diff options
| author | Kent Overstreet <kmo@daterainc.com> | 2013-08-21 18:36:45 -0700 |
|---|---|---|
| committer | Kent Overstreet <kmo@daterainc.com> | 2013-11-10 21:56:42 -0800 |
| commit | 098fb25498214069e6bbf908515f2952dd7654d0 (patch) | |
| tree | f170596e7abb95f78a6b89044945aaea81919214 /drivers/md/bcache | |
| parent | 28935ab5163c49ca5c199d67335e5e3c72c50853 (diff) | |
bcache: Delete some slower inline asm
Never saw a profile of bset_search_tree() where it wasn't bottlenecked
on memory until I got my new Haswell machine, but when I tried it there
it was suddenly burning 20% of the cpu in the inner loop on shrd...
Turns out, the version of shrd that takes 64 bit operands has a 9 cycle
latency. hah.
Signed-off-by: Kent Overstreet <kmo@daterainc.com>
Diffstat (limited to 'drivers/md/bcache')
| -rw-r--r-- | drivers/md/bcache/bset.c | 8 |
1 files changed, 0 insertions, 8 deletions
diff --git a/drivers/md/bcache/bset.c b/drivers/md/bcache/bset.c index 14573391206b..7d388b8bb50e 100644 --- a/drivers/md/bcache/bset.c +++ b/drivers/md/bcache/bset.c @@ -481,16 +481,8 @@ static struct bkey *table_to_bkey(struct bset_tree *t, unsigned cacheline) static inline uint64_t shrd128(uint64_t high, uint64_t low, uint8_t shift) { -#ifdef CONFIG_X86_64 - asm("shrd %[shift],%[high],%[low]" - : [low] "+Rm" (low) - : [high] "R" (high), - [shift] "ci" (shift) - : "cc"); -#else low >>= shift; low |= (high << 1) << (63U - shift); -#endif return low; } |