abel-bernabeu created this revision.
Herald added subscribers: jobnoorman, luke, VincentWu, vkmr, frasercrmck,
luismarques, apazos, sameer.abuasal, s.egerton, Jim, benna, psnobl, jocewei,
PkmX, the_o, brucehoult, MartinMosbeck, rogfer01, edward-jones, zzheng, jrtc27,
shiva0217, kito-cheng, niosHD, sabuasal, simoncook, johnrusso, rbar, asb,
hiraditya, kristof.beyls, arichardson.
Herald added a project: All.
abel-bernabeu requested review of this revision.
Herald added subscribers: llvm-commits, cfe-commits, wangpc, eopXD, MaskRay.
Herald added projects: clang, LLVM.
It has been reported by one of Esperanto's customers that slash-start
comments ("/*") within inline assembly were only allowed before the
first instruction operand or at the end of the lines. Those comments
were, however, not allowed when interleaved within the operands.
An example follows:
unsigned long int dst;
__asm__ __volatile__(
"li /* this was fine */ %[dst], /* this was NOT fine */ 0x1234\n"
"add zero, %[dst], %[dst]\n"
: [ dst ] "=r"(dst)
:
:);
A code review of the top level parser (AsmParser class) showed that
when comments were placed before the instruction operand or at end of
a line, then they were gracefully handled irrespective of the backend.
When the comments were interleaved within the instruction operands it
was the backend's responsibility to handle the comments.
RISC-V's backend did not handle the comments in any way.
Beyond the obvious solution of explicitly handling the comments within
the RISC-V backend, another, easier to maintain, was suggested by Sergei
Barannikov in a Discourse discussion thread:
https://discourse.llvm.org/t/interleaving-several-c-style-comments-in-the-same-inline-assembly-line/71353/8
In summary, all backends, including the RISC-V's, should switch
from getLexer().Lex() to getParser().Lex() in their ParseInstruction
implementation.
The getLexer().Lex() approach relies on the user to explicitly handle
the comments, whereas the suggested getParser().Lex() alternive already
handles the comments in the same way as done for non-target-specific
assembly directives.
Here we just do the RISC-V work. Other backends should also do their own
review.
In addition to Sergei Barannikov, I would also like to thank David Spikett
from Arm's community for pointing out where to start looking within the
LLVM code base, and also the patch reviewers.
Repository:
rG LLVM Github Monorepo
https://reviews.llvm.org/D153150
Files:
clang/test/CodeGen/RISCV/riscv-inline-asm-gcc-commenting.c
llvm/lib/Target/RISCV/AsmParser/RISCVAsmParser.cpp
Index: llvm/lib/Target/RISCV/AsmParser/RISCVAsmParser.cpp
===================================================================
--- llvm/lib/Target/RISCV/AsmParser/RISCVAsmParser.cpp
+++ llvm/lib/Target/RISCV/AsmParser/RISCVAsmParser.cpp
@@ -1617,7 +1617,7 @@
Operands.push_back(RISCVOperand::createToken("(", FirstS));
SMLoc S = getLoc();
SMLoc E = SMLoc::getFromPointer(S.getPointer() + Name.size());
- getLexer().Lex();
+ getParser().Lex();
Operands.push_back(RISCVOperand::createReg(RegNo, S, E));
}
@@ -1978,11 +1978,11 @@
return MatchOperand_Success;
case AsmToken::Plus:
Opcode = MCBinaryExpr::Add;
- getLexer().Lex();
+ getParser().Lex();
break;
case AsmToken::Minus:
Opcode = MCBinaryExpr::Sub;
- getLexer().Lex();
+ getParser().Lex();
break;
}
@@ -2131,11 +2131,11 @@
MaskAgnostic))
return MatchOperand_NoMatch;
- getLexer().Lex();
+ getParser().Lex();
while (getLexer().is(AsmToken::Comma)) {
// Consume comma.
