#archlinux32 | Logs for 2018-03-25

Back
[00:41:02] <buildmaster> firefox-developer-edition is broken (says buildknecht3).
[01:04:38] <buildmaster> gobject-introspection is broken (says buildknecht2).
[05:51:15] <buildmaster> unarchiver is broken (says rechenknecht).
[07:37:40] <buildmaster> gnome-desktop is broken (says buildknecht3).
[09:31:49] -!- abaumann has joined #archlinux32
[09:31:55] -!- deep42thought has joined #archlinux32
[09:32:03] <deep42thought> good morning, abaumann
[09:32:08] <abaumann> morning.
[09:32:15] <abaumann> a new day.. a new spam cleaning day. :-)
[09:32:30] <deep42thought> yeah, my time stamping didn't improve a thing :-(
[09:33:24] <abaumann> I'm really all for: Russian -> Spam.
[09:35:31] <deep42thought> I'd like to filter them out during the registrations
[09:35:51] <deep42thought> russian => spam
[09:35:54] <deep42thought> bitcoin => spam
[09:35:59] <deep42thought> casino => spam
[09:36:05] <deep42thought> this would solve 98% of our spam
[09:36:13] <abaumann> right.
[09:36:33] <abaumann> and a puzzle: what does 'cat /proc/cpuinfo' do.
[09:36:38] <abaumann> also keeps away newbies. :-)
[09:37:04] <abaumann> identical posts.
[09:38:09] <deep42thought> I wouldn't block newbies
[09:39:20] <abaumann> scorpion26.
[09:39:32] <abaumann> this is the kind of people who should read a man page first.
[09:40:06] <abaumann> he has legit problems.. but I fear, any help is futile..
[09:43:06] <buildmaster> gst-plugins-good is broken (says buildknecht3).
[09:56:27] <deep42thought> abaumann: I'd guess scorpion26 is just impatient plus inexperienced - both can be cured :-)
[09:57:14] <deep42thought> this reminds me of someone, who - at the slightest sign of friction - reinstalled everything from scratch
[11:49:49] <deep42thought> tyzoid: What do you think of abaumann's idea to disallow posts with many/only cyrillic letters and such containing too many "casino"s and/or "bitcoin"s?
[14:35:48] <buildmaster> haskell-hspec-wai is broken (says buildknecht2). - I rescheduled: haskell-wai-extra.
[14:54:12] -!- deep42thought has quit [Quit: Leaving.]
[15:28:13] <buildmaster> hoogle is broken (says buildknecht). - I rescheduled: haskell-wai-logger.
[16:35:37] -!- abaumann has quit [Quit: leaving]
[16:46:36] <buildmaster> haskell-wai-extra is broken (says buildknecht2).
[16:57:10] <buildmaster> haskell-wai-logger is broken (says buildknecht2).
[17:01:00] <tyzoid> deep42thought: Sounds good, did you find a plugin that does that?
[18:15:48] -!- abaumann has joined #archlinux32
[18:24:05] <abaumann> tyzoid: if( preg_match( '/[\p{Cyrillic}]/u', $text) ) { # don't create post } :-)
[18:24:20] <abaumann> Are there filter plugins for fluxbb?
[18:25:29] <abaumann> which version of fluxbb is installed?
[18:27:21] <abaumann> I would check for standard ways in fluxbb to fight spam.
[18:27:23] <abaumann> https://github.com
[18:27:33] <abaumann> https://fluxbb.org
[18:35:36] <abaumann> girls: the first user with 0 posts complained he can no longer log in.
[18:36:12] <abaumann> maybe we should put a message somewhere explaining people, they should re-register, if they are unable to log in?
[18:44:40] -!- deep42thought has joined #archlinux32
[18:44:46] <deep42thought> abaumann: oops
[18:45:05] <abaumann> No problem. I answered him by email. :-)
[18:46:20] <deep42thought> great - it seems upstream starts using decimal pkgrels now *grrrr*
[18:46:30] <abaumann> oh no!
[18:46:33] <deep42thought> I'll have a closer look into this after dinner
[19:21:13] -!- belanthor has joined #archlinux32
[19:37:52] -!- arora has joined #archlinux32
[19:42:04] -!- Mart|n has joined #archlinux32
[20:05:41] <deep42thought> abaumann: I have some more data points with my freezing archlinux32 box
[20:06:08] <deep42thought> it does not freeze if the archiso is booted (some 4.13'ish kernel)
[20:06:09] <abaumann> ah.
