New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Test failures on Alpine Linux building notcurses 2.2.1 #1344
Comments
egads! please feel free to disable/clear it for now, and thanks for the heads-up. |
i've got an Alpine builder in my CI pipeline, but this didn't trigger on my setup. i'm looking into it, thanks and apologies. |
you know, i haven't updated it recently -- there may well be a new musl or something. let me check that out. |
ok, there were a healthy number of updates on my CI, so hopefully this'll flush it out on my side:
because i certainly don't see anything that could have caused this breakage between 2.2.0 and 2.2.1! there were only like four commits, lol. we'll know soon... |
bah, still passes on my alpine builder :( https://drone.dsscaw.com:4443/dankamongmen/notcurses/5516/2/2. so let's see what we can derive from your pasted output. i don't believe i've ever seen |
"Child crashed with a numerical exception." in |
@PureTryOut , I'm pretty bewildered by this failure report:
So I can't reproduce this, and moreover can't at all comprehend how this went down. I hate to ask, but were there any other weird failures on this particular autobuilder recently that might point to hardware issues? Was it changed recently? If I can get access to the machine, or even a coredump, I'm sure I can track this down, but right now I'm sitting with a few hours invested and zilch results. =[ Very much want to get this tracked down, though. |
I've just made changes to |
I sadly don't have the rights to give you any access to the builders, and neither do I have the knowledge of the state of the machine. You'll have to ask some other dev on #alpine-devel on Freenode, sorry. |
no worries, i'll jump in there and ask for some help later this evening. thanks a ton! |
The latest issue was caused by the x86_64 builder using screen-256color. Notcurses assumes that the |
I worked around it by setting TERM=vt100 when running tests, but I think solving the |
definitely, how embarrassing. thanks @kaniini !!!! |
hrmmm, weren't we seeing failures in direct mode? it looks pretty well protected against |
definitely a missing check in |
i fixed up |
The deficiency is in
Those fprintf were at the beginning of the function, middle after sgr/sgr0 were checked, and return back to ncdirect_set_styles. |
that's correct behavior, though. we only emit int r = -1;
if(stylebits == 0 && n->tcache.sgr0){
r = term_emit(n->tcache.sgr0, n->ttyfp, false);
}else if(n->tcache.sgr){
r = term_emit(tiparm(n->tcache.sgr, stylebits & NCSTYLE_STANDOUT,
stylebits & NCSTYLE_UNDERLINE,
stylebits & NCSTYLE_REVERSE,
stylebits & NCSTYLE_BLINK,
stylebits & NCSTYLE_DIM,
stylebits & NCSTYLE_BOLD,
stylebits & NCSTYLE_INVIS,
stylebits & NCSTYLE_PROTECT, 0), out, false);
} it does not take the first conditional because |
The problem is that there is not a fallback for the case where |
oh assertion problems i got you. i thought we were dereferencing |
That should be fixed. There will be further breakage in |
the sgr issues are likely related to the emoji case too. or it is similar. as soon as i forced TERM=vt100, the tests passed. |
ok in the rendered mode, lack of |
wasn't there some manner of timeout on s390x recently? |
i'm now able to succeed despite lacking |
so
w00000000000000000000000000t. yay! couldn't have done it without you @kaniini , this was pretty much your fix. |
https://build.alpinelinux.org/buildlogs/build-edge-x86_64/community/notcurses/notcurses-2.2.1-r0.log
@dankamongmen seems you maintain our Alpine Linux package, could you look into this? It is currently blocking our x86_64 builder 😉
The text was updated successfully, but these errors were encountered: