Messages in this thread | | | Date | Tue, 30 Apr 2013 18:30:32 +0200 | From | Ingo Molnar <> | Subject | Re: [GIT PULL] scheduler changes for v3.10 |
| |
* Linus Torvalds <torvalds@linux-foundation.org> wrote:
> On Tue, Apr 30, 2013 at 9:23 AM, Ingo Molnar <mingo@kernel.org> wrote: > > > > Linus, would you like me to revert d9a3c9823a2e and re-send the pull > > request? > > I took the pull request, but I'd like to see the non-64-bit divide > version before the merge window ends. [...]
Yeah, wanted to apply it later today or tomorrow, and push it to you this week - well before the end of the merge window.
> [...] And with no more div_rem users I think we should at least > double-check that the new div64_u64() (that uses the new div64_u64_rem) > isn't noticeably slower than the old one. If it is, we should revert > that new (and then unused) div64 code too.
Correct, this is what Stanislaw's series from earlier today does:
[PATCH -tip 1/4] sched: Avoid cputime scaling overflow [PATCH -tip 2/4] sched: Do not account bogus utime [PATCH -tip 3/4] sched: Avoid prev->stime underflow [PATCH -tip 4/4] Revert "math64: New div64_u64_rem helper"
so it gets rid of the div64_u64_rem() bits as well.
Thanks,
Ingo
| |