Hi Hans, On 26/08/15 20:31, Hans Petter Selasky wrote: > On 08/26/15 20:14, Julien Charbon wrote: >> Let say that if nobody screams until Friday 8/28, I will put back >> r284245 and revert this change_and_ I will make this case clear in the >> man page. > > If you can update the manual page about this special case for MPSAFE > callouts only I presume, then its totally fine.
Good idea, I am going to update the callout(9) man page to make clear that in mpsafe case if a callout is pending and currently being serviced callout_stop() returns 0 (fail). > Then I can update my projects/hps_head to follow that new change, > which now is a bit broken. You might also want to check existing > MPSAFE consumers in the kernel, if this API change makes any > difference. I checked all the mpsafe callouts that check callout_stop() return value (actually only a few are testing callout_stop() return value), and none relies on the old behavior (i.e. before r286880). > I was thinking if user-space TCP might be affected by this. CC'ing Adrian. Interesting. -- Julien
signature.asc
Description: OpenPGP digital signature