[00:33] *** _franck__ has quit IRC (Ping timeout: 250 seconds) [02:25] *** pinchartl has quit IRC (Ping timeout: 244 seconds) [08:11] <Renault> hverkuil: I'm the GS1662 driver developer, when you want to plan the meeting? [08:11] <Renault> (hello all) [08:30] *** DaRock has quit IRC (Quit: DaRock) [09:00] <Renault> hverkuil: and, I'm in contact with another developper to add GS2972 which is very closed to GS1662 component [10:11] <hverkuil> Renault: I am very, very busy I'm afraid. I'd also like to involve nohous in the discussion and he's not online. I have to wait until I see him here. The annoying thing is that I don't have nohous' email, otherwise we can arrange a date and time for a meeting via email. [10:12] <hverkuil> Should you see him here before me, please ping him and ask for his email address. [10:12] <Renault> ok :) [10:12] <hverkuil> Realistically I won't have time this week. [10:13] <Renault> no problem, the next week I'm not here monday and friday (in case of the meeting will be planned the next week) [10:31] *** aballier has quit IRC (Quit: leaving) [10:44] <pinchartl> hverkuil: I assume you won't have time to discuss my patch that adds a few new functions then ? [10:48] <hverkuil> pinchartl: ping me Friday. [10:50] <pinchartl> ok thanks [10:51] <pinchartl> it's down to nitpicking now, it will be easier to handle it over e-mail [10:51] <pinchartl> sorry, I meant on IRC instead of over e-mail [13:13] <pinchartl> mchehab: ping [13:15] <mchehab> pong [13:16] <pinchartl> how are you doing ? [13:16] <mchehab> fine, and you? [13:16] <pinchartl> I'm doing good, back from vacation, and tempted to leave again :-D [13:16] <pinchartl> (I wish I could...) [13:17] <mchehab> :) [13:17] <pinchartl> I'm about to send you a pull request that includes two patches touching drivers/gpu/drm/rcar-du [13:17] * mchehab doesn't know the meaning of vacations... [13:17] <mchehab> :-D [13:17] <pinchartl> Dave has acked them, he's fine with you merging them through your tree [13:17] <pinchartl> I'll CC him on the pull request, is there anything else you'd like me to do ? [13:18] <mchehab> pinchartl: except if Linus sends a -rc8 or it is a serious bug fix, we're done for 4.7 merge window [13:19] <pinchartl> are we ? :-/ [13:19] <mchehab> yes [13:19] <pinchartl> those patches will generate conflict for v4.8 [13:19] <pinchartl> they've been out for some time now [13:20] <mchehab> why dave don't submit them? [13:20] <pinchartl> because they need to be submitted with corresponding V4L2 patches [13:20] <pinchartl> for the VSP1 driver [13:20] <mchehab> let's then apply those after the merge window [13:21] <pinchartl> yes but they will conflict [13:21] <mchehab> ? [13:21] <pinchartl> there will be other patches for v4.8 on the KMS side [13:21] <pinchartl> let me send the pull request, it will be easier to explain with the list of patches [13:24] <pinchartl> "[GIT PULL FOR v4.7] Renesas VSP updates" [13:24] <pinchartl> there's 13 patches [13:25] <pinchartl> the rcar-du-drm patches are number 11 and 12 [13:25] <pinchartl> they depend on previous vsp1 patches in the series [13:25] <mchehab> ok [13:25] <pinchartl> and patch 13 for the vsp1 driver further depend on them [13:25] <pinchartl> they're thus sandwiched between vsp1 patches [13:25] <pinchartl> that's why Dave agreed to have them merged through your tree [13:25] <pinchartl> they don't conflict with anything queued in Dave's tree for v4.7 [13:26] <pinchartl> but it won't be true anymore for v4.8 [13:26] <pinchartl> as people are working on patches for v4.8 that will be merged through Dave's tree and will touch the rcar-du-drm driver [13:26] <mchehab> well, nothing prevents putting those things on a topic branch [13:26] <pinchartl> so we have an opportunity now to merge the patches without conflicts [13:26] <mchehab> and let both gpu and media branches depend on it [13:27] <pinchartl> I'd need a topic branch from you that Dave would then merge in his tree, on top of which he would need to merge my patches, and then create a topic branch for you that you would merge back to merge the last patch in the series [13:27] <pinchartl> it's the back-and-forth that is messy [13:27] <pinchartl> if the dependency was in a single direction that's what I would have done [13:27] <pinchartl> why are we closed for v4.7 ? v4.6 hasn't been released yet [13:28] <mchehab> I mean, if we put this 13 patches on a topic branch, it won't be different than merging for 4.7 [13:28] <mchehab> both gpu and media topic branches would be based on such "common" topic branch [13:28] <pinchartl> it would need to be merged by Dave before the current work in progress to avoid conflicts there [13:29] <pinchartl> if the conflicting patches get merged in Dave's tree concurrently with the topic branch being merged in your tree we'll have conflicts when Linus will pull from you and Dave [13:31] <pinchartl> why are we closed for v4.7 ? [13:32] <mchehab> we close the week before rc7 [13:32] <mchehab> to allow the Kernel genitors to review and submit fixes [13:33] <mchehab> they made such request years ago, and we're doing such workflow for a while [13:33] <pinchartl> is that documented somewhere ? [13:33] <mchehab> there were some e-mails [13:34] <pinchartl> any chance for an exception ? [13:34] <pinchartl> it's driver code only ;-) [13:34] <mchehab> still I use to remember people that submit late pull requests via IRC [13:34] <pinchartl> I would have submitted it last week but I was away [13:34] <pinchartl> the patches were posted before that and I was waiting for review [13:35] <mchehab> last week I warned hverkuil that two of his pull requests would likely be delayed [13:35] <mchehab> due to that [13:35] <mchehab> on IRC [13:35] <mchehab> one is pure driver changes [13:36] <pinchartl> I'm fine with "likely" if it's not "certainly" ;-) [13:39] <hverkuil> mchehab: did the rcar-vin rewrite make the cut for 4.7? The others (cec framework, mediatek) were late so I'm not surprised they didn't make it. [13:45] <pinchartl> it hasn't been merged yet so I assume it's delayed too [13:47] <pinchartl> mchehab: how will you synchronize with Dave for this if it can't be merged for v4.7 ? [14:41] <mchehab> hverkuil: no, sorry. I got sick last week and was unable to handle it [14:42] <mchehab> I barely finished handling the individual patches for existing drivers [15:50] <pinchartl> mchehab: are you feeling better ? [15:50] <mchehab> still recovering [15:50] <mchehab> under medicine [15:53] <mchehab> nothing serious... just a severe cold [16:07] <mchehab> sailus: I think one of the VB2 fixes you wrote caused a regression for DVB [16:07] *** benjiG has left [16:08] <mchehab> some change at __verify_planes_array [16:08] <mchehab> at videobuf2-v4l2.c [16:10] <mchehab> maybe this patch [16:10] <mchehab> commit 2c1f6951a8a82e6de0d82b1158b5e493fc6c54ab [16:10] <mchehab> Author: Sakari Ailus <sakari.ailus@linux.intel.com> [16:10] <mchehab> Date: Sun Apr 3 16:31:03 2016 -0300 [16:10] <mchehab> [media] videobuf2-v4l2: Verify planes array in buffer dequeueing [16:10] <mchehab> [16:10] <mchehab> I'll revert and see if it solves the issue [16:22] *** egavinc has quit IRC (Quit: Leaving) [16:29] <mchehab> sailus: indeed it caused regressions at DVB [16:29] <mchehab> just sent a revert patch [16:34] *** stoth has quit IRC (Quit: stoth) [18:02] *** awalls has left [21:08] <sailus> mchehab: Ping? [21:13] <mchehab> pong [21:14] <nohous> gleep [21:15] <sailus> I missed that pb could actually be NULL in some cases. [21:15] <sailus> File I/O apparently as well besides DVB. [21:15] <sailus> So avoiding the check seems like the right thing to do. [21:23] <mchehab> sailus: for now, I'm just reverting it... it caused breakages on stable kernels [21:23] <mchehab> please prepare a next version and test it for all cases before re-submitting [21:26] <mchehab> I discovered the bug by accident, just because I needed to test a DVB device and decided to rebase my test tree [21:26] <mchehab> but people were already complaining about it at LKML [21:28] <sailus> I'll test file I/O but I have no convenient access to DVB devices. [21:30] <sailus> I entirely missed DVB was using videobuf2-v4l2. [21:31] <headless> pinchartl: do you remember about the TCON patches? [21:31] <headless> oh, it's 0:30 again :-) [23:37] *** kaspter has quit IRC (Remote host closed the connection)