Messages in this thread | | | Date | Thu, 3 Jan 2008 09:55:25 +0100 | From | Ingo Molnar <> | Subject | Re: [patch 02/20] make the inode i_mmap_lock a reader/writer lock |
| |
* Nick Piggin <nickpiggin@yahoo.com.au> wrote:
> > Have you done anything more with allowing > 256 CPUS in this > > spinlock patch? We've been testing with 1k cpus and to verify with > > -mm kernel, we need to "unpatch" these spinlock changes. > > Hi Mike, > > Actually I had it in my mind that 64 bit used single-byte locking like > i386, so I didn't think I'd caused a regression there. > > I'll take a look at fixing that up now.
thanks - this is a serious showstopper for the ticket spinlock patch.
( which has otherwise been performing very well in x86.git so far - it has passed a few thousand bootup tests on 64-bit and 32-bit as well, so we are close to it being in a mergable state. Would be a pity to lose it due to the 256 cpus limit. )
Ingo
| |