AuxXxilium/linux_dsm_epyc7002
1
0
Fork 0
mirror of https://github.com/AuxXxilium/linux_dsm_epyc7002.git synced 2024-12-28 11:18:45 +07:00
Code Issues Actions Packages Projects Releases Wiki Activity
5f54c8b2d4
linux_dsm_epyc7002/arch/s390/include/asm/vx-insn.h
Martin Schwidefsky 474fd6e80f RAID/s390: add SIMD implementation for raid6 gen/xor 
Using vector registers is slightly faster:

raid6: vx128x8  gen() 19705 MB/s
raid6: vx128x8  xor() 11886 MB/s
raid6: using algorithm vx128x8 gen() 19705 MB/s
raid6: .... xor() 11886 MB/s, rmw enabled

vs the software algorithms:

raid6: int64x1  gen()  3018 MB/s
raid6: int64x1  xor()  1429 MB/s
raid6: int64x2  gen()  4661 MB/s
raid6: int64x2  xor()  3143 MB/s
raid6: int64x4  gen()  5392 MB/s
raid6: int64x4  xor()  3509 MB/s
raid6: int64x8  gen()  4441 MB/s
raid6: int64x8  xor()  3207 MB/s
raid6: using algorithm int64x4 gen() 5392 MB/s
raid6: .... xor() 3509 MB/s, rmw enabled

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2016-08-29 11:05:04 +02:00

561 lines
11 KiB
C
Raw Blame History

/*
* Support for Vector Instructions
*
* Assembler macros to generate .byte/.word code for particular
* vector instructions that are supported by recent binutils (>= 2.26) only.
*
* Copyright IBM Corp. 2015
* Author(s): Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
*/
#ifndef __ASM_S390_VX_INSN_H
#define __ASM_S390_VX_INSN_H
#ifdef __ASSEMBLY__
/* Macros to generate vector instruction byte code */
/* GR_NUM - Retrieve general-purpose register number
*
* @opd: Operand to store register number
* @r64: String designation register in the format "%rN"
*/
.macro GR_NUM opd gr
\opd = 255
.ifc \gr,%r0
\opd = 0
.endif
.ifc \gr,%r1
\opd = 1
.endif
.ifc \gr,%r2
\opd = 2
.endif
.ifc \gr,%r3
\opd = 3
.endif
.ifc \gr,%r4
\opd = 4
.endif
.ifc \gr,%r5
\opd = 5
.endif
.ifc \gr,%r6
\opd = 6
.endif
.ifc \gr,%r7
\opd = 7
.endif
.ifc \gr,%r8
\opd = 8
.endif
.ifc \gr,%r9
\opd = 9
.endif
.ifc \gr,%r10
\opd = 10
.endif
.ifc \gr,%r11
\opd = 11
.endif
.ifc \gr,%r12
\opd = 12
.endif
.ifc \gr,%r13
\opd = 13
.endif
.ifc \gr,%r14
\opd = 14
.endif
.ifc \gr,%r15
\opd = 15
.endif
.if \opd == 255
\opd = \gr
.endif
.endm
/* VX_NUM - Retrieve vector register number
*
* @opd: Operand to store register number
* @vxr: String designation register in the format "%vN"
*
* The vector register number is used for as input number to the
* instruction and, as well as, to compute the RXB field of the
* instruction.
*/
.macro VX_NUM opd vxr
\opd = 255
.ifc \vxr,%v0
\opd = 0
.endif
.ifc \vxr,%v1
\opd = 1
.endif
.ifc \vxr,%v2
\opd = 2
.endif
.ifc \vxr,%v3
\opd = 3
.endif
.ifc \vxr,%v4
\opd = 4
.endif
.ifc \vxr,%v5
\opd = 5
.endif
.ifc \vxr,%v6
\opd = 6
.endif
.ifc \vxr,%v7
\opd = 7
.endif
.ifc \vxr,%v8
\opd = 8
.endif
.ifc \vxr,%v9
\opd = 9
.endif
.ifc \vxr,%v10
\opd = 10
.endif
.ifc \vxr,%v11
\opd = 11
.endif
.ifc \vxr,%v12
\opd = 12
