On Sun, Jan 02, 2011 at 12:41:33AM +0100, Gábor Stefanik wrote:
> On Fri, Dec 31, 2010 at 8:28 PM, Javier Martinez Canillas
> <[email protected]> wrote:
> > + DWORD err;
> > + err = correct_data(buf, redundant_ecc, *(calculate_ecc+1),
> > + *(calculate_ecc), *(calculate_ecc+2));
>
> Any reason why you didn't unify these 2 lines? Like this: DWORD err =
> correct_data(...);
>
These kind of things aren't described in CodingStyle so they're up to
whoever writes the code to decide. Or if the maintainer is a
micromanager the maintainer can decide.
But personally I much prefer to put anything complicated on separate
lines. No one reads the initializers. In my work with Smatch I see a
lot of bugs like this:
int x = foo->bar;
if (!foo)
return -EINVAL;
It's astounding how many. The famous tun.c security bug was one of
these.
But there should have been a blank line between the initializers and
the code. Otherwise people will think the code is initiliazation and
ignore it. That is in CodingStyle I think. We can fix that when we get
rid of the DWORD data type in a later patch (don't resend).
regards,
dan carpenter
_______________________________________________
devel mailing list
[email protected]
http://driverdev.linuxdriverproject.org/mailman/listinfo/devel
