search for: macros_armv4

One other minor thing: I notice that in the inline assembly the result (rd) is constrained as an earlyclobber operand. What was the reason for that?

...On Nov 12, 2015, at 12:23 PM, John Ridges <jridges at masque.com> wrote: >> >> One other minor thing: I notice that in the inline assembly the result (rd) is constrained as an earlyclobber operand. What was the reason for that? > Possibly an error? Probably from modeling it on macros_armv4.h, which I guess does require earlyclobber. I?ll ask the developer.

...3 PM, John Ridges <jridges at masque.com> wrote: >>>> >>>> One other minor thing: I notice that in the inline assembly the result (rd) is constrained as an earlyclobber operand. What was the reason for that? >>> Possibly an error? Probably from modeling it on macros_armv4.h, which I guess does require earlyclobber. I?ll ask the developer. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.xiph.org/pipermail/opus/attachments/20151113/ea13efaf/attachment.htm

> On Nov 12, 2015, at 12:23 PM, John Ridges <jridges at masque.com> wrote: > > One other minor thing: I notice that in the inline assembly the result (rd) is constrained as an earlyclobber operand. What was the reason for that? Possibly an error? Probably from modeling it on macros_armv4.h, which I guess does require earlyclobber. I?ll ask the developer.

..._AARCH64_NEON_INTR +#include "arm/macros_arm64.h" +#endif + #endif /* SILK_MACROS_H */ diff --git a/silk_headers.mk b/silk_headers.mk index c74ab81..f8bf1d2 100644 --- a/silk_headers.mk +++ b/silk_headers.mk @@ -24,6 +24,7 @@ silk/SigProc_FIX.h \ silk/x86/SigProc_FIX_sse.h \ silk/arm/macros_armv4.h \ silk/arm/macros_armv5e.h \ +silk/arm/macros_arm64.h \ silk/arm/SigProc_FIX_armv4.h \ silk/arm/SigProc_FIX_armv5e.h \ silk/arm/NSQ_neon.h \ -- 2.4.9 (Apple Git-60)

...15, at 12:23 PM, John Ridges <jridges at masque.com> wrote: >>> >>> One other minor thing: I notice that in the inline assembly the result (rd) is constrained as an earlyclobber operand. What was the reason for that? >> Possibly an error? Probably from modeling it on macros_armv4.h, which I guess does require earlyclobber. I?ll ask the developer. >

...at 12:23 PM, John Ridges <jridges at masque.com><mailto:jridges at masque.com> wrote: One other minor thing: I notice that in the inline assembly the result (rd) is constrained as an earlyclobber operand. What was the reason for that? Possibly an error? Probably from modeling it on macros_armv4.h, which I guess does require earlyclobber. I?ll ask the developer. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.xiph.org/pipermail/opus/attachments/20151116/8710f6f9/attachment.htm

..._ARM64_INLINE_ADM +#include "arm/macros_arm64.h" +#endif + #endif /* SILK_MACROS_H */ diff --git a/silk_headers.mk b/silk_headers.mk index 679ff8f..6676133 100644 --- a/silk_headers.mk +++ b/silk_headers.mk @@ -23,8 +23,10 @@ silk/SigProc_FIX.h \ silk/x86/SigProc_FIX_sse.h \ silk/arm/macros_armv4.h \ silk/arm/macros_armv5e.h \ +silk/arm/macros_arm64.h \ silk/arm/SigProc_FIX_armv4.h \ silk/arm/SigProc_FIX_armv5e.h \ +silk/arm/SigProc_FIX_arm64.h \ silk/fixed/main_FIX.h \ silk/fixed/structs_FIX.h \ silk/fixed/mips/noise_shape_analysis_FIX_mipsr1.h \ -- 2.3.2 (Apple Git-55)

--- configure.ac | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+) diff --git a/configure.ac b/configure.ac index 90a06c8..adcb969 100644 --- a/configure.ac +++ b/configure.ac @@ -503,6 +503,26 @@ AS_IF([test x"$enable_intrinsics" = x"yes"],[ [rtcd_support="$rtcd_support (NE10)"]) ]) + OPUS_CHECK_INTRINSICS( +

...0644 --- a/silk/macros.h +++ b/silk/macros.h @@ -138,11 +138,11 @@ static OPUS_INLINE opus_int32 silk_CLZ32(opus_int32 in32) (*((Matrix_base_adr) + ((row)+(M)*(column)))) #endif -#ifdef OPUS_ARM_INLINE_ASM +#if defined(OPUS_ARM_INLINE_ASM) && defined(FIXED_POINT) #include "arm/macros_armv4.h" #endif -#ifdef OPUS_ARM_INLINE_EDSP +#if defined(OPUS_ARM_INLINE_EDSP) && defined(FIXED_POINT) #include "arm/macros_armv5e.h" #endif -- 1.7.9.5

