=== gerald is now known as Guest98822 === mwhudson is now known as zz_mwhudson === FreezingAlt is now known as FreezingCold [11:45] ogra_, this syslog vomit, which kernel is this [11:46] uh, ask jdstrand, i think it is mako [11:47] root@ubuntu-phablet:/# tail -1 /var/log/syslog [11:47] Feb 14 12:47:09 ubuntu-phablet kernel: [53723.068756] adjust_soc: ibat_ua = 0, vbat_uv = 4326558, soc = 100, batt_temp=290 [11:47] yep, looks like mako [11:47] you can tell it is mako from that line ? [11:47] seems to write that every 20sec here for me [11:48] no, i can tell by following the cable that goes from my laptop to the device ;) [11:48] thats actually my output above [11:48] oh heh [11:49] it wrote this line 5 times since we speak (i have a tail -f running) [11:50] * ogra_ takes a look at maguro [11:50] ogra_, so that is device specific i suspect [11:50] same thing on maguro [11:50] Feb 14 12:49:33 ubuntu-phablet kernel: [ 4449.108245] max17040 4-0036: online = 1 vcell = 3767500 soc = 29 status = 1 health = 1 temp = 250 charger status = 0 [11:50] Feb 14 12:50:24 ubuntu-phablet kernel: [ 4499.191436] max17040 4-0036: online = 1 vcell = 3852500 soc = 29 status = 1 health = 1 temp = 260 charger status = 0 [11:50] yeah but a totally different mesage [11:50] even on a shorter cycle [11:50] sigh, these people are full of crap [11:51] as i told you :) [11:51] * apw does wonder why we have syslog running though on such small disks [11:51] android drivers ... [11:51] apw, we plan to restrict it [11:51] i assume they don't which is why they get away with it [11:51] currently we need it for debugging indeed [11:51] enduser devices will have limits set for logging [11:52] ogra_, which is maguro, the galaxy nexus ? [11:52] (and logrotate throw away logs instead of keeping them compressed) [11:52] yeah [11:53] i thought that one was dead a buried [11:54] here is manta: [11:54] Feb 14 12:53:42 ubuntu-phablet kernel: [ 488.998610] battery: l=100 v=4275 c=-123 temp=29.7 h=1 st=1 [11:54] same thing [11:54] (N10) [11:54] manta does it once a minute ... less frequent than the others [11:55] apw, maguro will be dead and buried soon (once we default to a 4.4 base) but isnt until then [11:56] (not sure you need to care for it though) [11:56] (i wouldnt :P ) [11:57] my fo (N7 2013) is sadly out of battery ... but i bet it does the same too [11:57] *flo [12:06] ogra_, could yu find out what the message is in that one, when it has charged :) [12:06] and i can get some bits uploaded to the ppa for these [12:06] yep, just charging [12:07] ogra_, though first i'll get you a test kernel with it moved to pr_debug and make sure they don't ned up in the logs still [12:07] i'll add excerpts for all the devices we care about to the bug [12:11] ogra_, thanks [12:21] ogra_: apw: they are really verbose, even more when you try to suspend it [12:22] which happens like a lot when you use it as a phone :-) [12:22] rsalveti, right, we need to get rid of that [12:22] well, I think it's expected from the kernel perspective to have such logs [12:22] as it's very useful for debug [12:22] we just shouldn't store them [12:23] or at least put a small limit [12:23] rsalveti, well, if we out a small limit in you will only see these messages in your syslog [12:23] i dont think we want charger status reported every 20sec [12:24] (no matter if the device is on or not) [12:24] well, then do you really want to patch every kernel we had to remove those messages? [12:26] rsalveti, in the long term we will i fear [12:26] or skip kernel logging at all and rely on the ringbuffer [12:26] logs are usually a good thing (that's why android is so verbose), we just shouldn't store them [12:26] but that will then only have the battery/charger messages [12:27] yes, but that means using the rigbuffer (dmesg) [12:27] *ring [12:27] that's fine, at least it's still useful if you're debugging [12:27] if we remove the messages, it'll just be harder to debug issues [12:28] well, i still dont think every 20sec a battery ping is relevant to debugging [12:29] i agree that ril messages might be ... but for that we have the radio logcat log ... [12:44] ogra_, http://people.canonical.com/~apw/mako-trusty/ is that message moved to debug, be interested if that makes it not end up in the logs [12:45] rsalveti, how goes testing those yama kernels [12:45] apw, no mako for testing here [12:46] apw: testing all the kernels as we speak [12:46] (thats my main phone ... i dont do tests on it ... company needs to supply me one :P ) [12:46] * apw calls this whole thing stupid [12:46] people who have kit need to have all of the supported kit or not bother [12:46] else we are just wasting peoples time [12:46] rsalveti, thanks ! [12:49] apw, well, i have flo, maguro an manta ... the mako is my private one and i dont want to have to set up calendar and my accounts multiple times a week (or have to install the click packages i use on a daily basis) [14:00] * smb wants a sushi now [14:05] apw, remember to upload to the c-k-t PPA per bjf request [14:08] rtg ack === ara is now known as Guest55777 === tyhicks` is now known as tyhicks [15:58] rtg, fire in the hole [18:14] apw: rtg: missing a new kernel release for manta [18:15] after you updated the tree with the yama changes [18:15] that would still be ppa-only (as it's on top of 4.4.2), but just noticed it's not there yet [18:17] rsalveti, k, lemme figure out what happened. I think apw was doing that at the time. [18:25] rsalveti, looks like we (likely me) didn't finish packaging and uploading. I'll git 'er done here in a bit. [18:28] rtg: great, thanks [18:35] rsalveti, linux-manta_3.4.0-5.21 uploaded to the c-k-t PPA [18:40] rtg: awesome, thanks [19:39] apw: hi, any comments on the overlayfs unprivileged whiteout patch? (stgraber will be poking me soon about it I'm sure :) [22:32] what kernel trace tool produces output like this ? [22:32] ip-21638 [020] .N.. 37431.042253: lg_global_unlock <-do_umount [22:32] ip-21638 [020] .N.. 37431.042265: lg_global_lock <-release_mounts