On 07/09/2018 09:47, Jakub Jelinek wrote:
On Fri, Sep 07, 2018 at 08:57:25AM +0200, David Brown wrote:
I am always wary of saying there might be a compiler bug - usually it is a
bug in the user code. But this time I am very suspicious. The example here
comes from a discussion in the comp.lang.c Usenet group.
Here is the code I have been testing:
unsigned char foo_u(unsigned int v) {
return (v & 0x7f) | (v & ~0x7f ? 0x80 : 0);
}
unsigned char foo_s(int v) {
return (v & 0x7f) | (v & ~0x7f ? 0x80 : 0);
}
Seems to have started with http://gcc.gnu.org/r134376, I'll have a look.
That said, can you file a bug in http://gcc.gnu.org/bugzilla/ for this, that
is where it should be tracked.
I didn't want to bother the developers with a bug report until someone
agreed it probably is a bug, rather than my misunderstanding. But since
it is at least worth investigating, I've filed one now (with the typo
fixed).
<https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87248>
unsigned char bar_s(int v) {
int x = v & 0x7f;
return (v & 0x7f) | (x ? 0x80 : 0);
}
This does something different (did you mean int x = v & ~0x7f; ?).
Yes. The effect is still the same, but it helps to be accurate in bug
reports!
Jakub
