=== JanC is now known as Guest5754 === JanC_ is now known as JanC [06:52] Additionally: If I look at linux/Documentation/workqueue.txt and do "echo workqueue:workqueue_queue_work > /sys/kernel/debug/tracing/set_event" and "cat /sys/kernel/debug/tracing/trace_pipe > out.txt" [06:53] with the issue, I get somwhere between 10,000 and 20,000 occurances of memcg_kmem_cache_create_func in the file. [06:54] without the issue, I get 21, and an overall file size about 50 times smaller, for otherwise similar conditions. [06:55] Using a simplified method to create the issue that I made. === Guest8435 is now known as sts [09:54] Hello, just an FYI : makedumpfile is failing on 4.8 kernels (i.e. Yakkety). I'm on the issue and should fix it today [09:56] caribou, thanks, if you have a bug for that could you tag it kernel-4.8 so we see it too [09:56] apw: sure : LP: #1626269 [09:56] Error: Could not gather data from Launchpad for bug #1626269 (https://launchpad.net/bugs/1626269). The error has been logged [09:56] ubot you are just a waste of electrons [10:09] apw, caribou I added the tag [10:09] smb: thanks! [12:19] * xnox hates EFI [12:19] so much respins, so much to test =) [12:41] xnox: they'll be more yet [12:43] dsmythies, any luck on changing CONFIG_NR_CPUS to something less then 8192 ? [13:08] rtg, that was not succesful, cking is debugging [13:49] dsmythies, I've been informed that http://bugs.launchpad.net/bugs/1626564 is likely the root of your kworker issue [13:49] Ubuntu bug 1626564 in linux (Ubuntu Yakkety) "4.8 regression: SLAB is being used instead of SLUB" [High,In progress] [14:16] rtg: Indeed, SLUB is a difference between 4.7-rc4 and 4.7-rc5, and not something that I had tried reverting. I have an inredibly simple way to demonstrate this issue. I'll add to the bug report. [14:17] dsmythies, I credit apw and cking for finding root cause [14:18] cking, did all the heavy lifting ... [14:23] I'm experiencing system "slowdown" in yakkety -- that's a known issue? [14:24] mterry, particularly in early boot yes [14:24] Ah. Maybe different then. I'm experiencing system unresponsiveness if anything is happening (like compiling, browsers using cpu) [14:25] Desktop experience gets quite bad, can't barely type [14:25] rtg, perhaps we should have a dirty build with the curent fixes so we can see if things like this is the same or a new issue [14:30] mterry also mentioned seeing “crazy high occasional loads” [14:31] Yeah. Browsers and such tend to go mental in top, which helps trigger the unresponsiveness. But I don't know if they are going mental because of initial unresponsiveness problem that spirals out of control? [14:32] Yeah. Browsers and such tend to go mental in top, which helps trigger the unresponsiveness. But I don't know if they are going mental because of initial unresponsiveness problem that spirals out of control? [15:11] rtg: do you want SRU bugs opened for pull requests for yakkety at this point - if not, when does that change? [15:13] dannf, yes .. bugs, please [15:13] bjf: ack [15:15] dannf, please do [15:35] Confirmed: reverting SLAB /SLUB kernel configuration changes solves the over 2000 kworker threads issue. As a side effect, and as expected, it also solves rediculous load average values after boot. i.e. "load average: 430.90, 99.40, 32.78" [15:38] Is this Ask Ubuntu Accurate (http://askubuntu.com/questions/804111/is-no-reboot-kernel-patching-enabled-in-16-04)? Is there now live kernel patching in Ubuntu 16.04? I asked on #ubuntu and they said they didn't think so but I should check with you guys. [15:45] chalbersma, the statements in there are accurate in that the facility to _do_ live kernel patching is included, there is no stream of such patches though [15:47] Has there been any live patches released? [15:50] not to my knowledge no [15:51] * ogra_ notes that hot-patching does not necessarily mean no reboots ... if you patch changes something that lives in the initrd you most likely want to re-generate it and actually also ask for rebooting [15:51] (and there are surely other cases where a hot patch will still cause reboots) [15:59] Thanks guys just wanted to clear that up. [16:04] .b 6 [16:30] <_ami_> which driver is responsible for serial over ip in linux? [16:35] that is normally called sol in the kernel side if it is coming via a bmc, though in my experience sol is handled by the device and faked to the machine as a real serial port [17:01] <_ami_> apw: ok, thanks [18:10] Hello! It seems the 4.8 Update on Yakkety have brought in some troubles. 1. My system becomes sluggish during load. 2. There is a lag in screen brightness keystrokes actually taking effect. [18:34] om26er, yes indeed, the first one we know the fix, the second we are investigating [19:31] apw, hi there! we have some sort of mismatching headers between powerpc/ppc64el and the other archs that is causing a few packages to FTBFS on powerpc/ppc64el, I stumbled upon freevo's FTBFS and slangasek said he saw other packages failing due to a similar reason... he said you might be able to help me on that, the current bug is LP: #1619446 [19:31] Launchpad bug 1619446 in linux (Ubuntu) "mismatching headers between powerpc/ppc64el and other archs" [Medium,Triaged] https://launchpad.net/bugs/1619446 [19:32] please let me know when it is a good time to talk about that [19:44] tdaitx, there should be a new linux-libc-dev hitting the archive about now ... could you confirm it is the same there and add that to the bug [19:44] apw, sure, I will check that, thanks for the info [20:43] dannf, I've pulled your changes. Please build yourself a test kernel to check if I've broken any arm64 config settings. I've been syncing with Xenial after modularizing everything. [20:52] rtg: sure and thx [22:24] apw, did you refer to version linux-libc-dev 4.8.0-15.16? it still FTBFS with that one