Messages in this thread | | | Date | Tue, 22 Mar 2022 12:07:13 +0100 | From | Ingo Molnar <> | Subject | Re: [GIT PULL] scheduler updates for v5.18 |
| |
* Huang, Ying <ying.huang@intel.com> wrote:
> Hi, Ingo, > > Ingo Molnar <mingo@kernel.org> writes: > > > * Ingo Molnar <mingo@kernel.org> wrote: > > > >> > >> * Linus Torvalds <torvalds@linux-foundation.org> wrote: > >> > >> > On Mon, Mar 21, 2022 at 6:13 AM Qian Cai <quic_qiancai@quicinc.com> wrote: > >> > > > >> > > On Mon, Mar 21, 2022 at 11:54:09AM +0100, Ingo Molnar wrote: > >> > > > Huang Ying (3): > >> > > > sched/numa-balancing: Move some document to make it consistent with the code > >> > > > sched/numa: Fix NUMA topology for systems with CPU-less nodes > >> > > > sched/numa: Avoid migrating task to CPU-less node > >> > > > >> > > Linus, I don't think you want to merge this as-is. This will introduce a > >> > > kernel crash on arm64 NUMA as mentioned in this thread, > >> > > >> > Ok, dropped from my queue. Thanks, > >> > >> I've reverted the broken commit & will send another pull request after > >> some testing. Sorry about that! > > > > Ended up using the fix below instead - it's tested already on the affected > > system. > > > > Thanks a lot for your help! > > Is it a general rule to send a fixing patch for the bug of a patch > merged by tip tree? Or a new version of the patch is acceptable too?
Depends on how recently it got merged. This one was merged a few weeks ago, and the fix patch is simple enough, so I went for that approach.
Thanks,
Ingo
| |