On Sat, Jan 28, 2012 at 12:43:39AM +0100, Jesper Juhl wrote: > On Wed, 25 Jan 2012, Dan Carpenter wrote: > > > This is probably going to need to be redone on top of the other bcm > > cleanup patches anyway. > > > If that turns out to be so, well, so be it. I was not aware of other > patches, so my bad... > > > On Tue, Jan 24, 2012 at 11:46:31PM +0100, Jesper Juhl wrote: > > > + pstClassifierEntry->ucDestPortRangeLength > > > + = psfCSType->cCPacketClassificationRule.u8ProtocolDestPortRangeLength/4; > > > > Put the equals on the first line when you break something up like > > this. Also the real fix here is to choose shorter and better names > > than "cCPacketClassificationRule.u8ProtocolDestPortRangeLength". > > > > If you're breaking up a condition normally the && or || go on the > > first line as well. > > if (foo && > > bar && > > baz) { ... > > > > Although that's not in CodingStyle so it's not a "redo the patch" > > situation. > > > As you say, "not in CodingStyle", so I guess it's down to your personal > preference vs my personal preference vs what's most common in the kernel > vs maintainers preference. > > I did it the way I did for a reason. > I personally find that if one puts the '=', '==', '&&', '||', '+', > at the end of each line on a multi-line statement, then they > are easy to overlook when just reading the code casually. If, however, one > puts them at the start of each line, they stand out and are easier to see > (harder to miss). That's why I did as I did. You may not agree, but at > least now you know my reason for doing it the way I did. > The '=' certainly should go on the first line. That one doesn't need to be in CodingStyle because it's just obvious. ;) Anyway it is the overwhelmingly prefered way in the kernel with 31449 cases on the first line and 603 on the second. For the others, I don't care, but in the kernel the clear preference is to put them on the first line. I've redone at least one patch because someone complained about && placement. regards, dan carpenter