[FFmpeg-devel] [PATCH v4 5/5] avcodec/x86/diracdsp_init: remove unused macro

2024-11-14 Thread Kyosuke Kawakami
PIXFUNC macro is unused since d29a9c2aa68fc3eb6d61ff95c698e29316037583. Signed-off-by: Kyosuke Kawakami --- libavcodec/x86/diracdsp_init.c | 5 - 1 file changed, 5 deletions(-) diff --git a/libavcodec/x86/diracdsp_init.c b/libavcodec/x86/diracdsp_init.c index 08247133e1..ef01ebdf2e 100644

[FFmpeg-devel] [PATCH v4 3/5] avcodec/x86/diracdsp: cast stride argument

2024-11-14 Thread Kyosuke Kawakami
Signed-off-by: Kyosuke Kawakami --- libavcodec/x86/diracdsp.asm | 1 + 1 file changed, 1 insertion(+) diff --git a/libavcodec/x86/diracdsp.asm b/libavcodec/x86/diracdsp.asm index 03b929da76..a653fa04de 100644 --- a/libavcodec/x86/diracdsp.asm +++ b/libavcodec/x86/diracdsp.asm @@ -218,6 +218,7

[FFmpeg-devel] [PATCH v4 4/5] avcodec/x86/diracdsp: migrate last remaining MMX function to SSE2

2024-11-14 Thread Kyosuke Kawakami
: Kyosuke Kawakami --- libavcodec/x86/diracdsp.asm| 24 libavcodec/x86/diracdsp_init.c | 10 +++--- 2 files changed, 23 insertions(+), 11 deletions(-) diff --git a/libavcodec/x86/diracdsp.asm b/libavcodec/x86/diracdsp.asm index a653fa04de..6ae7f888b3 100644 --- a

[FFmpeg-devel] [PATCH v4 2/5] avcodec/x86/diracdsp: fix wrong register load/use count

2024-11-14 Thread Kyosuke Kawakami
--- libavcodec/x86/diracdsp.asm | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/libavcodec/x86/diracdsp.asm b/libavcodec/x86/diracdsp.asm index e5e2b11846..03b929da76 100644 --- a/libavcodec/x86/diracdsp.asm +++ b/libavcodec/x86/diracdsp.asm @@ -216,7 +216,7 @@ cglobal add_rect

[FFmpeg-devel] [PATCH v4 1/5] checkasm/diracdsp: test add_dirac_obmc

2024-11-14 Thread Kyosuke Kawakami
Signed-off-by: Kyosuke Kawakami --- tests/checkasm/Makefile | 1 + tests/checkasm/checkasm.c | 3 ++ tests/checkasm/checkasm.h | 1 + tests/checkasm/diracdsp.c | 91 +++ tests/fate/checkasm.mak | 1 + 5 files changed, 97 insertions(+) create mode

[FFmpeg-devel] [PATCH v4 0/5] avcodec/x86/diracdsp: migrate last remaining MMX function to SSE2

2024-11-14 Thread Kyosuke Kawakami
This series of patch migrates the last remaining MMX function in diracdsp to SSE2. Changes from v3 are: - Use correct register load/use counts - Fix garbage value issue on Windows - Use constant yblen on checkasm benchmark - Test that functions accept unaligned buffer Thanks to James and Ronald f

Re: [FFmpeg-devel] [PATCH v3 0/3] avcodec/x86/diracdsp: migrate last remaining MMX function to SSE2

2024-11-14 Thread Kyosuke Kawakami
Sorry, I sent a top message twice by mistake. Please ignore this one. Kyosuke ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmpeg.o

[FFmpeg-devel] [PATCH v3 3/3] avcodec/x86/diracdsp_init: remove unused macro

2024-11-14 Thread Kyosuke Kawakami
PIXFUNC macro is unused since d29a9c2aa68fc3eb6d61ff95c698e29316037583. Signed-off-by: Kyosuke Kawakami --- libavcodec/x86/diracdsp_init.c | 5 - 1 file changed, 5 deletions(-) diff --git a/libavcodec/x86/diracdsp_init.c b/libavcodec/x86/diracdsp_init.c index 08247133e1..ef01ebdf2e 100644

[FFmpeg-devel] [PATCH v3 2/3] avcodec/x86/diracdsp: migrate last remaining MMX function to SSE2

2024-11-14 Thread Kyosuke Kawakami
: Kyosuke Kawakami --- libavcodec/x86/diracdsp.asm| 23 +++ libavcodec/x86/diracdsp_init.c | 10 +++--- 2 files changed, 22 insertions(+), 11 deletions(-) diff --git a/libavcodec/x86/diracdsp.asm b/libavcodec/x86/diracdsp.asm index e5e2b11846..e708400b66 100644 --- a

[FFmpeg-devel] [PATCH v3 1/3] checkasm/diracdsp: test add_dirac_obmc

2024-11-14 Thread Kyosuke Kawakami
Signed-off-by: Kyosuke Kawakami --- tests/checkasm/Makefile | 1 + tests/checkasm/checkasm.c | 3 ++ tests/checkasm/checkasm.h | 1 + tests/checkasm/diracdsp.c | 86 +++ tests/fate/checkasm.mak | 1 + 5 files changed, 92 insertions(+) create mode

