[Bug middle-end/64928] [8/9/10/11 Regression] Inordinate cpu time and memory usage in "phase opt and generate" with -ftest-coverage -fprofile-arcs

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64928

--- Comment #31 from Richard Biener <rguenth at gcc dot gnu.org> ---
(In reply to lucier from comment #30)
> I'm coming back to this project.
> 
> I naively thought "Well, I don't need arc profiling, I'll just set
> -ftest-coverage without -fprofile-arcs" but it appears that I can't do that,
> the gcda files are generated by -fprofile-arcs.
> 
> It seems to me that test coverage could be implemented simply by
> instrumenting each basic block in an algorithm that's linear in the number
> of basic blocks.  Is it possible to do this?
> 
> Brad

I don't think the instrumentation itself is the problem - it's already
doing better than one counter per block.  It's simply that the large
source runs into multiple non-linearities in core pieces of the compiler
that cannot be turned off ...