Thanks, r166522. On Tue, Oct 23, 2012 at 4:19 PM, David Blaikie <[email protected]> wrote: > On Tue, Oct 23, 2012 at 1:48 PM, Matt Beaumont-Gay <[email protected]> > wrote: >> The motivating case here is that OpenSSL has some code like this: >> >> static long bio_ctrl(BIO *bio, int cmd, long num, void *ptr); >> ... >> #define BIO_flush(b) (int)BIO_ctrl(b,BIO_CTRL_FLUSH,0,NULL) >> >> That cast to int causes client code which calls BIO_flush and ignores >> the result to get a -Wunused-value warning. > > LGTM. > > (I don't suppose there's any way to narrow this down to ensure it's > logically at the outer most scope of the macro?)
I guess we could try to be more clever, but I'd want to see some evidence that the effort would be worthwhile. _______________________________________________ cfe-commits mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
