[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH v2] xen/sched: fix cpu offlining with core scheduling

On 04.03.20 10:53, Jan Beulich wrote:
On 03.03.2020 18:39, Juergen Gross wrote:
--- a/xen/common/sched/core.c
+++ b/xen/common/sched/core.c
@@ -2299,6 +2299,10 @@ void sched_context_switched(struct vcpu *vprev, struct 
vcpu *vnext)
+ * Switch to a new context or keep the current one running.
+ * On x86 it won't return, so it will drop the still held sched_res_rculock.
+ */
  static void sched_context_switch(struct vcpu *vprev, struct vcpu *vnext,
                                   bool reset_idle_unit, s_time_t now)

I don't follow the comment: There's

         return continue_running(vprev);

in the function which afaict can happen on all architectures.
The lock gets dropped there too. I see no path through this
function where the lock would not get dropped.

It was meant as reasoning: due to the fact that sched_context_switch()
won't return on x86, it is required that it will drop sched_res_rculock.

@@ -2408,6 +2412,9 @@ static struct vcpu *sched_force_context_switch(struct 
vcpu *vprev,
   * zero do_schedule() is called and the rendezvous counter for leaving
   * context_switch() is set. All other members will wait until the counter is
   * becoming zero, dropping the schedule lock in between.
+ * Either returns the new unit to run, or NULL if no context switch is
+ * required or (on ARM) has already been performed. If NULL is returned
+ * sched_res_rculock has been dropped.

I guess official Arm folks would like Arm to not be spelled all
upper case anymore.


@@ -2482,6 +2490,21 @@ static struct sched_unit 
*sched_wait_rendezvous_in(struct sched_unit *prev,
              atomic_set(&prev->next_task->rendezvous_out_cnt, 0);
              prev->rendezvous_in_cnt = 0;
+        /*
+         * Check for scheduling resourced switched. This happens when we are
+         * moved away from our cpupool and cpus are subject of the idle
+         * scheduler now.
+         */

The 'd' on both "resourced" and "switched" are odd to read at
least to me, and hence make me uncertain whether I actually
correctly understand what is meant here.

Should be "resources", of course.


Xen-devel mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.