search for: silk_lpc_inverse_pred_gain

Hi, Attached is a patch with arm neon optimizations for silk_LPC_inverse_pred_gain(). Please review. Thanks, Linfeng -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.xiph.org/pipermail/opus/attachments/20170207/5c5ab508/attachment-0001.html> -------------- next part -------------- A non-text attachment was scrubbed... Name: 0001...

...could be triggered by the > following input vector: > A_QA[] = { 46596096, -72118272, 78532608, -69447680, 52707328, > -22073344, -19890176, 50507776, -54829056, 45518848, -33939456, > 21086208, -7127040, -4136960, 3993600, -1699840 } > > 2) I'm not quite sure what to make of silk_LPC_inverse_pred_gain_Q24(). > It seems to never be called anywhere -- except from the MIPS code. Maybe > it should just stay as it is (not renamed to _c()) but I need to check. > Any thoughts? > > Cheers, > > Jean-Marc > > On 07/02/17 04:06 PM, Linfeng Zhang wrote: > > Hi, >...

...cit unit test. IIRC, the case could be triggered by the following input vector: A_QA[] = { 46596096, -72118272, 78532608, -69447680, 52707328, -22073344, -19890176, 50507776, -54829056, 45518848, -33939456, 21086208, -7127040, -4136960, 3993600, -1699840 } 2) I'm not quite sure what to make of silk_LPC_inverse_pred_gain_Q24(). It seems to never be called anywhere -- except from the MIPS code. Maybe it should just stay as it is (not renamed to _c()) but I need to check. Any thoughts? Cheers, Jean-Marc On 07/02/17 04:06 PM, Linfeng Zhang wrote: > Hi, > > Attached is a patch with arm neon optimizations...

...y the > following input vector: > A_QA[] = { 46596096, -72118272, 78532608, -69447680, 52707328, > -22073344, -19890176, 50507776, -54829056, 45518848, -33939456, > 21086208, -7127040, -4136960, 3993600, -1699840 } > > 2) I'm not quite sure what to make of silk_LPC_inverse_pred_gain_Q24(). > It seems to never be called anywhere -- except from the MIPS code. Maybe > it should just stay as it is (not renamed to _c()) but I need to check. > Any thoughts? > > Cheers, > > Jean-Marc > > On 07/02/17 04:06 PM, Linfeng Zhang...

...A_QA[] = { 46596096, -72118272, 78532608, -69447680, 52707328, > > -22073344, -19890176, 50507776, -54829056, 45518848, -33939456, > > 21086208, -7127040, -4136960, 3993600, -1699840 } > > > > 2) I'm not quite sure what to make of > silk_LPC_inverse_pred_gain_Q24(). > > It seems to never be called anywhere -- except from the MIPS > code. Maybe > > it should just stay as it is (not renamed to _c()) but I need > to check. > > Any thoughts? > > > > Cheers, > > >...

...input vector: > > A_QA[] = { 46596096, -72118272, 78532608, -69447680, 52707328, > > -22073344, -19890176, 50507776, -54829056, 45518848, -33939456, > > 21086208, -7127040, -4136960, 3993600, -1699840 } > > > > 2) I'm not quite sure what to make of silk_LPC_inverse_pred_gain_ > Q24(). > > It seems to never be called anywhere -- except from the MIPS code. > Maybe > > it should just stay as it is (not renamed to _c()) but I need to > check. > > Any thoughts? > > > > Cheers, > > > > Jean-Marc &...

Hi, Attached is a patch with silk/tests/test_unit_optimization_LPC_inv_pred_gain which does the unit test of silk_LPC_inverse_pred_gain() optimizations. Please review. The testing loop number is set to 10,000, since all branches in this function get hit after 9,085 loops of random inputs. Thanks, Linfeng -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.xiph.org/pipermail/opus/attach...

...biquad_alt_stride2_c, /* ARMv4 */ silk_biquad_alt_stride2_c, /* EDSP */ silk_biquad_alt_stride2_c, /* Media */ - MAY_HAVE_NEON(silk_biquad_alt_stride2), /* Neon */ + silk_biquad_alt_stride2_neon, /* Neon */ }; opus_int32 (*const SILK_LPC_INVERSE_PRED_GAIN_IMPL[OPUS_ARCHMASK + 1])( /* O Returns inverse prediction gain in energy domain, Q30 */ @@ -58,7 +58,7 @@ opus_int32 (*const SILK_LPC_INVERSE_PRED_GAIN_IMPL[OPUS_ARCHMASK + 1])( /* O R silk_LPC_inverse_pred_gain_c, /* ARMv4 */ silk_LPC_inverse_pred_gain_c,...

Thank Jonathan! I'll fix the MAY_HAVE_NEON() in silk/arm/arm_silk_map.c Linfeng On Thu, Jun 1, 2017 at 3:34 PM, Jonathan Lennox <jonathan at vidyo.com> wrote: > Semantically, OPUS_ARM_MAY_HAVE_NEON is supposed to mean the compiler > supports, and the CPU may support, Neon assembly code, which isn’t > necessarily the same thing as the compiler supporting Neon intrinsics. >

I rebased my previous 3 patches to the current master with minor changes. Patches 1 to 3 replace all my previous submitted patches. Patches 4 and 5 are new. Thanks, Linfeng Zhang

...N /* I Number of input samples */ ); +#if !defined(OVERRIDE_SILK_LPC_ANALYSIS_FILTER) +#define silk_LPC_analysis_filter(out, in, B, len, d, arch) (silk_LPC_analysis_filter_c(out, in, B, len, d, arch)) +#endif + #if !defined(OVERRIDE_silk_LPC_inverse_pred_gain) #define silk_LPC_inverse_pred_gain(A_Q12, order, arch) ((void)(arch),silk_LPC_inverse_pred_gain_c(A_Q12, order)) #endif diff --git a/silk/arm/LPC_analysis_filter_arm.h b/silk/arm/LPC_analysis_filter_arm.h new file mode 100644 index 0000000..96615d2 --- /dev/null +++ b/silk/arm/LPC_analysis_f...

...+ } + + /* Save inverse gain */ + NSQ->prev_gain_Q16 = Gains_Q16[ subfr ]; + } +} diff --git a/silk/arm/arm_silk_map.c b/silk/arm/arm_silk_map.c index b1783c7..8deaf99 100644 --- a/silk/arm/arm_silk_map.c +++ b/silk/arm/arm_silk_map.c @@ -60,6 +60,29 @@ opus_int32 (*const SILK_LPC_INVERSE_PRED_GAIN_IMPL[OPUS_ARCHMASK + 1])( /* O R MAY_HAVE_NEON(silk_LPC_inverse_pred_gain), /* Neon */ }; +void (*const SILK_NSQ_DEL_DEC_IMPL[OPUS_ARCHMASK + 1])( + const silk_encoder_state *psEncC, /* I Encoder State */ + silk_n...

NSQ_LPC_BUF_LENGTH is independent of DECISION_DELAY. --- silk/define.h | 4 ---- 1 file changed, 4 deletions(-) diff --git a/silk/define.h b/silk/define.h index 781cfdc..1286048 100644 --- a/silk/define.h +++ b/silk/define.h @@ -173,11 +173,7 @@ extern "C" #define MAX_MATRIX_SIZE MAX_LPC_ORDER /* Max of LPC Order and LTP order */ -#if( MAX_LPC_ORDER >