- getLexer().Lex();
+ getParser().Lex();
if (getLexer().isNot(AsmToken::Identifier))
break;
@@ -2146,7 +2146,7 @@
MaskAgnostic))
break;
- getLexer().Lex();
+ getParser().Lex();
}
if (getLexer().is(AsmToken::EndOfStatement) && State == VTypeState_Done) {
@@ -2186,7 +2186,7 @@
return MatchOperand_NoMatch;
SMLoc S = getLoc();
SMLoc E = SMLoc::getFromPointer(S.getPointer() + Name.size());
- getLexer().Lex();
+ getParser().Lex();
Operands.push_back(RISCVOperand::createReg(RegNo, S, E));
return MatchOperand_Success;
}
@@ -2202,7 +2202,7 @@
return MatchOperand_NoMatch;
SMLoc S = getLoc();
SMLoc E = SMLoc::getFromPointer(S.getPointer() + Name.size());
- getLexer().Lex();
+ getParser().Lex();
Operands.push_back(RISCVOperand::createReg(
RegNo, S, E, !getSTI().hasFeature(RISCV::FeatureStdExtF)));
return MatchOperand_Success;
@@ -2391,11 +2391,11 @@
Error(getLoc(), "register list must start from 'ra' or 'x1'");
return MatchOperand_ParseFail;
}
- getLexer().Lex();
+ getParser().Lex();
// parse case like ,s0
if (getLexer().is(AsmToken::Comma)) {
- getLexer().Lex();
+ getParser().Lex();
if (getLexer().isNot(AsmToken::Identifier)) {
Error(getLoc(), "invalid register");
return MatchOperand_ParseFail;
@@ -2410,12 +2410,12 @@
Error(getLoc(), "continuous register list must start from 's0' or 'x8'");
return MatchOperand_ParseFail;
}
- getLexer().Lex(); // eat reg
+ getParser().Lex(); // eat reg
}
// parse case like -s1
if (getLexer().is(AsmToken::Minus)) {
- getLexer().Lex();
+ getParser().Lex();
StringRef EndName = getLexer().getTok().getIdentifier();
// FIXME: the register mapping and checks of EABI is wrong
RegEnd = matchRegisterNameHelper(IsEABI, EndName);
@@ -2428,7 +2428,7 @@
"'x8-x9' pair");
return MatchOperand_ParseFail;
}
- getLexer().Lex();
+ getParser().Lex();
}
if (!IsEABI) {
@@ -2442,7 +2442,7 @@
}
// parse ', x18' for extra part
- getLexer().Lex();
+ getParser().Lex();
if (getLexer().isNot(AsmToken::Identifier)) {
Error(getLoc(), "invalid register");
return MatchOperand_ParseFail;
@@ -2453,11 +2453,11 @@
"must start from 'x18'");
return MatchOperand_ParseFail;
}
- getLexer().Lex();
+ getParser().Lex();
// parse '-x20' for extra part
if (getLexer().is(AsmToken::Minus)) {
- getLexer().Lex();
+ getParser().Lex();
if (getLexer().isNot(AsmToken::Identifier)) {
Error(getLoc(), "invalid register");
return MatchOperand_ParseFail;
@@ -2467,7 +2467,7 @@
Error(getLoc(), "invalid register");
return MatchOperand_ParseFail;
}
- getLexer().Lex();
+ getParser().Lex();
}
RegEnd = MatchRegisterName(EndName);
}
@@ -2497,7 +2497,7 @@
OperandMatchResultTy RISCVAsmParser::parseZcmpSpimm(OperandVector &Operands) {
if (getLexer().is(AsmToken::Minus))
- getLexer().Lex();
+ getParser().Lex();
SMLoc S = getLoc();
int64_t StackAdjustment = getLexer().getTok().getIntVal();
@@ -2508,7 +2508,7 @@
if (!RISCVZC::getSpimm(RlistVal, Spimm, StackAdjustment, isRV64(), IsEABI))
return MatchOperand_NoMatch;
Operands.push_back(RISCVOperand::createSpimm(Spimm << 4, S));
- getLexer().Lex();
+ getParser().Lex();
return MatchOperand_Success;
}
@@ -2576,7 +2576,7 @@
// Parse until end of statement, consuming commas between operands
while (getLexer().is(AsmToken::Comma)) {
// Consume comma token
- getLexer().Lex();
+ getParser().Lex();
// Parse next operand
if (parseOperand(Operands, Name))
Index: clang/test/CodeGen/RISCV/riscv-inline-asm-gcc-commenting.c
===================================================================
--- /dev/null
+++ clang/test/CodeGen/RISCV/riscv-inline-asm-gcc-commenting.c
@@ -0,0 +1,12 @@
+// RUN: %clang -fPIC --target=riscv64-unknown-elf -o - -S %s | FileCheck %s
+
+unsigned long int f() {
+ unsigned long int dst;
+ __asm__ ("addi /* this is fine */ %[dst], zero /* this is also fine */, /* this also */ 0 /* and last but not least */\n"
+ : [ dst ] "=r"(dst)
+ :
+ :);
+ return dst;
+}
+// CHECK: li a0, 0 # this is fine # this is also fine # this also # and last but not least
+// CHECK: ret
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
