Home freenode/#haskell: Logs Calendar

Logs: freenode/#haskell

←Prev  Next→
Page 1 .. 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 .. 5022
— 502,152 events total
2020-11-02 19:59:34 <tomsmeding> well, you can switch to a different tty :p
2020-11-02 19:59:42 <bqv> you'd think...
2020-11-02 19:59:44 <bqv> but no :p
2020-11-02 19:59:49 <tomsmeding> assuming the ctrl-alt-FN thing still goes through
2020-11-02 20:00:01 <bqv> had to ssh in via phone to kill gdb
2020-11-02 20:00:38 → justsomeguy joins (~justsomeg@2603:9000:960a:fc00:c4d2:5b73:c7b7:3a2b)
2020-11-02 20:00:38 × justsomeguy quits (~justsomeg@2603:9000:960a:fc00:c4d2:5b73:c7b7:3a2b) (Changing host)
2020-11-02 20:00:38 → justsomeguy joins (~justsomeg@unaffiliated/--/x-3805311)
2020-11-02 20:01:08 <bqv> is the class of bugs that vanish under strace/ltrace known?
2020-11-02 20:01:13 → vacm joins (~vacwm@70.23.92.191)
2020-11-02 20:01:23 <bqv> or will i just have to wait to see what gdb does
2020-11-02 20:01:30 × alx741 quits (~alx741@186.178.110.193) (Quit: alx741)
2020-11-02 20:01:44 bqv !
2020-11-02 20:01:48 <bqv> i can qemu this
2020-11-02 20:01:51 <sm[m]> space leaks often vanish when you add haskell tracing. Not sure about your situation there
2020-11-02 20:02:27 Gurkenglas__ is now known as Gurkenglas
2020-11-02 20:02:38 <sm[m]> but that's why I asked about memory usage
2020-11-02 20:02:38 <bqv> sm[m]: i've compiled with and without executable profiling enabled
2020-11-02 20:02:43 <bqv> i don't think it affected anything
2020-11-02 20:02:44 → meowphius joins (~meowphius@pool-173-76-247-88.bstnma.fios.verizon.net)
2020-11-02 20:02:46 × britva quits (~britva@2a02:aa13:7240:2980:292a:d07d:3ab9:26ba) (Quit: This computer has gone to sleep)
2020-11-02 20:02:55 <bqv> is that what you had in mind, or a specific kind?
2020-11-02 20:03:09 <tomsmeding> FWIW, you can simulate ctrl-alt-FN over ssh by running 'sudo chvt N'
2020-11-02 20:03:18 × ensyde quits (~ensyde@99-185-235-117.lightspeed.chrlnc.sbcglobal.net) (Ping timeout: 260 seconds)
2020-11-02 20:03:22 <bqv> ah, yes
2020-11-02 20:03:22 <sm[m]> bqv: no I haven't suggested haskell profiling (since it sounded like a FFI/c-related issue)
2020-11-02 20:03:40 <tomsmeding> run gdb in a tmux, start the thing, observe screen freeze, chvt over ssh, tmux a, and continue?
2020-11-02 20:04:16 <bqv> hmm
2020-11-02 20:04:24 <bqv> maybe worth a try
2020-11-02 20:04:33 → refried_ joins (~textual@pool-108-20-26-90.bstnma.fios.verizon.net)
2020-11-02 20:04:43 <tomsmeding> (I was honestly surprised that chvt over ssh works)
2020-11-02 20:05:21 <sm[m]> bqv: I don't think this is your issue but I was hinting to watch it in top once to make sure it's not being killed for using all memory - since this is pretty common with haskell apps
2020-11-02 20:06:00 <bqv> i imagine that would be logged in dmesg too, anyway, right?
2020-11-02 20:06:09 <tomsmeding> (and because I use an external display with the laptop lid shut, switching to a different tty over ssh made the systemd-inhibit process I have running nonfunctional, thus immediately sending my laptop to sleep...)
2020-11-02 20:06:26 <sm[m]> perhaps, probably so
2020-11-02 20:07:34 → nullheroes joins (~danielvu@168.235.66.22)
2020-11-02 20:08:40 <bqv> tomsmeding: nice! it worked
2020-11-02 20:08:45 <tomsmeding> :D
2020-11-02 20:08:45 × miklcct quits (quasselcor@2001:19f0:7001:5ad:5400:2ff:feb6:50d7) (Remote host closed the connection)
2020-11-02 20:08:48 <tomsmeding> and?
2020-11-02 20:09:01 <bqv> not helpful, because it works in gdb too :D
2020-11-02 20:09:01 <tomsmeding> convoluted hacks are best hacks
2020-11-02 20:09:09 <tomsmeding> 🎉
2020-11-02 20:09:15 <tomsmeding> #fixed
2020-11-02 20:09:30 <bqv> this is bizarre. am i just consigned to always run this with a debugger attached
2020-11-02 20:09:32 <tomsmeding> is it reproducible outside tracing tools?
2020-11-02 20:09:37 <bqv> to collapse the bug's wavefunction or something
2020-11-02 20:09:38 → geekosaur67 joins (82659a09@host154-009.vpn.uakron.edu)
2020-11-02 20:09:44 <tomsmeding> does it happen always or just sometimes