[20:06:21] <deep42thought> I tried to compile the current kernel with that live cd
[20:06:28] <deep42thought> compiled for >1d w/o crash
[20:06:40] <deep42thought> but the new kernel crashed within 15 minutes ...
[20:07:15] <abaumann> mmh. I have everywhere 4.15.6, and never had a crash.. so maybe a driver/module for your laptop has a problem?
[20:07:46] <deep42thought> it's not a laptop
[20:07:51] <deep42thought> but yeah, might be
[20:08:02] <deep42thought> on the other hand, I'm pretty much on "standard configuration"
[20:08:08] <deep42thought> because I don't have a desktop on that box
[20:08:13] <deep42thought> I do not need graphics
[20:08:36] <deep42thought> so it's basically base + grub + some cryptfs configuration
[20:08:45] <abaumann> cryptfs, aeh?
[20:08:57] <deep42thought> cryptsetup luks*
[20:08:58] <abaumann> your cpu has all it's SSE2s, I hope.
[20:08:59] <deep42thought> I mean
[20:09:04] <deep42thought> yes
[20:09:11] <abaumann> but then you would get an illegal opcode..
[20:09:15] <abaumann> and not a freeze.
[20:09:17] <deep42thought> it *claims* it has :-)
[20:09:35] <deep42thought> let me grab it before it crashes :-)
[20:09:38] <abaumann> BTW: I have systemd running on 486. not sure, if I should be happy about thatn..
[20:09:50] <deep42thought> :-D
[20:09:53] <deep42thought> be happy!
[20:09:57] <deep42thought> don't worry!
[20:10:12] <abaumann> actually.. I wonder how little memory I can give the virtual box now..
[20:10:36] <abaumann> systemd: Caught [QUIT], dumped core..
[20:10:38] <abaumann> mmm..
[20:11:40] <abaumann> KASLR disabled: no suitable memory region! and junk in compressed archive (initrd).
[20:11:46] <abaumann> ok. 32 MB doesn't suffice..
[20:12:05] <abaumann> I have to go with a less fat boot system for my real i486 box..
[20:12:28] <deep42thought> lib32-fontconfig/repos/multilib-x86_64/PKGBUILD in revision 6e94f95c549bebb56a9dc2adf9432b91b71ee202 has "pkgrel=1.1" *grrr*
[20:12:50] <deep42thought> there should be something like distsystemd
[20:13:00] <abaumann> there is uinit.
[20:13:02] <abaumann> and rc.
[20:13:17] <deep42thought> e.g. you run a thin systemd client on i486 and the actual systemd runs on an amazon cluster
[20:13:32] <abaumann> It seems, I can boot without any trouble with another boot system and link all packages aganist systemd.
[20:13:39] <abaumann> lol
[20:14:00] <abaumann> I already have to use NBD and iPXE tricks to get the kernel onto the machine..
[20:14:23] <deep42thought> well, that's just the next logical step, then, I guess ;-)
[20:15:02] <abaumann> As always, be very very quiet if utterning this kind of ideas, they risk to get implemented. ;-)
[20:15:30] <abaumann> fontconfig-2.12.6+5+g665584a. So it's a patched git version.
[20:15:57] <abaumann> very soon packages will all have a non-publiched repo hash as version number *sigh8
[20:16:08] <deep42thought> I would just delete the sub_pkgrel from upstreams lib32-* packages
[20:16:16] <deep42thought> lol
[20:16:34] <deep42thought> the nsa should have the content listed under the hash ...
[20:16:47] <deep42thought> no matter if it was public or "unpublished"
[20:17:26] <abaumann> publishing via prism..
[20:17:35] <deep42thought> exactly
[20:24:10] <abaumann> systemd[1]: Freezing execution.
[20:24:10] <abaumann> Mar 25 18:23:54 archlinux systemd[1]: Freezing execution.
[20:24:10] <abaumann> Mar 25 18:23:54 archlinux systemd-coredump[267]: Process 266 (systemd) of user 0 dumped core.
