIC=1'.

 - Avoid early exit in 'perf trace' due to running SIGCHLD handler
   before it makes sense to. It can happen when using a BPF source code
   event that have to be first built into an object file.

* tag 'perf-tools-fixes-for-v5.16-2022-01-09' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux:
  Revert "libtraceevent: Increase libtraceevent logging when verbose"
  perf trace: Avoid early exit due to running SIGCHLD handler before it makes sense to

Revert "drm/amdgpu: stop scheduler when calling hw_fini (v2)"

This reverts commit f7d6779df642720e22bffd449e683bb8690bd3bf.

This bisected regression has impacted suspend-resume stability
since 5.15-rc1. It regressed -stable via 5.14.10.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=215315
Fixes: f7d6779df64 ("drm/amdgpu: stop scheduler when calling hw_fini (v2)")
Cc: Guchun Chen <guchun.chen@amd.com>
Cc: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Cc: Christian Koenig <christian.koenig@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: <stable@vger.kernel.org> # 5.14+
Signed-off-by: Len Brown <len.brown@intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Input: zinitix - make sure the IRQ is allocated before it gets enabled

Since irq request is the last thing in the driver probe, it happens
later than the input device registration. This means that there is a
small time window where if the open method is called the driver will
attempt to enable not yet available irq.

Fix that by moving the irq request before the input device registration.

Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Fixes: 26822652c85e ("Input: add zinitix touchscreen driver")
Signed-off-by: Nikita Travkin <nikita@trvn.ru>
Link: https://lore.kernel.org/r/20220106072840.36851-2-nikita@trvn.ru
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

ARM: dts: gpio-ranges property is now required

Since [1], added in 5.7, the absence of a gpio-ranges property has
prevented GPIOs from being restored to inputs when released.
Add those properties for BCM283x and BCM2711 devices.

[1] commit 2ab73c6d8323 ("gpio: Support GPIO controllers without
    pin-ranges")

Link: https://lore.kernel.org/r/20220104170247.956760-1-linus.walleij@linaro.org
Fixes: 2ab73c6d8323 ("gpio: Support GPIO controllers without pin-ranges")
Fixes: 266423e60ea1 ("pinctrl: bcm2835: Change init order for gpio hogs")
Reported-by: Stefan Wahren <stefan.wahren@i2se.com>
Reported-by: Florian Fainelli <f.fainelli@gmail.com>
Reported-by: Jan Kiszka <jan.kiszka@web.de>
Signed-off-by: Phil Elwell <phil@raspberrypi.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20211206092237.4105895-3-phil@raspberrypi.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Olof Johansson <olof@lixom.net>

Merge tag 'soc-fixes-5.16-4' of git://git./linux/kernel/git/soc/soc

Pull ARM SoC fixes from Olof Johansson:
 "A few more fixes have come in, nothing overly severe but would be good
  to get in by final release:

   - More specific compatible fields on the qspi controller for socfpga,
     to enable quirks in the driver

   - A runtime PM fix for Renesas to fix mismatched reference counts on
     errors"

* tag 'soc-fixes-5.16-4' of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc:
  ARM: dts: socfpga: change qspi to "intel,socfpga-qspi"
  dt-bindings: spi: cadence-quadspi: document "intel,socfpga-qspi"
  reset: renesas: Fix Runtime PM usage

Merge branch 'i2c/for-current' of git://git./linux/kernel/git/wsa/linux

Pull i2c fixes from Wolfram Sang:
 "Fix the regression with AMD GPU suspend by reverting the
  handling of bus regulators in the I2C core.

  Also, there is a fix for the MPC driver to prevent an
  out-of-bound-access"

* 'i2c/for-current' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux:
  Revert "i2c: core: support bus regulator controlling in adapter"
  i2c: mpc: Avoid out of bounds memory access

Merge tag 'for-v5.16-rc' of git://git./linux/kernel/git/sre/linux-power-supply

Pull power supply fixes from Sebastian Reichel:
 "Three fixes for the 5.16 cycle:

   - Avoid going beyond last capacity in the power-supply core

   - Replace 1E6L with NSEC_PER_MSEC to avoid floating point calculation
     in LLVM resulting in a build failure

   - Fix ADC measurements in bq25890 charger driver"

* tag 'for-v5.16-rc' of git://git.kernel.org/pub/scm/linux/kernel/git/sre/linux-power-supply:
  power: reset: ltc2952: Fix use of floating point literals
  power: bq25890: Enable continuous conversion for ADC at charging
  power: supply: core: Break capacity loop

Merge tag 'xfs-5.16-fixes-4' of git://git./fs/xfs/xfs-linux

Pull xfs fix from Darrick Wong:

 - Make the old ALLOCSP ioctl behave in a consistent manner with newer
   syscalls like fallocate.

