I am as surprised as you. Perhaps it's just non-determinism somewhere in the x86 code. The regressions passed yesterday, and after you pushed the patch, they changed ever so slightly.
Andreas On 10/7/13 9:57 PM, "Andreas Sandberg" <[email protected]> wrote: >I'll have a look tomorrow, but I really don't think it's because of this >changeset. The code in question shouldn't even be executed when running >the regressions since all listeners are disabled (i.e., no sockets are >created and registered with the PollQueue). > >//Andreas > >On 10/07/2013 07:35 PM, Andreas Hansson wrote: >> This changeset affects the x86 full system (pc-timing/o3) regressions >> (+-1). Andreas, could you please dig into this and confirm whether this >>is >> expected (and bump the stats)? >> >> Thanks, >> >> Andreas >> >> >> On 07/10/2013 15:05, "Andreas Sandberg" <[email protected]> wrote: >> >>> changeset 2a7c8ec11cff in /z/repo/gem5 >>> details: http://repo.gem5.org/gem5?cmd=changeset;node=2a7c8ec11cff >>> description: >>> base: Fix a potential race in PollQueue::setupAsyncIO >>> >>> There is a potential race between enabling asynchronous IO and >>> selecting the target for the SIGIO signal. This changeset move >>>the >>> F_SETOWN call to before the F_SETFL call that enables SIGIO >>> delivery. This ensures that signals are always sent to the >>>correct >>> process. >>> >>> diffstat: >>> >>> src/base/pollevent.cc | 6 +++--- >>> 1 files changed, 3 insertions(+), 3 deletions(-) >>> >>> diffs (20 lines): >>> >>> diff -r e672a39fd426 -r 2a7c8ec11cff src/base/pollevent.cc >>> --- a/src/base/pollevent.cc Thu Oct 03 11:00:18 2013 +0200 >>> +++ b/src/base/pollevent.cc Mon Oct 07 16:03:15 2013 +0200 >>> @@ -213,13 +213,13 @@ >>> else >>> flags &= ~(FASYNC); >>> >>> - if (fcntl(fd, F_SETFL, flags) == -1) >>> - panic("Could not set up async IO"); >>> - >>> if (set) { >>> if (fcntl(fd, F_SETOWN, getpid()) == -1) >>> panic("Could not set up async IO"); >>> } >>> + >>> + if (fcntl(fd, F_SETFL, flags) == -1) >>> + panic("Could not set up async IO"); >>> } >>> >>> void >>> _______________________________________________ >>> gem5-dev mailing list >>> [email protected] >>> http://m5sim.org/mailman/listinfo/gem5-dev >>> >> >> >> -- IMPORTANT NOTICE: The contents of this email and any attachments are >>confidential and may also be privileged. If you are not the intended >>recipient, please notify the sender immediately and do not disclose the >>contents to any other person, use it for any purpose, or store or copy >>the information in any medium. Thank you. >> >> ARM Limited, Registered office 110 Fulbourn Road, Cambridge CB1 9NJ, >>Registered in England & Wales, Company No: 2557590 >> ARM Holdings plc, Registered office 110 Fulbourn Road, Cambridge CB1 >>9NJ, Registered in England & Wales, Company No: 2548782 >> >> _______________________________________________ >> gem5-dev mailing list >> [email protected] >> http://m5sim.org/mailman/listinfo/gem5-dev >> > >_______________________________________________ >gem5-dev mailing list >[email protected] >http://m5sim.org/mailman/listinfo/gem5-dev > -- IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not disclose the contents to any other person, use it for any purpose, or store or copy the information in any medium. Thank you. ARM Limited, Registered office 110 Fulbourn Road, Cambridge CB1 9NJ, Registered in England & Wales, Company No: 2557590 ARM Holdings plc, Registered office 110 Fulbourn Road, Cambridge CB1 9NJ, Registered in England & Wales, Company No: 2548782 _______________________________________________ gem5-dev mailing list [email protected] http://m5sim.org/mailman/listinfo/gem5-dev
