[PATCH] D58848: [DebugInfo] follow up for "add SectionedAddress to DebugInfo interfaces"

[Alexey Lapshin via Phabricator via llvm-commits]

avl added a comment.

In case Undef specified then there could be several sections with matching address ranges. 
It seems to me that DWARFDebugLineTable improper place to make such a decision - which of them should be taken as a result.
This in fact high level decision, not low level.

Previously that decision was made just by accident. Property of sorting algorithm defined which concrete section would be selected. 
I think that this decision should be unambiguous and currently for the same pair {Address, SectionIndex} there always would be returned the same result.

If DWARFDebugLineTable would be changed so that for Undef value there would be done search by one criteria then for overlapping address ranges there could be different results depending on concrete call to DWARFDebugLineTable. To prevent such random section selection there was implemented getModuleSectionIndexForAddress().

Do you think we need to omit that high level decision and allow random selection for sections with overlapping address ranges ?


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D58848/new/

https://reviews.llvm.org/D58848