* tag 'xfs-5.16-fixes-4' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux:
  xfs: map unwritten blocks in XFS_IOC_{ALLOC,FREE}SP just like fallocate

Merge branch 'for-5.16-fixes' of git://git./linux/kernel/git/tj/cgroup

Pull cgroup fixes from Tejun Heo:
 "This contains the cgroup.procs permission check fixes so that they use
  the credentials at the time of open rather than write, which also
  fixes the cgroup namespace lifetime bug"

* 'for-5.16-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup:
  selftests: cgroup: Test open-time cgroup namespace usage for migration checks
  selftests: cgroup: Test open-time credential usage for migration checks
  selftests: cgroup: Make cg_create() use 0755 for permission instead of 0644
  cgroup: Use open-time cgroup namespace for process migration perm checks
  cgroup: Allocate cgroup_file_ctx for kernfs_open_file->priv
  cgroup: Use open-time credentials for process migraton perm checks

Merge tag 'block-5.16-2022-01-07' of git://git.kernel.dk/linux-block

Pull block fix from Jens Axboe:
 "Just the md bitmap regression this time"

* tag 'block-5.16-2022-01-07' of git://git.kernel.dk/linux-block:
  md/raid1: fix missing bitmap update w/o WriteMostly devices

Merge tag 'edac_urgent_for_v5.16' of git://git./linux/kernel/git/ras/ras

Pull EDAC fix from Tony Luck:
 "Fix 10nm EDAC driver to release and unmap resources on systems without
  HBM"

* tag 'edac_urgent_for_v5.16' of git://git.kernel.org/pub/scm/linux/kernel/git/ras/ras:
  EDAC/i10nm: Release mdev/mbase when failing to detect HBM

Revert "i2c: core: support bus regulator controlling in adapter"

This largely reverts commit 5a7b95fb993ec399c8a685552aa6a8fc995c40bd. It
breaks suspend with AMD GPUs, and we couldn't incrementally fix it. So,
let's remove the code and go back to the drawing board. We keep the
header extension to not break drivers already populating the regulator.
We expect to re-add the code handling it soon.

Fixes: 5a7b95fb993e ("i2c: core: support bus regulator controlling in adapter")
Reported-by: "Tareque Md.Hanif" <tarequemd.hanif@yahoo.com>
Link: https://lore.kernel.org/r/1295184560.182511.1639075777725@mail.yahoo.com
Reported-by: Konstantin Kharlamov <hi-angel@yandex.ru>
Link: https://lore.kernel.org/r/7143a7147978f4104171072d9f5225d2ce355ec1.camel@yandex.ru
BugLink: https://gitlab.freedesktop.org/drm/amd/-/issues/1850
Tested-by: "Tareque Md.Hanif" <tarequemd.hanif@yahoo.com>
Tested-by: Konstantin Kharlamov <hi-angel@yandex.ru>
Signed-off-by: Wolfram Sang <wsa@kernel.org>
Cc: <stable@vger.kernel.org> # 5.14+

Revert "libtraceevent: Increase libtraceevent logging when verbose"

This reverts commit 08efcb4a638d260ef7fcbae64ecf7ceceb3f1841.

This breaks the build as it will prefer using libbpf-devel header files,
even when not using LIBBPF_DYNAMIC=1, breaking the build.

This was detected on OpenSuSE Tumbleweed with libtraceevent-devel 1.3.0,
as described by Jiri Slaby:

=======================================================================
It breaks build with LIBTRACEEVENT_DYNAMIC and version 1.3.0:
> util/debug.c: In function ‘perf_debug_option’:
> util/debug.c:243:17: error: implicit declaration of function
‘tep_set_loglevel’ [-Werror=implicit-function-declaration]
>   243 |                 tep_set_loglevel(TEP_LOG_INFO);
>       |                 ^~~~~~~~~~~~~~~~
> util/debug.c:243:34: error: ‘TEP_LOG_INFO’ undeclared (first use in this
function); did you mean ‘TEP_PRINT_INFO’?
>   243 |                 tep_set_loglevel(TEP_LOG_INFO);
>       |                                  ^~~~~~~~~~~~
>       |                                  TEP_PRINT_INFO
> util/debug.c:243:34: note: each undeclared identifier is reported only once
for each function it appears in
> util/debug.c:245:34: error: ‘TEP_LOG_DEBUG’ undeclared (first use in this
function)
>   245 |                 tep_set_loglevel(TEP_LOG_DEBUG);
>       |                                  ^~~~~~~~~~~~~
> util/debug.c:247:34: error: ‘TEP_LOG_ALL’ undeclared (first use in this
function)
>   247 |                 tep_set_loglevel(TEP_LOG_ALL);
>       |                                  ^~~~~~~~~~~

It is because the gcc's command line looks like:
gcc
...
-I/home/abuild/rpmbuild/BUILD/tools/lib/
...
-DLIBTRACEEVENT_VERSION=65790
...
=======================================================================

The proper way to fix this is more involved and so not suitable for this
late in the 5.16-rc stage.

Reported-by: Jiri Slaby <jirislaby@kernel.org>
Link: https://lore.kernel.org/lkml/bc2b0786-8965-1bcd-2316-9d9bb37b9c31@kernel.org
Cc: Andrii Nakryiko <andrii.nakryiko@gmail.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Song Liu <songliubraving@fb.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Link: https://lore.kernel.org/lkml/YddGjjmlMZzxUZbN@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf trace: Avoid early exit due to running SIGCHLD handler before it makes sense to

When running 'perf trace' with an BPF object like:

  # perf trace -e openat,tools/perf/examples/bpf/hello.c

the event parsing eventually calls llvm__get_kbuild_opts() that runs a
script and that ends up with SIGCHLD delivered to the 'perf trace'
handler, which assumes the workload process is done and quits 'perf
trace'.

Move the SIGCHLD handler setup directly to trace__run(), where the event
is parsed and the object is already compiled.

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Christy Lee <christyc.y.lee@gmail.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Michael Petlan <mpetlan@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lore.kernel.org/lkml/20220106222030.227499-1-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

Merge tag 'for-linus' of git://git./virt/kvm/kvm

Pull kvm fixes from Paolo Bonzini:
 "Two small fixes for x86:

   - lockdep WARN due to missing lock nesting annotation

   - NULL pointer dereference when accessing debugfs"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
  KVM: x86: Check for rmaps allocation
  KVM: SEV: Mark nested locking of kvm->lock

Merge tag 'drm-fixes-2022-01-07' of git://anongit.freedesktop.org/drm/drm

Pull drm fixes from Dave Airlie:
 "There is only the amdgpu runtime pm regression fix in here:

  amdgpu:

   - suspend/resume fix

   - fix runtime PM regression"

* tag 'drm-fixes-2022-01-07' of git://anongit.freedesktop.org/drm/drm:
  drm/amdgpu: disable runpm if we are the primary adapter
  fbdev: fbmem: add a helper to determine if an aperture is used by a fw fb
  drm/amd/pm: keep the BACO feature enabled for suspend

KVM: x86: Check for rmaps allocation

With TDP MMU being the default now, access to mmu_rmaps_stat debugfs
file causes following oops:

BUG: kernel NULL pointer dereference, address: 0000000000000000
PGD 0 P4D 0
Oops: 0000 [#1] PREEMPT SMP NOPTI
CPU: 7 PID: 3185 Comm: cat Not tainted 5.16.0-rc4+ #204
RIP: 0010:pte_list_count+0x6/0x40
 Call Trace:
  <TASK>
  ? kvm_mmu_rmaps_stat_show+0x15e/0x320
  seq_read_iter+0x126/0x4b0
  ? aa_file_perm+0x124/0x490
  seq_read+0xf5/0x140
  full_proxy_read+0x5c/0x80
  vfs_read+0x9f/0x1a0
  ksys_read+0x67/0xe0
  __x64_sys_read+0x19/0x20
  do_syscall_64+0x3b/0xc0
  entry_SYSCALL_64_after_hwframe+0x44/0xae
 RIP: 0033:0x7fca6fc13912

Return early when rmaps are not present.

Reported-by: Vasant Hegde <vasant.hegde@amd.com>
Tested-by: Vasant Hegde <vasant.hegde@amd.com>
Signed-off-by: Nikunj A Dadhania <nikunj@amd.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220105040337.4234-1-nikunj@amd.com>
Cc: stable@vger.kernel.org
Fixes: 3bcd0662d66f ("KVM: X86: Introduce mmu_rmaps_stat per-vm debugfs file")
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

KVM: SEV: Mark nested locking of kvm->lock

Both source and dest vms' kvm->locks are held in sev_lock_two_vms.
Mark one with a different subtype to avoid false positives from lockdep.

Fixes: c9d61dcb0bc26 (KVM: SEV: accept signals in sev_lock_two_vms)
Reported-by: Yiru Xu <xyru1999@gmail.com>
Tested-by: Jinrong Liang <cloudliang@tencent.com>
Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
Message-Id: <1641364863-26331-1-git-send-email-wanpengli@tencent.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Merge tag 'for-linus' of git://git./linux/kernel/git/rdma/rdma

Pull rdma fixes from Jason Gunthorpe:
 "Last pull for 5.16, the reversion has been known for a while now but
  didn't get a proper fix in time. Looks like we will have several
  info-leak bugs to take care of going foward.

