=== omarrmz is now known as omarrmz_
=== Guodong-Xu is now known as guodong
=== guodong is now known as Guest33218
=== Guest33218 is now known as Guodong-Xu
lagarnd: Are you ready for an initial review?09:32
arndlag: yes, but my time is a bit limited as we're ramping up to the merge window09:39
lagarnd: Any chance in these going into the merge window?09:41
arndlag: hmm, good point. If we are sure they don't break anything else, there is probably no harm doing it now, and it would be really nice to show some progress09:50
lagarnd: They're available on my tree: device-tree-snowball-for-review09:51
lagarnd: :)09:52
arndlag: ok, starting at the bottom:10:01
arndthe patch from Russell slightly complicates this because it's already in another tree and Russell really doesn't like patches to be merged multiple times (we all try to avoid it)10:01
arndit's also missing your "Signed-off-by:" -- you should put that below the others when you apply a patch from someone else10:02
arndI would suggest you just drop that patch from your series when we merge it, or rebase on something that has the original version10:02
arndThe "ARM: supply a DTB file to be appended" patch I believe was shot down in review already, but I don't remember the details10:05
arndnhe, npite or gcl might remember better10:05
arndit's not essential to your series, so I'd also suggest dropping that: if we want to have it, it should go through Russell's tree anyway, not be part of the ux500 series10:05
lagarnd: I only added Russell's commit 'cos I had to make the same patch, as it broke the driver - we don't have to send it10:06
lagarnd: "ARM: supply a DTB file to be appended" I liked it - its REALLY handy10:06
nhearnd, lag: That patch is not to be upstreamed. It is only for our own use...10:07
lagarnd: ... but if they don't want it upstream I'll just keep it locally10:07
lagarnd: Okay, two patches dropped10:07
arndin "ARM: ux500: Initial Device Tree support for Snowball", you should swap the order of the sob: arnd,nhe,lag not arnd,lag,nhe -- the person submitting the series should be last10:08
lagarnd: I need to pop out, can I pick this up when I get back?10:08
arndlag: sure, I'll just keep writing until my lunch, and will be away after that, too10:08
arndI mean will be back after lunch, in about 90 minutes10:09
lagarnd: Great, I'll start work when you're on lunch10:09
lagarnd: Thanks dude10:09
lagarnd: NP10:09
arnd"db8500: list most devices in the snowball device tree" needs to be renamed so the subject starts with "ARM: ux500:"10:10
arndit also needs you Signed-off-by in addition to mine10:10
arndIn patch "ARM: ux500: Enable Cortex-A9 GIC (Generic Interrupt Controller) in Device Tree" you can drop both the "#ifdef/#endif":10:11
arndwhen CONFIG_OF is disabled, of_have_populated_dt returns false anyway and the gcc will just drop all the other code10:12
arndthat would need to be tested though. If it doesn't work the way I say, you should leave the patch as is10:13
arndsame thing for "ARM: ux500: Provide local timer support for Device Tree"10:13
arndand for "ARM: ux500: Enable PL310 Level 2 Cache Controller in Device Tree"10:15
arndthe mmc patches look great!10:17
arndI would have put the u300 variant together with the ARM ones, not with ux500, but it's probably ok either way.10:18
arndThe mmci driver definitely needs binding documents, you should add Documentation/devicetree/bindings/mmc/mmci.txt and mmci-ux500.txt10:23
arnddescribing all the properties, similar to the other files in that directory10:23
arndI think we need to have some discussion about the property names for mmc then, which will presumably drag the discussion out for too long to make it into the merge window10:24
arndI'm happy to take the ten ARM:ux500 patches if you do the changes I listed above10:25
arndthe mmci patches need to go through the mmc tree, not arm-soc and have to be reviewed on the linux-mmc mailing list. I don't know whether Chris Ball is still accepting patches for 3.4, but I would definitely recommend doing the last patch in your series in 3.5, not 3.410:27
arndlag: please submit all the ux500 patches as a series to linux-arm-kernel, cc linusw,nhe,arnd,patches@linaro.org after you have done the small changes10:28
arndlunch for me now, see you later10:29
arndlag: one more thing: linusw will have the final word over the ux500 patches, of course, but I expect that he will like them11:42
lagarnd: Great :)11:42
lagarnd: You're a star - go and have lunch :)11:42
lagarnd: W00t!13:05
arndlag: yes, very good. As always, I find more details when having another look, and you missing one of my trivial comments above, so I commented on all of those by email14:07
arndnot sure about the spi any more, we should wait for gcl to reply14:07
arndI think the spi patch you did can't do any harm, but it seems incomplete14:07
lagarnd: Okay14:09
lagarnd: I'm in the middle of completing the mmci patch-set, I'll start on the others again in a bit14:09
lagarnd: What you want me to do about localtimer?15:47
lagMarc Zyngier already has a solution queued up15:48
lagarnd: Although, it doesn't look accepted?15:48
lagarnd: [PATCH v4 04/15] ARM: smp_twd: add device tree support15:49
=== omarrmz_ is now known as omarrmz
arndlag: isn't that in the next/timer branch of arm-soc?16:14
arndlag: you can use that branch as a base for your own patches if you like16:14
arndlag: or even better: provide one branch that has all the patches besides your localtimer patch16:15
arndlag: and another branch that has the first one merged with the next/timer branch, and your localtimer patch on top16:16
lagarnd: I can do that16:25
lagarnd: I get lots of errors from timer.c17:02
lagarnd: Have you tried to compile next/timer?17:02
arndlag: no, unfortunately not17:11
lagarnd: It's broken - should I comment on the ML?17:12
arndlag: yes, please reply to the pull request that Marc Zygnier sent for it17:12
lagI was going to reply to: [PATCH v4 09/15] ARM: ux500: convert to twd_local_timer_register() interface17:13
arndlag: that works, too, but do mention that you tested the version from arm-soc/for-next17:19
arndor arm-soc/next/timer17:19
arndbest test both versions before and say which one failed17:19
arndI have to go home now, see you later17:20
mounirbgcl: ping18:50
ukleinekarnd: that is an ack for my patch?22:07

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!