powerpc: Change vrX register defines to vX to match gcc and glibc

As our various loops (copy, string, crypto etc) get more complicated,
we want to share implementations between userspace (eg glibc) and
the kernel. We also want to write userspace test harnesses to put
in tools/testing/selftest.

One gratuitous difference between userspace and the kernel is the
VMX register definitions - the kernel uses vrX whereas both gcc and
glibc use vX.

Change the kernel to match userspace.

Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This commit is contained in:
Anton Blanchard 2015-02-10 09:51:22 +11:00 committed by Michael Ellerman
parent 06e5801b8c
commit c2ce6f9f3d
10 changed files with 352 additions and 385 deletions

View File

@ -637,38 +637,38 @@ END_FTR_SECTION_NESTED(CPU_FTR_HAS_PPR,CPU_FTR_HAS_PPR,945)
/* AltiVec Registers (VPRs) */ /* AltiVec Registers (VPRs) */
#define vr0 0 #define v0 0
#define vr1 1 #define v1 1
#define vr2 2 #define v2 2
#define vr3 3 #define v3 3
#define vr4 4 #define v4 4
#define vr5 5 #define v5 5
#define vr6 6 #define v6 6
#define vr7 7 #define v7 7
#define vr8 8 #define v8 8
#define vr9 9 #define v9 9
#define vr10 10 #define v10 10
#define vr11 11 #define v11 11
#define vr12 12 #define v12 12
#define vr13 13 #define v13 13
#define vr14 14 #define v14 14
#define vr15 15 #define v15 15
#define vr16 16 #define v16 16
#define vr17 17 #define v17 17
#define vr18 18 #define v18 18
#define vr19 19 #define v19 19
#define vr20 20 #define v20 20
#define vr21 21 #define v21 21
#define vr22 22 #define v22 22
#define vr23 23 #define v23 23
#define vr24 24 #define v24 24
#define vr25 25 #define v25 25
#define vr26 26 #define v26 26
#define vr27 27 #define v27 27
#define vr28 28 #define v28 28
#define vr29 29 #define v29 29
#define vr30 30 #define v30 30
#define vr31 31 #define v31 31
/* VSX Registers (VSRs) */ /* VSX Registers (VSRs) */

View File

@ -136,7 +136,7 @@ struct pt_regs {
#endif /* __powerpc64__ */ #endif /* __powerpc64__ */
/* /*
* Get/set all the altivec registers vr0..vr31, vscr, vrsave, in one go. * Get/set all the altivec registers v0..v31, vscr, vrsave, in one go.
* The transfer totals 34 quadword. Quadwords 0-31 contain the * The transfer totals 34 quadword. Quadwords 0-31 contain the
* corresponding vector registers. Quadword 32 contains the vscr as the * corresponding vector registers. Quadword 32 contains the vscr as the
* last word (offset 12) within that quadword. Quadword 33 contains the * last word (offset 12) within that quadword. Quadword 33 contains the

View File

@ -152,9 +152,9 @@ _GLOBAL(tm_reclaim)
addi r7, r3, THREAD_TRANSACT_VRSTATE addi r7, r3, THREAD_TRANSACT_VRSTATE
SAVE_32VRS(0, r6, r7) /* r6 scratch, r7 transact vr state */ SAVE_32VRS(0, r6, r7) /* r6 scratch, r7 transact vr state */
mfvscr vr0 mfvscr v0
li r6, VRSTATE_VSCR li r6, VRSTATE_VSCR
stvx vr0, r7, r6 stvx v0, r7, r6
dont_backup_vec: dont_backup_vec:
mfspr r0, SPRN_VRSAVE mfspr r0, SPRN_VRSAVE
std r0, THREAD_TRANSACT_VRSAVE(r3) std r0, THREAD_TRANSACT_VRSAVE(r3)
@ -359,8 +359,8 @@ _GLOBAL(__tm_recheckpoint)
addi r8, r3, THREAD_VRSTATE addi r8, r3, THREAD_VRSTATE
li r5, VRSTATE_VSCR li r5, VRSTATE_VSCR
lvx vr0, r8, r5 lvx v0, r8, r5
mtvscr vr0 mtvscr v0
REST_32VRS(0, r5, r8) /* r5 scratch, r8 ptr */ REST_32VRS(0, r5, r8) /* r5 scratch, r8 ptr */
dont_restore_vec: dont_restore_vec:
ld r5, THREAD_VRSAVE(r3) ld r5, THREAD_VRSAVE(r3)

