Michael Pratt

This is likely a bug internal to pprof, so I've filed https://github.com/google/pprof/issues/720.

Plain text of the HTML from [Pprof listing.zip](https://github.com/golang/go/files/8368908/Pprof.listing.zip) to see L268 referred above without opening the file: ``` image/png.filter C:\Program Files\Go\src\image\png\writer.go Total: 9.41s 13.07s (flat, cum) 74.56% 202 . ....

I notice that there are separate implementations of computing flat and cum counts in list vs weblist. If I understand correctly, the values in `list` come from https://github.com/google/pprof/blob/main/internal/report/source.go#L925, and `weblist`...

https://github.com/golang/go/issues/55257#issuecomment-1252358190 is #54906, closing. Watchflakes will reopen with new default matching crashes.

When the program is stuck, could you attach with gdb and see where all of the threads are (`thread apply all backtrace`)?

1.14 added asynchronous preemption. Could you try setting the `GODEBUG=asyncpreemptoff=1` environment variable on a later version to see if it has the same behavior as 1.13?

@mvdan The ability to collect profiles from a PGO-optimized binary is actually a core requirement in our design because it simplifies profile collection significantly (you can collect directly from production...

@josharian Having a list of optimization directives (or pragmas in the source) is something we've thought about, and definitely has pros and cons. I certainly think we want PGO to...

@klauspost > **Platforms** > > This will only cover platform independent code. Much optimized code has separate code paths for different platforms. As proposed I don't see any reasonable way...

Thanks for the clarification, that makes sense. Having an optimization list doesn't solve the problem of determining if a new profile has "good" results, but it is at least easier...