[08:57] <apw> hallyn, it sounds like you are conflating ubuntu abi numbers and upstream stable numbers
[12:25] <jsalisbury> tsimonq2, apw I'll build a test kernel for that bug.
[12:28] <tsimonq2> jsalisbury: Thanks.
[12:48] <mamarley> It looks like the v4.15-rc9 mainline kernel (with the 4.15.0-5.6 configuration) (inadvertently?) disables several options related to Broadcom B43 and B44 chips, causing these chips to not work with that kernel.
[12:49] <mamarley> And also a bunch of SSB-related configuration options, which are also related to these chips.
[12:58] <apw> mamarley, are they new?  new ones are taken at their defaults
[13:29] <mamarley> apw: I don't think so; these have been around forever.  I also checked the commit history and it doesn't look like anything related to ssb or b4x changed since -rc8.
[15:18] <hallyn> apw: coudl be, are you saying the kaiser patches were backported to 4.4.0-109-generic ?  except it's from jan 9, and the magic sysfile telling me i'm safe isn't there :)
[15:19] <apw> those sysfs files.are total rubbish in the main
[15:20] <apw> V1 reports bad always iirc even with the patches which fix it applied
[15:58] <jsalisbury> tsimonq2, I posted a test kernel to the bug.
[17:16] <tsimonq2> jsalisbury: Thank you :)
[20:09] <s10gopal> https://bugs.freedesktop.org/show_bug.cgi?id=104737 please help me
[20:18] <tomreyn> this is being discussed in #ubuntu now, where it belongs
[20:19] <TJ-> the user is inexperienced but persistent :)
[20:24] <hggdh> the shotgunned a lot of channel with the exact same Q
[21:57] <jackpot51> I am seeing linux-signed-generic 4.13.0.31.33 show up when doing debootstrap and then an apt update/upgrade. I do not see linux-signed-image-generic (= 4.13.0.31.33), and this is causing a failure to upgrade. Can someone help me out?
[23:07] <mamarley> apw: Even when I remove the parts of "0006-configs-based-on-Ubuntu-4.15.0-5.6.patch" that disable the SSB options, the options still end up getting disabled somehow when the kernel is compiled.  I'm trying to figure out why that is, but not having much luck.
[23:11] <apw> jackpot51: against which archive pocket?
[23:11] <apw> mamarley: that implies a config constraint is disabling them
[23:12] <apw> typically a select or depends influencing the value
[23:13] <mamarley> I figured it was something like that.  I'm not sure how to figure out which one though.
[23:22] <mamarley> I went through all the CONFIG_SSB* options that are defined and ensured that the config has them enabled.
[23:22] <mamarley> I mean that the config has their DEPENDS enabled.
[23:32] <mamarley> apw: I figured it out, https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=664eadd6f44b3d71dcc62d0a825319000de0d5c9 causes the problem.
[23:33] <mamarley> The problem is, as far as I can tell, CONFIG_PCI_DRIVERS_LEGACY is defined only on MIPS.
[23:34] <apw> and _LEGACY is that new, and off?
[23:35] <mamarley> apw: According to https://cateee.net/lkddb/web-lkddb/PCI_DRIVERS_LEGACY.html, it was introduced in 4.9.  There doesn't seem to be a switch for it anywhere in the x86_64 menuconfig though.
[23:36] <mamarley> And searching for that symbol in menuconfig shows "Symbol: PCI_DRIVERS_LEGACY [=PCI_DRIVERS_LEGACY]"/"Type  : unknown"
[23:39] <mamarley> Even if I set CONFIG_PCI_DRIVERS_LEGACY=y in the configuration, the next time I run menuconfig it just removes it.  I think the committer must have forgotten to try compiling b43/b44/ssb on anything but MIPS with that patch.  I will open an upstream kernel bug.
[23:45] <mamarley> Wait, it is actually nel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=58eae1416b804d900014d84feadda7195007cc30
[23:45] <mamarley> Oops, https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=58eae1416b804d900014d84feadda7195007cc30
[23:49] <mamarley> https://bugzilla.kernel.org/show_bug.cgi?id=198555