**** BEGIN LOGGING AT Mon Apr 27 02:59:59 2015 Apr 27 21:05:34 ant_home, hello Apr 27 21:06:16 hi Apr 27 21:06:39 so you mean there is a problem with mmc_spi ? Apr 27 21:07:37 yes, most probably. Apr 27 21:08:03 tested on 6 SD cards and 2 MMC Apr 27 21:08:13 next days I'll retry with my bunch of old cards Apr 27 21:08:52 I finally have serial again Apr 27 21:11:24 :) Apr 27 21:11:57 I had issues with poodle battery drivers (made mistakes during porting to 4.1-rc), but otherwise it looks good. Apr 27 21:13:08 And I again good review of iio 'DAC' driver. Apr 27 21:13:13 s/good/got/ Apr 27 21:13:39 Most probably I'll have to resubmit it. It's already v3. For quite a simple driver. Ugh. Apr 27 21:14:01 heh Apr 27 21:14:38 btw, it appears Robert is having hard time with lubbock/pxa, broken since 2011 :) Apr 27 21:15:18 I don't read lakml (heh). But I suppose lubbock should be not in the best shape. Apr 27 21:15:43 I read it now and then, too many msgs Apr 27 21:16:12 yep. Apr 27 21:16:28 today there are some sa1100/pxa irq_domain changes Apr 27 21:16:39 hope it doesn't break Apr 27 21:17:10 Hmm. Apr 27 21:17:18 on lakml? Apr 27 21:17:25 [PATCH] ARM: sa1100: Constify irq_domain_ops Apr 27 21:17:27 yes Apr 27 21:17:36 ah, n/p. should be Apr 27 21:17:53 pheww Apr 27 21:18:39 Hmm. Can't find on lakml in gmane Apr 27 21:18:54 But it really should not be a problem. Apr 27 21:20:36 I have few more patches for irq driver on sa1100, but that will probably wait. I'd like to get locomo first. Apr 27 21:20:38 is a series c/o Krzysztof Kozlowski Apr 27 21:25:36 do you think we should have linux-handheld_git? Apr 27 21:26:09 and perhaps keep the -yocto as stable/LTS? Apr 27 21:27:15 Hmm. I don't know. Apr 27 21:27:49 For long time I used my own kernel with all generated images. Apr 27 21:28:08 I have aleftover..LOCALVERSION = "-yocto" Apr 27 21:28:10 You'd better consult bluelightning. Apr 27 21:28:48 I agree we need a vanilla kernel Apr 27 21:29:11 for development/testing/contribution i.e. from you and linusw Apr 27 21:29:42 but it gets old when released...you're already on 4.1 Apr 27 21:29:52 Yep. On the other hand it you who are contributing patches to meta-hh. Apr 27 21:30:22 Yep. The issue with 4.0 -> 4.1-rc was in patches changing power supply interfaces. Apr 27 21:30:27 So I had to rebase. Apr 27 21:30:41 the big disadvantage of -yocto is that they remove regularly old versions Apr 27 21:30:54 so we get orphans Apr 27 21:31:12 and parse errors Apr 27 21:31:14 Hmm. Why is it a disadvantage? If we could update meta-hh in a timely manner. Apr 27 21:33:05 Regarding upstream/non-upstream kernels. Apr 27 21:33:10 on the other side a true -yocto could allow to split the config and standardize the fragments between machines Apr 27 21:34:06 plus it receives the regular versioned patchsets Apr 27 21:34:49 but for quick and dirty development nothing beats korgtarballs Apr 27 21:35:10 Ant it should be well-tested against other features. Like bluetooth, connman, wifi/wpa, etc. Apr 27 21:35:23 right Apr 27 21:35:44 but yocto-standard is oversized for handhelds Apr 27 21:36:14 and yocto-tiny is insufficient, maybe good for kexecboot Apr 27 21:36:53 about that, we could cook a kexecboot fragment (config + filesystems on boot) Apr 27 21:39:07 About kexecboot. I think we should find some way to enable printk in the kernel. Otherwise it might become very hard to debug init problems. Apr 27 21:39:40 I always used qemu, but recently qemu started to oops on pxa kernels. Apr 27 21:39:51 maybe if we use xz, otherwise it's too big for collie Apr 27 21:40:00 Yes, I know. Apr 27 21:40:02 printk and BUG Apr 27 21:40:33 Maybe we can drop pr_info/dev_info via a special patch/hack. Apr 27 21:40:53 but the rest might become necessary. Apr 27 21:43:22 Well. This is hopefully not the most urgent issue. Apr 27 21:43:41 heh, atm all machines are booting Apr 27 21:43:47 for the moment... **** ENDING LOGGING AT Tue Apr 28 02:59:59 2015