RE: __rcu_process_callbacks() in Linux 2.6

2007-11-26 Thread James Huang
> -Original Message- > From: James Huang [mailto:[EMAIL PROTECTED] > Sent: Monday, November 26, 2007 2:21 PM > To: James Huang > Subject: Fw: __rcu_process_callbacks() in Linux 2.6 > > - Forwarded Message > From: Manfred Spraul <[EMAIL PROTECTED]

Re: __rcu_process_callbacks() in Linux 2.6

2007-11-21 Thread James Huang
ail? Is it the one called by CPU 0, CPU 1, or CPU 2? Thanks, James Huang - Original Message From: Paul E. McKenney <[EMAIL PROTECTED]> To: James Huang <[EMAIL PROTECTED]> Cc: linux-kernel@vger.kernel.org; [EMAIL PROTECTED]; [EMAIL PROTECTED] Sent: Wednesday,

Re: __rcu_process_callbacks() in Linux 2.6

2007-11-20 Thread James Huang
t to curlist; (including the callback inserted at time t3) rdp->batch = rcp->cur + 1; <--- time t4 At time t4, CPU 2 might observe rcp->cpu as 100 and set rdp->batch to 101. So CPU 2 essentially

__rcu_process_callbacks() in Linux 2.6

2007-11-20 Thread James Huang
CPU 2 might observe rcp->cpu as 100 and set rdp->batch to 101. So CPU 2 essentially started its batch 101 (includes a callback inserted at time t3) after CPU 1 passed its quiescent state (at time t2) for batch 101. Isn't this an issue??? Am I missing something? Thanks for pointing me to the