[lkml]   [2016]   [Jul]   [14]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: [PATCH v2] locking/pvqspinlock: restore/set vcpu_hashed state after failing adaptive locking spinning
On 07/14/2016 07:26 AM, Peter Zijlstra wrote:
> On Thu, Jul 14, 2016 at 04:15:56PM +0800, Wanpeng Li wrote:
>> In this case, lock holder inserts the pv_node of queue head into the
>> hash table and set _Q_SLOW_VAL unnecessary. This patch avoids it by
>> restoring/setting vcpu_halted state after failing adaptive locking
> ^^^^^^
>> spinning.
>> - WRITE_ONCE(pn->state, vcpu_halted);
>> + WRITE_ONCE(pn->state, vcpu_hashed);
> ^^^^^^
> The Changelog meant so say vcpu_hashed, surely?

Do you have chance to review the "Fix missed PV wakeup & support PPC"
patch ( that I had sent out a while
ago? That patch addressed a similar pv_hash leakage and collision problem.


 \ /
  Last update: 2016-07-14 17:41    [W:0.035 / U:2.520 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site