[20:24:13] <abaumann>
[20:24:15] <abaumann> Stack trace of thread 266:
[20:24:18] <abaumann> #0 0x00000000b79e8591 kill (libc.so.6)
[20:24:20] <deep42thought> the forums mention (in contrast to the wiki) decimal pkgrel for differences between i686 and x86_64
[20:24:21] <abaumann> #1 0x00000000004b71a3 n/a (systemd)
[20:24:24] <abaumann> #2 0x00000000b7ed9ce0 __kernel_sigreturn (linux-gate.so.1)
[20:24:27] <abaumann> #3 0x00000000b7a9dd91 epoll_pwait (libc.so.6)
[20:24:30] <abaumann> #4 0x00000000b7a9e03a epoll_wait (libc.so.6)
[20:24:33] <abaumann> #5 0x00000000b7daf2f0 sd_event_wait (libsystemd-shared-237.so)
[20:24:36] <abaumann> #6 0x00000000b7db0252 sd_event_run (libsystemd-shared-237.so)
[20:24:39] <abaumann> #7 0x00000000004f0130 n/a (systemd)
[20:24:42] <abaumann> #8 0x00000000004b2865 n/a (systemd)
[20:24:45] <abaumann> #9 0x00000000b79d3683 __libc_start_main (libc.so.6)
[20:24:48] <abaumann> #10 0x00000000004b403d n/a (systemd)
[20:24:51] <abaumann> This sounds like some dbus probme.
[20:25:30] <deep42thought> abaumann: my screen is littered, please use ptpb.pw :-)
[20:25:37] <abaumann> yeah. sorry. :-)
[20:30:07] <deep42thought> btw: I'll, for now, remove upstreams sub_pkgrel from lib32-* packages
[20:30:27] <deep42thought> this approach might even be valid for _all_ packages - we will append unique sub_pkgrels in the end anyway
[20:34:08] <deep42thought> ah, crap, forgot to grab /proc/cpuinfo from my archlinux32 box and in the meantime it crashed again ...
[20:37:14] <abaumann> yeah. sounds like a good fix to drop the sub_pkgrels.
[20:37:29] <abaumann> hope nobody makes a depends= on a sub_pkgrel..
[20:38:25] <deep42thought> https://ptpb.pw
[20:39:07] <abaumann> well. somewhere around a P4, I would say.
[20:39:14] <abaumann> even PAE.
[20:40:05] <abaumann> An option is to compile in the MagicSys Oops backdoor into the kernel and see, where it freezes.
[20:40:19] <abaumann> It doesn't simply shut down, does it?
[20:42:05] <deep42thought> no, it simply freezes
[20:42:13] <deep42thought> I see the "user:" login promt
[20:42:18] <deep42thought> *prompt
[20:42:34] <deep42thought> I had "watch sensors" open one time and it just stopped refreshing
[20:42:36] <abaumann> You could run a top somewhere in a virtual screen.
[20:42:44] <deep42thought> and my music hangs in a ~.25 seconds loop
[20:42:50] <abaumann> If you can still swtich with Alt-Arrow keys..
[20:42:55] <deep42thought> cannot
[20:43:07] <deep42thought> even num is dead
[20:43:14] <abaumann> This could be the sound stuck in the chip repeating over and over again.
[20:43:26] <abaumann> so, this means, even interrupt handling is out..
[20:43:48] <abaumann> Hard to say, if you can enter the linux Magic area then at all.
[20:44:42] <deep42thought> yes, I guessed that too, with the sound
[20:44:52] <deep42thought> it just plays its buffer endlessly
[20:46:09] <deep42thought> could it be, that it's some hardware problem of the cf-card?
[20:46:22] <deep42thought> (/ is on a cf card)
[20:46:42] <deep42thought> so it only is visible, when booted normally, not from live cd?
[20:46:54] <abaumann> I doubt. Unless it's very low level and a CF operation freezes the driver system.
[20:47:07] <abaumann> Can you run from an USB stick?
[20:47:20] <deep42thought> yes
[20:47:27] <deep42thought> that's how I compiled the kernel
[20:47:35] <abaumann> ah. and there it runs stable.
[20:47:44] <deep42thought> it's a different system from usb
[20:47:49] <deep42thought> it's our live iso
[20:47:53] <abaumann> ah.
[20:48:01] <deep42thought> haven't tried the original system
[20:48:04] <deep42thought> should I?