.endif
.ifc \vxr,%v13
\opd = 13
.endif
.ifc \vxr,%v14
\opd = 14
.endif
.ifc \vxr,%v15
\opd = 15
.endif
.ifc \vxr,%v16
\opd = 16
.endif
.ifc \vxr,%v17
\opd = 17
.endif
.ifc \vxr,%v18
\opd = 18
.endif
.ifc \vxr,%v19
\opd = 19
.endif
.ifc \vxr,%v20
\opd = 20
.endif
.ifc \vxr,%v21
\opd = 21
.endif
.ifc \vxr,%v22
\opd = 22
.endif
.ifc \vxr,%v23
\opd = 23
.endif
.ifc \vxr,%v24
\opd = 24
.endif
.ifc \vxr,%v25
\opd = 25
.endif
.ifc \vxr,%v26
\opd = 26
.endif
.ifc \vxr,%v27
\opd = 27
.endif
.ifc \vxr,%v28
\opd = 28
.endif
.ifc \vxr,%v29
\opd = 29
.endif
.ifc \vxr,%v30
\opd = 30
.endif
.ifc \vxr,%v31
\opd = 31
.endif
.if \opd == 255
\opd = \vxr
.endif
.endm
/* RXB - Compute most significant bit used vector registers
*
* @rxb: Operand to store computed RXB value
* @v1: First vector register designated operand
* @v2: Second vector register designated operand
* @v3: Third vector register designated operand
* @v4: Fourth vector register designated operand
*/
.macro RXB rxb v1 v2=0 v3=0 v4=0
\rxb = 0
.if \v1 & 0x10
\rxb = \rxb | 0x08
.endif
.if \v2 & 0x10
\rxb = \rxb | 0x04
.endif
.if \v3 & 0x10
\rxb = \rxb | 0x02
.endif
.if \v4 & 0x10
\rxb = \rxb | 0x01
.endif
.endm
/* MRXB - Generate Element Size Control and RXB value
*
* @m: Element size control
* @v1: First vector register designated operand (for RXB)
* @v2: Second vector register designated operand (for RXB)
* @v3: Third vector register designated operand (for RXB)
* @v4: Fourth vector register designated operand (for RXB)
*/
.macro MRXB m v1 v2=0 v3=0 v4=0
rxb = 0
RXB rxb, \v1, \v2, \v3, \v4
.byte (\m << 4) | rxb
.endm
/* MRXBOPC - Generate Element Size Control, RXB, and final Opcode fields
*
* @m: Element size control
* @opc: Opcode
* @v1: First vector register designated operand (for RXB)
* @v2: Second vector register designated operand (for RXB)
* @v3: Third vector register designated operand (for RXB)
* @v4: Fourth vector register designated operand (for RXB)
*/
.macro MRXBOPC m opc v1 v2=0 v3=0 v4=0
MRXB \m, \v1, \v2, \v3, \v4
.byte \opc
.endm
/* Vector support instructions */
/* VECTOR GENERATE BYTE MASK */
.macro VGBM vr imm2
VX_NUM v1, \vr
.word (0xE700 | ((v1&15) << 4))
.word \imm2
MRXBOPC 0, 0x44, v1
.endm
.macro VZERO vxr
VGBM \vxr, 0
.endm
.macro VONE vxr
VGBM \vxr, 0xFFFF
.endm
/* VECTOR LOAD VR ELEMENT FROM GR */
.macro VLVG v, gr, disp, m
VX_NUM v1, \v
GR_NUM b2, "%r0"
GR_NUM r3, \gr
.word 0xE700 | ((v1&15) << 4) | r3
.word (b2 << 12) | (\disp)
MRXBOPC \m, 0x22, v1
.endm
.macro VLVGB v, gr, index, base
VLVG \v, \gr, \index, \base, 0
.endm
.macro VLVGH v, gr, index
VLVG \v, \gr, \index, 1
.endm
.macro VLVGF v, gr, index
VLVG \v, \gr, \index, 2
.endm
.macro VLVGG v, gr, index
VLVG \v, \gr, \index, 3
.endm
/* VECTOR LOAD REGISTER */
.macro VLR v1, v2
VX_NUM v1, \v1
VX_NUM v2, \v2
.word 0xE700 | ((v1&15) << 4) | (v2&15)
.word 0
MRXBOPC 0, 0x56, v1, v2
.endm
/* VECTOR LOAD */