View File

@ -24,8 +24,8 @@ _GLOBAL(do_load_up_transact_altivec)
stw r4,THREAD_USED_VR(r3) stw r4,THREAD_USED_VR(r3)
li r10,THREAD_TRANSACT_VRSTATE+VRSTATE_VSCR li r10,THREAD_TRANSACT_VRSTATE+VRSTATE_VSCR
lvx vr0,r10,r3 lvx v0,r10,r3
mtvscr vr0 mtvscr v0
addi r10,r3,THREAD_TRANSACT_VRSTATE addi r10,r3,THREAD_TRANSACT_VRSTATE
REST_32VRS(0,r4,r10) REST_32VRS(0,r4,r10)
@ -52,8 +52,8 @@ _GLOBAL(vec_enable)
*/ */
_GLOBAL(load_vr_state) _GLOBAL(load_vr_state)
li r4,VRSTATE_VSCR li r4,VRSTATE_VSCR
lvx vr0,r4,r3 lvx v0,r4,r3
mtvscr vr0 mtvscr v0
REST_32VRS(0,r4,r3) REST_32VRS(0,r4,r3)
blr blr
@ -63,9 +63,9 @@ _GLOBAL(load_vr_state)
*/ */
_GLOBAL(store_vr_state) _GLOBAL(store_vr_state)
SAVE_32VRS(0, r4, r3) SAVE_32VRS(0, r4, r3)
mfvscr vr0 mfvscr v0
li r4, VRSTATE_VSCR li r4, VRSTATE_VSCR
stvx vr0, r4, r3 stvx v0, r4, r3
blr blr
/* /*
@ -104,9 +104,9 @@ _GLOBAL(load_up_altivec)
addi r4,r4,THREAD addi r4,r4,THREAD
addi r6,r4,THREAD_VRSTATE addi r6,r4,THREAD_VRSTATE
SAVE_32VRS(0,r5,r6) SAVE_32VRS(0,r5,r6)
mfvscr vr0 mfvscr v0
li r10,VRSTATE_VSCR li r10,VRSTATE_VSCR
stvx vr0,r10,r6 stvx v0,r10,r6
/* Disable VMX for last_task_used_altivec */ /* Disable VMX for last_task_used_altivec */
PPC_LL r5,PT_REGS(r4) PPC_LL r5,PT_REGS(r4)
toreal(r5) toreal(r5)
@ -142,8 +142,8 @@ _GLOBAL(load_up_altivec)
li r4,1 li r4,1
li r10,VRSTATE_VSCR li r10,VRSTATE_VSCR
stw r4,THREAD_USED_VR(r5) stw r4,THREAD_USED_VR(r5)
lvx vr0,r10,r6 lvx v0,r10,r6
mtvscr vr0 mtvscr v0
REST_32VRS(0,r4,r6) REST_32VRS(0,r4,r6)
#ifndef CONFIG_SMP #ifndef CONFIG_SMP
/* Update last_task_used_altivec to 'current' */ /* Update last_task_used_altivec to 'current' */
@ -186,9 +186,9 @@ _GLOBAL(giveup_altivec)
addi r7,r3,THREAD_VRSTATE addi r7,r3,THREAD_VRSTATE
2: PPC_LCMPI 0,r5,0 2: PPC_LCMPI 0,r5,0
SAVE_32VRS(0,r4,r7) SAVE_32VRS(0,r4,r7)
mfvscr vr0 mfvscr v0
li r4,VRSTATE_VSCR li r4,VRSTATE_VSCR
stvx vr0,r4,r7 stvx v0,r4,r7
beq 1f beq 1f
PPC_LL r4,_MSR-STACK_FRAME_OVERHEAD(r5) PPC_LL r4,_MSR-STACK_FRAME_OVERHEAD(r5)
#ifdef CONFIG_VSX #ifdef CONFIG_VSX

View File

@ -83,23 +83,23 @@ _GLOBAL(copypage_power7)
li r12,112 li r12,112
.align 5 .align 5
1: lvx vr7,r0,r4 1: lvx v7,r0,r4
lvx vr6,r4,r6 lvx v6,r4,r6
lvx vr5,r4,r7 lvx v5,r4,r7
lvx vr4,r4,r8 lvx v4,r4,r8
lvx vr3,r4,r9 lvx v3,r4,r9
lvx vr2,r4,r10 lvx v2,r4,r10
lvx vr1,r4,r11 lvx v1,r4,r11
lvx vr0,r4,r12 lvx v0,r4,r12
addi r4,r4,128 addi r4,r4,128
stvx vr7,r0,r3 stvx v7,r0,r3
stvx vr6,r3,r6 stvx v6,r3,r6
stvx vr5,r3,r7 stvx v5,r3,r7
stvx vr4,r3,r8 stvx v4,r3,r8
stvx vr3,r3,r9 stvx v3,r3,r9
stvx vr2,r3,r10 stvx v2,r3,r10
stvx vr1,r3,r11 stvx v1,r3,r11
stvx vr0,r3,r12 stvx v0,r3,r12
addi r3,r3,128 addi r3,r3,128
bdnz 1b bdnz 1b