[20:48:17] <abaumann> to eliminate the possibility of the CF being faulty you need an identical system on an USB stick for instance.
[20:48:21] <abaumann> same kernel.
[20:48:34] <abaumann> that or bisecting kernel changelogs and changes.. :-)
[20:48:57] <deep42thought> bisecting is not an option - it takes ~1 day to compile on that box ...
[20:49:10] <abaumann> urghh.
[20:49:23] <deep42thought> "cpu MHz : 400.060"
[20:49:31] <abaumann> underclocked.
[20:49:43] <deep42thought> it's already 60kHz overclocked ;-P
[20:49:49] -!- Mart|n has quit [Remote host closed the connection]
[20:51:21] <deep42thought> I'll take "dd if=/dev/sda of=/dev/null" as a cf-card test ...
[20:51:27] <abaumann> mmh.
[20:51:30] <abaumann> https://forums.gentoo.org
[20:51:34] <abaumann> http://lukasz.dk
[20:51:40] <abaumann> CPU scaling.
[20:51:59] <abaumann> ok, it's a 1.2 GHZ.
[20:52:33] <abaumann> CF could freeze, if it tries to relocate storage and cannot find some. But I would expect the driver to handle something like that as I/O error.
[20:53:19] <abaumann> BIOS version and update?
[20:53:37] <deep42thought> umm
[20:53:43] <deep42thought> haven't had a look into the bios
[20:53:51] <deep42thought> I'm not even sure how to enter it :-/
[20:54:14] <abaumann> oh. one of those lapopts. I know the feeling.. had a Toshiba with that effect..
[20:54:30] <abaumann> SMP.
[20:54:50] <deep42thought> it's not a laptop, it's a thin client
[20:55:10] <abaumann> yeah. SMP and ACPI are good candidates to cause trouble..
[20:56:26] <abaumann> Well.. I'm just guessing right now. :-)
[20:57:08] <deep42thought> what makes me wonder, is, that our kernel from the live cd boots, but the newer, installed one does not ...
[20:57:21] <deep42thought> hmm, I could bisect with our archive
[20:57:49] <deep42thought> and then have a look, what changed in the kernel config
[20:58:18] <abaumann> yepp. good idea.. maybe this gives a hint..
[20:58:46] * deep42thought starts the hunt ...
[20:59:37] <deep42thought> um, hunt is over
[20:59:54] <abaumann> short hunt.. or slow deer.. ;-)
[21:00:08] <deep42thought> the iso has linux-4.14.13, which runs fine - and I had compiled and run linux-4.13.11 which froze within minutes
[21:00:33] <deep42thought> but this was the self-compiled kernel ...
[21:00:39] <deep42thought> let me check an official one :-)
[21:00:44] <eschwartz> So "something was fixed" then?
[21:00:55] <eschwartz> Seems like a good bug to have
[21:01:00] <deep42thought> well, not exactly
[21:01:23] <deep42thought> it's more like "if you compile the kernel with the newer gcc, it won't run"
[21:01:26] <deep42thought> or something like that
[21:01:35] <deep42thought> (I'm not sure, gcc was updated in the meantime)
[21:01:41] <eschwartz> Ouch.
[21:02:07] <eschwartz> Well is that a gcc 7.3.1 thing then?
[21:02:21] <deep42thought> I don't know
[21:02:38] <deep42thought> I just see a freezing box, which runs fine with the older iso (with a newer kernel)
[21:15:05] <deep42thought> cool
[21:15:28] <deep42thought> I made a mistake when installing the new kernel (somehow the modules were not installed) and now the machine froze even in the rescue console ...
[21:17:05] -!- abaumann has quit [Quit: leaving]
[21:25:38] -!- Mart|n has joined #archlinux32
[21:33:49] <deep42thought> ok, turning off cpu frequency scaling doesn't help either
[21:33:51] <deep42thought> :-(
[21:34:49] <deep42thought> and linux-4.14.15 also freezes (it's the oldest one of us in the archive)
[21:42:10] -!- arora has parted #archlinux32
[21:51:43] -!- deep42thought has quit [Quit: Leaving.]
[22:45:50] -!- Mart|n has quit [Remote host closed the connection]
[22:57:01] -!- Mart|n has joined #archlinux32