.macro VL v, disp, index="%r0", base
VX_NUM v1, \v
GR_NUM x2, \index
GR_NUM b2, \base
.word 0xE700 | ((v1&15) << 4) | x2
.word (b2 << 12) | (\disp)
MRXBOPC 0, 0x06, v1
.endm
/* VECTOR LOAD ELEMENT */
.macro VLEx vr1, disp, index="%r0", base, m3, opc
VX_NUM v1, \vr1
GR_NUM x2, \index
GR_NUM b2, \base
.word 0xE700 | ((v1&15) << 4) | x2
.word (b2 << 12) | (\disp)
MRXBOPC \m3, \opc, v1
.endm
.macro VLEB vr1, disp, index="%r0", base, m3
VLEx \vr1, \disp, \index, \base, \m3, 0x00
.endm
.macro VLEH vr1, disp, index="%r0", base, m3
VLEx \vr1, \disp, \index, \base, \m3, 0x01
.endm
.macro VLEF vr1, disp, index="%r0", base, m3
VLEx \vr1, \disp, \index, \base, \m3, 0x03
.endm
.macro VLEG vr1, disp, index="%r0", base, m3
VLEx \vr1, \disp, \index, \base, \m3, 0x02
.endm
/* VECTOR LOAD ELEMENT IMMEDIATE */
.macro VLEIx vr1, imm2, m3, opc
VX_NUM v1, \vr1
.word 0xE700 | ((v1&15) << 4)
.word \imm2
MRXBOPC \m3, \opc, v1
.endm
.macro VLEIB vr1, imm2, index
VLEIx \vr1, \imm2, \index, 0x40
.endm
.macro VLEIH vr1, imm2, index
VLEIx \vr1, \imm2, \index, 0x41
.endm
.macro VLEIF vr1, imm2, index
VLEIx \vr1, \imm2, \index, 0x43
.endm
.macro VLEIG vr1, imm2, index
VLEIx \vr1, \imm2, \index, 0x42
.endm
/* VECTOR LOAD GR FROM VR ELEMENT */
.macro VLGV gr, vr, disp, base="%r0", m
GR_NUM r1, \gr
GR_NUM b2, \base
VX_NUM v3, \vr
.word 0xE700 | (r1 << 4) | (v3&15)
.word (b2 << 12) | (\disp)
MRXBOPC \m, 0x21, v3
.endm
.macro VLGVB gr, vr, disp, base="%r0"
VLGV \gr, \vr, \disp, \base, 0
.endm
.macro VLGVH gr, vr, disp, base="%r0"
VLGV \gr, \vr, \disp, \base, 1
.endm
.macro VLGVF gr, vr, disp, base="%r0"
VLGV \gr, \vr, \disp, \base, 2
.endm
.macro VLGVG gr, vr, disp, base="%r0"
VLGV \gr, \vr, \disp, \base, 3
.endm
/* VECTOR LOAD MULTIPLE */
.macro VLM vfrom, vto, disp, base
VX_NUM v1, \vfrom
VX_NUM v3, \vto
GR_NUM b2, \base /* Base register */
.word 0xE700 | ((v1&15) << 4) | (v3&15)
.word (b2 << 12) | (\disp)
MRXBOPC 0, 0x36, v1, v3
.endm
/* VECTOR STORE MULTIPLE */
.macro VSTM vfrom, vto, disp, base
VX_NUM v1, \vfrom
VX_NUM v3, \vto
GR_NUM b2, \base /* Base register */
.word 0xE700 | ((v1&15) << 4) | (v3&15)
.word (b2 << 12) | (\disp)
MRXBOPC 0, 0x3E, v1, v3
.endm
/* VECTOR PERMUTE */
.macro VPERM vr1, vr2, vr3, vr4
VX_NUM v1, \vr1
VX_NUM v2, \vr2
VX_NUM v3, \vr3
VX_NUM v4, \vr4
.word 0xE700 | ((v1&15) << 4) | (v2&15)
.word ((v3&15) << 12)
MRXBOPC (v4&15), 0x8C, v1, v2, v3, v4
.endm
/* VECTOR UNPACK LOGICAL LOW */
.macro VUPLL vr1, vr2, m3
VX_NUM v1, \vr1
VX_NUM v2, \vr2
.word 0xE700 | ((v1&15) << 4) | (v2&15)
.word 0x0000
MRXBOPC \m3, 0xD4, v1, v2
.endm
.macro VUPLLB vr1, vr2
VUPLL \vr1, \vr2, 0
.endm
.macro VUPLLH vr1, vr2
VUPLL \vr1, \vr2, 1
.endm
.macro VUPLLF vr1, vr2
VUPLL \vr1, \vr2, 2
.endm
/* Vector integer instructions */
/* VECTOR AND */
.macro VN vr1, vr2, vr3
VX_NUM v1, \vr1
VX_NUM v2, \vr2
VX_NUM v3, \vr3
.word 0xE700 | ((v1&15) << 4) | (v2&15)
