Uros Bizjak 95ece48165 locking/atomic/x86: Rewrite x86_32 arch_atomic64_{,fetch}_{and,or,xor}() functions
Rewrite x86_32 arch_atomic64_{,fetch}_{and,or,xor}() functions to
use arch_atomic64_try_cmpxchg().  This implementation avoids one extra
trip through the CMPXCHG loop.

The value preload before the cmpxchg loop does not need to be atomic.
Use arch_atomic64_read_nonatomic(v) to load the value from atomic_t
location in a non-atomic way.

The generated code improves from:

  1917d5:	31 c9                	xor    %ecx,%ecx
  1917d7:	31 db                	xor    %ebx,%ebx
  1917d9:	89 4c 24 3c          	mov    %ecx,0x3c(%esp)
  1917dd:	8b 74 24 24          	mov    0x24(%esp),%esi
  1917e1:	89 c8                	mov    %ecx,%eax
  1917e3:	89 5c 24 34          	mov    %ebx,0x34(%esp)
  1917e7:	8b 7c 24 28          	mov    0x28(%esp),%edi
  1917eb:	21 ce                	and    %ecx,%esi
  1917ed:	89 74 24 4c          	mov    %esi,0x4c(%esp)
  1917f1:	21 df                	and    %ebx,%edi
  1917f3:	89 de                	mov    %ebx,%esi
  1917f5:	89 7c 24 50          	mov    %edi,0x50(%esp)
  1917f9:	8b 54 24 4c          	mov    0x4c(%esp),%edx
  1917fd:	8b 7c 24 2c          	mov    0x2c(%esp),%edi
  191801:	8b 4c 24 50          	mov    0x50(%esp),%ecx
  191805:	89 d3                	mov    %edx,%ebx
  191807:	89 f2                	mov    %esi,%edx
  191809:	f0 0f c7 0f          	lock cmpxchg8b (%edi)
  19180d:	89 c1                	mov    %eax,%ecx
  19180f:	8b 74 24 34          	mov    0x34(%esp),%esi
  191813:	89 d3                	mov    %edx,%ebx
  191815:	89 44 24 4c          	mov    %eax,0x4c(%esp)
  191819:	8b 44 24 3c          	mov    0x3c(%esp),%eax
  19181d:	89 df                	mov    %ebx,%edi
  19181f:	89 54 24 44          	mov    %edx,0x44(%esp)
  191823:	89 ca                	mov    %ecx,%edx
  191825:	31 de                	xor    %ebx,%esi
  191827:	31 c8                	xor    %ecx,%eax
  191829:	09 f0                	or     %esi,%eax
  19182b:	75 ac                	jne    1917d9 <...>

to:

  1912ba:	8b 06                	mov    (%esi),%eax
  1912bc:	8b 56 04             	mov    0x4(%esi),%edx
  1912bf:	89 44 24 3c          	mov    %eax,0x3c(%esp)
  1912c3:	89 c1                	mov    %eax,%ecx
  1912c5:	23 4c 24 34          	and    0x34(%esp),%ecx
  1912c9:	89 d3                	mov    %edx,%ebx
  1912cb:	23 5c 24 38          	and    0x38(%esp),%ebx
  1912cf:	89 54 24 40          	mov    %edx,0x40(%esp)
  1912d3:	89 4c 24 2c          	mov    %ecx,0x2c(%esp)
  1912d7:	89 5c 24 30          	mov    %ebx,0x30(%esp)
  1912db:	8b 5c 24 2c          	mov    0x2c(%esp),%ebx
  1912df:	8b 4c 24 30          	mov    0x30(%esp),%ecx
  1912e3:	f0 0f c7 0e          	lock cmpxchg8b (%esi)
  1912e7:	0f 85 f3 02 00 00    	jne    1915e0 <...>

Signed-off-by: Uros Bizjak <ubizjak@gmail.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: https://lore.kernel.org/r/20240410062957.322614-3-ubizjak@gmail.com
2024-04-10 15:04:55 +02:00
2022-09-28 09:02:20 +02:00
2024-04-07 13:22:46 -07:00
2024-03-18 03:36:32 -06:00

Linux kernel
============

There are several guides for kernel developers and users. These guides can
be rendered in a number of formats, like HTML and PDF. Please read
Documentation/admin-guide/README.rst first.

In order to build the documentation, use ``make htmldocs`` or
``make pdfdocs``.  The formatted documentation can also be read online at:

    https://www.kernel.org/doc/html/latest/

There are various text files in the Documentation/ subdirectory,
several of them using the reStructuredText markup notation.

Please read the Documentation/process/changes.rst file, as it contains the
requirements for building and running the kernel, and information about
the problems which may result by upgrading your kernel.
Description
No description provided
Readme 3.5 GiB
Languages
C 97.1%
Assembly 1%
Shell 0.6%
Rust 0.4%
Python 0.4%
Other 0.3%