[03:59] <duflu> RAOF: Can you check and confirm Mesa is fixable, and not intrinsically doing the right thing already?... https://bugs.launchpad.net/ubuntu/+source/mesa/+bug/1480755/comments/5
[04:02] <RAOF> Checking...
[05:48] <RAOF> Oh, damn.
[05:48] <RAOF> *Of course* it passes under ThreadSanitizer and normally but not under valgrind.
[05:50] <RAOF> *Of course* it passes under ThreadSanitizer and normally but not under valgrind.
[05:50] <RAOF> ...but only when tracking fds?
[05:50] <RAOF> No?
[05:51] <RAOF> Oh, right. --trace-children
[06:24] <duflu> RAOF: Yeah that's a gotcha from binary wrapping. I sent out an email but we've all forgotten now.
[06:24] <duflu> RAOF: P.S. I'm using this tag for the pixel format related issues (and more?): https://bugs.launchpad.net/ubuntu/+source/mesa/+bugs?field.tag=egl-platform-mir
[06:50] <RAOF> Hurray for valgrind accidentally finding race conditions.
[06:51] <RAOF> Also, it'd be *really* nice if we could move into our lambdas.
[07:04] <duflu> RAOF: shared/unique_ptr should do
[07:05] <RAOF> Well, unique_ptr clearly doesn't, because you can't move it into a lambda :)
[07:06] <duflu> RAOF: Parameter by value?
[07:06] <duflu> Or capture
[07:06] <RAOF> Can't by value; it's not copyable.
[07:08] <RAOF> I can emulate what I want with a shared_ptr, but it'd be cleaner without.
[07:09] <duflu> Sweet. In other news, switching my new monitor to DP 1.2 mode Mir now sees 4 DP outputs (multi-stream)
[07:09] <RAOF> Superb
[07:10] <duflu> Still max 3 simultaneous outputs??
[07:10] <duflu> Not sure about that
[07:14] <duflu> Hmm, or was it just a kernel upgrade and I haven't looked in a while
[07:14] <duflu> ?
[07:28] <duflu> RAOF: Yeah confirmed putting the monitor in DP1.2 mode is what gives me twice the outputs available. Unfortunately DP1.2 mode is not understood by the Linux framebuffer and it (along with Plymouth) is in a low-res BIOS mode
[07:28] <duflu> X and Mir both handle it well tho
[07:29] <duflu> I would have thought kernel 4.1 would be more current than that
[07:29] <RAOF> Time for kmscon!
[07:29] <duflu> !
[07:37] <RAOF> I am apparently going to need to fix each and every race in Mir before I can land this thing...
[09:36] <duflu> anpok_...?
[09:36] <duflu> anpok_ ?
[09:39] <anpok_> duflu: tuesday!
[14:46] <anpok> AlbertA: should we land your fix or cherry pick christophers?
[14:48] <AlbertA> anpok: heh I'm about to push the cherry picked solution from Chris
[14:48] <anpok> +1
[19:08] <camako> https://code.launchpad.net/~cemil-azizoglu/mir/add-packaging-for-mir-on-x/+merge/266915/comments/670098
[19:09] <camako> vogons, ^^ known failure?
[19:10] <kdub> not known to me at least
[19:11] <anpok> camako: a fix is about to land
[19:11] <AlbertA> camako: yes: https://code.launchpad.net/~albaguirre/mir/fix-1481034/+merge/266781
[19:12] <anpok> this the shutdown race during the synthetic exception in the input thread
[19:12] <camako> ok thanks... I remember seeing it on other MPs
[23:24] <RAOF> GAH! Why does http://bazaar.launchpad.net/~albaguirre/mir/fix-1481034/revision/2808 compile?
[23:26] <RAOF> Alternatively, why did my attempts to do the same fail to compile?
[23:47] <RAOF> Oh! By moving the std::shared_ptr<promise> rather than the std::promise.
[23:47] <RAOF> Yeah, that'd work ☺