[PATCH] Entry point function should never be dead-stripped.

Rui Ueyama ruiu at google.com
Mon Aug 26 20:05:57 PDT 2013


  I'm not convinced that this is not a clean way, as if we have to inevitably call addDeadStripRoot() after setEntrySymbolName(), separation of two is error prone and redundant. However, that is not an important point of this patch, so I think I don't want to push this view. I modified the patch not to implicitly add a symbol name to _deadStripRoot. PTAL.

http://llvm-reviews.chandlerc.com/D1524



More information about the llvm-commits mailing list