View File

@ -388,29 +388,29 @@ err3; std r0,0(r3)
li r11,48 li r11,48
bf cr7*4+3,5f bf cr7*4+3,5f
err3; lvx vr1,r0,r4 err3; lvx v1,r0,r4
addi r4,r4,16 addi r4,r4,16
err3; stvx vr1,r0,r3 err3; stvx v1,r0,r3
addi r3,r3,16 addi r3,r3,16
5: bf cr7*4+2,6f 5: bf cr7*4+2,6f
err3; lvx vr1,r0,r4 err3; lvx v1,r0,r4
err3; lvx vr0,r4,r9 err3; lvx v0,r4,r9
addi r4,r4,32 addi r4,r4,32
err3; stvx vr1,r0,r3 err3; stvx v1,r0,r3
err3; stvx vr0,r3,r9 err3; stvx v0,r3,r9
addi r3,r3,32 addi r3,r3,32
6: bf cr7*4+1,7f 6: bf cr7*4+1,7f
err3; lvx vr3,r0,r4 err3; lvx v3,r0,r4
err3; lvx vr2,r4,r9 err3; lvx v2,r4,r9
err3; lvx vr1,r4,r10 err3; lvx v1,r4,r10
err3; lvx vr0,r4,r11 err3; lvx v0,r4,r11
addi r4,r4,64 addi r4,r4,64
err3; stvx vr3,r0,r3 err3; stvx v3,r0,r3
err3; stvx vr2,r3,r9 err3; stvx v2,r3,r9
err3; stvx vr1,r3,r10 err3; stvx v1,r3,r10
err3; stvx vr0,r3,r11 err3; stvx v0,r3,r11
addi r3,r3,64 addi r3,r3,64
7: sub r5,r5,r6 7: sub r5,r5,r6
@ -433,23 +433,23 @@ err3; stvx vr0,r3,r11
*/ */
.align 5 .align 5
8: 8:
err4; lvx vr7,r0,r4 err4; lvx v7,r0,r4
err4; lvx vr6,r4,r9 err4; lvx v6,r4,r9
err4; lvx vr5,r4,r10 err4; lvx v5,r4,r10
err4; lvx vr4,r4,r11 err4; lvx v4,r4,r11
err4; lvx vr3,r4,r12 err4; lvx v3,r4,r12
err4; lvx vr2,r4,r14 err4; lvx v2,r4,r14
err4; lvx vr1,r4,r15 err4; lvx v1,r4,r15
err4; lvx vr0,r4,r16 err4; lvx v0,r4,r16
addi r4,r4,128 addi r4,r4,128
err4; stvx vr7,r0,r3 err4; stvx v7,r0,r3
err4; stvx vr6,r3,r9 err4; stvx v6,r3,r9
err4; stvx vr5,r3,r10 err4; stvx v5,r3,r10
err4; stvx vr4,r3,r11 err4; stvx v4,r3,r11
err4; stvx vr3,r3,r12 err4; stvx v3,r3,r12
err4; stvx vr2,r3,r14 err4; stvx v2,r3,r14
err4; stvx vr1,r3,r15 err4; stvx v1,r3,r15
err4; stvx vr0,r3,r16 err4; stvx v0,r3,r16
addi r3,r3,128 addi r3,r3,128
bdnz 8b bdnz 8b
@ -463,29 +463,29 @@ err4; stvx vr0,r3,r16
mtocrf 0x01,r6 mtocrf 0x01,r6
bf cr7*4+1,9f bf cr7*4+1,9f
err3; lvx vr3,r0,r4 err3; lvx v3,r0,r4
err3; lvx vr2,r4,r9 err3; lvx v2,r4,r9
err3; lvx vr1,r4,r10 err3; lvx v1,r4,r10
err3; lvx vr0,r4,r11 err3; lvx v0,r4,r11
addi r4,r4,64 addi r4,r4,64
err3; stvx vr3,r0,r3 err3; stvx v3,r0,r3
err3; stvx vr2,r3,r9 err3; stvx v2,r3,r9
err3; stvx vr1,r3,r10 err3; stvx v1,r3,r10
err3; stvx vr0,r3,r11 err3; stvx v0,r3,r11
addi r3,r3,64 addi r3,r3,64
9: bf cr7*4+2,10f 9: bf cr7*4+2,10f
err3; lvx vr1,r0,r4 err3; lvx v1,r0,r4
err3; lvx vr0,r4,r9 err3; lvx v0,r4,r9
addi r4,r4,32 addi r4,r4,32
err3; stvx vr1,r0,r3 err3; stvx v1,r0,r3
err3; stvx vr0,r3,r9 err3; stvx v0,r3,r9
addi r3,r3,32 addi r3,r3,32
10: bf cr7*4+3,11f 10: bf cr7*4+3,11f
err3; lvx vr1,r0,r4 err3; lvx v1,r0,r4
addi r4,r4,16 addi r4,r4,16
err3; stvx vr1,r0,r3 err3; stvx v1,r0,r3
addi r3,r3,16 addi r3,r3,16
/* Up to 15B to go */ /* Up to 15B to go */
@ -560,42 +560,42 @@ err3; stw r7,4(r3)
li r10,32 li r10,32
li r11,48 li r11,48
LVS(vr16,0,r4) /* Setup permute control vector */ LVS(v16,0,r4) /* Setup permute control vector */
err3; lvx vr0,0,r4 err3; lvx v0,0,r4
addi r4,r4,16 addi r4,r4,16
bf cr7*4+3,5f bf cr7*4+3,5f
err3; lvx vr1,r0,r4 err3; lvx v1,r0,r4
VPERM(vr8,vr0,vr1,vr16) VPERM(v8,v0,v1,v16)
addi r4,r4,16 addi r4,r4,16
err3; stvx vr8,r0,r3 err3; stvx v8,r0,r3
addi r3,r3,16 addi r3,r3,16
vor vr0,vr1,vr1 vor v0,v1,v1
5: bf cr7*4+2,6f 5: bf cr7*4+2,6f
err3; lvx vr1,r0,r4 err3; lvx v1,r0,r4
VPERM(vr8,vr0,vr1,vr16) VPERM(v8,v0,v1,v16)
