Lines Matching refs:r8
115 ldr r8, [r12, #offW1] @ R8=W1
117 mul r0, r8, r7 @ R0=W1*ROWr16[1]=b0 (ROWr16[1] must be the second arg, to have the possibility to save 1 cycle)
128 mlane r5, r8, r2, r5 @ R5-=W1*ROWr16[3]=b2 (ROWr16[3] must be the second arg, to have the possibility to save 1 cycle)
157 mlane r1, r8, r3, r1 @ R7-=W1*ROWr16[5]=b1
163 mlane r7, r8, r4, r7 @ R7-=W1*ROWr16[7]=b3
184 ldr r8, [r12, #offW2] @ R8=W2
192 mul r11, r8, r4 @ R11=W2*ROWr16[2]
219 mulne r10, r8, r9 @ R10=W2*ROWr16[6]
240 add r8, r6, r0 @ R8=a0+b0
247 and r8, r11, r8, asr #ROW_SHIFT @ R8=0x0000FFFF & ((a0+b0)>>11)
248 orr r8, r8, r9
249 str r8, [r14, #0]
251 add r8, r3, r5 @ R8=a2+b2
254 and r8, r11, r8, asr #ROW_SHIFT @ R8=0x0000FFFF & ((a2+b2)>>11)
255 orr r8, r8, r9
256 str r8, [r14, #4]
258 sub r8, r4, r7 @ R8=a3-b3
261 and r8, r11, r8, asr #ROW_SHIFT @ R8=0x0000FFFF & ((a3-b3)>>11)
262 orr r8, r8, r9
263 str r8, [r14, #8]
265 sub r8, r2, r1 @ R8=a1-b1
268 and r8, r11, r8, asr #ROW_SHIFT @ R8=0x0000FFFF & ((a1-b1)>>11)
269 orr r8, r8, r9
270 str r8, [r14, #12]
279 mov r8, #0x10000 @ R8=0xFFFF (2 steps needed!) it saves a ldr call (because of delay run).
280 sub r8, r8, #1 @ R8 is now ready.
281 and r5, r8, r6, lsl #3 @ R5=R8 & (R6<<3)= (ROWr16[0]<<3) & 0xFFFF
315 ldr r8, [r12, #offW1] @ R8=W1
317 mul r0, r8, r7 @ R0=W1*ROWr16[1]=b0 (ROWr16[1] must be the second arg, to have the possibility to save 1 cycle)
329 mlane r5, r8, r2, r5 @ R5-=W1*ROWr16[3]=b2 (ROWr16[3] must be the second arg, to have the possibility to save 1 cycle)
350 mlane r1, r8, r3, r1 @ R7-=W1*ROWr16[5x8]=b1
356 mlane r7, r8, r4, r7 @ R7-=W1*ROWr16[7x8]=b3
377 ldr r8, [r12, #offW2] @ R8=W2
380 mul r11, r8, r4 @ R11=W2*ROWr16[2]
404 mulne r10, r8, r9 @ R10=W2*ROWr16[6]
425 add r8, r6, r0 @ R8=a0+b0
427 mov r8, r8, asr #COL_SHIFT
429 strh r8, [r14, #0]
431 add r8, r3, r5 @ R8=a2+b2
433 mov r8, r8, asr #COL_SHIFT
435 strh r8, [r14, #32]
437 sub r8, r4, r7 @ R8=a3-b3
439 mov r8, r8, asr #COL_SHIFT
441 strh r8, [r14, #64]
443 sub r8, r2, r1 @ R8=a1-b1
445 mov r8, r8, asr #COL_SHIFT
447 strh r8, [r14, #96]
471 mul r11, r8, r4 @ R11=W2*ROWr16[2]