.word ((v3&15) << 12)
MRXBOPC 0, 0x68, v1, v2, v3
.endm
/* VECTOR EXCLUSIVE OR */
.macro VX vr1, vr2, vr3
VX_NUM v1, \vr1
VX_NUM v2, \vr2
VX_NUM v3, \vr3
.word 0xE700 | ((v1&15) << 4) | (v2&15)
.word ((v3&15) << 12)
MRXBOPC 0, 0x6D, v1, v2, v3
.endm
/* VECTOR GALOIS FIELD MULTIPLY SUM */
.macro VGFM vr1, vr2, vr3, m4
VX_NUM v1, \vr1
VX_NUM v2, \vr2
VX_NUM v3, \vr3
.word 0xE700 | ((v1&15) << 4) | (v2&15)
.word ((v3&15) << 12)
MRXBOPC \m4, 0xB4, v1, v2, v3
.endm
.macro VGFMB vr1, vr2, vr3
VGFM \vr1, \vr2, \vr3, 0
.endm
.macro VGFMH vr1, vr2, vr3
VGFM \vr1, \vr2, \vr3, 1
.endm
.macro VGFMF vr1, vr2, vr3
VGFM \vr1, \vr2, \vr3, 2
.endm
.macro VGFMG vr1, vr2, vr3
VGFM \vr1, \vr2, \vr3, 3
.endm
/* VECTOR GALOIS FIELD MULTIPLY SUM AND ACCUMULATE */
.macro VGFMA vr1, vr2, vr3, vr4, m5
VX_NUM v1, \vr1
VX_NUM v2, \vr2
VX_NUM v3, \vr3
VX_NUM v4, \vr4
.word 0xE700 | ((v1&15) << 4) | (v2&15)
.word ((v3&15) << 12) | (\m5 << 8)
MRXBOPC (v4&15), 0xBC, v1, v2, v3, v4
.endm
.macro VGFMAB vr1, vr2, vr3, vr4
VGFMA \vr1, \vr2, \vr3, \vr4, 0
.endm
.macro VGFMAH vr1, vr2, vr3, vr4
VGFMA \vr1, \vr2, \vr3, \vr4, 1
.endm
.macro VGFMAF vr1, vr2, vr3, vr4
VGFMA \vr1, \vr2, \vr3, \vr4, 2
.endm
.macro VGFMAG vr1, vr2, vr3, vr4
VGFMA \vr1, \vr2, \vr3, \vr4, 3
.endm
/* VECTOR SHIFT RIGHT LOGICAL BY BYTE */
.macro VSRLB vr1, vr2, vr3
VX_NUM v1, \vr1
VX_NUM v2, \vr2
VX_NUM v3, \vr3
.word 0xE700 | ((v1&15) << 4) | (v2&15)
.word ((v3&15) << 12)
MRXBOPC 0, 0x7D, v1, v2, v3
.endm
/* VECTOR REPLICATE IMMEDIATE */
.macro VREPI vr1, imm2, m3
VX_NUM v1, \vr1
.word 0xE700 | ((v1&15) << 4)
.word \imm2
MRXBOPC \m3, 0x45, v1
.endm
.macro VREPIB vr1, imm2
VREPI \vr1, \imm2, 0
.endm
.macro VREPIH vr1, imm2
VREPI \vr1, \imm2, 1
.endm
.macro VREPIF vr1, imm2
VREPI \vr1, \imm2, 2
.endm
.macro VREPIG vr1, imm2
VREP \vr1, \imm2, 3
.endm
/* VECTOR ADD */
.macro VA vr1, vr2, vr3, m4
VX_NUM v1, \vr1
VX_NUM v2, \vr2
VX_NUM v3, \vr3
.word 0xE700 | ((v1&15) << 4) | (v2&15)
.word ((v3&15) << 12)
MRXBOPC \m4, 0xF3, v1, v2, v3
.endm
.macro VAB vr1, vr2, vr3
VA \vr1, \vr2, \vr3, 0
.endm
.macro VAH vr1, vr2, vr3
VA \vr1, \vr2, \vr3, 1
.endm
.macro VAF vr1, vr2, vr3
VA \vr1, \vr2, \vr3, 2
.endm
.macro VAG vr1, vr2, vr3
VA \vr1, \vr2, \vr3, 3
.endm
.macro VAQ vr1, vr2, vr3
VA \vr1, \vr2, \vr3, 4
.endm
/* VECTOR ELEMENT SHIFT RIGHT ARITHMETIC */
.macro VESRAV vr1, vr2, vr3, m4
VX_NUM v1, \vr1
VX_NUM v2, \vr2
VX_NUM v3, \vr3
.word 0xE700 | ((v1&15) << 4) | (v2&15)
.word ((v3&15) << 12)
MRXBOPC \m4, 0x7A, v1, v2, v3
.endm
.macro VESRAVB vr1, vr2, vr3
VESRAV \vr1, \vr2, \vr3, 0
.endm
.macro VESRAVH vr1, vr2, vr3
VESRAV \vr1, \vr2, \vr3, 1
.endm
.macro VESRAVF vr1, vr2, vr3
VESRAV \vr1, \vr2, \vr3, 2
.endm
.macro VESRAVG vr1, vr2, vr3
VESRAV \vr1, \vr2, \vr3, 3
.endm
#endif /* __ASSEMBLY__ */
#endif /* __ASM_S390_VX_INSN_H */
Reference in New Issue View Git Blame Copy Permalink