[llvm-commits] [llvm] r75806 - in /llvm/trunk: include/llvm/CodeGen/AsmPrinter.h utils/TableGen/AsmWriterEmitter.cpp
Sanjiv Gupta
sanjiv.gupta at microchip.com
Fri Jul 17 11:58:15 PDT 2009
David A. Greene wrote:
> On Friday 17 July 2009 12:26, Sanjiv Gupta wrote:
>
>>> // Do the post instruction processing and print the final newline
>>> - UniqueOperandCommands.push_back(" postInstructionAction(*MI);\n
>>> O << \"\\n\";\n return true;\n"); +
>>> UniqueOperandCommands.push_back(" EmitComments(*MI);\n O <<
>>> \"\\n\";\n return true;\n");
>>>
>> This "return true" here is too early, it is missing the last fragment in
>> case of MOVF_INSN in PIC16. I think this is broken since your commit
>> 75490 itself.
>>
>
> Hmm. The original code was:
>
> 29196 lattner // For the first operand check, add a default value for
> instructions with
> 29196 lattner // just opcode strings to use.
> 29183 lattner if (isFirst) {
> 29196 lattner UniqueOperandCommands.push_back(" return true;\n");
> 29183 lattner isFirst = false;
> 29183 lattner }
>
> The current code is:
>
> 29196 lattner // For the first operand check, add a default value for
> instructions with
> 29196 lattner // just opcode strings to use.
> 29183 lattner if (isFirst) {
> 75490 greened // Do the post instruction processing and print the
> final newline
> 75806 greened UniqueOperandCommands.push_back("
> EmitComments(*MI);\n O << \"\\n\";\n return true;\n");
> 29183 lattner isFirst = false;
> 29183 lattner }
>
> So the "return true" was always there. Is it possible this is a bug that's
> been in AsmWriterEmitter.cpp for a long time?
>
>
No. Try the .bc file with 75489, it works.
with 75490 it doesn't.
- Sanjiv
> -Dave
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>
More information about the llvm-commits
mailing list