On Sun, Aug 26, 2012 at 07:17:07PM -0300, Antonio Terceiro wrote: > >shawn escreveu isso aí: >> On Wed, 2012-08-08 at 17:19 +0200, Cédric Boutillier wrote: >> > Hi! >> > >> > ruby-gherkin does not build on the current architectures: armel, armhf, >> > powerpc, s390 and s390x. Since I do not have access to any of these >> > architectures, I was wondering if someone else had any clue about why >> > the compilation fails (the failing tests are the same on all these >> > architectures). >> >> Without having looked at the code, this list of failing architectures >> makes me almost certain it is a problem of assuming the C type "char" is >> signed. The easy fix is to change the offending "char" declaration >> "signed char".
Shawn is exactly right, AFAICS. It *is* an invalid assumption about signed-char. I don't know what happened to anybody else's attempts to debug this, but I just did a build on harris (armhf porter box). The current unstable source fails to build with the same symptoms as described. Running the following command: $ find . -name '*.rb' | xargs grep -l CFLAGS | xargs sed -i 's/Wall/Wall -fsigned-char/g' fixes the issue - the build succeeds after that. The equivalent patch to make that change in the source is quite simple to generate for inclusion in debian/patches, but messy because of the big set of separate extconf.rb files. This package is a *mess* as shipped by upstream, with lots of needless repetition. In fact, looking at the .c source files here, I'm tempted to say that we don't actually *have* real source in this package. 93sam@harris:~/build/ruby-gherkin/ruby-gherkin-2.11.1$ grep ^#line ext/gherkin_lexer_fi/gherkin_lexer_fi.c #line 1 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 254 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 89 "ext/gherkin_lexer_fi/gherkin_lexer_fi.c" #line 258 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 819 "ext/gherkin_lexer_fi/gherkin_lexer_fi.c" #line 425 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 826 "ext/gherkin_lexer_fi/gherkin_lexer_fi.c" #line 83 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 89 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 93 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 98 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 108 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 112 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 116 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 120 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 124 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 128 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 132 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 136 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 143 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 148 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 153 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 158 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 162 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 168 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 175 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 179 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 185 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 189 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 203 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 207 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 1116 "ext/gherkin_lexer_fi/gherkin_lexer_fi.c" #line 207 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" #line 1179 "ext/gherkin_lexer_fi/gherkin_lexer_fi.c" #line 426 "/Users/ahellesoy/github/gherkin/tasks/../ragel/i18n/fi.c.rl" Most of the "source" here is being generated from fi.c.rl, and we don't have that file in this package, nor anywhere else in the Debian archive. Ick. :-( Is there a good reason why? -- Steve McIntyre, Cambridge, UK. st...@einval.com "It's actually quite entertaining to watch ag129 prop his foot up on the desk so he can get a better aim." [ seen in ucam.chat ] -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org