[FFmpeg-devel] [PATCH v3 0/3] avcodec/x86/diracdsp: migrate last remaining MMX function to SSE2

2024-11-14 Thread Kyosuke Kawakami
This series of patch migrates the last remaining MMX function in diracdsp to SSE2. Changes from v2 are: - Rewrite tests with a normal function instead of a macro - Fix typo in a commit message GIT: [PATCH v3 1/3] checkasm/diracdsp: test add_dirac_obmc GIT: [PATCH v3 2/3] avcodec/x86/diracdsp: mi

[FFmpeg-devel] [PATCH v3 0/3] avcodec/x86/diracdsp: migrate last remaining MMX function to SSE2

2024-11-14 Thread Kyosuke Kawakami
This series of patch migrates the last remaining MMX function in diracdsp to SSE2. Changes from v2 are: - Rewrite tests with a normal function instead of a macro - Fix typo in a commit message ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https:

Re: [FFmpeg-devel] [PATCH v2 1/3] checkasm/diracdsp: test add_dirac_obmc

2024-11-14 Thread Kyosuke Kawakami
. Kyosuke On Thu, Nov 14, 2024 at 9:40 PM Ronald S. Bultje wrote: > > Hi, > > thanks for adding the test! This looks pretty good. Minor suggestion: > > On Wed, Nov 13, 2024 at 5:39 PM Kyosuke Kawakami > wrote: >> >> +#define RANDOMIZE_DESTS(name, size)

[FFmpeg-devel] [PATCH v2 3/3] avcodec/x86/diracdsp_init: remove unused macro

2024-11-13 Thread Kyosuke Kawakami
PIXFUNC macro is unused since d29a9c2aa68fc3eb6d61ff95c698e29316037583. Signed-off-by: Kyosuke Kawakami --- libavcodec/x86/diracdsp_init.c | 5 - 1 file changed, 5 deletions(-) diff --git a/libavcodec/x86/diracdsp_init.c b/libavcodec/x86/diracdsp_init.c index 08247133e1..ef01ebdf2e 100644

[FFmpeg-devel] [PATCH v2 2/3] avcodec/x86/diracdsp: migrate last remaining MMX function to SSE2

2024-11-13 Thread Kyosuke Kawakami
: Kyosuke Kawakami --- libavcodec/x86/diracdsp.asm| 23 +++ libavcodec/x86/diracdsp_init.c | 10 +++--- 2 files changed, 22 insertions(+), 11 deletions(-) diff --git a/libavcodec/x86/diracdsp.asm b/libavcodec/x86/diracdsp.asm index e5e2b11846..e708400b66 100644 --- a

[FFmpeg-devel] [PATCH v2 1/3] checkasm/diracdsp: test add_dirac_obmc

2024-11-13 Thread Kyosuke Kawakami
Signed-off-by: Kyosuke Kawakami --- tests/checkasm/Makefile | 1 + tests/checkasm/checkasm.c | 3 ++ tests/checkasm/checkasm.h | 1 + tests/checkasm/diracdsp.c | 81 +++ tests/fate/checkasm.mak | 1 + 5 files changed, 87 insertions(+) create mode

[FFmpeg-devel] [PATCH v2 0/3] avcodec/x86/diracdsp: migrate last remaining MMX function to SSE2

2024-11-13 Thread Kyosuke Kawakami
This series of patch migrates the last remaining MMX function in diracdsp to SSE2. Changes from v1 are basically two things: - fix fate test failure - add a checkasm test case (thanks Ronald for suggestion!) ___ ffmpeg-devel mailing list ffmpeg-devel@ff

Re: [FFmpeg-devel] [PATCH 2/2] avcodec/x86/diracdsp: Migrate last remaining MMX function to SSE2

2024-11-10 Thread Kyosuke Kawakami
On Sat, Nov 9, 2024 at 9:07 PM Michael Niedermayer wrote: > > this seems to break fate: Ouch, I didn't know fate needs samples to be downloaded separately. I'll work on patch v2 tomorrow. Thanks for the review. Kyosuke ___ ffmpeg-devel mailing list ffm

[FFmpeg-devel] [PATCH 1/2] avcodec/x86/diracdsp_init: Remove unused macro

2024-11-07 Thread Kyosuke Kawakami
PIXFUNC macro is unused since d29a9c2aa68fc3eb6d61ff95c698e29316037583. Signed-off-by: Kyosuke Kawakami --- libavcodec/x86/diracdsp_init.c | 5 - 1 file changed, 5 deletions(-) diff --git a/libavcodec/x86/diracdsp_init.c b/libavcodec/x86/diracdsp_init.c index f678759dc0..6a31d3921f 100644

[FFmpeg-devel] [PATCH 2/2] avcodec/x86/diracdsp: Migrate last remaining MMX function to SSE2

2024-11-07 Thread Kyosuke Kawakami
The add_dirac_obmc8_mmx function was the only MMX function left. This patch migrates it to SSE2. Signed-off-by: Kyosuke Kawakami --- libavcodec/x86/diracdsp.asm| 4 +--- libavcodec/x86/diracdsp_init.c | 10 +++--- 2 files changed, 4 insertions(+), 10 deletions(-) diff --git a