err3; lvx vr0,r4,r9 err3; lvx v0,r4,r9
VPERM(vr9,vr1,vr0,vr16) VPERM(v9,v1,v0,v16)
addi r4,r4,32 addi r4,r4,32
err3; stvx vr8,r0,r3 err3; stvx v8,r0,r3
err3; stvx vr9,r3,r9 err3; stvx v9,r3,r9
addi r3,r3,32 addi r3,r3,32
6: bf cr7*4+1,7f 6: bf cr7*4+1,7f
err3; lvx vr3,r0,r4 err3; lvx v3,r0,r4
VPERM(vr8,vr0,vr3,vr16) VPERM(v8,v0,v3,v16)
err3; lvx vr2,r4,r9 err3; lvx v2,r4,r9
VPERM(vr9,vr3,vr2,vr16) VPERM(v9,v3,v2,v16)
err3; lvx vr1,r4,r10 err3; lvx v1,r4,r10
VPERM(vr10,vr2,vr1,vr16) VPERM(v10,v2,v1,v16)
err3; lvx vr0,r4,r11 err3; lvx v0,r4,r11
VPERM(vr11,vr1,vr0,vr16) VPERM(v11,v1,v0,v16)
addi r4,r4,64 addi r4,r4,64
err3; stvx vr8,r0,r3 err3; stvx v8,r0,r3
err3; stvx vr9,r3,r9 err3; stvx v9,r3,r9
err3; stvx vr10,r3,r10 err3; stvx v10,r3,r10
err3; stvx vr11,r3,r11 err3; stvx v11,r3,r11
addi r3,r3,64 addi r3,r3,64
7: sub r5,r5,r6 7: sub r5,r5,r6
@ -618,31 +618,31 @@ err3; stvx vr11,r3,r11
*/ */
.align 5 .align 5
8: 8:
err4; lvx vr7,r0,r4 err4; lvx v7,r0,r4
VPERM(vr8,vr0,vr7,vr16) VPERM(v8,v0,v7,v16)
err4; lvx vr6,r4,r9 err4; lvx v6,r4,r9
VPERM(vr9,vr7,vr6,vr16) VPERM(v9,v7,v6,v16)
err4; lvx vr5,r4,r10 err4; lvx v5,r4,r10
VPERM(vr10,vr6,vr5,vr16) VPERM(v10,v6,v5,v16)
err4; lvx vr4,r4,r11 err4; lvx v4,r4,r11
VPERM(vr11,vr5,vr4,vr16) VPERM(v11,v5,v4,v16)
err4; lvx vr3,r4,r12 err4; lvx v3,r4,r12
VPERM(vr12,vr4,vr3,vr16) VPERM(v12,v4,v3,v16)
err4; lvx vr2,r4,r14 err4; lvx v2,r4,r14
VPERM(vr13,vr3,vr2,vr16) VPERM(v13,v3,v2,v16)
err4; lvx vr1,r4,r15 err4; lvx v1,r4,r15
VPERM(vr14,vr2,vr1,vr16) VPERM(v14,v2,v1,v16)
err4; lvx vr0,r4,r16 err4; lvx v0,r4,r16
VPERM(vr15,vr1,vr0,vr16) VPERM(v15,v1,v0,v16)
addi r4,r4,128 addi r4,r4,128
err4; stvx vr8,r0,r3 err4; stvx v8,r0,r3
err4; stvx vr9,r3,r9 err4; stvx v9,r3,r9
err4; stvx vr10,r3,r10 err4; stvx v10,r3,r10
err4; stvx vr11,r3,r11 err4; stvx v11,r3,r11
err4; stvx vr12,r3,r12 err4; stvx v12,r3,r12
err4; stvx vr13,r3,r14 err4; stvx v13,r3,r14
err4; stvx vr14,r3,r15 err4; stvx v14,r3,r15
err4; stvx vr15,r3,r16 err4; stvx v15,r3,r16
addi r3,r3,128 addi r3,r3,128
bdnz 8b bdnz 8b
@ -656,36 +656,36 @@ err4; stvx vr15,r3,r16
mtocrf 0x01,r6 mtocrf 0x01,r6
bf cr7*4+1,9f bf cr7*4+1,9f
err3; lvx vr3,r0,r4 err3; lvx v3,r0,r4
VPERM(vr8,vr0,vr3,vr16) VPERM(v8,v0,v3,v16)
err3; lvx vr2,r4,r9 err3; lvx v2,r4,r9
VPERM(vr9,vr3,vr2,vr16) VPERM(v9,v3,v2,v16)
err3; lvx vr1,r4,r10 err3; lvx v1,r4,r10
VPERM(vr10,vr2,vr1,vr16) VPERM(v10,v2,v1,v16)
err3; lvx vr0,r4,r11 err3; lvx v0,r4,r11
VPERM(vr11,vr1,vr0,vr16) VPERM(v11,v1,v0,v16)
addi r4,r4,64 addi r4,r4,64
err3; stvx vr8,r0,r3 err3; stvx v8,r0,r3
err3; stvx vr9,r3,r9 err3; stvx v9,r3,r9
err3; stvx vr10,r3,r10 err3; stvx v10,r3,r10
err3; stvx vr11,r3,r11 err3; stvx v11,r3,r11
addi r3,r3,64 addi r3,r3,64
9: bf cr7*4+2,10f 9: bf cr7*4+2,10f
err3; lvx vr1,r0,r4 err3; lvx v1,r0,r4
VPERM(vr8,vr0,vr1,vr16) VPERM(v8,v0,v1,v16)
err3; lvx vr0,r4,r9 err3; lvx v0,r4,r9
VPERM(vr9,vr1,vr0,vr16) VPERM(v9,v1,v0,v16)
addi r4,r4,32 addi r4,r4,32
err3; stvx vr8,r0,r3 err3; stvx v8,r0,r3
err3; stvx vr9,r3,r9 err3; stvx v9,r3,r9
addi r3,r3,32 addi r3,r3,32
10: bf cr7*4+3,11f 10: bf cr7*4+3,11f
err3; lvx vr1,r0,r4 err3; lvx v1,r0,r4
VPERM(vr8,vr0,vr1,vr16) VPERM(v8,v0,v1,v16)
addi r4,r4,16 addi r4,r4,16
err3; stvx vr8,r0,r3 err3; stvx v8,r0,r3
addi r3,r3,16 addi r3,r3,16
/* Up to 15B to go */ /* Up to 15B to go */

