powerpc: Tweak copy selection parameter in __copy_tofrom_user_power7()
Experiments with the netperf benchmark indicated that the size selecting VMX-based copies in __copy_tofrom_user_power7() was suboptimal on POWER8. Measurements showed that parity was in the neighbourhood of 3328 bytes, rather than greater than 4096. The change gives a 1.5-2.0% improvement in performance for 4096-byte buffers, reducing the relative time spent in __copy_tofrom_user_power7() from approximately 7% to approximately 5% in the TCP_RR benchmark. Signed-off-by: Andrew Jeffery <andrew@aj.id.au> Acked-by: Anton Blanchard <anton@samba.org> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This commit is contained in:
committed by
Michael Ellerman
parent
09b6c1129f
commit
a3f952df3c
@@ -82,14 +82,14 @@
|
|||||||
_GLOBAL(__copy_tofrom_user_power7)
|
_GLOBAL(__copy_tofrom_user_power7)
|
||||||
#ifdef CONFIG_ALTIVEC
|
#ifdef CONFIG_ALTIVEC
|
||||||
cmpldi r5,16
|
cmpldi r5,16
|
||||||
cmpldi cr1,r5,4096
|
cmpldi cr1,r5,3328
|
||||||
|
|
||||||
std r3,-STACKFRAMESIZE+STK_REG(R31)(r1)
|
std r3,-STACKFRAMESIZE+STK_REG(R31)(r1)
|
||||||
std r4,-STACKFRAMESIZE+STK_REG(R30)(r1)
|
std r4,-STACKFRAMESIZE+STK_REG(R30)(r1)
|
||||||
std r5,-STACKFRAMESIZE+STK_REG(R29)(r1)
|
std r5,-STACKFRAMESIZE+STK_REG(R29)(r1)
|
||||||
|
|
||||||
blt .Lshort_copy
|
blt .Lshort_copy
|
||||||
bgt cr1,.Lvmx_copy
|
bge cr1,.Lvmx_copy
|
||||||
#else
|
#else
|
||||||
cmpldi r5,16
|
cmpldi r5,16
|
||||||
|
|
||||||
|
|||||||
Reference in New Issue
Block a user