cmd/compile: declare no output register for loong64 LoweredAtomic{And,Or}32 ops

The ICE seen on loong64 while compiling the `(*gcWork).tryStealSpan`
function was due to an `LoweredAtomicAnd32` op (inlined from the
`(pMask).clear` implementation) being incorrectly assigned an output
register while it shouldn't have. Because the op is of mem type, it has
needRegister() == false; hence in the shuffle phase of regalloc, its
bogus output register has no associated `orig` value recorded. The bug
was introduced in CL 482756, but only recently exposed by CL 696035.

Since the old-style atomic ops need no return value (and is even
documented so besides the loong64 ssa op definition), just fix the
register info for both.

While at it, add a note in the ssa op definition file about the
architectural necessity of resultNotInArgs for loong64 atomic ops,
because the practice is not seen in several other arches I have
checked.

Updates #75776

Change-Id: I087f51b8a2825d7b00fc3965b0afcc8b02cad277
Reviewed-on: https://go-review.googlesource.com/c/go/+/710475
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Reviewed-by: abner chenc <chenguoqi@loongson.cn>
Reviewed-by: Cherry Mui <cherryyz@google.com>
This commit is contained in:
WANG Xuerui 2025-10-09 15:52:02 +08:00 committed by abner chenc
parent 19a30ea3f2
commit 954fdcc51a
2 changed files with 9 additions and 8 deletions

View file

@ -143,6 +143,7 @@ func init() {
gp2load = regInfo{inputs: []regMask{gpspsbg, gpg}, outputs: []regMask{gp}}
gpstore = regInfo{inputs: []regMask{gpspsbg, gpg}}
gpstore2 = regInfo{inputs: []regMask{gpspsbg, gpg, gpg | rz}}
gpoldatom = regInfo{inputs: []regMask{gpspsbg, gpg}}
gpxchg = regInfo{inputs: []regMask{gpspsbg, gpg}, outputs: []regMask{gp}}
gpcas = regInfo{inputs: []regMask{gpspsbg, gpg, gpg}, outputs: []regMask{gp}}
preldreg = regInfo{inputs: []regMask{gpspg}}
@ -431,6 +432,12 @@ func init() {
faultOnNilArg1: true,
},
// Atomic operations.
//
// resultNotInArgs is needed by all ops lowering to LoongArch
// atomic memory access instructions, because these instructions
// are defined to require rd != rj && rd != rk per the ISA spec.
// atomic loads.
// load from arg0. arg1=mem.
// returns <value,memory> so they can be properly ordered with other loads.
@ -500,8 +507,8 @@ func init() {
// Atomic 32 bit AND/OR.
// *arg0 &= (|=) arg1. arg2=mem. returns nil.
{name: "LoweredAtomicAnd32", argLength: 3, reg: gpxchg, asm: "AMANDDBW", resultNotInArgs: true, faultOnNilArg0: true, hasSideEffects: true},
{name: "LoweredAtomicOr32", argLength: 3, reg: gpxchg, asm: "AMORDBW", resultNotInArgs: true, faultOnNilArg0: true, hasSideEffects: true},
{name: "LoweredAtomicAnd32", argLength: 3, reg: gpoldatom, asm: "AMANDDBW", resultNotInArgs: true, faultOnNilArg0: true, hasSideEffects: true},
{name: "LoweredAtomicOr32", argLength: 3, reg: gpoldatom, asm: "AMORDBW", resultNotInArgs: true, faultOnNilArg0: true, hasSideEffects: true},
// Atomic 32,64 bit AND/OR.
// *arg0 &= (|=) arg1. arg2=mem. returns <old content of *arg0, memory>. auxint must be zero.

View file

@ -26392,9 +26392,6 @@ var opcodeTable = [...]opInfo{
{1, 1073741816}, // R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 g R23 R24 R25 R26 R27 R28 R29 R31
{0, 4611686019501129724}, // SP R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 g R23 R24 R25 R26 R27 R28 R29 R31 SB
},
outputs: []outputInfo{
{0, 1071644664}, // R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R23 R24 R25 R26 R27 R28 R29 R31
},
},
},
{
@ -26409,9 +26406,6 @@ var opcodeTable = [...]opInfo{
{1, 1073741816}, // R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 g R23 R24 R25 R26 R27 R28 R29 R31
{0, 4611686019501129724}, // SP R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 g R23 R24 R25 R26 R27 R28 R29 R31 SB
},
outputs: []outputInfo{
{0, 1071644664}, // R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R23 R24 R25 R26 R27 R28 R29 R31
},
},
},
{