[racket] weird build failed after changing one character in docs that shouldn't affect anything

From: Matthew Flatt (mflatt at cs.utah.edu)
Date: Mon Jul 14 01:24:01 EDT 2014

Thanks for the report! I haven't been able to replicate the crash so
far, and there's a small chance that commit c72f441d93 in the Racket
repo was a relevant repair, but I will keep an eye out for more
problems like this.

At Sun, 13 Jul 2014 12:00:17 -0400, "Alexander D. Knauth" wrote:
> Not sure if this is related (I could be my fault this time), but I got another 
> build failed:
> $ /usr/racket/bin/raco test -x .
> raco test: (submod "./alist.rkt" test)
> raco test: (submod "./conditionals.rkt" test)
> SIGSEGV MAPERR si_code 1 fault on addr 0x10040
> conditionals.rkt: test: non-zero exit: 134
> raco test: (submod "./dict.rkt" test)
> raco test: (submod "./egal.rkt" test)
> raco test: (submod "./lambda-reader.rkt" test)
> raco test: (submod "./str.rkt" test)
> raco test: (submod "./test.rkt" test)
> raco test: (submod "./threading.rkt" test)
> raco test: (submod "./utils.rkt" test)
> 1/9 test failures
> The command "/usr/racket/bin/raco test -x ." exited with 1.
> Done. Your build exited with 1.
> What I did was change the docs so that I could show a #fn anonymous function 
> literal within a @racket[] form and changed the #fn examples to use that.  I 
> don’t see how that could affect the conditionals tests, so I thought I’d post 
> this.  
> https://travis-ci.org/greghendershott/rackjure/jobs/29779055#L84
> https://github.com/greghendershott/rackjure/pull/42
> https://github.com/AlexKnauth/rackjure/commit/7602d631112e15d33bab3356f34818f0af
> ec6682
> 
> On Jul 11, 2014, at 8:49 PM, Greg Hendershott <greghendershott at gmail.com> wrote:
> 
> > Before noticing this email, I noticed that the build had failed on
> > Racket HEAD with a segfault. Builds for all other Racket versions were
> > OK.
> > 
> > I figured it was some temporary weirdness on HEAD, and without
> > thinking much, I restarted the build on Travis.
> > 
> > This time it succeeded fine, even on HEAD.
> > 
> > 
> > If I'd noticed this email first, I would have copied the exact
> > segfault message. Unfortunately now I can't find it on Travis.
> > Although Travis keeps a build history, that doesn't seem to include
> > errored builds that you manually restart.
> > 
> > 
> > On Fri, Jul 11, 2014 at 7:15 PM, Alexander D. Knauth
> > <alexander at knauth.org> wrote:
> >> I was trying to make docs for ~>% and ~>_ in my fork of rackjure, and the 
> travis cl builds were passing,
> >> but then I changed one character in the documentation to fix a typo and then 
> all of a sudden the build failed.
> >> All I did was change one % to an _ within a @racketblock so this:
> >> @racketblock[
> >> (~>% #”foobar” bytes-length (number->string _ 16) string->bytes/utf-8)
> >> ]}
> >> to this:
> >> @racketblock[
> >> (~>_ #”foobar” bytes-length (number->string _ 16) string->bytes/utf-8)
> >> ]}
> >> And then all of a sudden the build failed.
> >> Is there anything scribble or anything else does that would make this not 
> work?
> >> 
> >> https://github.com/greghendershott/rackjure/pull/42
> >> 
> https://github.com/AlexKnauth/rackjure/commit/39473e967a186d9a1d53aede461ba95d5b
> ef16ac
> >> 
> >> 
> >> ____________________
> >>  Racket Users list:
> >>  http://lists.racket-lang.org/users
> 
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users


Posted on the users mailing list.