linux-stable.git/fs/binfmt_elf.c, branch linux-6.3.y

mm: always expand the stack with the mmap write lock held

2023-07-01T11:14:46+00:00

commit 8d7071af890768438c14db6172cc8f9f4d04e184 upstream.

This finishes the job of always holding the mmap write lock when
extending the user stack vma, and removes the 'write_locked' argument
from the vm helper functions again.

For some cases, we just avoid expanding the stack at all: drivers and
page pinning really shouldn't be extending any stacks.  Let's see if any
strange users really wanted that.

It's worth noting that architectures that weren't converted to the new
lock_mm_and_find_vma() helper function are left using the legacy
"expand_stack()" function, but it has been changed to drop the mmap_lock
and take it for writing while expanding the vma.  This makes it fairly
straightforward to convert the remaining architectures.

As a result of dropping and re-taking the lock, the calling conventions
for this function have also changed, since the old vma may no longer be
valid.  So it will now return the new vma if successful, and NULL - and
the lock dropped - if the area could not be extended.

Tested-by: Vegard Nossum 
Tested-by: John Paul Adrian Glaubitz  # ia64
Tested-by: Frank Scheiner  # ia64
Signed-off-by: Linus Torvalds 
Signed-off-by: Greg Kroah-Hartman

mm: make find_extend_vma() fail if write lock not held

2023-07-01T11:14:46+00:00

commit f440fa1ac955e2898893f9301568435eb5cdfc4b upstream.

Make calls to extend_vma() and find_extend_vma() fail if the write lock
is required.

To avoid making this a flag-day event, this still allows the old
read-locking case for the trivial situations, and passes in a flag to
say "is it write-locked".  That way write-lockers can say "yes, I'm
being careful", and legacy users will continue to work in all the common
cases until they have been fully converted to the new world order.

Co-Developed-by: Matthew Wilcox (Oracle) 
Signed-off-by: Matthew Wilcox (Oracle) 
Signed-off-by: Liam R. Howlett 
Signed-off-by: Linus Torvalds 
Signed-off-by: Greg Kroah-Hartman

Merge tag 'v6.2-rc6' into sched/core, to pick up fixes

2023-01-31T14:01:20+00:00

Pick up fixes before merging another batch of cpuidle updates.

Signed-off-by: Ingo Molnar

elfcore: Add a cprm parameter to elf_core_extra_{phdrs,data_size}

2023-01-05T15:12:12+00:00

A subsequent fix for arm64 will use this parameter to parse the vma
information from the snapshot created by dump_vma_snapshot() rather than
traversing the vma list without the mmap_lock.

Fixes: 6dd8b1a0b6cb ("arm64: mte: Dump the MTE tags in the core file")
Cc:  # 5.18.x
Signed-off-by: Catalin Marinas 
Reported-by: Seth Jenkins 
Suggested-by: Seth Jenkins 
Cc: Will Deacon 
Cc: Eric Biederman 
Cc: Kees Cook 
Link: https://lore.kernel.org/r/20221222181251.1345752-3-catalin.marinas@arm.com
Signed-off-by: Will Deacon

rseq: Introduce feature size and alignment ELF auxiliary vector entries

2022-12-27T11:52:10+00:00

Export the rseq feature size supported by the kernel as well as the
required allocation alignment for the rseq per-thread area to user-space
through ELF auxiliary vector entries.

This is part of the extensible rseq ABI.

Signed-off-by: Mathieu Desnoyers 
Signed-off-by: Peter Zijlstra (Intel) 
Link: https://lore.kernel.org/r/20221122203932.231377-3-mathieu.desnoyers@efficios.com

Merge tag 'pull-elfcore' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs

2022-12-13T02:18:34+00:00

Pull elf coredumping updates from Al Viro:
 "Unification of regset and non-regset sides of ELF coredump handling.

  Collecting per-thread register values is the only thing that needs to
  be ifdefed there..."

* tag 'pull-elfcore' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
  [elf] get rid of get_note_info_size()
  [elf] unify regset and non-regset cases
  [elf][non-regset] use elf_core_copy_task_regs() for dumper as well
  [elf][non-regset] uninline elf_core_copy_task_fpregs() (and lose pt_regs argument)
  elf_core_copy_task_regs(): task_pt_regs is defined everywhere
  [elf][regset] simplify thread list handling in fill_note_info()
  [elf][regset] clean fill_note_info() a bit
  kill extern of vsyscall32_sysctl
  kill coredump_params->regs
  kill signal_pt_regs()

[elf] get rid of get_note_info_size()

2022-11-25T04:26:04+00:00

it's trivial now...

Signed-off-by: Al Viro

[elf] unify regset and non-regset cases

2022-11-25T04:26:04+00:00

The only real difference is in filling per-thread notes - getting
the values of registers.   And this is the only part that is worth
an ifdef - we don't need to duplicate the logics regarding gathering
threads, filling other notes, etc.

It would've been hard to do back when regset-based variant had been
introduced, mostly due to sharing bits and pieces of helpers with
aout coredumps.  As the result, too much had been duplicated and
the copies had drifted away since then.  Now it can be done cleanly...

Signed-off-by: Al Viro

[elf][non-regset] use elf_core_copy_task_regs() for dumper as well

2022-11-25T04:26:04+00:00

elf_core_copy_regs() is equivalent to elf_core_copy_task_regs() of
current on all architectures.

Signed-off-by: Al Viro

[elf][non-regset] uninline elf_core_copy_task_fpregs() (and lose pt_regs argument)

2022-11-25T04:24:23+00:00

Don't bother with pointless macros - we are not sharing it with aout coredumps
anymore.  Just convert the underlying functions to the same arguments (nobody
uses regs, actually) and call them elf_core_copy_task_fpregs().  And unexport
the entire bunch, while we are at it.

[added missing includes in arch/{csky,m68k,um}/kernel/process.c to avoid extra
warnings about the lack of externs getting added to huge piles for those
files.  Pointless, but...]

Signed-off-by: Al Viro