View File

@ -236,78 +236,78 @@ _GLOBAL(_rest32gpr_31_x)
_GLOBAL(_savevr_20) _GLOBAL(_savevr_20)
li r11,-192 li r11,-192
stvx vr20,r11,r0 stvx v20,r11,r0
_GLOBAL(_savevr_21) _GLOBAL(_savevr_21)
li r11,-176 li r11,-176
stvx vr21,r11,r0 stvx v21,r11,r0
_GLOBAL(_savevr_22) _GLOBAL(_savevr_22)
li r11,-160 li r11,-160
stvx vr22,r11,r0 stvx v22,r11,r0
_GLOBAL(_savevr_23) _GLOBAL(_savevr_23)
li r11,-144 li r11,-144
stvx vr23,r11,r0 stvx v23,r11,r0
_GLOBAL(_savevr_24) _GLOBAL(_savevr_24)
li r11,-128 li r11,-128
stvx vr24,r11,r0 stvx v24,r11,r0
_GLOBAL(_savevr_25) _GLOBAL(_savevr_25)
li r11,-112 li r11,-112
stvx vr25,r11,r0 stvx v25,r11,r0
_GLOBAL(_savevr_26) _GLOBAL(_savevr_26)
li r11,-96 li r11,-96
stvx vr26,r11,r0 stvx v26,r11,r0
_GLOBAL(_savevr_27) _GLOBAL(_savevr_27)
li r11,-80 li r11,-80
stvx vr27,r11,r0 stvx v27,r11,r0
_GLOBAL(_savevr_28) _GLOBAL(_savevr_28)
li r11,-64 li r11,-64
stvx vr28,r11,r0 stvx v28,r11,r0
_GLOBAL(_savevr_29) _GLOBAL(_savevr_29)
li r11,-48 li r11,-48
stvx vr29,r11,r0 stvx v29,r11,r0
_GLOBAL(_savevr_30) _GLOBAL(_savevr_30)
li r11,-32 li r11,-32
stvx vr30,r11,r0 stvx v30,r11,r0
_GLOBAL(_savevr_31) _GLOBAL(_savevr_31)
li r11,-16 li r11,-16
stvx vr31,r11,r0 stvx v31,r11,r0
blr blr
_GLOBAL(_restvr_20) _GLOBAL(_restvr_20)
li r11,-192 li r11,-192
lvx vr20,r11,r0 lvx v20,r11,r0
_GLOBAL(_restvr_21) _GLOBAL(_restvr_21)
li r11,-176 li r11,-176
lvx vr21,r11,r0 lvx v21,r11,r0
_GLOBAL(_restvr_22) _GLOBAL(_restvr_22)
li r11,-160 li r11,-160
lvx vr22,r11,r0 lvx v22,r11,r0
_GLOBAL(_restvr_23) _GLOBAL(_restvr_23)
li r11,-144 li r11,-144
lvx vr23,r11,r0 lvx v23,r11,r0
_GLOBAL(_restvr_24) _GLOBAL(_restvr_24)
li r11,-128 li r11,-128
lvx vr24,r11,r0 lvx v24,r11,r0
_GLOBAL(_restvr_25) _GLOBAL(_restvr_25)
li r11,-112 li r11,-112
lvx vr25,r11,r0 lvx v25,r11,r0
_GLOBAL(_restvr_26) _GLOBAL(_restvr_26)
li r11,-96 li r11,-96
lvx vr26,r11,r0 lvx v26,r11,r0
_GLOBAL(_restvr_27) _GLOBAL(_restvr_27)
li r11,-80 li r11,-80
lvx vr27,r11,r0 lvx v27,r11,r0
_GLOBAL(_restvr_28) _GLOBAL(_restvr_28)
li r11,-64 li r11,-64
lvx vr28,r11,r0 lvx v28,r11,r0
_GLOBAL(_restvr_29) _GLOBAL(_restvr_29)
li r11,-48 li r11,-48
lvx vr29,r11,r0 lvx v29,r11,r0
_GLOBAL(_restvr_30) _GLOBAL(_restvr_30)
li r11,-32 li r11,-32
lvx vr30,r11,r0 lvx v30,r11,r0
_GLOBAL(_restvr_31) _GLOBAL(_restvr_31)
li r11,-16 li r11,-16
lvx vr31,r11,r0 lvx v31,r11,r0
blr blr
#endif /* CONFIG_ALTIVEC */ #endif /* CONFIG_ALTIVEC */
@ -443,101 +443,101 @@ _restgpr0_31:
.globl _savevr_20 .globl _savevr_20
_savevr_20: _savevr_20:
li r12,-192 li r12,-192
stvx vr20,r12,r0 stvx v20,r12,r0
.globl _savevr_21 .globl _savevr_21
_savevr_21: _savevr_21:
li r12,-176 li r12,-176
stvx vr21,r12,r0 stvx v21,r12,r0
.globl _savevr_22 .globl _savevr_22
_savevr_22: _savevr_22:
li r12,-160 li r12,-160
stvx vr22,r12,r0 stvx v22,r12,r0
.globl _savevr_23 .globl _savevr_23
_savevr_23: _savevr_23:
li r12,-144 li r12,-144
stvx vr23,r12,r0 stvx v23,r12,r0
.globl _savevr_24 .globl _savevr_24
_savevr_24: _savevr_24:
li r12,-128 li r12,-128
stvx vr24,r12,r0 stvx v24,r12,r0
.globl _savevr_25 .globl _savevr_25
_savevr_25: _savevr_25:
li r12,-112 li r12,-112
stvx vr25,r12,r0 stvx v25,r12,r0
.globl _savevr_26 .globl _savevr_26
_savevr_26: _savevr_26:
li r12,-96 li r12,-96
stvx vr26,r12,r0 stvx v26,r12,r0
.globl _savevr_27 .globl _savevr_27
_savevr_27: _savevr_27:
li r12,-80 li r12,-80
stvx vr27,r12,r0 stvx v27,r12,r0
.globl _savevr_28 .globl _savevr_28
_savevr_28: _savevr_28:
li r12,-64 li r12,-64
stvx vr28,r12,r0 stvx v28,r12,r0
.globl _savevr_29 .globl _savevr_29
_savevr_29: _savevr_29:
li r12,-48 li r12,-48
stvx vr29,r12,r0 stvx v29,r12,r0
.globl _savevr_30 .globl _savevr_30
_savevr_30: _savevr_30:
li r12,-32 li r12,-32
stvx vr30,r12,r0 stvx v30,r12,r0
.globl _savevr_31 .globl _savevr_31
_savevr_31: _savevr_31:
li r12,-16 li r12,-16
stvx vr31,r12,r0 stvx v31,r12,r0
blr blr
.globl _restvr_20 .globl _restvr_20
_restvr_20: _restvr_20:
li r12,-192 li r12,-192
lvx vr20,r12,r0 lvx v20,r12,r0
.globl _restvr_21 .globl _restvr_21
_restvr_21: _restvr_21:
li r12,-176 li r12,-176
lvx vr21,r12,r0 lvx v21,r12,r0
.globl _restvr_22 .globl _restvr_22
_restvr_22: _restvr_22:
li r12,-160 li r12,-160
lvx vr22,r12,r0 lvx v22,r12,r0
.globl _restvr_23 .globl _restvr_23
_restvr_23: _restvr_23:
li r12,-144 li r12,-144
lvx vr23,r12,r0 lvx v23,r12,r0
.globl _restvr_24 .globl _restvr_24
_restvr_24: _restvr_24:
li r12,-128 li r12,-128
lvx vr24,r12,r0 lvx v24,r12,r0
.globl _restvr_25 .globl _restvr_25
_restvr_25: _restvr_25:
li r12,-112 li r12,-112
lvx vr25,r12,r0 lvx v25,r12,r0
.globl _restvr_26 .globl _restvr_26
_restvr_26: _restvr_26:
li r12,-96 li r12,-96
lvx vr26,r12,r0 lvx v26,r12,r0
.globl _restvr_27 .globl _restvr_27
_restvr_27: _restvr_27:
li r12,-80 li r12,-80
lvx vr27,r12,r0 lvx v27,r12,r0
.globl _restvr_28 .globl _restvr_28
_restvr_28: _restvr_28:
li r12,-64 li r12,-64
lvx vr28,r12,r0 lvx v28,r12,r0
.globl _restvr_29 .globl _restvr_29
_restvr_29: _restvr_29:
li r12,-48 li r12,-48
lvx vr29,r12,r0 lvx v29,r12,r0
.globl _restvr_30 .globl _restvr_30
_restvr_30: _restvr_30:
li r12,-32 li r12,-32
lvx vr30,r12,r0 lvx v30,r12,r0
.globl _restvr_31 .globl _restvr_31
_restvr_31: _restvr_31:
li r12,-16 li r12,-16
lvx vr31,r12,r0 lvx v31,r12,r0
blr blr
#endif /* CONFIG_ALTIVEC */ #endif /* CONFIG_ALTIVEC */

