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.

The attached patch 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.)

-- 
John Reiser
Index: libmpeg2/motion_comp_arm_s.S
===================================================================
--- libmpeg2/motion_comp_arm_s.S	(revision 1205)
+++ libmpeg2/motion_comp_arm_s.S	(working copy)
@@ -29,9 +29,13 @@
 	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}
@@ -75,11 +79,6 @@
 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
@@ -89,9 +88,14 @@
 	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
@@ -133,12 +137,6 @@
 	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
@@ -157,12 +155,17 @@
 	@@ 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)
@@ -235,12 +238,6 @@
 	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
@@ -249,12 +246,17 @@
 	@@ 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)
@@ -315,9 +317,3 @@
 	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
------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Libmpeg2-devel mailing list
Libmpeg2-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/libmpeg2-devel

Reply via email to