On 20251107@21:03, John Paul Adrian Glaubitz wrote:
> Source: fim
> Version: 0.7.1-4
> Severity: normal
> User: [email protected]
> Usertags: sparc64
> X-Debbugs-Cc: [email protected],[email protected]
> 
> Hi,
> 
> fim FTBFS on sparc64 due to unaligned access (Bus error):
> 
> ! ( echo /build/reproducible-path/fim-0.7.1 | grep '[\\ :"'"'"']' ) # reject 
> unreasonable base directory names
> ! grep TODO doc/fimrc.man
> ulimit -c unlimited || true
> ulimit -v 32000 || true
> ulimit -t 1000 || true
> timeout --foreground 30 ./src/fim --no-rc-file --no-etc-fimrc-file 
> --no-history -o dumb --sanity-check
> Using the default "dumb" display device instead.
> s 98 b 1 1 1 c 99.000000 
> s 98 b 1 1 1 c 99.000000 
> fim display check : please be patient
> fim display check : 2.54588e+07 clears/s
> fim browser push check: please be patient
> fim browser push check: 149561.156250 push_path()/s
> fim browser pop check: please be patient
> fim browser pop check: 230909 pop()/s
> fim console check: please be patient
> fim console random variables set/get test: 32280.5 set/get /s
> timeout --foreground 30 ./src/fim --no-rc-file --no-etc-fimrc-file 
> --no-history -o dumb -c 'stdout random ; quit 0;' # make sure no leak from 
> 'random'
> Using the default "dumb" display device instead.
> make[2]: *** [Makefile:1306: tests] Bus error
> 
> This is usually easy to debug and fix with the help of GDB.
> 
> Full build log available in [1]. Access to a SPARC machine is either
> available through Debian [2] or the GCC Compile Farm [3].
> 
> Thanks,
> Adrian
> 
> > [1] 
> > https://buildd.debian.org/status/fetch.php?pkg=fim&arch=sparc64&ver=0.7.1-4&stamp=1762545347&raw=0
> > [2] https://db.debian.org/machines.cgi?host=stadler
> > [3] https://gcc.gnu.org/wiki/CompileFarm
> 
> --
>  .''`.  John Paul Adrian Glaubitz
> : :' :  Debian Developer
> `. `'   Physicist
>   `-    GPG: 62FF 8A75 84E0 2956 9546  0006 7426 3B37 F5B5 F913

Thanks for telling of the problem Adrian!

I added a patch for this:
 
https://salsa.debian.org/debian/fim/-/commit/1facb3d64e59320aea42f40cfbd8d42de9b978b9
and I will patch it upstream too -- perhaps I'll need a compile-time
mechanism to know when unaligned access is problematic. 

cheers,
Michele

Attachment: signature.asc
Description: PGP signature

Reply via email to