**** BEGIN LOGGING AT Thu Dec 24 02:59:59 2015 Dec 24 05:31:16 fishbulb should use their 3 N900s to run Doom with a 270-degree FOV: http://toastytech.com/dooma/doom099.html Dec 24 10:37:28 moin :) Dec 24 11:37:44 Pali: the last message in qemu is "Serial console enabled", do I miss something? Dec 24 11:38:05 hm? Dec 24 11:38:42 kernel doesn;t seem to boot in qemu Dec 24 11:39:15 freemangordon: are you using my git tree? Dec 24 11:39:25 yes Dec 24 11:40:05 hmm Dec 24 11:40:27 seems it was save_atags() that made it hang Dec 24 11:40:55 after commenting it, the kernel booted Dec 24 11:41:32 Pali: wait, why I have data in proc/atags? Dec 24 11:43:21 freemangordon: http://mg.pov.lt/maemo-irclog/#maemo.2015-12-23.log.html#t2015-12-23T16:17:13 Dec 24 11:44:45 on branch v4.2-rc2-n900 revert 97ea449c49fbe2d4b4abda1d65b5b68ef3ad29a2 and 1b0eeec5a0b43954aeeb126dbf8d89df32449fb6 Dec 24 11:45:00 oh, I see Dec 24 11:45:01 ok Dec 24 11:57:26 Pali: seems to work :) Dec 24 11:58:19 ok :-) Dec 24 12:00:28 Pali: http://pastebin.com/E8vJijki Dec 24 12:01:45 Pali: and http://pastebin.com/wmaGZ91h Dec 24 12:02:13 now I only need to figure out how to include kernel/atags.h and I'll send a patch Dec 24 12:03:46 freemangordon: you need to also send 5a4a299f59b17bc6dc694865e1d007abfe211e7a Dec 24 12:13:53 Pali: what about moving atags.h to include/asm? Dec 24 12:16:48 freemangordon: rather ask these questions on ML Dec 24 12:16:58 I'm really not somebody who can decide Dec 24 12:17:13 I was thinking whether you want to send those patches Dec 24 12:17:48 the problem is that I've never sent patch series Dec 24 12:18:10 not sure how to do it, will need to google a lot before doing it Dec 24 12:51:50 Pali: ^^^ Dec 24 12:52:00 Pali: and what about /proc/cpuinfo? Dec 24 12:52:36 freemangordon: just do git format-patch and send it Dec 24 12:52:40 same as sending one patch Dec 24 12:53:01 but I need to send 3 patches Dec 24 12:53:14 anyway, will look at docs how to do it Dec 24 12:53:21 /proc/cpuinfo is semi-fixed Dec 24 12:53:31 problem with name is fixed in upstream Dec 24 12:53:46 missing part not in upstream is revision number Dec 24 12:54:00 it is in ATAGS? Dec 24 12:54:12 this is [PATCH 1/5] and [PATCH 2/5] in that big stream :-) Dec 24 12:54:17 yes it is in ATAGs Dec 24 12:54:30 but I understood that those 2 patches will be accepted Dec 24 12:54:46 just patch [PATCH 1/5] needs documentation Dec 24 12:55:10 but I guess you need to send anothe series with patch 5/5 dropped Dec 24 12:55:31 also drop 4/5 Dec 24 12:55:39 and move 3/5 into your patches :-) Dec 24 12:55:41 ah, yes Dec 24 12:56:00 yes, I'll fix that one too Dec 24 12:56:15 so just 1/5 and 2/5 Dec 24 12:56:31 ok, lets see if it boots with linux 4.4 :) Dec 24 12:56:34 anyway, if you have time you can try to write documentation for 1/5 and 2/5 patches Dec 24 12:57:00 never did any DT doc so far :) Dec 24 12:57:02 documentation for "/revision" DT entry Dec 24 12:58:51 hmm - "mmc0: error -22 whilst initialising SDIO card" with upstream 4.4 Dec 24 12:59:14 dammit, is it possible sometimes new releases to not break what is already working :( Dec 24 13:10:29 freemangordon: sd card is broken in qemu or kernel for a long time Dec 24 13:10:35 special patch is in my tree Dec 24 13:10:40 great Dec 24 13:11:03 ok, save_atags() work in upstream too Dec 24 13:11:16 going to send the patches Dec 24 13:11:24 first it try SDIO which fails with -22 and then try MMC which success Dec 24 13:11:29 ok Dec 24 13:47:27 Pali: how to format the "cover