   - Revert the patch fixing the DM related crash causing a widespread
     regression for kernel ULPs. A proper fix just didn't appear this
     cycle due to the holidays

   - Missing NULL check on alloc in uverbs

   - Double free in rxe error paths

   - Fix a new kernel-infoleak report when forming ah_attr's without
     GRH's in ucma"

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:
  RDMA/core: Don't infoleak GRH fields
  RDMA/uverbs: Check for null return of kmalloc_array
  Revert "RDMA/mlx5: Fix releasing unallocated memory in dereg MR flow"
  RDMA/rxe: Prevent double freeing rxe_map_set()

Merge tag 'trace-v5.16-rc8' of git://git./linux/kernel/git/rostedt/linux-trace

Pull tracing fixes from Steven Rostedt:
 "Three minor tracing fixes:

   - Fix missing prototypes in sample module for direct functions

   - Fix check of valid buffer in get_trace_buf()

   - Fix annotations of percpu pointers"

* tag 'trace-v5.16-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:
  tracing: Tag trace_percpu_buffer as a percpu pointer
  tracing: Fix check for trace_percpu_buffer validity in get_trace_buf()
  ftrace/samples: Add missing prototypes direct functions

selftests: cgroup: Test open-time cgroup namespace usage for migration checks

When a task is writing to an fd opened by a different task, the perm check
should use the cgroup namespace of the latter task. Add a test for it.

Tested-by: Michal Koutný <mkoutny@suse.com>
Signed-off-by: Tejun Heo <tj@kernel.org>

selftests: cgroup: Test open-time credential usage for migration checks

When a task is writing to an fd opened by a different task, the perm check
should use the credentials of the latter task. Add a test for it.

Tested-by: Michal Koutný <mkoutny@suse.com>
Signed-off-by: Tejun Heo <tj@kernel.org>

selftests: cgroup: Make cg_create() use 0755 for permission instead of 0644

0644 is an odd perm to create a cgroup which is a directory. Use the regular
0755 instead. This is necessary for euid switching test case.

Reviewed-by: Michal Koutný <mkoutny@suse.com>
Signed-off-by: Tejun Heo <tj@kernel.org>

cgroup: Use open-time cgroup namespace for process migration perm checks

cgroup process migration permission checks are performed at write time as
whether a given operation is allowed or not is dependent on the content of
the write - the PID. This currently uses current's cgroup namespace which is
a potential security weakness as it may allow scenarios where a less
privileged process tricks a more privileged one into writing into a fd that
it created.

This patch makes cgroup remember the cgroup namespace at the time of open
and uses it for migration permission checks instad of current's. Note that
this only applies to cgroup2 as cgroup1 doesn't have namespace support.

This also fixes a use-after-free bug on cgroupns reported in

 https://lore.kernel.org/r/00000000000048c15c05d0083397@google.com

Note that backporting this fix also requires the preceding patch.

Reported-by: "Eric W. Biederman" <ebiederm@xmission.com>
Suggested-by: Linus Torvalds <torvalds@linuxfoundation.org>
Cc: Michal Koutný <mkoutny@suse.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Reviewed-by: Michal Koutný <mkoutny@suse.com>
Reported-by: syzbot+50f5cf33a284ce738b62@syzkaller.appspotmail.com
Link: https://lore.kernel.org/r/00000000000048c15c05d0083397@google.com
Fixes: 5136f6365ce3 ("cgroup: implement "nsdelegate" mount option")
Signed-off-by: Tejun Heo <tj@kernel.org>

cgroup: Allocate cgroup_file_ctx for kernfs_open_file->priv

of->priv is currently used by each interface file implementation to store
private information. This patch collects the current two private data usages
into struct cgroup_file_ctx which is allocated and freed by the common path.
This allows generic private data which applies to multiple files, which will
be used to in the following patch.

Note that cgroup_procs iterator is now embedded as procs.iter in the new
cgroup_file_ctx so that it doesn't need to be allocated and freed
separately.

v2: union dropped from cgroup_file_ctx and the procs iterator is embedded in
    cgroup_file_ctx as suggested by Linus.

v3: Michal pointed out that cgroup1's procs pidlist uses of->priv too.
    Converted. Didn't change to embedded allocation as cgroup1 pidlists get
    stored for caching.

Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Reviewed-by: Michal Koutný <mkoutny@suse.com>

cgroup: Use open-time credentials for process migraton perm checks

cgroup process migration permission checks are performed at write time as
whether a given operation is allowed or not is dependent on the content of
the write - the PID. This currently uses current's credentials which is a
potential security weakness as it may allow scenarios where a less
privileged process tricks a more privileged one into writing into a fd that
it created.

This patch makes both cgroup2 and cgroup1 process migration interfaces to
use the credentials saved at the time of open (file->f_cred) instead of
current's.

Reported-by: "Eric W. Biederman" <ebiederm@xmission.com>
Suggested-by: Linus Torvalds <torvalds@linuxfoundation.org>
Fixes: 187fe84067bd ("cgroup: require write perm on common ancestor when moving processes on the default hierarchy")
Reviewed-by: Michal Koutný <mkoutny@suse.com>
Signed-off-by: Tejun Heo <tj@kernel.org>

Merge tag 'amd-drm-fixes-5.16-2021-12-31' of ssh://gitlab.freedesktop.org/agd5f/linux into drm-fixes

amd-drm-fixes-5.16-2021-12-31:

amdgpu:
- Suspend/resume fix
- Restore runtime pm behavior with efifb

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20211231143825.11479-1-alexander.deucher@amd.com

i2c: mpc: Avoid out of bounds memory access

When performing an I2C transfer where the last message was a write KASAN
would complain:

  BUG: KASAN: slab-out-of-bounds in mpc_i2c_do_action+0x154/0x630
  Read of size 2 at addr c814e310 by task swapper/2/0

  CPU: 2 PID: 0 Comm: swapper/2 Tainted: G    B             5.16.0-rc8 #1
  Call Trace:
  [e5ee9d50] [c08418e8] dump_stack_lvl+0x4c/0x6c (unreliable)
  [e5ee9d70] [c02f8a14] print_address_description.constprop.13+0x64/0x3b0
  [e5ee9da0] [c02f9030] kasan_report+0x1f0/0x204
  [e5ee9de0] [c0c76ee4] mpc_i2c_do_action+0x154/0x630
  [e5ee9e30] [c0c782c4] mpc_i2c_isr+0x164/0x240
  [e5ee9e60] [c00f3a04] __handle_irq_event_percpu+0xf4/0x3b0
  [e5ee9ec0] [c00f3d40] handle_irq_event_percpu+0x80/0x110
  [e5ee9f40] [c00f3e48] handle_irq_event+0x78/0xd0
  [e5ee9f60] [c00fcfec] handle_fasteoi_irq+0x19c/0x370
  [e5ee9fa0] [c00f1d84] generic_handle_irq+0x54/0x80
  [e5ee9fc0] [c0006b54] __do_irq+0x64/0x200
  [e5ee9ff0] [c0007958] __do_IRQ+0xe8/0x1c0
  [c812dd50] [e3eaab20] 0xe3eaab20
  [c812dd90] [c0007a4c] do_IRQ+0x1c/0x30
  [c812dda0] [c0000c04] ExternalInput+0x144/0x160
  --- interrupt: 500 at arch_cpu_idle+0x34/0x60
  NIP:  c000b684 LR: c000b684 CTR: c0019688
  REGS: c812ddb0 TRAP: 0500   Tainted: G    B              (5.16.0-rc8)
  MSR:  00029002 <CE,EE,ME>  CR: 22000488  XER: 20000000

  GPR00: c10ef7fc c812de90 c80ff200 c2394718 00000001 00000001 c10e3f90 00000003
  GPR08: 00000000 c0019688 c2394718 fc7d625b 22000484 00000000 21e17000 c208228c
  GPR16: e3e99284 00000000 ffffffff c2390000 c001bac0 c2082288 c812df60 c001ba60
  GPR24: c23949c0 00000018 00080000 00000004 c80ff200 00000002 c2348ee4 c2394718
  NIP [c000b684] arch_cpu_idle+0x34/0x60
  LR [c000b684] arch_cpu_idle+0x34/0x60
  --- interrupt: 500
  [c812de90] [c10e3f90] rcu_eqs_enter.isra.60+0xc0/0x110 (unreliable)
  [c812deb0] [c10ef7fc] default_idle_call+0xbc/0x230
  [c812dee0] [c00af0e8] do_idle+0x1c8/0x200
  [c812df10] [c00af3c0] cpu_startup_entry+0x20/0x30
  [c812df20] [c001e010] start_secondary+0x5d0/0xba0
  [c812dff0] [c00028a0] __secondary_start+0x90/0xdc

This happened because we would overrun the i2c->msgs array on the final
interrupt for the I2C STOP. This didn't happen if the last message was a
read because there is no interrupt in that case. Ensure that we only
access the current message if we are not processing a I2C STOP
condition.

