cybercyst/go
1
0
Fork 0
mirror of https://github.com/golang/go.git synced 2025-05-18 05:44:35 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

runtime: optimise gcWriteBarrier for riscv64

Browse Source 
Avoid saving unnecessary registers in gcWriteBarrier on riscv64, which also
removes references to X4 and X27 (TP and g registers).

Change-Id: I2854161dcdf0c6047a45347165371827dcf8c1cf
Reviewed-on: https://go-review.googlesource.com/c/go/+/265518
Trust: Joel Sing <joel@sing.id.au>
Reviewed-by: Cherry Zhang <cherryyz@google.com>
Run-TryBot: Cherry Zhang <cherryyz@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
This commit is contained in:
Joel Sing 2020-10-27 23:30:14 +11:00
parent 15131caeaa
commit ae82ee4016
1 changed files with 59 additions and 75 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

134
src/runtime/asm_riscv64.s
View File

@ -488,10 +488,10 @@ TEXT ·checkASM(SB),NOSPLIT,$0-1
// The act of CALLing gcWriteBarrier will clobber RA (LR). // The act of CALLing gcWriteBarrier will clobber RA (LR).
// It does not clobber any other general-purpose registers, // It does not clobber any other general-purpose registers,
// but may clobber others (e.g., floating point registers). // but may clobber others (e.g., floating point registers).
TEXT runtime·gcWriteBarrier(SB),NOSPLIT,$296 TEXT runtime·gcWriteBarrier(SB),NOSPLIT,$216
// Save the registers clobbered by the fast path. // Save the registers clobbered by the fast path.
MOV A0, 280(X2) MOV A0, 25*8(X2)
MOV A1, 288(X2) MOV A1, 26*8(X2)
MOV g_m(g), A0 MOV g_m(g), A0
MOV m_p(A0), A0 MOV m_p(A0), A0
MOV (p_wbBuf+wbBuf_next)(A0), A1 MOV (p_wbBuf+wbBuf_next)(A0), A1
@ -507,8 +507,8 @@ TEXT runtime·gcWriteBarrier(SB),NOSPLIT,$296
// Is the buffer full? // Is the buffer full?
BEQ A1, T6, flush BEQ A1, T6, flush
ret: ret:
MOV 280(X2), A0 MOV 25*8(X2), A0
MOV 288(X2), A1 MOV 26*8(X2), A1
// Do the write. // Do the write.
MOV T1, (T0) MOV T1, (T0)
RET RET
@ -516,84 +516,68 @@ ret:
flush: flush:
// Save all general purpose registers since these could be // Save all general purpose registers since these could be
// clobbered by wbBufFlush and were not saved by the caller. // clobbered by wbBufFlush and were not saved by the caller.
MOV T0, 8(X2) // Also first argument to wbBufFlush MOV T0, 1*8(X2) // Also first argument to wbBufFlush
MOV T1, 16(X2) // Also second argument to wbBufFlush MOV T1, 2*8(X2) // Also second argument to wbBufFlush
// X0 is zero register
// TODO: Optimise // X1 is LR, saved by prologue
// X5 already saved (T0) // X2 is SP
// X6 already saved (T1) MOV X3, 3*8(X2)
// X4 is TP
// X5 is first arg to wbBufFlush (T0)
// X6 is second arg to wbBufFlush (T1)
MOV X7, 4*8(X2)
MOV X8, 5*8(X2)
MOV X9, 6*8(X2)
// X10 already saved (A0) // X10 already saved (A0)
// X11 already saved (A1) // X11 already saved (A1)
MOV X12, 7*8(X2)
MOV X13, 8*8(X2)
MOV X14, 9*8(X2)
MOV X15, 10*8(X2)
MOV X16, 11*8(X2)
MOV X17, 12*8(X2)
MOV X18, 13*8(X2)
MOV X19, 14*8(X2)
MOV X20, 15*8(X2)
MOV X21, 16*8(X2)
MOV X22, 17*8(X2)
MOV X23, 18*8(X2)
MOV X24, 19*8(X2)
MOV X25, 20*8(X2)
MOV X26, 21*8(X2)
// X27 is g. // X27 is g.
MOV X28, 22*8(X2)
MOV X29, 23*8(X2)
MOV X30, 24*8(X2)
// X31 is tmp register. // X31 is tmp register.
MOV X0, 24(X2)
MOV X1, 32(X2)
MOV X2, 40(X2)
MOV X3, 48(X2)
MOV X4, 56(X2)
MOV X5, 64(X2)
MOV X6, 72(X2)
MOV X7, 80(X2)
MOV X8, 88(X2)
MOV X9, 96(X2)
MOV X10, 104(X2)
MOV X11, 112(X2)
MOV X12, 120(X2)
MOV X13, 128(X2)
MOV X14, 136(X2)
MOV X15, 144(X2)
MOV X16, 152(X2)
MOV X17, 160(X2)
MOV X18, 168(X2)
MOV X19, 176(X2)
MOV X20, 184(X2)
MOV X21, 192(X2)
MOV X22, 200(X2)
MOV X23, 208(X2)
MOV X24, 216(X2)
MOV X25, 224(X2)
MOV X26, 232(X2)
MOV X27, 240(X2)
MOV X28, 248(X2)
MOV X29, 256(X2)
MOV X30, 264(X2)
MOV X31, 272(X2)
// This takes arguments T0 and T1. // This takes arguments T0 and T1.
CALL runtime·wbBufFlush(SB) CALL runtime·wbBufFlush(SB)
MOV 24(X2), X0 MOV 1*8(X2), T0
MOV 32(X2), X1 MOV 2*8(X2), T1
MOV 40(X2), X2 MOV 3*8(X2), X3
MOV 48(X2), X3 MOV 4*8(X2), X7
MOV 56(X2), X4 MOV 5*8(X2), X8
MOV 64(X2), X5 MOV 6*8(X2), X9
MOV 72(X2), X6 MOV 7*8(X2), X12
MOV 80(X2), X7 MOV 8*8(X2), X13
MOV 88(X2), X8 MOV 9*8(X2), X14
MOV 96(X2), X9 MOV 10*8(X2), X15
MOV 104(X2), X10 MOV 11*8(X2), X16
MOV 112(X2), X11 MOV 12*8(X2), X17
MOV 120(X2), X12 MOV 13*8(X2), X18
MOV 128(X2), X13 MOV 14*8(X2), X19
MOV 136(X2), X14 MOV 15*8(X2), X20
MOV 144(X2), X15 MOV 16*8(X2), X21
MOV 152(X2), X16 MOV 17*8(X2), X22
MOV 160(X2), X17 MOV 18*8(X2), X23
MOV 168(X2), X18 MOV 19*8(X2), X24
MOV 176(X2), X19 MOV 20*8(X2), X25
MOV 184(X2), X20 MOV 21*8(X2), X26
MOV 192(X2), X21 MOV 22*8(X2), X28
MOV 200(X2), X22 MOV 23*8(X2), X29
MOV 208(X2), X23 MOV 24*8(X2), X30
MOV 216(X2), X24
MOV 224(X2), X25
MOV 232(X2), X26
MOV 240(X2), X27
MOV 248(X2), X28
MOV 256(X2), X29
MOV 264(X2), X30
MOV 272(X2), X31
JMP ret JMP ret