diff options
author | Anton Blanchard <anton@samba.org> | 2015-02-10 09:51:22 +1100 |
---|---|---|
committer | Michael Ellerman <mpe@ellerman.id.au> | 2015-03-16 18:32:11 +1100 |
commit | c2ce6f9f3dc00daca5714ef070a9a2d4e78eb336 (patch) | |
tree | c008f72ced83ffd950f0920566c378b4809780cf /arch/powerpc/lib/memcpy_power7.S | |
parent | 06e5801b8cb3fc057d88cb4dc03c0b64b2744cda (diff) | |
download | linux-stable-c2ce6f9f3dc00daca5714ef070a9a2d4e78eb336.tar.gz linux-stable-c2ce6f9f3dc00daca5714ef070a9a2d4e78eb336.tar.bz2 linux-stable-c2ce6f9f3dc00daca5714ef070a9a2d4e78eb336.zip |
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>
Diffstat (limited to 'arch/powerpc/lib/memcpy_power7.S')
-rw-r--r-- | arch/powerpc/lib/memcpy_power7.S | 226 |
1 files changed, 113 insertions, 113 deletions
diff --git a/arch/powerpc/lib/memcpy_power7.S b/arch/powerpc/lib/memcpy_power7.S index 0830587df16e..786234fd4e91 100644 --- a/arch/powerpc/lib/memcpy_power7.S +++ b/arch/powerpc/lib/memcpy_power7.S @@ -321,29 +321,29 @@ _GLOBAL(memcpy_power7) li r11,48 bf cr7*4+3,5f - lvx vr1,r0,r4 + lvx v1,r0,r4 addi r4,r4,16 - stvx vr1,r0,r3 + stvx v1,r0,r3 addi r3,r3,16 5: bf cr7*4+2,6f - lvx vr1,r0,r4 - lvx vr0,r4,r9 + lvx v1,r0,r4 + lvx v0,r4,r9 addi r4,r4,32 - stvx vr1,r0,r3 - stvx vr0,r3,r9 + stvx v1,r0,r3 + stvx v0,r3,r9 addi r3,r3,32 6: bf cr7*4+1,7f - lvx vr3,r0,r4 - lvx vr2,r4,r9 - lvx vr1,r4,r10 - lvx vr0,r4,r11 + lvx v3,r0,r4 + lvx v2,r4,r9 + lvx v1,r4,r10 + lvx v0,r4,r11 addi r4,r4,64 - stvx vr3,r0,r3 - stvx vr2,r3,r9 - stvx vr1,r3,r10 - stvx vr0,r3,r11 + stvx v3,r0,r3 + stvx v2,r3,r9 + stvx v1,r3,r10 + stvx v0,r3,r11 addi r3,r3,64 7: sub r5,r5,r6 @@ -366,23 +366,23 @@ _GLOBAL(memcpy_power7) */ .align 5 8: - lvx vr7,r0,r4 - lvx vr6,r4,r9 - lvx vr5,r4,r10 - lvx vr4,r4,r11 - lvx vr3,r4,r12 - lvx vr2,r4,r14 - lvx vr1,r4,r15 - lvx vr0,r4,r16 + lvx v7,r0,r4 + lvx v6,r4,r9 + lvx v5,r4,r10 + lvx v4,r4,r11 + lvx v3,r4,r12 + lvx v2,r4,r14 + lvx v1,r4,r15 + lvx v0,r4,r16 addi r4,r4,128 - stvx vr7,r0,r3 - stvx vr6,r3,r9 - stvx vr5,r3,r10 - stvx vr4,r3,r11 - stvx vr3,r3,r12 - stvx vr2,r3,r14 - stvx vr1,r3,r15 - stvx vr0,r3,r16 + stvx v7,r0,r3 + stvx v6,r3,r9 + stvx v5,r3,r10 + stvx v4,r3,r11 + stvx v3,r3,r12 + stvx v2,r3,r14 + stvx v1,r3,r15 + stvx v0,r3,r16 addi r3,r3,128 bdnz 8b @@ -396,29 +396,29 @@ _GLOBAL(memcpy_power7) mtocrf 0x01,r6 bf cr7*4+1,9f - lvx vr3,r0,r4 - lvx vr2,r4,r9 - lvx vr1,r4,r10 - lvx vr0,r4,r11 + lvx v3,r0,r4 + lvx v2,r4,r9 + lvx v1,r4,r10 + lvx v0,r4,r11 addi r4,r4,64 - stvx vr3,r0,r3 - stvx vr2,r3,r9 - stvx vr1,r3,r10 - stvx vr0,r3,r11 + stvx v3,r0,r3 + stvx v2,r3,r9 + stvx v1,r3,r10 + stvx v0,r3,r11 addi r3,r3,64 9: bf cr7*4+2,10f - lvx vr1,r0,r4 - lvx vr0,r4,r9 + lvx v1,r0,r4 + lvx v0,r4,r9 addi r4,r4,32 - stvx vr1,r0,r3 - stvx vr0,r3,r9 + stvx v1,r0,r3 + stvx v0,r3,r9 addi r3,r3,32 10: bf cr7*4+3,11f - lvx vr1,r0,r4 + lvx v1,r0,r4 addi r4,r4,16 - stvx vr1,r0,r3 + stvx v1,r0,r3 addi r3,r3,16 /* Up to 15B to go */ @@ -494,42 +494,42 @@ _GLOBAL(memcpy_power7) li r10,32 li r11,48 - LVS(vr16,0,r4) /* Setup permute control vector */ - lvx vr0,0,r4 + LVS(v16,0,r4) /* Setup permute control vector */ + lvx v0,0,r4 addi r4,r4,16 bf cr7*4+3,5f - lvx vr1,r0,r4 - VPERM(vr8,vr0,vr1,vr16) + lvx v1,r0,r4 + VPERM(v8,v0,v1,v16) addi r4,r4,16 - stvx vr8,r0,r3 + stvx v8,r0,r3 addi r3,r3,16 - vor vr0,vr1,vr1 + vor v0,v1,v1 5: bf cr7*4+2,6f - lvx vr1,r0,r4 - VPERM(vr8,vr0,vr1,vr16) - lvx vr0,r4,r9 - VPERM(vr9,vr1,vr0,vr16) + lvx v1,r0,r4 + VPERM(v8,v0,v1,v16) + lvx v0,r4,r9 + VPERM(v9,v1,v0,v16) addi r4,r4,32 - stvx vr8,r0,r3 - stvx vr9,r3,r9 + stvx v8,r0,r3 + stvx v9,r3,r9 addi r3,r3,32 6: bf cr7*4+1,7f - lvx vr3,r0,r4 - VPERM(vr8,vr0,vr3,vr16) - lvx vr2,r4,r9 - VPERM(vr9,vr3,vr2,vr16) - lvx vr1,r4,r10 - VPERM(vr10,vr2,vr1,vr16) - lvx vr0,r4,r11 - VPERM(vr11,vr1,vr0,vr16) + lvx v3,r0,r4 + VPERM(v8,v0,v3,v16) + lvx v2,r4,r9 + VPERM(v9,v3,v2,v16) + lvx v1,r4,r10 + VPERM(v10,v2,v1,v16) + lvx v0,r4,r11 + VPERM(v11,v1,v0,v16) addi r4,r4,64 - stvx vr8,r0,r3 - stvx vr9,r3,r9 - stvx vr10,r3,r10 - stvx vr11,r3,r11 + stvx v8,r0,r3 + stvx v9,r3,r9 + stvx v10,r3,r10 + stvx v11,r3,r11 addi r3,r3,64 7: sub r5,r5,r6 @@ -552,31 +552,31 @@ _GLOBAL(memcpy_power7) */ .align 5 8: - lvx vr7,r0,r4 - VPERM(vr8,vr0,vr7,vr16) - lvx vr6,r4,r9 - VPERM(vr9,vr7,vr6,vr16) - lvx vr5,r4,r10 - VPERM(vr10,vr6,vr5,vr16) - lvx vr4,r4,r11 - VPERM(vr11,vr5,vr4,vr16) - lvx vr3,r4,r12 - VPERM(vr12,vr4,vr3,vr16) - lvx vr2,r4,r14 - VPERM(vr13,vr3,vr2,vr16) - lvx vr1,r4,r15 - VPERM(vr14,vr2,vr1,vr16) - lvx vr0,r4,r16 - VPERM(vr15,vr1,vr0,vr16) + lvx v7,r0,r4 + VPERM(v8,v0,v7,v16) + lvx v6,r4,r9 + VPERM(v9,v7,v6,v16) + lvx v5,r4,r10 + VPERM(v10,v6,v5,v16) + lvx v4,r4,r11 + VPERM(v11,v5,v4,v16) + lvx v3,r4,r12 + VPERM(v12,v4,v3,v16) + lvx v2,r4,r14 + VPERM(v13,v3,v2,v16) + lvx v1,r4,r15 + VPERM(v14,v2,v1,v16) + lvx v0,r4,r16 + VPERM(v15,v1,v0,v16) addi r4,r4,128 - stvx vr8,r0,r3 - stvx vr9,r3,r9 - stvx vr10,r3,r10 - stvx vr11,r3,r11 - stvx vr12,r3,r12 - stvx vr13,r3,r14 - stvx vr14,r3,r15 - stvx vr15,r3,r16 + stvx v8,r0,r3 + stvx v9,r3,r9 + stvx v10,r3,r10 + stvx v11,r3,r11 + stvx v12,r3,r12 + stvx v13,r3,r14 + stvx v14,r3,r15 + stvx v15,r3,r16 addi r3,r3,128 bdnz 8b @@ -590,36 +590,36 @@ _GLOBAL(memcpy_power7) mtocrf 0x01,r6 bf cr7*4+1,9f - lvx vr3,r0,r4 - VPERM(vr8,vr0,vr3,vr16) - lvx vr2,r4,r9 - VPERM(vr9,vr3,vr2,vr16) - lvx vr1,r4,r10 - VPERM(vr10,vr2,vr1,vr16) - lvx vr0,r4,r11 - VPERM(vr11,vr1,vr0,vr16) + lvx v3,r0,r4 + VPERM(v8,v0,v3,v16) + lvx v2,r4,r9 + VPERM(v9,v3,v2,v16) + lvx v1,r4,r10 + VPERM(v10,v2,v1,v16) + lvx v0,r4,r11 + VPERM(v11,v1,v0,v16) addi r4,r4,64 - stvx vr8,r0,r3 - stvx vr9,r3,r9 - stvx vr10,r3,r10 - stvx vr11,r3,r11 + stvx v8,r0,r3 + stvx v9,r3,r9 + stvx v10,r3,r10 + stvx v11,r3,r11 addi r3,r3,64 9: bf cr7*4+2,10f - lvx vr1,r0,r4 - VPERM(vr8,vr0,vr1,vr16) - lvx vr0,r4,r9 - VPERM(vr9,vr1,vr0,vr16) + lvx v1,r0,r4 + VPERM(v8,v0,v1,v16) + lvx v0,r4,r9 + VPERM(v9,v1,v0,v16) addi r4,r4,32 - stvx vr8,r0,r3 - stvx vr9,r3,r9 + stvx v8,r0,r3 + stvx v9,r3,r9 addi r3,r3,32 10: bf cr7*4+3,11f - lvx vr1,r0,r4 - VPERM(vr8,vr0,vr1,vr16) + lvx v1,r0,r4 + VPERM(v8,v0,v1,v16) addi r4,r4,16 - stvx vr8,r0,r3 + stvx v8,r0,r3 addi r3,r3,16 /* Up to 15B to go */ |