linux/bitmap: Force inlining of bitmap weight functions
authorDenys Vlasenko <dvlasenk@redhat.com>
Tue, 4 Aug 2015 14:15:14 +0000 (16:15 +0200)
committerIngo Molnar <mingo@kernel.org>
Wed, 5 Aug 2015 07:38:08 +0000 (09:38 +0200)
commit1a1d48a4a8fde49aedc045d894efe67173d59fe0
treeb1e4dc0bd4c7621c98610947d3baa417d73ae6ef
parentc2f3ba745d1c2013811cac6308c4abf9527c478e
linux/bitmap: Force inlining of bitmap weight functions

With this config:

  http://busybox.net/~vda/kernel_config_OPTIMIZE_INLINING_and_Os

gcc-4.7.2 generates many copies of these tiny functions:

bitmap_weight (55 copies):
55                      push   %rbp
48 89 e5                mov    %rsp,%rbp
e8 3f 3a 8b 00          callq  __bitmap_weight
5d                      pop    %rbp
c3                      retq

hweight_long (23 copies):
55                      push   %rbp
e8 b5 65 8e 00          callq  __sw_hweight64
48 89 e5                mov    %rsp,%rbp
5d                      pop    %rbp
c3                      retq

See https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66122

This patch fixes this via s/inline/__always_inline/

While at it, replaced two "__inline__" with usual "inline"
(the rest of the source file uses the latter).

    text     data      bss       dec  filename
86971357 17195880 36659200 140826437  vmlinux.before
86971120 17195912 36659200 140826232  vmlinux

Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: David Rientjes <rientjes@google.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Thomas Graf <tgraf@suug.ch>
Cc: linux-kernel@vger.kernel.org
Link: http://lkml.kernel.org/r/1438697716-28121-1-git-send-email-dvlasenk@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
include/linux/bitmap.h
include/linux/bitops.h