From 0cafb99b57f43cf6ac2c6208718e49ad2dbe462d Mon Sep 17 00:00:00 2001
From: Jan Schmidt <thaytan@noraisin.net>
Date: Sun, 10 Nov 2013 00:49:52 +1100
Subject: [PATCH] Import revision 1206 from upstream to fix PIE build.

Functions MC_put_o_16_arm, MC_put_o_8_arm, MC_put_x_16_arm, MC_put_x_8_arm
in libmpeg2/motion_comp_arm_s.S have addresses in .text, which is bad
for shared libraries.  Some environments demand that .text actually be
read-only all the time, yet MC_put_o_16_arm etc require that the addresses
be modified by the dynamic linking mechanism (dlopen, LoadLibrary, etc.)
Even in those environments which permit the dynamic linker to modify the
.text segment, the runtime cost of doing the relocation can be noticeable.

This commit rewrites the linkage, discarding the tables of addresses
in favor of tables of offsets.  All transfers are local within each individual
function, so there can be no interference by processing that occurs
after assembly, such as link-time re-ordering (even of individual functions.)

Patch by John Reiser <jreiser@bitwagon.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Upstream-Status: Backport [https://code.videolan.org/videolan/libmpeg2/commit/946bf4b518aacc224f845e73708f99e394744499]
---
 libmpeg2/motion_comp_arm_s.S | 70 +++++++++++++++++-------------------
 1 file changed, 33 insertions(+), 37 deletions(-)

diff --git a/libmpeg2/motion_comp_arm_s.S b/libmpeg2/motion_comp_arm_s.S
index c921f7c..82143f8 100644
--- a/libmpeg2/motion_comp_arm_s.S
+++ b/libmpeg2/motion_comp_arm_s.S
@@ -30,9 +30,13 @@ MC_put_o_16_arm:
 	pld [r1]
         stmfd sp!, {r4-r11, lr} @ R14 is also called LR
 	and r4, r1, #3
-	adr r5, MC_put_o_16_arm_align_jt
-	add r5, r5, r4, lsl #2
-	ldr pc, [r5]
+	ldrb r4, [pc, r4]
+	add pc, pc, r4, lsl #2
+	.byte (MC_put_o_16_arm_align0 - 0f)>>2
+	.byte (MC_put_o_16_arm_align1 - 0f)>>2
+	.byte (MC_put_o_16_arm_align2 - 0f)>>2
+	.byte (MC_put_o_16_arm_align3 - 0f)>>2
+0:
 
 MC_put_o_16_arm_align0:
 	ldmia r1, {r4-r7}
@@ -76,11 +80,6 @@ MC_put_o_16_arm_align3:
 1:	PROC(24)
 	bne 1b
         ldmfd sp!, {r4-r11, pc} @@ update PC with LR content.
-MC_put_o_16_arm_align_jt:
-	.word MC_put_o_16_arm_align0
-	.word MC_put_o_16_arm_align1
-	.word MC_put_o_16_arm_align2
-	.word MC_put_o_16_arm_align3
 
 @ ----------------------------------------------------------------
 	.align
@@ -91,9 +90,14 @@ MC_put_o_8_arm:
 	pld [r1]
         stmfd sp!, {r4-r10, lr} @ R14 is also called LR
 	and r4, r1, #3
-	adr r5, MC_put_o_8_arm_align_jt
-	add r5, r5, r4, lsl #2
-	ldr pc, [r5]
+	ldrb r4, [pc, r4]
+	add pc, pc, r4, lsl #2
+	.byte (MC_put_o_8_arm_align0 - 0f)>>2
+	.byte (MC_put_o_8_arm_align1 - 0f)>>2
+	.byte (MC_put_o_8_arm_align2 - 0f)>>2
+	.byte (MC_put_o_8_arm_align3 - 0f)>>2
+0:
+
 MC_put_o_8_arm_align0:
 	ldmia r1, {r4-r5}
 	add r1, r1, r2
@@ -135,12 +139,6 @@ MC_put_o_8_arm_align3:
 	bne 1b
         ldmfd sp!, {r4-r10, pc} @@ update PC with LR content.
 
-MC_put_o_8_arm_align_jt:
-	.word MC_put_o_8_arm_align0
-	.word MC_put_o_8_arm_align1
-	.word MC_put_o_8_arm_align2
-	.word MC_put_o_8_arm_align3
-
 @ ----------------------------------------------------------------
 .macro	AVG_PW rW1, rW2
 	mov \rW2, \rW2, lsl #24
@@ -160,12 +158,17 @@ MC_put_x_16_arm:
 	@@ void func(uint8_t * dest, const uint8_t * ref, int stride, int height)
 	pld [r1]
         stmfd sp!, {r4-r11,lr} @ R14 is also called LR
+	ldr r11, 0f
 	and r4, r1, #3
-	adr r5, MC_put_x_16_arm_align_jt
-	ldr r11, [r5]
 	mvn r12, r11
-	add r5, r5, r4, lsl #2
-	ldr pc, [r5, #4]
+	ldrb r4, [pc, r4]
+	add pc, pc, r4, lsl #2
+	.byte (MC_put_x_16_arm_align0 - 0f)>>2
+	.byte (MC_put_x_16_arm_align1 - 0f)>>2
+	.byte (MC_put_x_16_arm_align2 - 0f)>>2
+	.byte (MC_put_x_16_arm_align3 - 0f)>>2
+0:
+	  .word 0x01010101
 
 .macro	ADJ_ALIGN_QW shift, R0, R1, R2, R3, R4
 	mov \R0, \R0, lsr #(\shift)
@@ -238,12 +241,6 @@ MC_put_x_16_arm_align3:
 	add r0, r0, r2
 	bne 1b
         ldmfd sp!, {r4-r11,pc} @@ update PC with LR content.
-MC_put_x_16_arm_align_jt:
-	.word 0x01010101
-	.word MC_put_x_16_arm_align0
-	.word MC_put_x_16_arm_align1
-	.word MC_put_x_16_arm_align2
-	.word MC_put_x_16_arm_align3
 
 @ ----------------------------------------------------------------
 	.align
@@ -253,12 +250,17 @@ MC_put_x_8_arm:
 	@@ void func(uint8_t * dest, const uint8_t * ref, int stride, int height)
 	pld [r1]
         stmfd sp!, {r4-r11,lr} @ R14 is also called LR
+	ldr r11, 0f
 	and r4, r1, #3
-	adr r5, MC_put_x_8_arm_align_jt
-	ldr r11, [r5]
 	mvn r12, r11
-	add r5, r5, r4, lsl #2
-	ldr pc, [r5, #4]
+	ldrb r4, [pc, r4]
+	add pc, pc, r4, lsl #2
+	.byte (MC_put_x_8_arm_align0 - 0f)>>2
+	.byte (MC_put_x_8_arm_align1 - 0f)>>2
+	.byte (MC_put_x_8_arm_align2 - 0f)>>2
+	.byte (MC_put_x_8_arm_align3 - 0f)>>2
+0:
+	  .word 0x01010101
 
 .macro	ADJ_ALIGN_DW shift, R0, R1, R2
 	mov \R0, \R0, lsr #(\shift)
@@ -319,9 +321,3 @@ MC_put_x_8_arm_align3:
 	add r0, r0, r2
 	bne 1b
         ldmfd sp!, {r4-r11,pc} @@ update PC with LR content.
-MC_put_x_8_arm_align_jt:
-	.word 0x01010101
-	.word MC_put_x_8_arm_align0
-	.word MC_put_x_8_arm_align1
-	.word MC_put_x_8_arm_align2
-	.word MC_put_x_8_arm_align3