letter"? Dec 24 13:47:44 I know about --compose, but what about diffstat etc Dec 24 13:47:52 git format-patch --cover-letter Dec 24 13:47:58 oh :) Dec 24 13:49:10 Pali: shold I edit it by hand? Dec 24 13:49:32 yes, generate all patches with --cover-letter and then edit it Dec 24 13:49:40 you can also edit patch files if you need Dec 24 13:52:31 ok, finally the patch series seems ok :) Dec 24 13:54:13 Pali: get_maintainer spit about 15 people, shall I cc them all? Dec 24 13:55:53 send email to people who are marged as maintainers Dec 24 13:55:57 and to mailing lists Dec 24 13:56:15 and ideally also to people active in current discussion! Dec 24 13:56:51 ok Dec 24 14:01:35 Pali: done Dec 24 14:02:38 wait, what happened with the last patch? Dec 24 14:03:11 ah, it is there :0 Dec 24 14:04:24 all patches are there Dec 24 14:05:01 yeah, the last one came later here Dec 24 14:05:24 freemangordon: for future, when you have commits which moving files, generate diffs with param -M Dec 24 14:05:50 -M detect moved files and so generate smaller and more readable diffs Dec 24 14:06:06 you mean a parameter to format-patch? Dec 24 14:06:10 but diff generation is slower! so this is reason why is not -M enabled by default Dec 24 14:06:13 yes Dec 24 14:06:20 ok Dec 24 14:06:28 hope to remember it Dec 24 14:06:31 also git diff -M Dec 24 14:06:42 I use git-gui for commits Dec 24 14:06:51 maybe there is an option Dec 24 14:06:52 every git command which show diff should accept -M param Dec 24 14:07:40 hmm, "additional diff parameters" :) Dec 24 14:08:06 -M is running in some O(n^2) time where n is number possible renamed files Dec 24 14:08:26 so for big diffs on lot of files it is sloow Dec 24 14:08:49 so be it, I dun't ususally change lots of files before making a commit Dec 24 14:08:57 *don't usually Dec 24 14:09:33 Pali: Are ya'll talking new KP stuff? =) Dec 24 14:09:51 hm... what? Dec 24 14:10:25 Kernel Dec 24 14:10:32 You're "the kernel" person =) Dec 24 14:10:43 I see you talking, and I'm like, hmmm... What's Pali cookin up Dec 24 14:11:12 http://lkml.org/lkml/2015/12/24/134 Dec 24 14:12:08 :) Dec 24 14:12:25 do you think it will be included ? Dec 24 14:13:37 yes Dec 24 14:14:04 the only possible proble is moving that header Dec 24 14:14:08 yeah Dec 24 14:14:12 precisely Dec 24 14:14:19 but if they reject it, they should provide a solution Dec 24 14:14:44 so the worst that can happen is /me re-sending the series again Dec 24 14:15:05 :) Dec 24 14:15:44 Pali: is elinux.org accurate/up-to-date by the way? Dec 24 14:15:52 but having in mind that this ATAGS problem persist for ~2 years, this is more than acceptable Dec 24 14:16:18 http://elinux.org/N900 should be up-to-date Dec 24 14:18:10 anyway, do you know POSIX way how to get all ipv6 addresses assigned to network adapter? Dec 24 14:18:19 in linux we have getifaddrs() function Dec 24 14:18:31 but it is not portable Dec 24 14:19:07 which maemo version isnt based on linux? Dec 24 14:19:26 this is not for maemo, general network question :-) Dec 24 14:20:02 getifaddrs exists on osx Dec 24 14:20:11 well, the man page at least. and the man says "BSD" Dec 24 14:20:49 I have it on freebsd as well (in /usr/include/ifaddrs.h) Dec 24 14:21:11 same on openbsd Dec 24 14:21:29 it doesnt look too bad Dec 24 14:22:38 structure definition seems pretty similar too (apart from the union on linux) Dec 24 14:24:13 looks like a pretty sane choice Dec 24 14:32:41 hm... I just need to get link-local address of adapter Dec 24 14:33:10 is not there