View File

@ -184,16 +184,16 @@ _GLOBAL(do_stfd)
extab 2b,3b extab 2b,3b
#ifdef CONFIG_ALTIVEC #ifdef CONFIG_ALTIVEC
/* Get the contents of vrN into vr0; N is in r3. */ /* Get the contents of vrN into v0; N is in r3. */
_GLOBAL(get_vr) _GLOBAL(get_vr)
mflr r0 mflr r0
rlwinm r3,r3,3,0xf8 rlwinm r3,r3,3,0xf8
bcl 20,31,1f bcl 20,31,1f
blr /* vr0 is already in vr0 */ blr /* v0 is already in v0 */
nop nop
reg = 1 reg = 1
.rept 31 .rept 31
vor vr0,reg,reg /* assembler doesn't know vmr? */ vor v0,reg,reg /* assembler doesn't know vmr? */
blr blr
reg = reg + 1 reg = reg + 1
.endr .endr
@ -203,16 +203,16 @@ reg = reg + 1
mtlr r0 mtlr r0
bctr bctr
/* Put the contents of vr0 into vrN; N is in r3. */ /* Put the contents of v0 into vrN; N is in r3. */
_GLOBAL(put_vr) _GLOBAL(put_vr)
mflr r0 mflr r0
rlwinm r3,r3,3,0xf8 rlwinm r3,r3,3,0xf8
bcl 20,31,1f bcl 20,31,1f
blr /* vr0 is already in vr0 */ blr /* v0 is already in v0 */
nop nop
reg = 1 reg = 1
.rept 31 .rept 31
vor reg,vr0,vr0 vor reg,v0,v0
blr blr
reg = reg + 1 reg = reg + 1
.endr .endr
@ -234,13 +234,13 @@ _GLOBAL(do_lvx)
MTMSRD(r7) MTMSRD(r7)
isync isync
beq cr7,1f beq cr7,1f
stvx vr0,r1,r8 stvx v0,r1,r8
1: li r9,-EFAULT 1: li r9,-EFAULT
2: lvx vr0,0,r4 2: lvx v0,0,r4
li r9,0 li r9,0
3: beq cr7,4f 3: beq cr7,4f
bl put_vr bl put_vr
lvx vr0,r1,r8 lvx v0,r1,r8
4: PPC_LL r0,STKFRM+PPC_LR_STKOFF(r1) 4: PPC_LL r0,STKFRM+PPC_LR_STKOFF(r1)
mtlr r0 mtlr r0
MTMSRD(r6) MTMSRD(r6)
@ -262,13 +262,13 @@ _GLOBAL(do_stvx)
MTMSRD(r7) MTMSRD(r7)
isync isync
beq cr7,1f beq cr7,1f
stvx vr0,r1,r8 stvx v0,r1,r8
bl get_vr bl get_vr
1: li r9,-EFAULT 1: li r9,-EFAULT
2: stvx vr0,0,r4 2: stvx v0,0,r4
li r9,0 li r9,0
3: beq cr7,4f 3: beq cr7,4f
lvx vr0,r1,r8 lvx v0,r1,r8
4: PPC_LL r0,STKFRM+PPC_LR_STKOFF(r1) 4: PPC_LL r0,STKFRM+PPC_LR_STKOFF(r1)
mtlr r0 mtlr r0
MTMSRD(r6) MTMSRD(r6)
@ -304,7 +304,7 @@ _GLOBAL(put_vsr)
mflr r0 mflr r0
rlwinm r3,r3,3,0x1f8 rlwinm r3,r3,3,0x1f8
bcl 20,31,1f bcl 20,31,1f
blr /* vr0 is already in vr0 */ blr /* v0 is already in v0 */
nop nop
reg = 1 reg = 1
.rept 63 .rept 63

