**** BEGIN LOGGING AT Tue Oct 30 02:59:58 2012 Oct 30 07:14:53 howdy Oct 30 08:18:29 janimo: ping Oct 30 08:18:53 janimo: where's the git tree for the nexus7 kernel? are you mantining it, right? Oct 30 08:18:59 *mintining Oct 30 08:19:02 argh Oct 30 08:19:04 ... Oct 30 08:23:09 ppisati, https://wiki.ubuntu.com/Nexus7/SourceCode Oct 30 08:23:45 lilstevie: cool, thanks Oct 30 08:26:27 np Oct 30 08:28:29 ppisati, we should shortlyh decide at UDS where to place it so it can be comaintained Oct 30 08:28:59 janimo: ack, iirc there was a session about it Oct 30 08:29:29 ppisati, right. Looking for something specific or just checking it out? Oct 30 08:29:39 I have a few more power related commits that are not yet pushed out Oct 30 08:29:39 janimo: just checking it out Oct 30 08:29:42 ok Oct 30 08:30:44 janimo, I was just looking briefly at what you have done, what else does the battery driver have that it can export Oct 30 08:32:58 lilstevie, I just received a patch yesterday about that. I'll push it out shortly Oct 30 08:36:16 ok cool, the few patches I have tried to get them exposed with the tf101/tf201 just break the driver :( Oct 30 11:06:30 ogra_: ping, your presence is requested Oct 30 19:57:53 is it possible to install a softfloat compiler on precise (on arm)? to build e.g. a kernel? Oct 30 19:57:58 armhf* Oct 30 20:01:02 steev: basically you pass -mfloat-abi=softfp to gcc when building your kernel Oct 30 20:01:21 export CFLAGS="-mfloat-abi=softfp" Oct 30 20:01:40 export CXXFLAGS="-mfloat-abi=softfp" Oct 30 20:01:42 shoudl do it Oct 30 20:05:08 xranby_ac100: Uhm. The kernel doesn't need any such thing done. Oct 30 20:05:49 infinity: you are right .. Oct 30 20:05:51 steev: Compiling kernels should Just Work, regardless. Oct 30 20:06:58 steev: the kernel interface is the same for softfloat and hardfloat (the kernel do not use any userspace system librarys) Oct 30 20:19:32 yeah, you're right, except there's a minor issue with the 4.6 compiler and hardfloat, so i'd prefer to tell it to use softfloat, or use a softfloat compiler Oct 30 20:19:45 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52445 Oct 30 20:19:48 gcc.gnu.org bug 52445 in tree-optimization "[4.6 Regression] conditional store replacement causes segfault in generated code" [Normal,Assigned] Oct 30 20:23:28 infinity: btw.. are you at uds right now? Oct 30 20:23:42 xranby_ac100: Aye. Oct 30 20:23:55 in the testcase writing seminar? Oct 30 20:24:13 xranby_ac100: Nope. Oct 30 20:24:47 ok well see you around.. i am at the testcase hackatron right now Oct 30 20:27:32 steev: Hrm. That bug doesn't seem to imply it's HF-only (or even ARM-only). We can definitely look at backporting the fix to precise, though. Oct 30 20:28:37 infinity: you're right, i was going from experience with gentoo which has the issue only with hf compiler (4.6.3) Oct 30 20:29:03 i have a bug open on the gentoo end, but sadly, i can't seem to build a debian package of the kernel on my gentoo box Oct 30 20:31:05 steev: doko's got a precise SRU for gcc-4.6 pending, I'll ask him to add this backport. Can you file a launchpad bug for reference? Oct 30 20:31:57 infinity: i would prefer not to - launchpad loves to spam me with bugmails. i still get mails from a bug that was opened in 2005 Oct 30 20:32:30 Unsubscribe from it, if it's still getting comments? Oct 30 20:33:00 steev: Anyhow, you could also just build your kernels with gcc-4.7 in a quantal chroot, they'll install just fine on precise. Oct 30 20:33:02 If you are the filer, you can't unsubscribe. I've tried. Oct 30 20:33:12 GrueMaster: Oh, fair point. Oct 30 20:33:26 infinity: heh, actually, it's for maverick Oct 30 20:33:29 * infinity filters all his bug mail anyway, so doesn't really notice or care. Oct 30 20:33:41 but yeah, the arm box doesn't really have enough space to add a quantal chroot Oct 30 20:34:32 steev: I'll poke doko later about adding PR52445 to his precise SRU. We'll need someone to open a bug for tracking/validating it, though. :/ Oct 30 20:34:49 * steev points at GrueMaster Oct 30 20:35:16 i'll be more than willing to test it, but no, i'm sorry but i refuse to use launchpad Oct 30 20:35:28 steev: Do you know which driver(s) get miscompiled? The upstream bug is vague. Oct 30 20:39:19 My only real complaint is that a lot of the bugs I filed got ignored until I was no longer doing arm work. Then suddently there appeared to be fixes for me to test, but they were reverted because I no longer do testing. Oct 30 20:40:10 off the top of my head the flexcan driver, but if you search gcc 4.6.3 miscompiling arm-linux Oct 30 20:40:18 there's a long thread on lakml Oct 30 20:40:39 GrueMaster: ditto, except i'm still doing arm - and more specifically, bugs were filed with patches that were ignored until irrelevant Oct 30 20:41:21 and by irrelevant, i mean we're no longer using that software because we've forked it because we couldn't wait for the fixes to be added Oct 30 20:41:44 Yep. Oct 30 20:41:59 it's not a whole lot, but when you appear to not maintain the software, users and developers really develop a distate Oct 30 20:43:24 and no offense, but if i've already done the work, and provided a patch, i don't really see why i should also have to pay canonical to get it applied Oct 30 20:43:47 steev: Who's asking you to pay to get pathes applied? Oct 30 20:43:51 Or patches. Oct 30 20:44:12 no one, directly Oct 30 20:44:19 I realise it's frustrating to not have every bug addressed, but hyperbole doesn't help. Oct 30 20:45:01 (For the record, of all the bugs I've fixed over the years in Ubuntu, exactly one of them was escalated from a paid customer, and it was a bug that I was working on upstream regardless) Oct 30 22:20:25 how'd i add a window manager to the list show at the login screen? Oct 31 00:05:15 I'm looking for a GNU/Linux-friendly tablet computer and someone who got good advice here suggested I also ask here. **** ENDING LOGGING AT Wed Oct 31 02:59:58 2012