search for: vextq_f32

Hi, Optimizes celt_pitch_xcorr for ARM NEON floating point. Changes from RFCv3: - celt_neon_intr.c - removed warnings due to not having constant pointers - Put simpler loop to take care of corner cases. Unrolling using intrinsics was not really mapping well to what was done in celt_pitch_xcorr_arm.s - Makefile.am Removed explicit -O3 optimization - test_unit_mathops.c,

...*/ + while (len > 8) { + yi += 4; + YY[1] = vld1q_f32(yi); + yi += 4; + YY[2] = vld1q_f32(yi); + + XX[0] = vld1q_f32(xi); + xi += 4; + XX[1] = vld1q_f32(xi); + xi += 4; + + SUMM = vmlaq_lane_f32(SUMM, YY[0], vget_low_f32(XX[0]), 0); + YEXT[0] = vextq_f32(YY[0], YY[1], 1); + SUMM = vmlaq_lane_f32(SUMM, YEXT[0], vget_low_f32(XX[0]), 1); + YEXT[1] = vextq_f32(YY[0], YY[1], 2); + SUMM = vmlaq_lane_f32(SUMM, YEXT[1], vget_high_f32(XX[0]), 0); + YEXT[2] = vextq_f32(YY[0], YY[1], 3); + SUMM = vmlaq_lane_f32(SUMM, YEXT[2], vget_hig...

...*/ + while (len > 8) { + yi += 4; + YY[1] = vld1q_f32(yi); + yi += 4; + YY[2] = vld1q_f32(yi); + + XX[0] = vld1q_f32(xi); + xi += 4; + XX[1] = vld1q_f32(xi); + xi += 4; + + SUMM = vmlaq_lane_f32(SUMM, YY[0], vget_low_f32(XX[0]), 0); + YEXT[0] = vextq_f32(YY[0], YY[1], 1); + SUMM = vmlaq_lane_f32(SUMM, YEXT[0], vget_low_f32(XX[0]), 1); + YEXT[1] = vextq_f32(YY[0], YY[1], 2); + SUMM = vmlaq_lane_f32(SUMM, YEXT[1], vget_high_f32(XX[0]), 0); + YEXT[2] = vextq_f32(YY[0], YY[1], 3); + SUMM = vmlaq_lane_f32(SUMM, YEXT[2], vget_hig...

...*/ + while (len > 8) { + yi += 4; + YY[1] = vld1q_f32(yi); + yi += 4; + YY[2] = vld1q_f32(yi); + + XX[0] = vld1q_f32(xi); + xi += 4; + XX[1] = vld1q_f32(xi); + xi += 4; + + SUMM = vmlaq_lane_f32(SUMM, YY[0], vget_low_f32(XX[0]), 0); + YEXT[0] = vextq_f32(YY[0], YY[1], 1); + SUMM = vmlaq_lane_f32(SUMM, YEXT[0], vget_low_f32(XX[0]), 1); + YEXT[1] = vextq_f32(YY[0], YY[1], 2); + SUMM = vmlaq_lane_f32(SUMM, YEXT[1], vget_high_f32(XX[0]), 0); + YEXT[2] = vextq_f32(YY[0], YY[1], 3); + SUMM = vmlaq_lane_f32(SUMM, YEXT[2], vget_hig...

...YY[1] = vld1q_f32(yi); > + yi += 4; > + YY[2] = vld1q_f32(yi); > + > + XX[0] = vld1q_f32(xi); > + xi += 4; > + XX[1] = vld1q_f32(xi); > + xi += 4; > + > + SUMM = vmlaq_lane_f32(SUMM, YY[0], vget_low_f32(XX[0]), 0); > + YEXT[0] = vextq_f32(YY[0], YY[1], 1); > + SUMM = vmlaq_lane_f32(SUMM, YEXT[0], vget_low_f32(XX[0]), 1); > + YEXT[1] = vextq_f32(YY[0], YY[1], 2); > + SUMM = vmlaq_lane_f32(SUMM, YEXT[1], vget_high_f32(XX[0]), 0); > + YEXT[2] = vextq_f32(YY[0], YY[1], 3); > + SUMM = vmlaq_lane_f3...

