News for hackers. Stuff that matters.

A comment on Frame pointers vs. DWARF – my verdict. Parent comment

Re:Frame pointers vs. DWARF – my verdict by DaveFlater on Wednesday February 15, 2023 @10:53AM (#34805455)
AFAIK, the optimization undermined the only use case that -fno-omit-frame-pointer actually had on x86_64. Is there a real use case that benefitted from allowing the frame pointer push to wander? Why why why
[ Parent | on HN ]
Re:Frame pointers vs. DWARF – my verdict by sitkack on Wednesday February 15, 2023 @05:31PM (#34811749)
That is too bad, it does mention `-fno-schedule-insns2` in the comments. I'll have to see how well that works in practice.

Having high quality low cost stack straces is important for continuous profiling. ref Knuth.

[ Parent | on HN ]