Lines Matching refs:ptr
669 my @ptr=map("%r$_",(8..15));
737 $ptr_reg=&pointer_register($flavour,@ptr[$i]);
752 cmovle %rsp,@ptr[$i] # cancel input
753 sub @ptr[$i],$temp # distance between input and output
765 vpxor (@ptr[0]),$zero,@inp[0] # load inputs and xor with 0-round
767 vpxor (@ptr[1]),$zero,@inp[1]
768 vpxor (@ptr[2]),$zero,@inp[2]
769 vpxor (@ptr[3]),$zero,@inp[3]
771 vpxor (@ptr[4]),$zero,@inp[0]
773 vpxor (@ptr[5]),$zero,@inp[1]
775 vpxor (@ptr[6]),$zero,@inp[2]
777 vpxor (@ptr[7]),$zero,@inp[3]
799 prefetcht0 31(@ptr[$i]) # prefetch input
803 prefetcht0 15(@ptr[$i-2]) # prefetch output
807 lea (@ptr[$i],$offset),$offset
808 cmovge %rsp,@ptr[$i] # cancel input
812 sub @ptr[$i],$offset
814 vpxor 16(@ptr[$i]),$zero,@inp[$i%4] # load input and xor with 0-round
818 lea 16(@ptr[$i],$offset),@ptr[$i] # switch to output
826 prefetcht0 15(@ptr[$i-2]) # prefetch output
827 prefetcht0 15(@ptr[$i-1])
904 vmovups @out[0],-16(@ptr[0]) # write output
905 sub $offset,@ptr[0] # switch to input
907 vmovups @out[1],-16(@ptr[1])
908 sub `64+1*8`(%rsp),@ptr[1]
910 vmovups @out[2],-16(@ptr[2])
911 sub `64+2*8`(%rsp),@ptr[2]
913 vmovups @out[3],-16(@ptr[3])
914 sub `64+3*8`(%rsp),@ptr[3]
916 vmovups @out[4],-16(@ptr[4])
917 sub `64+4*8`(%rsp),@ptr[4]
919 vmovups @out[5],-16(@ptr[5])
920 sub `64+5*8`(%rsp),@ptr[5]
922 vmovups @out[6],-16(@ptr[6])
923 sub `64+6*8`(%rsp),@ptr[6]
925 vmovups @out[7],-16(@ptr[7])
926 sub `64+7*8`(%rsp),@ptr[7]
1036 $ptr_reg=&pointer_register($flavour,@ptr[$i]);
1051 cmovle %rsp,@ptr[$i] # cancel input
1052 sub @ptr[$i],$temp # distance between input and output
1066 vmovdqu (@ptr[0]),@out[0] # load inputs
1067 vmovdqu (@ptr[1]),@out[1]
1068 vmovdqu (@ptr[2]),@out[2]
1069 vmovdqu (@ptr[3]),@out[3]
1070 vmovdqu (@ptr[4]),@out[4]
1071 vmovdqu (@ptr[5]),@out[5]
1072 vmovdqu (@ptr[6]),@out[6]
1073 vmovdqu (@ptr[7]),@out[7]
1108 prefetcht0 31(@ptr[$i]) # prefetch input
1112 prefetcht0 15(@ptr[$i-2]) # prefetch output
1116 lea (@ptr[$i],$offset),$offset
1117 cmovge %rsp,@ptr[$i] # cancel input
1121 sub @ptr[$i],$offset
1123 vmovdqu 16(@ptr[$i]),@inp[$i%4] # load input
1127 lea 16(@ptr[$i],$offset),@ptr[$i] # switch to output
1135 prefetcht0 15(@ptr[$i-2]) # prefetch output
1136 prefetcht0 15(@ptr[$i-1])
1220 vmovups @out[0],-16(@ptr[0]) # write output
1221 sub $offset,@ptr[0] # switch to input
1224 vmovups @out[1],-16(@ptr[1])
1225 sub `64+1*8`(%rsp),@ptr[1]
1229 vmovups @out[2],-16(@ptr[2])
1230 sub `64+2*8`(%rsp),@ptr[2]
1234 vmovups @out[3],-16(@ptr[3])
1235 sub `64+3*8`(%rsp),@ptr[3]
1239 vmovups @out[4],-16(@ptr[4])
1240 sub `64+4*8`(%rsp),@ptr[4]
1245 vmovups @out[5],-16(@ptr[5])
1246 sub `64+5*8`(%rsp),@ptr[5]
1249 vmovups @out[6],-16(@ptr[6])
1250 sub `64+6*8`(%rsp),@ptr[6]
1253 vmovups @out[7],-16(@ptr[7])
1254 sub `64+7*8`(%rsp),@ptr[7]