something better as enumerarate all network interfaces via getifaddrs()? Dec 24 14:37:12 hmm, I think there is, lemme check Dec 24 14:41:35 nevermind, the example I had in mind for network adapters used ... netlink Dec 24 14:41:44 definitely not portable Dec 24 14:47:34 buZz: none Dec 24 14:51:56 oh Dec 24 16:45:21 bencoh: toldya :) Dec 24 16:45:45 "I'd think I'd prefer moving just the above save_atags declaration to arch/arm/include/asm/setup.h so we're not allowing the rest of this to be exposed to anyone who includes asm/atags.h." Dec 24 16:47:26 :) Dec 24 16:55:28 Pali: are you going to resend your patches, without what I've already send? Dec 24 16:55:49 freemangordon: after your (new) patches will be accepted Dec 24 16:56:21 Pali: also, don;t count on me re DT docs, I don;t really know where to start from neither what am I supposed to write in :) Dec 24 16:56:29 ok Dec 24 16:56:40 ok Dec 24 16:57:00 BTW I think you can send driver patches without docs Dec 24 16:57:16 or rather - docs can follow-up Dec 24 16:57:28 DT that is Dec 24 16:59:27 PATCH 1/5 just needs documentation what is "/revision" Dec 24 16:59:42 everything other (driver parts) are done Dec 24 17:01:19 iiuc, you only need to resend 1/5 and 2/5 Dec 24 18:37:47 happy xmas Dec 24 18:59:02 Pali: Russell's patch system?!? Dec 24 18:59:17 reading that email too... Dec 24 18:59:18 DocScrutinizer05: happy :) Dec 24 18:59:26 merry Dec 24 18:59:35 whatever Dec 24 18:59:40 :-) Dec 24 18:59:42 no idea what is that system Dec 24 18:59:51 rather ask Dec 24 18:59:53 going to ask Dec 24 19:00:17 I do not want to see this patch to be missed just because some policy or system Dec 24 19:00:29 folks, shouldn't you burn trees or sth? :-) Dec 24 19:01:48 the real hacker doesn't sing xmas songs, he tries to finish that patch in time to get the datestamp YY-12-24 Dec 24 19:01:58 ;) Dec 24 19:02:10 :) Dec 24 19:03:04 :-) Dec 24 19:03:11 or in my case: that whitepaper Dec 24 19:11:11 burning a whitepaper ? Dec 24 19:25:34 Pali: http://www.arm.linux.org.uk/developer/patches/info.php Dec 24 19:26:02 this is it. I created an account, will wait a bit for more comments and will send those 2 patches there Dec 24 19:26:31 ok Dec 24 19:29:26 wait... that's not mainline at all. do you need to go through them first? Dec 24 19:29:34 (arm.linux.org.uk) Dec 24 20:01:50 bencoh: it is supposed to happen faster Dec 24 20:02:01 oh, okay Dec 24 20:02:49 bencoh: https://lkml.org/lkml/2015/12/24/297 Dec 24 20:03:37 :) Dec 24 20:04:14 DocScrutinizer05: so you're back from your holiday? Dec 24 20:04:25 DocScrutinizer05: merry xmas 2 Dec 24 20:07:30 freemangordon: yep, though I still try to catch up and get pace Dec 24 21:21:15 if somebody has thinkpad with backlight keyboard, here is patch for linux which can control backlight level from OS (usefull for scripting): https://lkml.org/lkml/2015/12/24/295 Dec 24 21:21:16 :-) Dec 24 21:22:09 i have one with the led light in the display frame Dec 24 21:29:58 KotCzarny: no, this is for keyboard Dec 24 21:30:27 i know, just saying the original name was 'keyboard light' Dec 24 21:30:37 ah, ok Dec 24 21:47:48 hey, someone here had a link to n900hackers.org Dec 24 21:47:55 is there anyone running that image? Dec 24 21:48:31 I tried pwnie express years ago, it lasted about 1 hour before the phone was totally dead Dec 24 21:54:22 "dead"? Dec 24 21:55:40 probably empty battery .. likely that image does some custom wifi things with no respect for power savings **** ENDING LOGGING AT Fri Dec 25 02:59:58 2015