--- configure.ac | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/configure.ac b/configure.ac index f52d2c2..e1a6e9b 100644 --- a/configure.ac +++ b/configure.ac @@ -190,7 +190,7 @@ AC_ARG_ENABLE([rtcd], [enable_rtcd=yes]) AC_ARG_ENABLE([intrinsics], - [AS_HELP_STRING([--disable-intrinsics], [Disable intrinsics optimizations for ARM(float) X86(fixed)])],, +

...headers.mk index 679ff8f..c74ab81 100644 --- a/silk_headers.mk +++ b/silk_headers.mk @@ -15,6 +15,7 @@ silk/Inlines.h \ silk/MacroCount.h \ silk/MacroDebug.h \ silk/macros.h \ +silk/NSQ.h \ silk/pitch_est_defines.h \ silk/resampler_private.h \ silk/resampler_rom.h \ @@ -25,6 +26,7 @@ silk/arm/macros_armv4.h \ silk/arm/macros_armv5e.h \ silk/arm/SigProc_FIX_armv4.h \ silk/arm/SigProc_FIX_armv5e.h \ +silk/arm/NSQ_neon.h \ silk/fixed/main_FIX.h \ silk/fixed/structs_FIX.h \ silk/fixed/mips/noise_shape_analysis_FIX_mipsr1.h \ diff --git a/silk_sources.mk b/silk_sources.mk index 7cfb7d3..79ac6f0 100...

...mk b/silk_headers.mk index 10a169a..d3dce08 100644 --- a/silk_headers.mk +++ b/silk_headers.mk @@ -22,6 +22,7 @@ silk/resampler_rom.h \ silk/resampler_structs.h \ silk/SigProc_FIX.h \ silk/x86/SigProc_FIX_sse.h \ +silk/arm/LPC_analysis_filter_arm.h \ silk/arm/LPC_inv_pred_gain_arm.h \ silk/arm/macros_armv4.h \ silk/arm/macros_armv5e.h \ diff --git a/silk_sources.mk b/silk_sources.mk index ac5647b..d2d5b35 100644 --- a/silk_sources.mk +++ b/silk_sources.mk @@ -85,6 +85,7 @@ silk/x86/VQ_WMat_EC_sse.c SILK_SOURCES_ARM_NEON_INTR = \ silk/arm/arm_silk_map.c \ +silk/arm/LPC_analysis_filter_neon_intr.c...

...761488 100644 --- a/silk_headers.mk +++ b/silk_headers.mk @@ -22,6 +22,7 @@ silk/resampler_rom.h \ silk/resampler_structs.h \ silk/SigProc_FIX.h \ silk/x86/SigProc_FIX_sse.h \ +silk/arm/inner_prod_aligned_arm.h \ silk/arm/LPC_analysis_filter_arm.h \ silk/arm/LPC_inv_pred_gain_arm.h \ silk/arm/macros_armv4.h \ diff --git a/silk_sources.mk b/silk_sources.mk index 9dcfe83..1823d30 100644 --- a/silk_sources.mk +++ b/silk_sources.mk @@ -85,6 +85,7 @@ silk/x86/VQ_WMat_EC_sse.c SILK_SOURCES_ARM_NEON_INTR = \ silk/arm/arm_silk_map.c \ +silk/arm/inner_prod_aligned_neon_intr.c \ silk/arm/LPC_analysis_fil...

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

Following Tim's comments, here are my reworked patches for the Neon intrinsic function patches of of my Aarch64 patchset, i.e. replacing patches 5-8 of the v2 series. Patches 1-4 and 9-18 of the old series still apply unmodified. The one new (as opposed to changed) patch is the first one in this series, to add named constants for the ARM architecture variants. There are also some minor code

As promised, here's a re-send of all my Aarch64 patches, following comments by John Ridges. Note that they actually affect more than just Aarch64 -- other than the ones specifically guarded by AARCH64_NEON defines, the Neon intrinsics all also apply on armv7; and the OPUS_FAST_INT64 patches apply on any 64-bit machine. The patches should largely be independent and independently useful, other

This sequence of patches provides arm64 support for Opus. Tested on iOS, Android, and Ubuntu 14.04. The patch sequence was written on top of Viswanath Puttagunta's Ne10 patches, but all but the second ("Reorganize pitch_arm.h") should, I think, apply independently of it. It does depends on my previous intrinsics configury reorganization, however. Comments welcome. With this and

Here are my aarch64 patches rebased to the current tip of Opus master. They're largely the same as my previous patch set, with the addition of the final one (the Neon fixed-point implementation of xcorr_kernel). This replaces Viswanath's Neon fixed-point celt_pitch_xcorr, since xcorr_kernel is used in celt_fir and celt_iir as well. These have been tested for correctness under qemu

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