vector-t.c failures with Forte
rra at stanford.edu
Wed Aug 28 23:22:50 UTC 2002
Alex Kiernan <alexk at demon.net> writes:
> Huh, wierd... here's what we build with:
> -fast -g -xs -v -xstrconst -xdepend -xtarget=ultra2 -xarch=v8plusa
> -xprefetch -xildoff
> I haven't gone through each of them to figure out where it goes wrong
I played a hunch. It's -xstrconst. Looks like the lifting of static
strings into the rodata section breaks when the strings contain embedded
nuls (which isn't unreasonable, since that makes them not valid C strings;
I haven't checked the standard to see if this is really allowed).
I turned them into initialized static const char arrays, which seems to
have made Forte happy.
Russ Allbery (rra at stanford.edu) <http://www.eyrie.org/~eagle/>
Please send questions to the list rather than mailing me directly.
<http://www.eyrie.org/~eagle/faqs/questions.html> explains why.
More information about the inn-workers