Lines Matching refs:loop
45 .loop:
53 jl .loop
82 .loop:
100 jl .loop
120 .loop%1:
157 jg .loop%1
253 .loop:
315 jae .loop
348 .loop:
385 jl .loop
397 %%.loop:
422 jl %%.loop
466 .loop:
471 js .loop
491 ; jump to the slow loop functions
493 ; jump to the fast loop functions
497 ; ... and then the same for left/right extend also. See below for loop
625 ; opcode, and then the loop fits in 128 bytes)
885 ; width specified in a register, and run a loop to copy the full amount
888 ; bytes per loop iteration. It could be considered to use xmm for x86-64
1060 ; %2 = number of inline load/process/store loops per asm loop
1061 ; %3 = process 4*mmsize (%3=0) or 8*mmsize (%3=1) bytes per loop
1075 .loop:
1113 jg .loop
1151 .loop:
1168 jl .loop