View File

@ -321,29 +321,29 @@ _GLOBAL(memcpy_power7)
li r11,48 li r11,48
bf cr7*4+3,5f bf cr7*4+3,5f
lvx vr1,r0,r4 lvx v1,r0,r4
addi r4,r4,16 addi r4,r4,16
stvx vr1,r0,r3 stvx v1,r0,r3
addi r3,r3,16 addi r3,r3,16
5: bf cr7*4+2,6f 5: bf cr7*4+2,6f
lvx vr1,r0,r4 lvx v1,r0,r4
lvx vr0,r4,r9 lvx v0,r4,r9
addi r4,r4,32 addi r4,r4,32
stvx vr1,r0,r3 stvx v1,r0,r3
stvx vr0,r3,r9 stvx v0,r3,r9
addi r3,r3,32 addi r3,r3,32
6: bf cr7*4+1,7f 6: bf cr7*4+1,7f
lvx vr3,r0,r4 lvx v3,r0,r4
lvx vr2,r4,r9 lvx v2,r4,r9
lvx vr1,r4,r10 lvx v1,r4,r10
lvx vr0,r4,r11 lvx v0,r4,r11
addi r4,r4,64 addi r4,r4,64
stvx vr3,r0,r3 stvx v3,r0,r3
stvx vr2,r3,r9 stvx v2,r3,r9
stvx vr1,r3,r10 stvx v1,r3,r10
stvx vr0,r3,r11 stvx v0,r3,r11
addi r3,r3,64 addi r3,r3,64
7: sub r5,r5,r6 7: sub r5,r5,r6
@ -366,23 +366,23 @@ _GLOBAL(memcpy_power7)
*/ */
.align 5 .align 5
8: 8:
lvx vr7,r0,r4 lvx v7,r0,r4
lvx vr6,r4,r9 lvx v6,r4,r9
lvx vr5,r4,r10 lvx v5,r4,r10
lvx vr4,r4,r11 lvx v4,r4,r11
lvx vr3,r4,r12 lvx v3,r4,r12
lvx vr2,r4,r14 lvx v2,r4,r14
lvx vr1,r4,r15 lvx v1,r4,r15
lvx vr0,r4,r16 lvx v0,r4,r16
addi r4,r4,128 addi r4,r4,128
stvx vr7,r0,r3 stvx v7,r0,r3
stvx vr6,r3,r9 stvx v6,r3,r9
stvx vr5,r3,r10 stvx v5,r3,r10
stvx vr4,r3,r11 stvx v4,r3,r11
stvx vr3,r3,r12 stvx v3,r3,r12
stvx vr2,r3,r14 stvx v2,r3,r14
stvx vr1,r3,r15 stvx v1,r3,r15
stvx vr0,r3,r16 stvx v0,r3,r16
addi r3,r3,128 addi r3,r3,128
bdnz 8b bdnz 8b
@ -396,29 +396,29 @@ _GLOBAL(memcpy_power7)
mtocrf 0x01,r6 mtocrf 0x01,r6
bf cr7*4+1,9f bf cr7*4+1,9f
lvx vr3,r0,r4 lvx v3,r0,r4
lvx vr2,r4,r9 lvx v2,r4,r9
lvx vr1,r4,r10 lvx v1,r4,r10
lvx vr0,r4,r11 lvx v0,r4,r11
addi r4,r4,64 addi r4,r4,64
stvx vr3,r0,r3 stvx v3,r0,r3
stvx vr2,r3,r9 stvx v2,r3,r9
stvx vr1,r3,r10 stvx v1,r3,r10
stvx vr0,r3,r11 stvx v0,r3,r11
addi r3,r3,64 addi r3,r3,64
9: bf cr7*4+2,10f 9: bf cr7*4+2,10f
lvx vr1,r0,r4 lvx v1,r0,r4
lvx vr0,r4,r9 lvx v0,r4,r9
addi r4,r4,32 addi r4,r4,32
stvx vr1,r0,r3 stvx v1,r0,r3
stvx vr0,r3,r9 stvx v0,r3,r9
addi r3,r3,32 addi r3,r3,32
10: bf cr7*4+3,11f 10: bf cr7*4+3,11f
lvx vr1,r0,r4 lvx v1,r0,r4
addi r4,r4,16 addi r4,r4,16
stvx vr1,r0,r3 stvx v1,r0,r3
addi r3,r3,16 addi r3,r3,16
/* Up to 15B to go */ /* Up to 15B to go */
@ -494,42 +494,42 @@ _GLOBAL(memcpy_power7)
li r10,32 li r10,32
li r11,48 li r11,48
LVS(vr16,0,r4) /* Setup permute control vector */ LVS(v16,0,r4) /* Setup permute control vector */
lvx vr0,0,r4 lvx v0,0,r4
addi r4,r4,16 addi r4,r4,16
bf cr7*4+3,5f bf cr7*4+3,5f
lvx vr1,r0,r4 lvx v1,r0,r4
VPERM(vr8,vr0,vr1,vr16) VPERM(v8,v0,v1,v16)
addi r4,r4,16 addi r4,r4,16
stvx vr8,r0,r3 stvx v8,r0,r3
addi r3,r3,16 addi r3,r3,16
vor vr0,vr1,vr1 vor v0,v1,v1
5: bf cr7*4+2,6f 5: bf cr7*4+2,6f
lvx vr1,r0,r4 lvx v1,r0,r4
VPERM(vr8,vr0,vr1,vr16) VPERM(v8,v0,v1,v16)
lvx vr0,r4,r9 lvx v0,r4,r9
VPERM(vr9,vr1,vr0,vr16) VPERM(v9,v1,v0,v16)
addi r4,r4,32 addi r4,r4,32
stvx vr8,r0,r3 stvx v8,r0,r3
stvx vr9,r3,r9 stvx v9,r3,r9
addi r3,r3,32 addi r3,r3,32
6: bf cr7*4+1,7f 6: bf cr7*4+1,7f
lvx vr3,r0,r4 lvx v3,r0,r4
VPERM(vr8,vr0,vr3,vr16) VPERM(v8,v0,v3,v16)
lvx vr2,r4,r9 lvx v2,r4,r9
VPERM(vr9,vr3,vr2,vr16) VPERM(v9,v3,v2,v16)
