I'm a bit lost. Is the problem that mmap_again is considered to be unused since the compiler has shown all gotos are unreachable?
@DavidEichmann will have a look at sorting this out in the coming weeks.
@sjakobi, what is your plan for this?
Mon, Oct 15
@dfeuer, what is the status of this? It's not clear that there was ever any performance testing of this done.
What is the plan for this, @shayan-najd? I agree that we will really need to find a way to break this up a bit if we want meaningful review.
The patch itself looks good.
@DavidEichmann, let's try to finish this up soon.
Gently pinging @AnthonySuper.
Quite useful; this confused me in the past as well!
Thanks for reviewing @ppk!
@tdammers, what is the status of this? It seems to be failing CI.
As another NixOS user, I agree that there should be some way to disable this warning.
This looks reasonable to @simonmar and me.
@harpocrates, you might consider splitting this into two differentials: one for the documentation and the other for the proposed addition.
Oh dear, I was a bit hasty in concluding that this was stale; it looks like I was wrong.
Bumping out of the review queue. This has been broken up into several pieces. See Trac #14880 for details.
@osa1, can you rebase this? I am wondering whether the haddock.base regression is real.
Looks reasonable enough.
@simonpj, when you get a chance it would be nice to have a second set of eyes on the note.
Bumping out of the review queue until acceptance.
Sun, Oct 14
Sat, Oct 13
As noted in Trac #15969 (comment 80) this approach is a bit of a compromise since we still end up with conservative demand analysis results. However, it's perhaps good enough for now.
Yes, looks reasonable.