search for: pv_set_head

...have been waiting for the + * node->next store above before setting node->locked. + */ + if (ppn->mcs.locked) + return; + + cpu_relax(); + } + /* + * If we interleave such that the store from pv_set_head() happens + * between our load and store we could have over-written the new head + * value. + * + * Therefore use cmpxchg to only propagate the old head value if our + * head value is unmodified. + */ + (void)cmpxchg(&pn->head, NULL, READ_ONCE(...

...have been waiting for the + * node->next store above before setting node->locked. + */ + if (ppn->mcs.locked) + return; + + cpu_relax(); + } + /* + * If we interleave such that the store from pv_set_head() happens + * between our load and store we could have over-written the new head + * value. + * + * Therefore use cmpxchg to only propagate the old head value if our + * head value is unmodified. + */ + (void)cmpxchg(&pn->head, NULL, READ_ONCE(...

On 03/16/2015 09:16 AM, Peter Zijlstra wrote: > Implement simple paravirt support for the qspinlock. > > Provide a separate (second) version of the spin_lock_slowpath for > paravirt along with a special unlock path. > > The second slowpath is generated by adding a few pv hooks to the > normal slowpath, but where those will compile away for the native > case, they expand

On 03/16/2015 09:16 AM, Peter Zijlstra wrote: > Implement simple paravirt support for the qspinlock. > > Provide a separate (second) version of the spin_lock_slowpath for > paravirt along with a special unlock path. > > The second slowpath is generated by adding a few pv hooks to the > normal slowpath, but where those will compile away for the native > case, they expand