2020-11-02 20:09:48 <bqv> it is, yes, it's perfectly reproducible
2020-11-02 20:09:53 × coot quits (~coot@37.30.60.135.nat.umts.dynamic.t-mobile.pl) (Remote host closed the connection)
2020-11-02 20:09:57 → miklcct joins (quasselcor@2001:19f0:7001:5ad:5400:2ff:feb6:50d7)
2020-11-02 20:10:03 <bqv> without strace/ltrace/gdb, the program always exits within a second
2020-11-02 20:10:21 <bqv> exitcode 0, no relevant error logged
2020-11-02 20:10:35 <tomsmeding> exitcode 0 also of your haskell process, or only of the launcher around it?
2020-11-02 20:10:37 → ensyde joins (~ensyde@99-185-235-117.lightspeed.chrlnc.sbcglobal.net)
2020-11-02 20:10:47 <bqv> oh, that's a good point! i hadn't checked.
2020-11-02 20:11:01 × justsomeguy quits (~justsomeg@unaffiliated/--/x-3805311) ()
2020-11-02 20:11:09 <tomsmeding> c.f. my remark above that I bet something is getting a signal and the exit code is getting swallowed :p
2020-11-02 20:11:13 → elliott_ joins (~elliott_@pool-108-51-141-12.washdc.fios.verizon.net)
2020-11-02 20:11:16 <tomsmeding> I've seen this before
2020-11-02 20:11:27 → mmohammadi9812 joins (~mmohammad@188.210.118.100)
2020-11-02 20:11:29 <tomsmeding> (I think)
2020-11-02 20:11:47 <geekosaur67> +RTS -v0 ?
2020-11-02 20:12:03 <geekosaur67> (C call interrupted by RTS itimer)
2020-11-02 20:12:09 × geekosaur quits (82659a09@host154-009.vpn.uakron.edu) (Ping timeout: 245 seconds)
2020-11-02 20:12:17 <bqv> ha
2020-11-02 20:12:19 <bqv> you were right
2020-11-02 20:12:23 <bqv> it's getting USR1
2020-11-02 20:12:27 <bqv> but it is exiting with 0
2020-11-02 20:12:54 → aarvar joins (~foewfoiew@50.35.43.33)
2020-11-02 20:12:56 <bqv> does haskell die on SIGUSR1?
2020-11-02 20:13:37 <geekosaur67> anything that doesn't specifically handle it will die
2020-11-02 20:13:56 <tomsmeding> $ man 7 signal
2020-11-02 20:14:01 <tomsmeding> SIGUSR1 has 'Term' as the default action
2020-11-02 20:14:13 <bqv> that seems extreme
2020-11-02 20:14:17 <tomsmeding> USR1 is a really weird one to get though
2020-11-02 20:14:25 <tomsmeding> most signals have Term as default, only a few don't
2020-11-02 20:14:57 <bqv> i feel like it's the launcher sending it
2020-11-02 20:15:02 <tomsmeding> SIGHUP, or SIGBUS/SIGSEGV, SIGPIPE, I would've expected
2020-11-02 20:15:17 × ensyde quits (~ensyde@99-185-235-117.lightspeed.chrlnc.sbcglobal.net) (Ping timeout: 260 seconds)
2020-11-02 20:15:50 <bqv> i'll try adding a System.Posix.Signals handler
2020-11-02 20:16:49 → merijn joins (~merijn@83-160-49-249.ip.xs4all.nl)
2020-11-02 20:17:53 <tomsmeding> if the process doesn't actively catch USR1 somehow, it shouldn't exit with code 0 due to it, but rather code 138
2020-11-02 20:17:53 × Tario quits (~Tario@201.192.165.173) (Read error: Connection reset by peer)
2020-11-02 20:18:00 <tomsmeding> (128 + (10 == SIGUSR1))
2020-11-02 20:18:21 <bqv> i checked by using bash -c "prog; echo $?"
2020-11-02 20:18:30 <bqv> it echoed 0, but maybe that's a bash quirk
2020-11-02 20:18:37 → Tario joins (~Tario@201.192.165.173)
2020-11-02 20:19:06 <bqv> hang on, maybe it wasn't recieving it, maybe it was sending it
2020-11-02 20:19:13 → rprije joins (~rprije@194-193-168-77.tpgi.com.au)
2020-11-02 20:19:15 <monochrom> I don't know whether it's relevant. It may be safer to use the threaded runtime system (give -threaded to GHC at link time). This eliminates a SIGVTALRM that may interfere with some syscalls.
2020-11-02 20:19:36 <bqv> i'm already using -threaded by necessity
2020-11-02 20:19:45 <merijn> oh, oh, oh
2020-11-02 20:19:52 <tomsmeding> bqv: is your program starting X? maybe https://www.x.org/archive/X11R6.8.1/doc/Xserver.1.html#sect11 ?
2020-11-02 20:19:52 <monochrom> Ah OK.
2020-11-02 20:19:56 <merijn> People are talking posix signals! My time to shine!
2020-11-02 20:19:59 × flatmap quits (~flatmap@p200300dd370a47007ce3847efe0685c7.dip0.t-ipconnect.de) (Ping timeout: 268 seconds)
2020-11-02 20:19:59 <merijn> You're all wrong
2020-11-02 20:20:06 <bqv> tomsmeding: wayland, unfortunately
←Prev  Next→
Page 1 .. 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 .. 5022

All times are in UTC.