begin quoting Carlos R. Mafra as of Fri, Mar 26, 2010 at 09:36:07PM +0100:
> On Fri, 26 Mar 2010 at 20:35:02 +0100, Tamas TEVESZ wrote:
> > +
> > +#define RETRY( x ) do { \
> > + x; \
> > + } while (errno == EINTR);
> >
I was under the impression that good C style was to leave off the semicolons
in macros, so that using the macro uses a semicolon like a normal statement.
> [...]
>
> >
> > - if ((src = open(srcFile, O_RDONLY | O_BINARY)) == 0) {
> > + RETRY( src = fopen(srcFile, "rb") )
Would be instead
RETRY( src = fopen(srcFile, "rb") );
...and thus not suprise the reader of the code.
> > + if (src == NULL) {
> > wsyserror(_("Could not open %s"), srcFile);
> > return -1;
> > }
>
> Using this RETRY() macro is a bit ugly. I didn't like from the look
> of it.
Shouldn't there also be some sort of limit in the loop? If the
function so wrapped manages to have a bug that results in it causing
itself to be interrupted, this construct turns into an infinite loop.
RETRY_ON_EINTR seems like a better name. Still ugly, though.
> So let me ask a question. What bad thing would happen if you don't
> use this RETRY trick and do simply a
>
> src = fopen(srcFile, "rb");
> if (!src) {
> wsyserror(_("Could not open %s"), srcFile);
> return -1;
> }
>
> The old code didn't retry, why do you need it now?
Are signals really causing problems? Ignoring a signal errors seems
like a bad policy, especially when it's hidden in a macro.
-S.
--
To unsubscribe, send mail to [email protected].
