[BUG][GCC] *Serious* optimization bug: Any gcc specialists out there?
Stephan Rudlof
sr at evolgo.de
Fri Aug 4 14:31:32 UTC 2000
Dear Squeakers,
please compile
/* -------------------------------------------- */
#include <stdio.h>
#include <math.h>
testDoubles () {
double arg1 = 1.0e20;
printf("\nsin(arg1): %e", sin(arg1));
}
int main() { testDoubles(); return 0; }
/* -------------------------------------------- */
with
gcc -O testDoubles.c -lm
and without optimization:
gcc testDoubles.c -lm
;
the result strongly differs! (the optimized version gives the *wrong*
result)
At least for:
sr at Klaus:~/tmp > gcc -v
Reading specs from /usr/lib/gcc-lib/i486-linux/egcs-2.91.66/specs
gcc version egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)
.
This problem *seriously* affects Squeak, because this problem arises in
interp.c (e.g. in primitiveSine()) and gcc is widely used.
This makes me *really* nervous! (therefore so much CCs)
Is there anyone who follows the corresponding gcc mailing lists or
newsgroups and has some comments?
Or could post a bug report (after confirmation of this bug)?
Or has a newer gcc without this problem?
Greetings,
Stephan
--
Stephan Rudlof (sr at evolgo.de)
"Genius doesn't work on an assembly line basis.
You can't simply say, 'Today I will be brilliant.'"
-- Kirk, "The Ultimate Computer", stardate 4731.3
More information about the Squeak-dev
mailing list
|