[RFC+CFT] Use word operations in bitops

[linux@arm.linux.org.uk (Russell King - ARM Linux)]

On Tue, Jan 18, 2011 at 04:58:51PM +0100, Uwe Kleine-K?nig wrote:
> Hello Russell,
> 
> On Tue, Jan 18, 2011 at 03:43:44PM +0000, Russell King - ARM Linux wrote:
> > On Tue, Jan 18, 2011 at 04:32:57PM +0100, Uwe Kleine-K?nig wrote:
> > > On Mon, Jan 17, 2011 at 10:46:18AM +0000, Russell King - ARM Linux wrote:
> > I will continue to mail out patches which I want people to test and give
> > feedback on, because that is the _only_ way to do it.
> Note that I didn't suggest to stop sending out patches and collecting
> the acks and tested-bys there.  The automated tests are just a different
> test.
> 
> And IMHO it's not harder to reply with a git commit-id, but it makes it
> more likely that a reply to a wrong mail is detected.

No.  Take a moment and think.

Two people pull the commits one evening.  One gets in during GMT morning,
checks their tests, and reports "git commit X passed, have this Tested-by".
So, their tested-by is added to the commit, which changes the git ID to
Y.

Second person gets in during GMT afternoon, and reports "git commit X
passed."  Now instead of looking in the current branch, I'd have to
cross-reference it back to what was originally there, check that's
what they actually meant, look that up in the new branch, and apply
their tested-by to that instead.

More complicated means more things to go wrong.  No thank you.