02:38
guifa left
03:00
guifa joined
06:27
bartolin left
06:28
bartolin joined
08:38
sena_kun joined
08:52
sena_kun left
08:53
sena_kun joined
08:57
sena_kun left
|
|||
timo | a new challenger has appeared: ci.debian.net/packages/n/nqp/testi...6840/#L139 | 15:45 | |
lizmat | interesting :-) | 15:57 | |
repeatable ? | 15:59 | ||
otherwise it feels like a flapper? | 16:00 | ||
the "no plan in TAP output" indicates to me that something with loading a module went awry and segfaulted | 16:01 | ||
timo | i'm trying to reproduce it | 16:04 | |
patrickb | Just so I'm not lost, the are now three failures you were recently looking at, one on Windows, one on ARM and now another one on ARM? None of the three are solved, right? | 16:06 | |
timo | i've solved one on arm | 16:43 | |
this new one is a crash of some kind in an nqp test, the one i fixed was rakudo hanging - theoretically it could hang any time, but it occured during core setting compilation on debian builders | 16:44 | ||
non-zero wait status: 11 sounds like segfault | 16:45 | ||
lizmat | as I said, the "Parse errors: No plan found in TAP output" means it crashed *before* it could execute the plan | ||
which in that file, is like the very first thing | 16:46 | ||
timo | ah, so i shouldn't try to reproduce the crash from that one test file | 16:47 | |
but instead, with anything at all really? | |||
lizmat | that would be my guess, yes | ||
timo | ok i requested a rerun of the test for now | 16:48 | |
ci.debian.net/status/pending/?arch%5B%5D=armel has a bunch of packages in the queue before nqp gets tested | 16:50 | ||
lizmat | keeping fingers crossed for a reproducible error | 16:59 | |
timo | [Coke]: you're running Terminal::MakeRaw version 1.0.0, the fix for memory corruption because of an incorrect struct definition on mac and bsd is fixed in 1.0.1, please update and try again | 17:09 | |
[Coke] | ah. I remember the note that we didn't need to update terminal::lineeditor's deps - but if ther'es a bug fix, we should require the fixed version, at least | 17:21 | |
lizmat pings japhb | 17:22 | ||
[Coke] | timo++ | 17:23 | |
[Coke] opened a ticket, linked from moarvm ticket, closed moarvm version | |||
timo | the nqp test failure did not reproduce on the debian CI build job either | 17:24 | |
lizmat | tis what I feared :-( | 17:25 | |
timo | ummmmmm | ||
A fatal error internal to GDB has been detected, further | |||
debugging is not possible. GDB will now terminate. | |||
lizmat | oops | 17:26 | |
timo | couldn't repro the gdb crash either | 18:09 | |
lizmat | .oO( those damned cosmic rays ) |
18:14 | |
18:37
sena_kun joined
21:33
sena_kun left
|
|||
timo | what was that, 5 hours of running it on repeat, no repro | 21:46 | |
[Coke] | oof | 21:56 | |
timo | i'll see how long the porterbox guest account lasts by itself, when i would need to renew / rerequest it | 21:57 |