Skip to content
GitLab
Explore
Sign in
Primary navigation
Search or go to…
Project
F
FFmpeg
Manage
Activity
Members
Labels
Plan
Issues
Issue boards
Milestones
Wiki
Code
Merge requests
Repository
Branches
Commits
Tags
Repository graph
Compare revisions
Snippets
Build
Pipelines
Jobs
Pipeline schedules
Artifacts
Deploy
Releases
Container registry
Model registry
Operate
Environments
Monitor
Incidents
Service Desk
Analyze
Value stream analytics
Contributor analytics
CI/CD analytics
Repository analytics
Model experiments
Help
Help
Support
GitLab documentation
Compare GitLab plans
Community forum
Contribute to GitLab
Provide feedback
Keyboard shortcuts
?
Snippets
Groups
Projects
Show more breadcrumbs
libremedia
Tethys
FFmpeg
Commits
97ce1ba8
Commit
97ce1ba8
authored
13 years ago
by
Justin Ruggles
Browse files
Options
Downloads
Patches
Plain Diff
lavr: Add x86-optimized functions for s32 to flt conversion
parent
5904f25b
No related branches found
No related tags found
No related merge requests found
Changes
2
Hide whitespace changes
Inline
Side-by-side
Showing
2 changed files
libavresample/x86/audio_convert.asm
+37
-0
37 additions, 0 deletions
libavresample/x86/audio_convert.asm
libavresample/x86/audio_convert_init.c
+9
-0
9 additions, 0 deletions
libavresample/x86/audio_convert_init.c
with
46 additions
and
0 deletions
libavresample/x86/audio_convert.asm
+
37
−
0
View file @
97ce1ba8
...
@@ -26,6 +26,7 @@
...
@@ -26,6 +26,7 @@
SECTION
_RODATA
32
SECTION
_RODATA
32
pf_s32_inv_scale:
times
8
dd
0x30000000
pf_s16_inv_scale:
times
4
dd
0x38000000
pf_s16_inv_scale:
times
4
dd
0x38000000
SECTION
_TEXT
SECTION
_TEXT
...
@@ -121,6 +122,42 @@ CONV_S32_TO_S16
...
@@ -121,6 +122,42 @@ CONV_S32_TO_S16
INIT_XMM
ss
e2
INIT_XMM
ss
e2
CONV_S32_TO_S16
CONV_S32_TO_S16
;------------------------------------------------------------------------------
; void ff_conv_s32_to_flt(float *dst, const int32_t *src, int len);
;------------------------------------------------------------------------------
%macro CONV_S32_TO_FLT 0
cglobal
conv_s32_to_flt
,
3
,
3
,
3
,
ds
t
,
src
,
len
lea
lenq
,
[
4
*
lend
]
add
srcq
,
lenq
add
ds
tq
,
lenq
neg
lenq
mova
m0
,
[
pf_s32_inv_scale
]
ALIGN
16
.loop:
cvtdq2ps
m1
,
[
srcq
+
lenq
]
cvtdq2ps
m2
,
[
srcq
+
lenq
+
mmsize
]
mulps
m1
,
m1
,
m0
mulps
m2
,
m2
,
m0
mova
[
ds
tq
+
lenq
],
m1
mova
[
ds
tq
+
lenq
+
mmsize
],
m2
add
lenq
,
mmsize
*
2
jl
.loop
%if mmsize == 32
vzeroupper
RET
%else
REP_RET
%endif
%endmacro
INIT_XMM
ss
e2
CONV_S32_TO_FLT
%if HAVE_AVX
INIT_YMM
avx
CONV_S32_TO_FLT
%endif
;-----------------------------------------------------------------------------
;-----------------------------------------------------------------------------
; void ff_conv_fltp_to_flt_6ch(float *dst, float *const *src, int len,
; void ff_conv_fltp_to_flt_6ch(float *dst, float *const *src, int len,
; int channels);
; int channels);
...
...
This diff is collapsed.
Click to expand it.
libavresample/x86/audio_convert_init.c
+
9
−
0
View file @
97ce1ba8
...
@@ -30,6 +30,9 @@ extern void ff_conv_s16_to_flt_sse4(float *dst, const int16_t *src, int len);
...
