Re: [patch] pagecache SMP-scalability patch [was: spinlock usage]

2001-03-21 Thread Ingo Molnar
On Wed, 21 Mar 2001, Alexander Viro wrote: > ??? > > Nope, no calls into ext2/*. do_revalidate() seeing NULL ->i_revalidate > and doing nothing, lnamei() doing usual lookup, cp_old_stat() not touching > fs code at all... the problem was that it was calling lock_kernel(), not the lowlevel fs i

Re: [patch] pagecache SMP-scalability patch [was: spinlock usage]

2001-03-21 Thread Alexander Viro
On Wed, 21 Mar 2001, Ingo Molnar wrote: > > On Wed, 21 Mar 2001, Alexander Viro wrote: > > > > (about lstat(): IMO lstat() should not call into the lowlevel FS code.) > > > > a) revalidation on network filesystems > > b) just about anything layered would win from ability to replace the > > no

Re: [patch] pagecache SMP-scalability patch [was: spinlock usage]

2001-03-21 Thread Ingo Molnar
On Wed, 21 Mar 2001, Alexander Viro wrote: > > (about lstat(): IMO lstat() should not call into the lowlevel FS code.) > > a) revalidation on network filesystems > b) just about anything layered would win from ability to replace the > normal stat() behaviour (think of UI/GID replacement, etc.)

Re: [patch] pagecache SMP-scalability patch [was: spinlock usage]

2001-03-21 Thread Alexander Viro
On Wed, 21 Mar 2001, Ingo Molnar wrote: > > Anton, > > if you are doing SMP-intensive dbench runs, then check out the SMP > pagecache-scalability patch (against 2.4.2-ac20): > > http://people.redhat.com/~mingo/smp-pagecache-patches/pagecache-2.4.2-H1 > > this patch splits up the main scal

Re: [patch] pagecache SMP-scalability patch [was: spinlock usage]

2001-03-21 Thread Anton Blanchard
Hi, > http://people.redhat.com/~mingo/smp-pagecache-patches/pagecache-2.4.2-H1 > > this patch splits up the main scalability offender in non-RAM-limited > dbench runs, which is pagecache_lock. The patch was designed and written > by David Miller, and is being forward ported / maintained by m