From fc252eba02a8c95da75b48d26c8bb6a6de3e866e Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?M=C3=A5ns=20Rullg=C3=A5rd?= Date: Tue, 27 Jan 2009 16:06:41 +0000 Subject: [PATCH] ARM: change alignment of loops in put_pixels*_arm to 32 Originally committed as revision 16820 to svn://svn.ffmpeg.org/ffmpeg/trunk --- libavcodec/arm/dsputil_arm_s.S | 71 ++++++++++++++++------------------ 1 file changed, 33 insertions(+), 38 deletions(-) diff --git a/libavcodec/arm/dsputil_arm_s.S b/libavcodec/arm/dsputil_arm_s.S index 9425d33286..112830c9c4 100644 --- a/libavcodec/arm/dsputil_arm_s.S +++ b/libavcodec/arm/dsputil_arm_s.S @@ -91,7 +91,7 @@ function ff_prefetch_arm, export=1 .endm @ ---------------------------------------------------------------- - .align 8 + .align 5 function put_pixels16_arm, export=1 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h) @ block = word aligned, pixles = unaligned @@ -111,7 +111,7 @@ function put_pixels16_arm, export=1 add r0, r0, r2 bne 1b ldmfd sp!, {r4-r11, pc} - .align 8 + .align 5 2: ldmia r1, {r4-r8} add r1, r1, r2 @@ -122,7 +122,7 @@ function put_pixels16_arm, export=1 add r0, r0, r2 bne 2b ldmfd sp!, {r4-r11, pc} - .align 8 + .align 5 3: ldmia r1, {r4-r8} add r1, r1, r2 @@ -133,7 +133,7 @@ function put_pixels16_arm, export=1 add r0, r0, r2 bne 3b ldmfd sp!, {r4-r11, pc} - .align 8 + .align 5 4: ldmia r1, {r4-r8} add r1, r1, r2 @@ -144,7 +144,6 @@ function put_pixels16_arm, export=1 add r0, r0, r2 bne 4b ldmfd sp!, {r4-r11,pc} - .align 8 5: .word 1b .word 2b @@ -153,7 +152,7 @@ function put_pixels16_arm, export=1 .endfunc @ ---------------------------------------------------------------- - .align 8 + .align 5 function put_pixels8_arm, export=1 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h) @ block = word aligned, pixles = unaligned @@ -173,7 +172,7 @@ function put_pixels8_arm, export=1 add r0, r0, r2 bne 1b ldmfd sp!, {r4-r5,pc} - .align 8 + .align 5 2: ldmia r1, {r4-r5, r12} add r1, r1, r2 @@ -184,7 +183,7 @@ function put_pixels8_arm, export=1 add r0, r0, r2 bne 2b ldmfd sp!, {r4-r5,pc} - .align 8 + .align 5 3: ldmia r1, {r4-r5, r12} add r1, r1, r2 @@ -195,7 +194,7 @@ function put_pixels8_arm, export=1 add r0, r0, r2 bne 3b ldmfd sp!, {r4-r5,pc} - .align 8 + .align 5 4: ldmia r1, {r4-r5, r12} add r1, r1, r2 @@ -206,7 +205,6 @@ function put_pixels8_arm, export=1 add r0, r0, r2 bne 4b ldmfd sp!, {r4-r5,pc} - .align 8 5: .word 1b .word 2b @@ -215,7 +213,7 @@ function put_pixels8_arm, export=1 .endfunc @ ---------------------------------------------------------------- - .align 8 + .align 5 function put_pixels8_x2_arm, export=1 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h) @ block = word aligned, pixles = unaligned @@ -238,7 +236,7 @@ function put_pixels8_x2_arm, export=1 add r0, r0, r2 bne 1b ldmfd sp!, {r4-r10,pc} - .align 8 + .align 5 2: ldmia r1, {r4-r5, r10} add r1, r1, r2 @@ -251,7 +249,7 @@ function put_pixels8_x2_arm, export=1 add r0, r0, r2 bne 2b ldmfd sp!, {r4-r10,pc} - .align 8 + .align 5 3: ldmia r1, {r4-r5, r10} add r1, r1, r2 @@ -264,7 +262,7 @@ function put_pixels8_x2_arm, export=1 add r0, r0, r2 bne 3b ldmfd sp!, {r4-r10,pc} - .align 8 + .align 5 4: ldmia r1, {r4-r5, r10} add r1, r1, r2 @@ -276,7 +274,6 @@ function put_pixels8_x2_arm, export=1 add r0, r0, r2 bne 4b ldmfd sp!, {r4-r10,pc} @@ update PC with LR content. - .align 8 5: .word 0xFEFEFEFE .word 2b @@ -284,7 +281,7 @@ function put_pixels8_x2_arm, export=1 .word 4b .endfunc - .align 8 + .align 5 function put_no_rnd_pixels8_x2_arm, export=1 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h) @ block = word aligned, pixles = unaligned @@ -307,7 +304,7 @@ function put_no_rnd_pixels8_x2_arm, export=1 add r0, r0, r2 bne 1b ldmfd sp!, {r4-r10,pc} - .align 8 + .align 5 2: ldmia r1, {r4-r5, r10} add r1, r1, r2 @@ -320,7 +317,7 @@ function put_no_rnd_pixels8_x2_arm, export=1 add r0, r0, r2 bne 2b ldmfd sp!, {r4-r10,pc} - .align 8 + .align 5 3: ldmia r1, {r4-r5, r10} add r1, r1, r2 @@ -333,7 +330,7 @@ function put_no_rnd_pixels8_x2_arm, export=1 add r0, r0, r2 bne 3b ldmfd sp!, {r4-r10,pc} - .align 8 + .align 5 4: ldmia r1, {r4-r5, r10} add r1, r1, r2 @@ -345,7 +342,6 @@ function put_no_rnd_pixels8_x2_arm, export=1 add r0, r0, r2 bne 4b ldmfd sp!, {r4-r10,pc} @@ update PC with LR content. - .align 8 5: .word 0xFEFEFEFE .word 2b @@ -355,7 +351,7 @@ function put_no_rnd_pixels8_x2_arm, export=1 @ ---------------------------------------------------------------- - .align 8 + .align 5 function put_pixels8_y2_arm, export=1 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h) @ block = word aligned, pixles = unaligned @@ -386,7 +382,7 @@ function put_pixels8_y2_arm, export=1 add r0, r0, r2 bne 6b ldmfd sp!, {r4-r11,pc} - .align 8 + .align 5 2: ldmia r1, {r4-r6} add r1, r1, r2 @@ -409,7 +405,7 @@ function put_pixels8_y2_arm, export=1 add r0, r0, r2 bne 6b ldmfd sp!, {r4-r11,pc} - .align 8 + .align 5 3: ldmia r1, {r4-r6} add r1, r1, r2 @@ -432,7 +428,7 @@ function put_pixels8_y2_arm, export=1 add r0, r0, r2 bne 6b ldmfd sp!, {r4-r11,pc} - .align 8 + .align 5 4: ldmia r1, {r4-r6} add r1, r1, r2 @@ -456,7 +452,6 @@ function put_pixels8_y2_arm, export=1 bne 6b ldmfd sp!, {r4-r11,pc} - .align 8 5: .word 0xFEFEFEFE .word 2b @@ -464,7 +459,7 @@ function put_pixels8_y2_arm, export=1 .word 4b .endfunc - .align 8 + .align 5 function put_no_rnd_pixels8_y2_arm, export=1 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h) @ block = word aligned, pixles = unaligned @@ -495,7 +490,7 @@ function put_no_rnd_pixels8_y2_arm, export=1 add r0, r0, r2 bne 6b ldmfd sp!, {r4-r11,pc} - .align 8 + .align 5 2: ldmia r1, {r4-r6} add r1, r1, r2 @@ -518,7 +513,7 @@ function put_no_rnd_pixels8_y2_arm, export=1 add r0, r0, r2 bne 6b ldmfd sp!, {r4-r11,pc} - .align 8 + .align 5 3: ldmia r1, {r4-r6} add r1, r1, r2 @@ -541,7 +536,7 @@ function put_no_rnd_pixels8_y2_arm, export=1 add r0, r0, r2 bne 6b ldmfd sp!, {r4-r11,pc} - .align 8 + .align 5 4: ldmia r1, {r4-r6} add r1, r1, r2 @@ -564,7 +559,6 @@ function put_no_rnd_pixels8_y2_arm, export=1 add r0, r0, r2 bne 6b ldmfd sp!, {r4-r11,pc} - .align 8 5: .word 0xFEFEFEFE .word 2b @@ -637,7 +631,7 @@ function put_no_rnd_pixels8_y2_arm, export=1 ldmfd sp!, {r4-r11,pc} .endm - .align 8 + .align 5 function put_pixels8_xy2_arm, export=1 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h) @ block = word aligned, pixles = unaligned @@ -651,15 +645,15 @@ function put_pixels8_xy2_arm, export=1 1: RND_XY2_EXPAND 0 - .align 8 + .align 5 2: RND_XY2_EXPAND 1 - .align 8 + .align 5 3: RND_XY2_EXPAND 2 - .align 8 + .align 5 4: RND_XY2_EXPAND 3 @@ -673,7 +667,7 @@ function put_pixels8_xy2_arm, export=1 .word 0x0F0F0F0F .endfunc - .align 8 + .align 5 function put_no_rnd_pixels8_xy2_arm, export=1 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h) @ block = word aligned, pixles = unaligned @@ -687,15 +681,15 @@ function put_no_rnd_pixels8_xy2_arm, export=1 1: RND_XY2_EXPAND 0 - .align 8 + .align 5 2: RND_XY2_EXPAND 1 - .align 8 + .align 5 3: RND_XY2_EXPAND 2 - .align 8 + .align 5 4: RND_XY2_EXPAND 3 @@ -709,6 +703,7 @@ function put_no_rnd_pixels8_xy2_arm, export=1 .word 0x0F0F0F0F .endfunc + .align 5 @ void ff_add_pixels_clamped_ARM(int16_t *block, uint8_t *dest, int stride) function ff_add_pixels_clamped_ARM, export=1 push {r4-r10}