On Thu, Jun 28, 2012 at 12:24 PM, Liu Yuan <namei.unix at gmail.com> wrote: > On 06/25/2012 03:24 PM, Liu Yuan wrote: >> On 06/23/2012 09:40 PM, Yunkai Zhang wrote: >>> When EPOLLERR/EPOLLHUP events occur, other events(eg: EPOLLIN) may be >>> inclued at the same time, but we should priority to process them as the >>> connection was destroyed. >> >> Applied this series. >> > > This whole series need revisiting: > > 1) cause performance degradation, because it forces concurrent requests > from QEMU to be serialized Yes, this commit: 6575fdf332276939f006c1624359957558174b63 seems unnecessary. I'm going to revert it and do more testing. I'll give new patch later. > 2) Deadlock explained in patch 'sheep: take truns to monitor EPOLLIN and > EPOLLOUT events.' is questionable. I don't see a deadlock by code review. Agreed, it's my mistake. > > Thanks > Yuan > -- > sheepdog mailing list > sheepdog at lists.wpkg.org > http://lists.wpkg.org/mailman/listinfo/sheepdog -- Yunkai Zhang Work at Taobao |