The patch uses a common method to determine the appropriate form for the value of the attribute.
Details
Diff Detail
Event Timeline
Looks good
llvm/test/DebugInfo/X86/DW_AT_location-reference.ll | ||
---|---|---|
44 | Should this be dumped in a different format/more digits to match the number of bytes in the field? |
llvm/test/DebugInfo/X86/DW_AT_location-reference.ll | ||
---|---|---|
44 | That is how DWARFLocationTable::dumpLocationList() dumps the offset. At that moment, the information about the form of the value is lost. |
- Updated the test because the patch is being moved before fixing emitters for pre-v5 name lookup tables.
llvm/test/DebugInfo/X86/DW_AT_location-reference.ll | ||
---|---|---|
44 | Fair enough, just a thought to consider/keep in mind at some point. Consistency in dumping seems important if we're going to use longer fields to indicate longer encodings. Only using that strategy sometimes seems like it could be more confusing than helpful. |
Should this be dumped in a different format/more digits to match the number of bytes in the field?