The patch titled
Subject: include/asm-generic/atomic-long.h: force inlining of some atomic_long operations
has been added to the -mm tree. Its filename is
asm-generic-force-inlining-of-some-atomic_long-operations.patch
This patch should soon appear at
http://ozlabs.org/~akpm/mmots/broken-out/asm-generic-force-inlining-of-some-atomic_long-operations.patch
and later at
http://ozlabs.org/~akpm/mmotm/broken-out/asm-generic-force-inlining-of-some-atomic_long-operations.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/SubmitChecklist when testing your code ***
The -mm tree is included into linux-next and is updated
there every 3-4 working days
------------------------------------------------------
From: Denys Vlasenko <dvlasenk@xxxxxxxxxx>
Subject: include/asm-generic/atomic-long.h: force inlining of some atomic_long operations
Sometimes gcc mysteriously doesn't inline
very small functions we expect to be inlined. See
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66122
With this .config:
http://busybox.net/~vda/kernel_config_OPTIMIZE_INLINING_and_Os,
atomic_long_inc(), atomic_long_dec() and atomic_long_add()
functions get deinlined about 40 times. Examples of disassembly:
<atomic_long_inc> (21 copies, 147 calls):
55 push %rbp
48 89 e5 mov %rsp,%rbp
f0 48 ff 07 lock incq (%rdi)
5d pop %rbp
c3 retq
<atomic_long_dec> (4 copies, 14 calls) is similar to inc.
<atomic_long_add> (11 copies, 41 calls):
55 push %rbp
48 89 e5 mov %rsp,%rbp
f0 48 01 3e lock add %rdi,(%rsi)
5d pop %rbp
c3 retq
This patch fixes this via s/inline/__always_inline/.
Code size decrease after the patch is ~1.3k:
text data bss dec hex filename
92203657 20826112 36417536 149447305 8e86289 vmlinux
92202377 20826112 36417536 149446025 8e85d89 vmlinux4_atomiclong_after
Signed-off-by: Denys Vlasenko <dvlasenk@xxxxxxxxxx>
Cc: Ingo Molnar <mingo@xxxxxxxxxx>
Cc: Thomas Graf <tgraf@xxxxxxx>
Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Cc: David Rientjes <rientjes@xxxxxxxxxx>
Cc: Arnd Bergmann <arnd@xxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
include/asm-generic/atomic-long.h | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff -puN include/asm-generic/atomic-long.h~asm-generic-force-inlining-of-some-atomic_long-operations include/asm-generic/atomic-long.h
--- a/include/asm-generic/atomic-long.h~asm-generic-force-inlining-of-some-atomic_long-operations
+++ a/include/asm-generic/atomic-long.h
@@ -98,14 +98,14 @@ ATOMIC_LONG_ADD_SUB_OP(sub, _release)
#define atomic_long_xchg(v, new) \
(ATOMIC_LONG_PFX(_xchg)((ATOMIC_LONG_PFX(_t) *)(v), (new)))
-static inline void atomic_long_inc(atomic_long_t *l)
+static __always_inline void atomic_long_inc(atomic_long_t *l)
{
ATOMIC_LONG_PFX(_t) *v = (ATOMIC_LONG_PFX(_t) *)l;
ATOMIC_LONG_PFX(_inc)(v);
}
-static inline void atomic_long_dec(atomic_long_t *l)
+static __always_inline void atomic_long_dec(atomic_long_t *l)
{
ATOMIC_LONG_PFX(_t) *v = (ATOMIC_LONG_PFX(_t) *)l;
@@ -113,7 +113,7 @@ static inline void atomic_long_dec(atomi
}
#define ATOMIC_LONG_OP(op) \
-static inline void \
+static __always_inline void \
atomic_long_##op(long i, atomic_long_t *l) \
{ \
ATOMIC_LONG_PFX(_t) *v = (ATOMIC_LONG_PFX(_t) *)l; \
_
Patches currently in -mm which might be from dvlasenk@xxxxxxxxxx are
asm-generic-force-inlining-of-some-atomic_long-operations.patch
force-inlining-of-some-byteswap-operations.patch
force-inlining-of-unaligned-byteswap-operations.patch
--
To unsubscribe from this list: send the line "unsubscribe mm-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
