RepoMirrors
/
ffmpeg
mirror of https://git.ffmpeg.org/ffmpeg.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

avcodec/x86: modify simple_idct10 macros to add an action paramter

This commit is contained in:
James Darnley 2017-06-12 12:01:21 +02:00
parent 8781330d80
commit d2597fb0c1
3 changed files with 25 additions and 22 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
libavcodec/x86/proresdsp.asm
View File

@ -52,7 +52,7 @@ SECTION .text
%macro idct_fn 0 %macro idct_fn 0
cglobal prores_idct_put_10, 4, 4, 15, pixels, lsize, block, qmat cglobal prores_idct_put_10, 4, 4, 15, pixels, lsize, block, qmat
IDCT_FN pw_1, 15, pw_88, 18, pw_4, pw_1019, r3 IDCT_FN pw_1, 15, pw_88, 18, "put", pw_4, pw_1019, r3
RET RET
%endmacro %endmacro

8
libavcodec/x86/simple_idct10.asm
View File

@ -69,24 +69,24 @@ SECTION .text
%macro idct_fn 0 %macro idct_fn 0
cglobal simple_idct10, 1, 1, 16, block cglobal simple_idct10, 1, 1, 16, block
IDCT_FN "", 12, "", 19 IDCT_FN "", 12, "", 19, "store"
RET RET
cglobal simple_idct10_put, 3, 3, 16, pixels, lsize, block cglobal simple_idct10_put, 3, 3, 16, pixels, lsize, block
IDCT_FN "", 12, "", 19, 0, pw_1023 IDCT_FN "", 12, "", 19, "put", 0, pw_1023
RET RET
cglobal simple_idct12, 1, 1, 16, block cglobal simple_idct12, 1, 1, 16, block
; coeffs are already 15bits, adding the offset would cause ; coeffs are already 15bits, adding the offset would cause
; overflow in the input ; overflow in the input
IDCT_FN "", 15, pw_2, 16 IDCT_FN "", 15, pw_2, 16, "store"
RET RET
cglobal simple_idct12_put, 3, 3, 16, pixels, lsize, block cglobal simple_idct12_put, 3, 3, 16, pixels, lsize, block
; range isn't known, so the C simple_idct range is used ; range isn't known, so the C simple_idct range is used
; Also, using a bias on input overflows, so use the bias ; Also, using a bias on input overflows, so use the bias
; on output of the first butterfly instead ; on output of the first butterfly instead
IDCT_FN "", 15, pw_2, 16, 0, pw_4095 IDCT_FN "", 15, pw_2, 16, "put", 0, pw_4095
RET RET
%endmacro %endmacro

37
libavcodec/x86/simple_idct10_template.asm
View File

@ -218,11 +218,12 @@
; %2 = row bias macro ; %2 = row bias macro
; %3 = column shift ; %3 = column shift
; %4 = column bias macro ; %4 = column bias macro
; %5 = min pixel value ; %5 = final action (nothing, "store", "put", "add")
; %6 = max pixel value ; %6 = min pixel value
; %7 = qmat (for prores) ; %7 = max pixel value
; %8 = qmat (for prores)
%macro IDCT_FN 4-7 %macro IDCT_FN 4-8
; for (i = 0; i < 8; i++) ; for (i = 0; i < 8; i++)
; idctRowCondDC(block + i*8); ; idctRowCondDC(block + i*8);
mova m10,[blockq+ 0] ; { row[0] }[0-7] mova m10,[blockq+ 0] ; { row[0] }[0-7]
@ -230,13 +231,13 @@
mova m13,[blockq+64] ; { row[4] }[0-7] mova m13,[blockq+64] ; { row[4] }[0-7]
mova m12,[blockq+96] ; { row[6] }[0-7] mova m12,[blockq+96] ; { row[6] }[0-7]
%if %0 == 7 %if %0 == 8
pmullw m10,[%7+ 0] pmullw m10,[%8+ 0]
pmullw m8, [%7+32] pmullw m8, [%8+32]
pmullw m13,[%7+64] pmullw m13,[%8+64]
pmullw m12,[%7+96] pmullw m12,[%8+96]
IDCT_1D %1, %2, %7 IDCT_1D %1, %2, %8
%else %else
IDCT_1D %1, %2 IDCT_1D %1, %2
%endif %endif
@ -257,7 +258,8 @@
IDCT_1D %3, %4 IDCT_1D %3, %4
; clip/store ; clip/store
%if %0 == 4 %if %0 >= 5
%ifidn %5,"store"
; No clamping, means pure idct ; No clamping, means pure idct
mova [blockq+ 0], m8 mova [blockq+ 0], m8
mova [blockq+ 16], m0 mova [blockq+ 16], m0
@ -267,13 +269,13 @@
mova [blockq+ 80], m11 mova [blockq+ 80], m11
mova [blockq+ 96], m9 mova [blockq+ 96], m9
mova [blockq+112], m10 mova [blockq+112], m10
%else %elifidn %5,"put"
%ifidn %5, 0 %ifidn %6, 0
pxor m3, m3 pxor m3, m3
%else %else
mova m3, [%5] mova m3, [%6]
%endif %endif ; ifidn %6, 0
mova m5, [%6] mova m5, [%7]
pmaxsw m8, m3 pmaxsw m8, m3
pmaxsw m0, m3 pmaxsw m0, m3
pmaxsw m1, m3 pmaxsw m1, m3
@ -301,7 +303,8 @@
mova [r0+r1 ], m11 mova [r0+r1 ], m11
mova [r0+r1*2], m9 mova [r0+r1*2], m9
mova [r0+r2 ], m10 mova [r0+r2 ], m10
%endif %endif ; %5 action
%endif; if %0 >= 5
%endmacro %endmacro
%endif %endif