Re: Investigation of the failing build of gnat-4.9 using fakeroot-hurd
Hello :)
Quoting Svante Signell (2015-02-03 12:42:32)
> > In your particular case, did you check whether the error is right,
> > i.e. it tried to do something to the file on the ext2fs that the user
> > is not allowed to do ?
>
> Well, the printouts from libdiskfs shows a uid and gid of 1000, i.e they
> are not faked. Nevertheless, libdiskfs logs show:
It is *really* hard to interpret your log w/o seeing what you print
where, i.e. the diff with the printfs.
[...]
> As Block 2 above shows "dir_lookup flags & O_WRITE !=0" and the test
> if (!error && (flags & O_WRITE)) causes the EACCESS error.
> So the flags are not correctly set??
Apparently not:
> The files are moved from debian/tmp/usr/lib/ada/adalib/gnatvsn/ to
> debian/libgnatvsn4.9-dev/usr/lib/ada/adalib/gnatvsn with a previous
> dh_movefiles command. All files *.ali have file mode 10444 e.g.
> -r--r--r-- 1 srs srs 2267 Dec 18 00:38 alloc.ali
I would expect the test whether we may write to the file to fail if we
may not write to the file.
Justus
Reply to: