Michael James wrote:
And recompiled with the same flags. The assembly code for the loop
portion is identical to the one I posted above. Now though the code is
actually capable of producing NANs.
Just to be sure, I also tested this on my modified loop:
int main(void) {
printf("test(4, 6, 0) = %f\n", test(4,6,0));
printf("test(0, 2, 0) = %f\n", test(0,2,0));
printf("test(-2, 3, 0) = %f\n", test(-2,3,0));
return 0;
}
[EMAIL PROTECTED]:~/project/cf/util$ /home/james/local/gcc/bin/gcc -O2
-march=i686 -funsafe-math-optimizations -fno-math-errno uros-test.c -o
test
[EMAIL PROTECTED]:~/project/cf/util$ ./test
test(4, 6, 0) = 2.995732
test(0, 2, 0) = -inf
test(-2, 3, 0) = nan
[EMAIL PROTECTED]:~/project/cf/util$ /home/james/local/gcc/bin/gcc -O2
-march=i686 uros-test.c -o test -lm
[EMAIL PROTECTED]:~/project/cf/util$ ./uros
test(4, 6, 0) = 2.995732
test(0, 2, 0) = -inf
test(-2, 3, 0) = -inf
[EMAIL PROTECTED]:~/project/cf/util$ /home/james/local/gcc/bin/gcc -v
Using built-in specs.
Target: i686-pc-linux-gnu
Configured with: ../gcc-4-2/configure --prefix=/home/james/local/gcc
Thread model: posix
gcc version 4.2.0 20061103 (prerelease)
Perhaps I have not replicated your working environment closely enough,
or you have a different macro in place of the isnan call. I compiled
all code above both with and without include headers <math.h>,
<stdio.h>. I get the same results either way.
No, I'm working with gcc version 4.3.0 20061103 (experimental), and the
results are as expected:
gcc -O2 -funsafe-math-optimizations -fno-math-errno -march=i686
-mfpmath=387 -m32 mj.c
[-m32 is not needed on 32bit compiler]
test(4, 6, 0) = 2.995732
test(0, 2, 0) = -inf
test(-2, 3, 0) = -inf
This is the exact testcase I used:
--cut here--
double test(int i0, int n, double a)
{
double sum = 0.0;
int i;
for(i=i0; i<n; ++i)
{
float x = logf((float)i);
sum += isnan(x) ? 0 : x;
}
return sum;
}
int main(void) {
printf("test(4, 6, 0) = %f\n", test(4,6,0));
printf("test(0, 2, 0) = %f\n", test(0,2,0));
printf("test(-2, 3, 0) = %f\n", test(-2,3,0));
return 0;
}
--cut here--
Inspecting the code, there is fucomi insn present just after "log" insn
sequence.
Unfortunatelly, I have no 4.2 installed here, but it looks like a bug to
me. Perhaps could you open a bugreport for this issue?
Thanks,
Uros.
