har0ke/FFmpeg
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

libavcodec: Remove dynamic relocs from aarch64/h264idct_neon.S

Browse Source 
Some of the assembly functions e.g. ff_h264_idct_dc_add_neon
has code like:
        movrel          x14, X(ff_h264_idct_add_neon)

Linker cannot resolve them fully at link time and emits dynamic
relocations.
Use explicit labels instead so that no dynamic relocations are
needed at all.

This avoids lld complains about text relocations.

For background, see https://crbug.com/917919

Signed-off-by: Manoj Gupta <manojgupta@chromium.org>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
This commit is contained in:
Manoj Gupta 2018-12-28 15:12:53 -08:00 committed by Michael Niedermayer
parent bf96937a96
commit 6fcf813110
1 changed files with 12 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
libavcodec/aarch64/h264idct_neon.S
View File

@ -23,6 +23,7 @@
#include "neon.S" #include "neon.S"
function ff_h264_idct_add_neon, export=1 function ff_h264_idct_add_neon, export=1
.L_ff_h264_idct_add_neon:
ld1 {v0.4H, v1.4H, v2.4H, v3.4H}, [x1] ld1 {v0.4H, v1.4H, v2.4H, v3.4H}, [x1]
sxtw x2, w2 sxtw x2, w2
movi v30.8H, #0 movi v30.8H, #0
@ -77,6 +78,7 @@ function ff_h264_idct_add_neon, export=1
endfunc endfunc
function ff_h264_idct_dc_add_neon, export=1 function ff_h264_idct_dc_add_neon, export=1
.L_ff_h264_idct_dc_add_neon:
sxtw x2, w2 sxtw x2, w2
mov w3, #0 mov w3, #0
ld1r {v2.8H}, [x1] ld1r {v2.8H}, [x1]
@ -106,8 +108,8 @@ function ff_h264_idct_add16_neon, export=1
mov w9, w3 // stride mov w9, w3 // stride
movrel x7, scan8 movrel x7, scan8
mov x10, #16 mov x10, #16
movrel x13, X(ff_h264_idct_dc_add_neon) movrel x13, .L_ff_h264_idct_dc_add_neon
movrel x14, X(ff_h264_idct_add_neon) movrel x14, .L_ff_h264_idct_add_neon
1: mov w2, w9 1: mov w2, w9
ldrb w3, [x7], #1 ldrb w3, [x7], #1
ldrsw x0, [x5], #4 ldrsw x0, [x5], #4
@ -133,8 +135,8 @@ function ff_h264_idct_add16intra_neon, export=1
mov w9, w3 // stride mov w9, w3 // stride
movrel x7, scan8 movrel x7, scan8
mov x10, #16 mov x10, #16
movrel x13, X(ff_h264_idct_dc_add_neon) movrel x13, .L_ff_h264_idct_dc_add_neon
movrel x14, X(ff_h264_idct_add_neon) movrel x14, .L_ff_h264_idct_add_neon
1: mov w2, w9 1: mov w2, w9
ldrb w3, [x7], #1 ldrb w3, [x7], #1
ldrsw x0, [x5], #4 ldrsw x0, [x5], #4
@ -160,8 +162,8 @@ function ff_h264_idct_add8_neon, export=1
add x5, x1, #16*4 // block_offset add x5, x1, #16*4 // block_offset
add x9, x2, #16*32 // block add x9, x2, #16*32 // block
mov w19, w3 // stride mov w19, w3 // stride
movrel x13, X(ff_h264_idct_dc_add_neon) movrel x13, .L_ff_h264_idct_dc_add_neon
movrel x14, X(ff_h264_idct_add_neon) movrel x14, .L_ff_h264_idct_add_neon
movrel x7, scan8, 16 movrel x7, scan8, 16
mov x10, #0 mov x10, #0
mov x11, #16 mov x11, #16
@ -263,6 +265,7 @@ endfunc
.endm .endm
function ff_h264_idct8_add_neon, export=1 function ff_h264_idct8_add_neon, export=1
.L_ff_h264_idct8_add_neon:
movi v19.8H, #0 movi v19.8H, #0
sxtw x2, w2 sxtw x2, w2
ld1 {v24.8H, v25.8H}, [x1] ld1 {v24.8H, v25.8H}, [x1]
@ -326,6 +329,7 @@ function ff_h264_idct8_add_neon, export=1
endfunc endfunc
function ff_h264_idct8_dc_add_neon, export=1 function ff_h264_idct8_dc_add_neon, export=1
.L_ff_h264_idct8_dc_add_neon:
mov w3, #0 mov w3, #0
sxtw x2, w2 sxtw x2, w2
ld1r {v31.8H}, [x1] ld1r {v31.8H}, [x1]
@ -375,8 +379,8 @@ function ff_h264_idct8_add4_neon, export=1
mov w2, w3 mov w2, w3
movrel x7, scan8 movrel x7, scan8
mov w10, #16 mov w10, #16
movrel x13, X(ff_h264_idct8_dc_add_neon) movrel x13, .L_ff_h264_idct8_dc_add_neon
movrel x14, X(ff_h264_idct8_add_neon) movrel x14, .L_ff_h264_idct8_add_neon
1: ldrb w9, [x7], #4 1: ldrb w9, [x7], #4
ldrsw x0, [x5], #16 ldrsw x0, [x5], #16
ldrb w9, [x4, w9, UXTW] ldrb w9, [x4, w9, UXTW]