Fixes: 1538d82f4647 ("i2c: mpc: Interrupt driven transfer")
Reported-by: Maxime Bizon <mbizon@freebox.fr>
Signed-off-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
Signed-off-by: Wolfram Sang <wsa@kernel.org>

Merge tag 'socfpga_fix_for_v5.16_part_3' of git://git./linux/kernel/git/dinguyen/linux into arm/fixes

SoCFPGA dts updates for v5.16, part 3
- Change the SoCFPGA compatible to "intel,socfpga-qspi"
- Update dt-bindings document to include "intel,socfpga-qspi"

* tag 'socfpga_fix_for_v5.16_part_3' of git://git.kernel.org/pub/scm/linux/kernel/git/dinguyen/linux: (361 commits)
  ARM: dts: socfpga: change qspi to "intel,socfpga-qspi"
  dt-bindings: spi: cadence-quadspi: document "intel,socfpga-qspi"
  Linux 5.16-rc7
  mm/hwpoison: clear MF_COUNT_INCREASED before retrying get_any_page()
  mm/damon/dbgfs: protect targets destructions with kdamond_lock
  mm/page_alloc: fix __alloc_size attribute for alloc_pages_exact_nid
  mm: delete unsafe BUG from page_cache_add_speculative()
  mm, hwpoison: fix condition in free hugetlb page path
  MAINTAINERS: mark more list instances as moderated
  kernel/crash_core: suppress unknown crashkernel parameter warning
  mm: mempolicy: fix THP allocations escaping mempolicy restrictions
  kfence: fix memory leak when cat kfence objects
  platform/x86: intel_pmc_core: fix memleak on registration failure
  net: stmmac: dwmac-visconti: Fix value of ETHER_CLK_SEL_FREQ_SEL_2P5M
  r8152: sync ocp base
  r8152: fix the force speed doesn't work for RTL8156
  net: bridge: fix ioctl old_deviceless bridge argument
  net: stmmac: ptp: fix potentially overflowing expression
  net: dsa: tag_ocelot: use traffic class to map priority on injected header
  veth: ensure skb entering GRO are not cloned.
  ...

Link: https://lore.kernel.org/r/20211227103644.566694-1-dinguyen@kernel.org
Signed-off-by: Olof Johansson <olof@lixom.net>

Merge tag 'reset-fixes-for-v5.16-2' of git://git.pengutronix.de/pza/linux into arm/fixes

Reset controller fixes for v5.16, part 2

Fix pm_runtime_resume_and_get() error handling in the
reset-rzg2l-usbphy-ctrl driver.

* tag 'reset-fixes-for-v5.16-2' of git://git.pengutronix.de/pza/linux:
  reset: renesas: Fix Runtime PM usage
  reset: tegra-bpmp: Revert Handle errors in BPMP response

Link: https://lore.kernel.org/r/20220105172515.273947-1-p.zabel@pengutronix.de
Signed-off-by: Olof Johansson <olof@lixom.net>

tracing: Tag trace_percpu_buffer as a percpu pointer

Tag trace_percpu_buffer as a percpu pointer to resolve warnings
reported by sparse:
  /linux/kernel/trace/trace.c:3218:46: warning: incorrect type in initializer (different address spaces)
  /linux/kernel/trace/trace.c:3218:46:    expected void const [noderef] __percpu *__vpp_verify
  /linux/kernel/trace/trace.c:3218:46:    got struct trace_buffer_struct *
  /linux/kernel/trace/trace.c:3234:9: warning: incorrect type in initializer (different address spaces)
  /linux/kernel/trace/trace.c:3234:9:    expected void const [noderef] __percpu *__vpp_verify
  /linux/kernel/trace/trace.c:3234:9:    got int *

Link: https://lkml.kernel.org/r/ebabd3f23101d89cb75671b68b6f819f5edc830b.1640255304.git.naveen.n.rao@linux.vnet.ibm.com
Cc: stable@vger.kernel.org
Reported-by: kernel test robot <lkp@intel.com>
Fixes: 07d777fe8c398 ("tracing: Add percpu buffers for trace_printk()")
Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>

tracing: Fix check for trace_percpu_buffer validity in get_trace_buf()