@@ -30,6 +30,9 @@ extern void ff_conv_s16_to_flt_sse4(float *dst, const int16_t *src, int len);
extern
void
ff_conv_s32_to_s16_mmx
(
int16_t
*
dst
,
const
int32_t
*
src
,
int
len
);
extern
void
ff_conv_s32_to_s16_mmx
(
int16_t
*
dst
,
const
int32_t
*
src
,
int
len
);
extern
void
ff_conv_s32_to_s16_sse2
(
int16_t
*
dst
,
const
int32_t
*
src
,
int
len
);
extern
void
ff_conv_s32_to_s16_sse2
(
int16_t
*
dst
,
const
int32_t
*
src
,
int
len
);
extern
void
ff_conv_s32_to_flt_sse2
(
float
*
dst
,
const
int32_t
*
src
,
int
len
);
extern
void
ff_conv_s32_to_flt_avx
(
float
*
dst
,
const
int32_t
*
src
,
int
len
);
extern
void
ff_conv_fltp_to_flt_6ch_mmx
(
float
*
dst
,
float
*
const
*
src
,
int
len
);
extern
void
ff_conv_fltp_to_flt_6ch_mmx
(
float
*
dst
,
float
*
const
*
src
,
int
len
);
extern
void
ff_conv_fltp_to_flt_6ch_sse4
(
float
*
dst
,
float
*
const
*
src
,
int
len
);
extern
void
ff_conv_fltp_to_flt_6ch_sse4
(
float
*
dst
,
float
*
const
*
src
,
int
len
);
extern
void
ff_conv_fltp_to_flt_6ch_avx
(
float
*
dst
,
float
*
const
*
src
,
int
len
);
extern
void
ff_conv_fltp_to_flt_6ch_avx
(
float
*
dst
,
float
*
const
*
src
,
int
len
);
...
@@ -62,10 +65,16 @@ av_cold void ff_audio_convert_init_x86(AudioConvert *ac)
...
@@ -62,10 +65,16 @@ av_cold void ff_audio_convert_init_x86(AudioConvert *ac)
0
,
16
,
8
,
"SSE2"
,
ff_conv_s16_to_s32_sse2
);
0
,
16
,
8
,
"SSE2"
,
ff_conv_s16_to_s32_sse2
);
ff_audio_convert_set_func
(
ac
,
AV_SAMPLE_FMT_FLT
,
AV_SAMPLE_FMT_S16
,
ff_audio_convert_set_func
(
ac
,
AV_SAMPLE_FMT_FLT
,
AV_SAMPLE_FMT_S16
,
0
,
16
,
8
,
"SSE2"
,
ff_conv_s16_to_flt_sse2
);
0
,
16
,
8
,
"SSE2"
,
ff_conv_s16_to_flt_sse2
);
ff_audio_convert_set_func
(
ac
,
AV_SAMPLE_FMT_FLT
,
AV_SAMPLE_FMT_S32
,
0
,
16
,
8
,
"SSE2"
,
ff_conv_s32_to_flt_sse2
);
}
}
if
(
mm_flags
&
AV_CPU_FLAG_SSE4
&&
HAVE_SSE
)
{
if
(
mm_flags
&
AV_CPU_FLAG_SSE4
&&
HAVE_SSE
)
{
ff_audio_convert_set_func
(
ac
,
AV_SAMPLE_FMT_FLT
,
AV_SAMPLE_FMT_S16
,
ff_audio_convert_set_func
(
ac
,
AV_SAMPLE_FMT_FLT
,
AV_SAMPLE_FMT_S16
,
0
,
16
,
8
,
"SSE4"
,
ff_conv_s16_to_flt_sse4
);
0
,
16
,
8
,
"SSE4"
,
ff_conv_s16_to_flt_sse4
);
}
}
if
(
mm_flags
&
AV_CPU_FLAG_AVX
&&
HAVE_AVX
)
{
ff_audio_convert_set_func
(
ac
,
AV_SAMPLE_FMT_FLT
,
AV_SAMPLE_FMT_S32
,
0
,
32
,
16
,
"AVX"
,
ff_conv_s32_to_flt_avx
);
}
#endif
#endif
}
}
This diff is collapsed.
Click to expand it.
Preview
0%
Loading
Try again
or
attach a new file
.
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Save comment
Cancel
Please
register
or
sign in
to comment