/srv/irclogs.ubuntu.com/2019/08/08/#ubuntu-kernel.txt

tjaaltonlotuspsychje: which nvme does your machine have?07:42
tjaaltonssd07:42
lotuspsychjetjaalton: its an 860 EVO 250GB sata07:42
lotuspsychjesamsung07:42
tjaaltontry nvme_core.default_ps_max_latency_us=007:43
lotuspsychjeok07:43
tjaaltonwith 5.007:43
tjaaltondistro07:43
tjaaltonsata?07:43
lotuspsychjeits not an nvme one, classic sata07:44
tjaaltonsorry, not nvme07:44
tjaaltonright07:44
lotuspsychjestill try it?07:44
tjaaltonno, it won't do anything07:44
lotuspsychjeallrighty07:44
tjaaltonhttps://www.mail-archive.com/intel-gfx@lists.freedesktop.org/msg145779.html07:45
lotuspsychjetjaalton: noticed a new bug, but not sure its related to mine: https://bugs.launchpad.net/bugs/183935307:45
ubot5Ubuntu bug 1839353 in linux-signed-hwe (Ubuntu) "Login screen and desktop corruption (18.04.2)" [Undecided,New]07:45
tjaaltonthat's fairly old, but still weird07:45
tjaaltonradeon07:46
tjaaltonnot related07:46
lotuspsychjekk07:46
lotuspsychjetjaalton: padoka or oibaf ppa drivers wont do much neither i suppose?07:47
tjaaltonbut, another one where 19.04 works fine, backport is busted07:47
tjaaltonno07:47
lotuspsychjetjaalton: i also tryed the makelinuxfastagain kernel param, but no dice07:48
tjaaltonwhat's that?07:48
lotuspsychjenoibrs noibpb nopti nospectre_v2 nospectre_v1 l1tf=off nospec_store_bypass_disable no_stf_barrier mds=off mitigations=off 07:48
lotuspsychjehttps://make-linux-fast-again.com/07:48
tjaaltonaha07:49
lotuspsychjetjaalton: another weird fact is, i only had 1 call yet from an ubuntu customer updated to 5.0.0.23 yet07:50
lotuspsychjeand i installed surely alot on bionic hwe07:50
lotuspsychjepuzzle bug :p07:51
lotuspsychjeor maybe they didnt update yet..07:51
tjaaltonapw: can you kick the drm-intel-next/2019-07-31 mainline build now that there's space? it'd be nice to have that for future debugging07:51
tjaaltonlotuspsychje: apparently the tiniest change to power management can change things, which is why my build of 5.0 worked and distro doesn't07:52
tjaaltonwhich is probably why the bisect results were "weird"07:52
lotuspsychjei see07:53
tjaaltonmaybe try https://kernel.ubuntu.com/~kernel-ppa/mainline/drm-tip/current/07:53
lotuspsychjegot this one in my list, is it the same? vmlinuz-5.3.0-994-generic07:54
tjaaltonthis is built daily07:54
lotuspsychjeah ok, on it07:54
tjaaltonwhich version is yours?07:54
tjaaltonapt-cache policy linux-image-5.3.0-994-generic07:55
lotuspsychjenot a daily for sure07:55
tjaalton"994" is drm-tip07:55
tjaaltonthat part doesn't change07:55
lotuspsychjehmm, installed: none07:55
tjaalton997 is drm-intel-next07:55
tjaaltonok, removed already07:55
lotuspsychjelemme try your daily holdon07:56
lotuspsychjetjaalton: yeah, thats working Linux Rootbox 5.3.0-994-generic #201908072204 SMP Thu Aug 8 02:09:23 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux08:00
tjaaltonok08:00
=== lotus|i5 is now known as lotuspsychje
flingI got the linus tree!12:14
flingWhat kernel versions include shiftfs patch?12:15
sforsheefling: only ubuntu kernels have shiftfs currently12:16
flingBut what versions?12:16
flingI'm trying to extract the shiftfs patch :P12:16
sforshee19.04 onward, so 5.0+12:17
flingok12:17
flingSo how would I clone the repo containing 5.1 properly?12:18
sforsheewe don't have a tree that we currently maintain with 5.1, for 5.2 you can use https://git.launchpad.net/~ubuntu-kernel/ubuntu/+source/linux/+git/eoan12:19
sforsheeor for 5.0: https://git.launchpad.net/~ubuntu-kernel/ubuntu/+source/linux/+git/disco12:19
flinghmmm12:20
flingzfs does not work with 5.2 yet12:20
flingand no 5.0 sources in portage anymore :P12:20
flingCan still use 5.0…12:20
sforsheethe patches from either ought to work fine with 5.1 too, I don't recall any updates being needed between those versions12:21
flingare 5.0 and 5.2 shiftfs the same?12:21
flingI should probably compare12:22
sforsheeI'm fairly sure they are the same12:22
flingsforshee: can't figure out how to glue commits together to form a single patch14:32
sforsheefling: if you apply the whole set you can then use 'git rebase -i' and squash them into one14:42
cjwatsonOr if you're just trying to inspect a series in bulk then you can use "git diff" from the start to the end of the series14:43
amitk /quit16:07

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!