With the new osnoise tracer, we are seeing the below splat:
    Kernel attempted to read user page (c7d880000) - exploit attempt? (uid: 0)
    BUG: Unable to handle kernel data access on read at 0xc7d880000
    Faulting instruction address: 0xc0000000002ffa10
    Oops: Kernel access of bad area, sig: 11 [#1]
    LE PAGE_SIZE=64K MMU=Radix SMP NR_CPUS=2048 NUMA pSeries
    ...
    NIP [c0000000002ffa10] __trace_array_vprintk.part.0+0x70/0x2f0
    LR [c0000000002ff9fc] __trace_array_vprintk.part.0+0x5c/0x2f0
    Call Trace:
    [c0000008bdd73b80] [c0000000001c49cc] put_prev_task_fair+0x3c/0x60 (unreliable)
    [c0000008bdd73be0] [c000000000301430] trace_array_printk_buf+0x70/0x90
    [c0000008bdd73c00] [c0000000003178b0] trace_sched_switch_callback+0x250/0x290
    [c0000008bdd73c90] [c000000000e70d60] __schedule+0x410/0x710
    [c0000008bdd73d40] [c000000000e710c0] schedule+0x60/0x130
    [c0000008bdd73d70] [c000000000030614] interrupt_exit_user_prepare_main+0x264/0x270
    [c0000008bdd73de0] [c000000000030a70] syscall_exit_prepare+0x150/0x180
    [c0000008bdd73e10] [c00000000000c174] system_call_vectored_common+0xf4/0x278

osnoise tracer on ppc64le is triggering osnoise_taint() for negative
duration in get_int_safe_duration() called from
trace_sched_switch_callback()->thread_exit().

The problem though is that the check for a valid trace_percpu_buffer is
incorrect in get_trace_buf(). The check is being done after calculating
the pointer for the current cpu, rather than on the main percpu pointer.
Fix the check to be against trace_percpu_buffer.

Link: https://lkml.kernel.org/r/a920e4272e0b0635cf20c444707cbce1b2c8973d.1640255304.git.naveen.n.rao@linux.vnet.ibm.com
Cc: stable@vger.kernel.org
Fixes: e2ace001176dc9 ("tracing: Choose static tp_printk buffer by explicit nesting count")
Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>

ftrace/samples: Add missing prototypes direct functions

There's another compilation fail (first here [1]) reported by kernel
test robot for W=1 clang build:

  >> samples/ftrace/ftrace-direct-multi-modify.c:7:6: warning: no previous
  prototype for function 'my_direct_func1' [-Wmissing-prototypes]
     void my_direct_func1(unsigned long ip)

Direct functions in ftrace direct sample modules need to have prototypes
defined. They are already global in order to be visible for the inline
assembly, so there's no problem.

The kernel test robot reported just error for ftrace-direct-multi-modify,
but I got same errors also for the rest of the modules touched by this patch.

[1] 67d4f6e3bf5d ftrace/samples: Add missing prototype for my_direct_func

Link: https://lkml.kernel.org/r/20211219135317.212430-1-jolsa@kernel.org
Reported-by: kernel test robot <lkp@intel.com>
Fixes: e1067a07cfbc ("ftrace/samples: Add module to test multi direct modify interface")
Fixes: ae0cc3b7e7f5 ("ftrace/samples: Add a sample module that implements modify_ftrace_direct()")
Fixes: 156473a0ff4f ("ftrace: Add another example of register_ftrace_direct() use case")
Fixes: b06457c83af6 ("ftrace: Add sample module that uses register_ftrace_direct()")
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>

Merge tag 'net-5.16-final' of git://git./linux/kernel/git/netdev/net

Pull networking fixes from Jakub Kicinski"
 "Networking fixes, including fixes from bpf, and WiFi. One last pull
  request, turns out some of the recent fixes did more harm than good.

  Current release - regressions:

   - Revert "xsk: Do not sleep in poll() when need_wakeup set", made the
     problem worse

   - Revert "net: phy: fixed_phy: Fix NULL vs IS_ERR() checking in
     __fixed_phy_register", broke EPROBE_DEFER handling

   - Revert "net: usb: r8152: Add MAC pass-through support for more
     Lenovo Docks", broke setups without a Lenovo dock

  Current release - new code bugs:

   - selftests: set amt.sh executable

  Previous releases - regressions:

   - batman-adv: mcast: don't send link-local multicast to mcast routers

  Previous releases - always broken:

   - ipv4/ipv6: check attribute length for RTA_FLOW / RTA_GATEWAY

   - sctp: hold endpoint before calling cb in
     sctp_transport_lookup_process

   - mac80211: mesh: embed mesh_paths and mpp_paths into
     ieee80211_if_mesh to avoid complicated handling of sub-object
     allocation failures

   - seg6: fix traceroute in the presence of SRv6

   - tipc: fix a kernel-infoleak in __tipc_sendmsg()"