lvx vr1,r4,r10 lvx v1,r4,r10
VPERM(vr10,vr2,vr1,vr16) VPERM(v10,v2,v1,v16)
lvx vr0,r4,r11 lvx v0,r4,r11
VPERM(vr11,vr1,vr0,vr16) VPERM(v11,v1,v0,v16)
addi r4,r4,64 addi r4,r4,64
stvx vr8,r0,r3 stvx v8,r0,r3
stvx vr9,r3,r9 stvx v9,r3,r9
stvx vr10,r3,r10 stvx v10,r3,r10
stvx vr11,r3,r11 stvx v11,r3,r11
addi r3,r3,64 addi r3,r3,64
7: sub r5,r5,r6 7: sub r5,r5,r6
@ -552,31 +552,31 @@ _GLOBAL(memcpy_power7)
*/ */
.align 5 .align 5
8: 8:
lvx vr7,r0,r4 lvx v7,r0,r4
VPERM(vr8,vr0,vr7,vr16) VPERM(v8,v0,v7,v16)
lvx vr6,r4,r9 lvx v6,r4,r9
VPERM(vr9,vr7,vr6,vr16) VPERM(v9,v7,v6,v16)
lvx vr5,r4,r10 lvx v5,r4,r10
VPERM(vr10,vr6,vr5,vr16) VPERM(v10,v6,v5,v16)
lvx vr4,r4,r11 lvx v4,r4,r11
VPERM(vr11,vr5,vr4,vr16) VPERM(v11,v5,v4,v16)
lvx vr3,r4,r12 lvx v3,r4,r12
VPERM(vr12,vr4,vr3,vr16) VPERM(v12,v4,v3,v16)
lvx vr2,r4,r14 lvx v2,r4,r14
VPERM(vr13,vr3,vr2,vr16) VPERM(v13,v3,v2,v16)
lvx vr1,r4,r15 lvx v1,r4,r15
VPERM(vr14,vr2,vr1,vr16) VPERM(v14,v2,v1,v16)
lvx vr0,r4,r16 lvx v0,r4,r16
VPERM(vr15,vr1,vr0,vr16) VPERM(v15,v1,v0,v16)
addi r4,r4,128 addi r4,r4,128
stvx vr8,r0,r3 stvx v8,r0,r3
stvx vr9,r3,r9 stvx v9,r3,r9
stvx vr10,r3,r10 stvx v10,r3,r10
stvx vr11,r3,r11 stvx v11,r3,r11
stvx vr12,r3,r12 stvx v12,r3,r12
stvx vr13,r3,r14 stvx v13,r3,r14
stvx vr14,r3,r15 stvx v14,r3,r15
stvx vr15,r3,r16 stvx v15,r3,r16
addi r3,r3,128 addi r3,r3,128
bdnz 8b bdnz 8b
@ -590,36 +590,36 @@ _GLOBAL(memcpy_power7)
mtocrf 0x01,r6 mtocrf 0x01,r6
bf cr7*4+1,9f bf cr7*4+1,9f
lvx vr3,r0,r4 lvx v3,r0,r4
VPERM(vr8,vr0,vr3,vr16) VPERM(v8,v0,v3,v16)
lvx vr2,r4,r9 lvx v2,r4,r9
VPERM(vr9,vr3,vr2,vr16) VPERM(v9,v3,v2,v16)
lvx vr1,r4,r10 lvx v1,r4,r10
VPERM(vr10,vr2,vr1,vr16) VPERM(v10,v2,v1,v16)
lvx vr0,r4,r11 lvx v0,r4,r11
VPERM(vr11,vr1,vr0,vr16) VPERM(v11,v1,v0,v16)
addi r4,r4,64 addi r4,r4,64
stvx vr8,r0,r3 stvx v8,r0,r3
stvx vr9,r3,r9 stvx v9,r3,r9
stvx vr10,r3,r10 stvx v10,r3,r10
stvx vr11,r3,r11 stvx v11,r3,r11
addi r3,r3,64 addi r3,r3,64
9: bf cr7*4+2,10f 9: bf cr7*4+2,10f
lvx vr1,r0,r4 lvx v1,r0,r4
VPERM(vr8,vr0,vr1,vr16) VPERM(v8,v0,v1,v16)
lvx vr0,r4,r9 lvx v0,r4,r9
VPERM(vr9,vr1,vr0,vr16) VPERM(v9,v1,v0,v16)
addi r4,r4,32 addi r4,r4,32
stvx vr8,r0,r3 stvx v8,r0,r3
stvx vr9,r3,r9 stvx v9,r3,r9
addi r3,r3,32 addi r3,r3,32
10: bf cr7*4+3,11f 10: bf cr7*4+3,11f
lvx vr1,r0,r4 lvx v1,r0,r4
VPERM(vr8,vr0,vr1,vr16) VPERM(v8,v0,v1,v16)
addi r4,r4,16 addi r4,r4,16
stvx vr8,r0,r3 stvx v8,r0,r3
addi r3,r3,16 addi r3,r3,16
/* Up to 15B to go */ /* Up to 15B to go */

View File

@ -4,39 +4,6 @@
#define r1 1 #define r1 1
#define vr0 0
#define vr1 1
#define vr2 2
#define vr3 3
#define vr4 4
#define vr5 5
#define vr6 6
#define vr7 7
#define vr8 8
#define vr9 9
#define vr10 10
#define vr11 11
#define vr12 12
#define vr13 13
#define vr14 14
#define vr15 15
#define vr16 16
#define vr17 17
#define vr18 18
#define vr19 19
#define vr20 20
#define vr21 21
#define vr22 22
#define vr23 23
#define vr24 24
#define vr25 25
#define vr26 26
#define vr27 27
#define vr28 28
#define vr29 29
#define vr30 30
#define vr31 31
#define R14 r14 #define R14 r14
#define R15 r15 #define R15 r15
#define R16 r16 #define R16 r16