Is gcc's -Wsequence-point warning flag broken?

Question

I'm getting a warning for this line:

e = strtol(++s, (char **)&s, 10);

Moving the ++s to a separate statement makes the warning go away, but as far as I can tell, this warning is completely incorrect. The value of s is only used once, in the computation of ++s. The second use of s is only to take its address. Is there something I'm missing?

I cannot reproduce the problem using gcc 4.4.3 or 4.4.5. Post a full example.c that demonstrates it? — Kamal, Nov 21 '10 at 16:25
By the way, post that as an answer and I'll accept it, since "current gcc does not generate this warning" is a good answer to my question about whether the warning was incorrect. — R.. GitHub STOP HELPING ICE, Nov 21 '10 at 16:30

Kamal · Accepted Answer · 2010-11-21T16:38:50.423

4

Newer versions of gcc (4.4.3 or 4.4.5) do not generate any warning for the line you posted. The generation of the warning there appears to be a bug in older gcc versions.

edited Nov 21 '10 at 16:38

answered Nov 21 '10 at 16:30

Kamal

7,160
2
21
12

2

See [GCC bug 18050](http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18050) - fixed in 4.4.0. – ugoren Feb 13 '13 at 11:46

Is gcc's -Wsequence-point warning flag broken?

1 Answers1