IRC channel logs

2024-01-10.log

back to list of logs

<cow_2001>guile shouldn't end with a segmentation fault normally, right?
<ieure>lol no
<ieure>Were you calling out to C code?
<cow_2001>no, doing something in pure guile
<ieure>Nope, shouldn't happen.
<ieure>Can you reproduce the crash? If so, you should report a bug.
<cow_2001>it's a bit tricky to reproduce with all the code here: https://codeberg.org/kakafarm/guile-rsv/src/branch/segmentation-fault i'm running the line in "segmentation-fault" file
<cow_2001>maybe
<cow_2001>let's see
<cow_2001>using this branch as a little glass to keep this skittish bug in place
<cow_2001>oh wait
<cow_2001>sending zero arguments when the procedure needs 3
<ieure>That shouldn't segfault.
<cow_2001>no, i don't think i could reproduce it in a minimal example
<cow_2001>maybe with help?
<ieure>Your example is pretty minimal already, if you have a repo someone can pull, that's probably fine.
<ieure>(I am not a Guile developer, I have no authority, I'm just some guy)
<cow_2001>thank you
<cow_2001>sending a report
<cow_2001>ieure: aaaaa what i hate about this reporting thingamajig is that it's never good enough. i need to get it right the first time!
<cow_2001> https://debbugs.gnu.org/cgi/bugreport.cgi?bug=68362
<sneek>wb dsmith :)
<sneek>wb janneke!
<janneke>sneek: botsnack
<sneek>:)
<rekado>cow_2001: I don’t really understand the issue. “read-null” is a procedure with three arguments. You’re calling it without any arguments.
<rekado>is this on purpose?
<rekado>is the problem that “Wrong number of arguments to 254” should be “Wrong number of arguments to read-null”?
<rekado>(I don’t get a segmentation fault)
<cow_2001>run: guile -c "(define (f a) '()) (f)"
<cow_2001>you'll get a certain error
<cow_2001>it'll show you the procedure and how many arguments it needs
<cow_2001>guile -c "(define (f a b c) '()) (f)" # ice-9/eval.scm:339:13: Wrong number of arguments to #<procedure f (a b c)>
<cow_2001>maybe it's a segmentation fault that comes up only in a previous version of guile, but the error message shown in the new version is still a bug
<cow_2001>rekado: maybe i should add this to the bug report in another message?
<rekado>yes, perhaps also retitle it because recent versions don’t seem to trigger a segfault
<dsmith>sneek, botsnack
<sneek>:)
<dsmith>!uptime
<sneek>I've been faithfully serving for 2 months
<sneek>This system has been up 40 weeks, 3 days, 19 hours, 51 minutes
<mwette>I believe that error with missing name was reported and fixed.
<civodul>hey!
<civodul>thoughts about the patch at https://issues.guix.gnu.org/68087 ?
<civodul>i’d like to push it soon if there are no objections
<old>.
<ArneBab>mwette: as first step I sent a simpler patch that just checks for the presence of the GUILE_QUIET environment variable. Did you see my patch on guile-devel?