Re: [GIT PULL] x86/debug changes for v3.3

[Ingo Molnar <mingo@elte.hu>]

* Linus Torvalds <torvalds@linux-foundation.org> wrote:

> On Thu, Jan 5, 2012 at 6:47 AM, Ingo Molnar <mingo@elte.hu> wrote:
> >
> > Please pull the latest x86-debug-for-linus git tree from:
> >
> >   git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git x86-debug-for-linus
> 
> Hell no!
> 
> Why do you send me this sh*t? The "Use NMI instead of 
> REBOOT_VECTOR" commit has been reported to not work AT ALL.
> 
> It was totally broken - Don Zickus sent out the email to fix 
> the completely bogus parenthesis in that commit just a day ago 
> - yet you send me this KNOWN BROKEN CRAP.

My bad.

My primary mistake was that i could have noticed the original 
regression breaking kdump that Yinghai reported - i should have 
delayed the tree until a fix came in - like i delayed a couple 
of other trees that you didn't get sent.

Then, to compound it with another mistake, the fix for the kdump 
breakage was sent to me a couple of hours after i sent you the 
tree - and i havent processed all subsequent mails yet up until 
today - so you noticed it before i did. Double fail.

Anyway, i've queued up the fix and will send you the new tree
after some testing.

> And yes, I checked. The version you sent me is the f*cked one. 
> I was hoping that you would have fixed it up. But no.
> 
> In short, you didn't merge the fix, and yet you sent me a 
> patch series that was *known* to be broken for the last three+ 
> weeks! And I know you were cc'd on the report, which was in 
> the same thread as the automatic tip-bot message for the 
> particular broken commit, so the particular broken commit that 
> introduced it was well-known.
> 
> So that thing has been known broken since before Christmas, 
> and was due to a major typo that made a "atomic_cmpxchg()" not 
> work *at*all* - yet you blithely send me this crap.
> 
> Why? WHAT THE F*CK HAPPENED, INGO?
> 
> Yes, I'm angry as hell. Shit like this should NOT happen.  I 
> don't want people sending me known-buggy pull requests.

You are perfectly right, i messed up - sorry about that.

Thanks,

	Ingo