optimized dtrsm_logic_LT_16x4_power8.S and dtrsm_macros_LT_16x4_power8.S

This commit is contained in:
Werner Saar 2016-05-23 11:20:41 +02:00
parent bd06b246cc
commit 412bcd187a
2 changed files with 398 additions and 365 deletions

View File

@ -54,29 +54,26 @@ DSTRM_LT_L4x16_LOOP_START:
addic. L, KK, 0 addic. L, KK, 0
ble- DSTRM_LT_L4x16_SAVE ble- DSTRM_LT_L4x16_SAVE
mtctr L
DSTRM_LT_L4x16_LOOP: DSTRM_LT_L4x16_LOOP:
dcbt AO, PRE dcbt AO, PRE
dcbt BO, PRE dcbt BO, PRE
KERNEL_16x4 KERNEL_16x4
addic. L, L, -1 bdz- DSTRM_LT_L4x16_SAVE
ble- DSTRM_LT_L4x16_SAVE
dcbt AO, PRE dcbt AO, PRE
KERNEL_16x4 KERNEL_16x4
addic. L, L, -1 bdz- DSTRM_LT_L4x16_SAVE
ble- DSTRM_LT_L4x16_SAVE
dcbt AO, PRE dcbt AO, PRE
KERNEL_16x4 KERNEL_16x4
addic. L, L, -1 bdz- DSTRM_LT_L4x16_SAVE
ble- DSTRM_LT_L4x16_SAVE
dcbt AO, PRE dcbt AO, PRE
KERNEL_16x4 KERNEL_16x4
addic. L, L, -1 bdnz+ DSTRM_LT_L4x16_LOOP
bgt+ DSTRM_LT_L4x16_LOOP
DSTRM_LT_L4x16_SAVE: DSTRM_LT_L4x16_SAVE:

File diff suppressed because it is too large Load Diff