* tag 'net-5.16-final' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (36 commits)
  selftests: set amt.sh executable
  Revert "net: usb: r8152: Add MAC passthrough support for more Lenovo Docks"
  sfc: The RX page_ring is optional
  iavf: Fix limit of total number of queues to active queues of VF
  i40e: Fix incorrect netdev's real number of RX/TX queues
  i40e: Fix for displaying message regarding NVM version
  i40e: fix use-after-free in i40e_sync_filters_subtask()
  i40e: Fix to not show opcode msg on unsuccessful VF MAC change
  ieee802154: atusb: fix uninit value in atusb_set_extended_addr
  mac80211: mesh: embedd mesh_paths and mpp_paths into ieee80211_if_mesh
  mac80211: initialize variable have_higher_than_11mbit
  sch_qfq: prevent shift-out-of-bounds in qfq_init_qdisc
  netrom: fix copying in user data in nr_setsockopt
  udp6: Use Segment Routing Header for dest address if present
  icmp: ICMPV6: Examine invoking packet for Segment Route Headers.
  seg6: export get_srh() for ICMP handling
  Revert "net: phy: fixed_phy: Fix NULL vs IS_ERR() checking in __fixed_phy_register"
  ipv6: Do cleanup if attribute validation fails in multipath route
  ipv6: Continue processing multipath route even if gateway attribute is invalid
  net/fsl: Remove leftover definition in xgmac_mdio
  ...

RDMA/core: Don't infoleak GRH fields

If dst->is_global field is not set, the GRH fields are not cleared
and the following infoleak is reported.

=====================================================
BUG: KMSAN: kernel-infoleak in instrument_copy_to_user include/linux/instrumented.h:121 [inline]
BUG: KMSAN: kernel-infoleak in _copy_to_user+0x1c9/0x270 lib/usercopy.c:33
 instrument_copy_to_user include/linux/instrumented.h:121 [inline]
 _copy_to_user+0x1c9/0x270 lib/usercopy.c:33
 copy_to_user include/linux/uaccess.h:209 [inline]
 ucma_init_qp_attr+0x8c7/0xb10 drivers/infiniband/core/ucma.c:1242
 ucma_write+0x637/0x6c0 drivers/infiniband/core/ucma.c:1732
 vfs_write+0x8ce/0x2030 fs/read_write.c:588
 ksys_write+0x28b/0x510 fs/read_write.c:643
 __do_sys_write fs/read_write.c:655 [inline]
 __se_sys_write fs/read_write.c:652 [inline]
 __ia32_sys_write+0xdb/0x120 fs/read_write.c:652
 do_syscall_32_irqs_on arch/x86/entry/common.c:114 [inline]
 __do_fast_syscall_32+0x96/0xf0 arch/x86/entry/common.c:180
 do_fast_syscall_32+0x34/0x70 arch/x86/entry/common.c:205
 do_SYSENTER_32+0x1b/0x20 arch/x86/entry/common.c:248
 entry_SYSENTER_compat_after_hwframe+0x4d/0x5c

Local variable resp created at:
 ucma_init_qp_attr+0xa4/0xb10 drivers/infiniband/core/ucma.c:1214
 ucma_write+0x637/0x6c0 drivers/infiniband/core/ucma.c:1732

Bytes 40-59 of 144 are uninitialized
Memory access of size 144 starts at ffff888167523b00
Data copied to user address 0000000020000100

CPU: 1 PID: 25910 Comm: syz-executor.1 Not tainted 5.16.0-rc5-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
=====================================================

Fixes: 4ba66093bdc6 ("IB/core: Check for global flag when using ah_attr")
Link: https://lore.kernel.org/r/0e9dd51f93410b7b2f4f5562f52befc878b71afa.1641298868.git.leonro@nvidia.com
Reported-by: syzbot+6d532fa8f9463da290bc@syzkaller.appspotmail.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>

selftests: set amt.sh executable

amt.sh test script will not work because it doesn't have execution
permission. So, it adds execution permission.

Reported-by: Hangbin Liu <liuhangbin@gmail.com>
Fixes: c08e8baea78e ("selftests: add amt interface selftest script")
Signed-off-by: Taehee Yoo <ap420073@gmail.com>
Link: https://lore.kernel.org/r/20220105144436.13415-1-ap420073@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

RDMA/uverbs: Check for null return of kmalloc_array

Because of the possible failure of the allocation, data might be NULL
pointer and will cause the dereference of the NULL pointer later.
Therefore, it might be better to check it and return -ENOMEM.

Fixes: 6884c6c4bd09 ("RDMA/verbs: Store the write/write_ex uapi entry points in the uverbs_api")
Link: https://lore.kernel.org/r/20211231093315.1917667-1-jiasheng@iscas.ac.cn
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>