Hi, Optimizes celt_pitch_xcorr for floating point. Changes from RFCv2: - Changes recommended by Timothy for celt_neon_intr.c everything except, left the unrolled loop still unrolled - configure.ac - use AC_LINK_IFELSE instead of AC_COMPILE_IFELSE - Moved compile flags into Makefile.am - OPUS_ARM_NEON_INR --> typo --> OPUS_ARM_NEON_INTR Viswanath Puttagunta (1): armv7:

From: Viswanath Puttagunta <viswanath.puttagunta at linaro.org> Hi, Optimizes celt_pitch_xcorr for floating point. Changes from RFCv1: - Rebased on top of commit aad281878: Fix celt_pitch_xcorr_c signature. which got rid of ugly code around CELT_PITCH_XCORR_IMPL passing of "arch" parameter. - Unified with --enable-intrinsics used by x86 - Modified algorithm to be more

Hi, Optimizes celt_pitch_xcorr for floating point. Changes from RFCv1: - Rebased on top of commit aad281878: Fix celt_pitch_xcorr_c signature. which got rid of ugly code around CELT_PITCH_XCORR_IMPL passing of "arch" parameter. - Unified with --enable-intrinsics used by x86 - Modified algorithm to be more in-line with algorithm in celt_pitch_xcorr_arm.s Viswanath Puttagunta

...code only references y[0] up to y[len-1+3]). You need to end this loop early and fall back to another approach. See comments in celt_pitch_xcorr_arm.s for details and an example (there are other useful comments there that could shave another cycle or two from this inner loop). > + YY[1] = vextq_f32(YY[0], YY[4], 1); > + YY[2] = vextq_f32(YY[0], YY[4], 2); > + YY[3] = vextq_f32(YY[0], YY[4], 3); > + > + XX[0] = vld1q_dup_f32(xi++); > + XX[1] = vld1q_dup_f32(xi++); > + XX[2] = vld1q_dup_f32(xi++); > + XX[3] = vld1q_dup_f32(xi++); Don't do this. Do a single load a...

...e that > could shave another cycle or two from this inner loop). Analyzed the implementation in celt_pitch_xcorr_arm.s. I will re-do my implementation to follow same algorithm. It seems more elegant. This comment applies to rest of feedback on celt_neon_intr.c > >> + YY[1] = vextq_f32(YY[0], YY[4], 1); >> + YY[2] = vextq_f32(YY[0], YY[4], 2); >> + YY[3] = vextq_f32(YY[0], YY[4], 3); >> + >> + XX[0] = vld1q_dup_f32(xi++); >> + XX[1] = vld1q_dup_f32(xi++); >> + XX[2] = vld1q_dup_f32(xi+...

Hello, I received feedback from engineers working on NE10 [1] that it would be better to use NE10 [1] for FFT optimizations for opus use cases. However, these FFT patches are currently in review and haven't been integrated into NE10 yet. While the FFT functions in NE10 are getting baked, I wanted to optimize the celt_pitch_xcorr (floating point only) and use it to introduce ARM NEON

...UMM[0] = vdupq_n_f32(0); + SUMM[1] = vdupq_n_f32(0); + SUMM[2] = vdupq_n_f32(0); + SUMM[3] = vdupq_n_f32(0); + + YY[0] = vld1q_f32(yi); + + /* Each loop consumes 8 floats in y vector + * and 4 floats in x vector + */ + for (j = 0; j < cd; j++) { + yi += 4; + YY[4] = vld1q_f32(yi); + YY[1] = vextq_f32(YY[0], YY[4], 1); + YY[2] = vextq_f32(YY[0], YY[4], 2); + YY[3] = vextq_f32(YY[0], YY[4], 3); + + XX[0] = vld1q_dup_f32(xi++); + XX[1] = vld1q_dup_f32(xi++); + XX[2] = vld1q_dup_f32(xi++); + XX[3] = vld1q_dup_f32(xi++); + + SUMM[0] = vmlaq_f32(SUMM[0], XX[0], YY[0]); + SUMM[1] = vmlaq_f32(S...