AnotherFoxGuy/webrtc-audio-processing
1
0
Fork 0
Code Pull Requests Activity

Track patches to WebRTC code in patches/

Browse Source 
This should make it easier to not lose track during the next bump.
This commit is contained in:
Arun Raghavan
2024-12-30 14:05:31 -05:00
parent 297fd4f2ef
commit a1ccd6c700
5 changed files with 627 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

68
patches/0001-AECM-MIPS-Use-uintptr_t-for-pointer-arithmetic.patch Normal file
View File

@ -0,0 +1,68 @@
From 297fd4f2efc53b6d49433eaad91a8e09a0f9cbec Mon Sep 17 00:00:00 2001
From: Alper Nebi Yasak <alpernebiyasak@gmail.com>
Date: Fri, 25 Oct 2024 00:40:59 +0300
Subject: [PATCH] AECM: MIPS: Use uintptr_t for pointer arithmetic
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Trying to compile the MIPS-specific AECM audio processing file for
mips64el on Debian results in the following errors:
../webrtc/modules/audio_processing/aecm/aecm_core_mips.cc: In function ‘int webrtc::WebRtcAecm_ProcessBlock(AecmCore*, const int16_t*, const int16_t*, const int16_t*, int16_t*)’:
../webrtc/modules/audio_processing/aecm/aecm_core_mips.cc:955:30: error: cast from ‘int16_t*’ {aka ‘short int*’} to ‘uint32_t’ {aka ‘unsigned int’} loses precision [-fpermissive]
955 | int16_t* fft = (int16_t*)(((uint32_t)fft_buf + 31) & ~31);
| ^~~~~~~~~~~~~~~~~
../webrtc/modules/audio_processing/aecm/aecm_core_mips.cc:955:18: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast]
955 | int16_t* fft = (int16_t*)(((uint32_t)fft_buf + 31) & ~31);
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../webrtc/modules/audio_processing/aecm/aecm_core_mips.cc:956:36: error: cast from ‘int32_t*’ {aka ‘int*’} to ‘uint32_t’ {aka ‘unsigned int’} loses precision [-fpermissive]
956 | int32_t* echoEst32 = (int32_t*)(((uint32_t)echoEst32_buf + 31) & ~31);
| ^~~~~~~~~~~~~~~~~~~~~~~
../webrtc/modules/audio_processing/aecm/aecm_core_mips.cc:956:24: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast]
956 | int32_t* echoEst32 = (int32_t*)(((uint32_t)echoEst32_buf + 31) & ~31);
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../webrtc/modules/audio_processing/aecm/aecm_core_mips.cc:957:40: error: cast from ‘int32_t*’ {aka ‘int*’} to ‘uint32_t’ {aka ‘unsigned int’} loses precision [-fpermissive]
957 | ComplexInt16* dfw = (ComplexInt16*)(((uint32_t)dfw_buf + 31) & ~31);
| ^~~~~~~~~~~~~~~~~
../webrtc/modules/audio_processing/aecm/aecm_core_mips.cc:957:23: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast]
957 | ComplexInt16* dfw = (ComplexInt16*)(((uint32_t)dfw_buf + 31) & ~31);
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../webrtc/modules/audio_processing/aecm/aecm_core_mips.cc:958:40: error: cast from ‘int32_t*’ {aka ‘int*’} to ‘uint32_t’ {aka ‘unsigned int’} loses precision [-fpermissive]
958 | ComplexInt16* efw = (ComplexInt16*)(((uint32_t)efw_buf + 31) & ~31);
| ^~~~~~~~~~~~~~~~~
../webrtc/modules/audio_processing/aecm/aecm_core_mips.cc:958:23: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast]
958 | ComplexInt16* efw = (ComplexInt16*)(((uint32_t)efw_buf + 31) & ~31);
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Presumably, this file was written for 32-bit MIPS so the author used
uint32_t to do pointer arithmetic over these arrays. Fix the errors by
using uintptr_t to work with pointers.
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
---
webrtc/modules/audio_processing/aecm/aecm_core_mips.cc | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/webrtc/modules/audio_processing/aecm/aecm_core_mips.cc b/webrtc/modules/audio_processing/aecm/aecm_core_mips.cc
index 16b03cf..07c785e 100644
--- a/webrtc/modules/audio_processing/aecm/aecm_core_mips.cc
+++ b/webrtc/modules/audio_processing/aecm/aecm_core_mips.cc
@@ -952,10 +952,10 @@ int WebRtcAecm_ProcessBlock(AecmCore* aecm,
int32_t dfw_buf[PART_LEN2 + 8];
int32_t efw_buf[PART_LEN2 + 8];
- int16_t* fft = (int16_t*)(((uint32_t)fft_buf + 31) & ~31);
- int32_t* echoEst32 = (int32_t*)(((uint32_t)echoEst32_buf + 31) & ~31);
- ComplexInt16* dfw = (ComplexInt16*)(((uint32_t)dfw_buf + 31) & ~31);
- ComplexInt16* efw = (ComplexInt16*)(((uint32_t)efw_buf + 31) & ~31);
+ int16_t* fft = (int16_t*)(((uintptr_t)fft_buf + 31) & ~31);
+ int32_t* echoEst32 = (int32_t*)(((uintptr_t)echoEst32_buf + 31) & ~31);
+ ComplexInt16* dfw = (ComplexInt16*)(((uintptr_t)dfw_buf + 31) & ~31);
+ ComplexInt16* efw = (ComplexInt16*)(((uintptr_t)efw_buf + 31) & ~31);
int16_t hnl[PART_LEN1];
int16_t numPosCoef = 0;
--
2.47.1