xen-devel.lists.xenproject.org archive mirror
 help / color / mirror / Atom feed
From: George Dunlap <george.dunlap@citrix.com>
To: xen-devel@lists.xenproject.org
Cc: Dario Faggioli <dario.faggioli@citrix.com>,
	Jan Beulich <jbeulich@suse.com>,
	George Dunlap <george.dunlap@citrix.com>,
	Anshul Makkar <anshul.makkar@citrix.com>,
	Meng Xu <mengxu@cis.upenn.edu>
Subject: [PATCH v2 2/3] xen: Have schedulers revise initial placement
Date: Mon, 25 Jul 2016 16:28:12 +0100	[thread overview]
Message-ID: <1469460493-18842-2-git-send-email-george.dunlap@citrix.com> (raw)
In-Reply-To: <1469460493-18842-1-git-send-email-george.dunlap@citrix.com>

The generic domain creation logic in
xen/common/domctl.c:default_vcpu0_location() attempts to try to do
initial placement load-balancing by placing vcpu 0 on the least-busy
non-primary hyperthread available.  Unfortunately, the logic can end
up picking a pcpu that's not in the online mask.  When this is passed
to a scheduler such which assumes that the initial assignment is
valid, it causes a null pointer dereference looking up the runqueue.

Furthermore, this initial placement doesn't take into account hard or
soft affinity, or any scheduler-specific knowledge (such as historic
runqueue load, as in credit2).

To solve this, when inserting a vcpu, always call the per-scheduler
"pick" function to revise the initial placement.  This will
automatically take all knowledge the scheduler has into account.

csched2_cpu_pick ASSERTs that the vcpu's pcpu scheduler lock has been
taken.  Grab and release the lock to minimize time spend with irqs
disabled.

Signed-off-by: George Dunlap <george.dunlap@citrix.com>
---
v2:
- Actually grab lock before calling vcpu_schedule_lock() to avoid
  tripping over a new ASSERT

CC: Dario Faggioli <dario.faggioli@citrix.com>
CC: Anshul Makkar <anshul.makkar@citrix.com>
CC: Meng Xu <mengxu@cis.upenn.edu>
CC: Jan Beulich <jbeulich@suse.com>
---
 xen/common/sched_credit.c  |  3 +++
 xen/common/sched_credit2.c | 11 ++++++++++-
 xen/common/sched_rt.c      |  3 +++
 3 files changed, 16 insertions(+), 1 deletion(-)

diff --git a/xen/common/sched_credit.c b/xen/common/sched_credit.c
index d547716..220ff0d 100644
--- a/xen/common/sched_credit.c
+++ b/xen/common/sched_credit.c
@@ -998,6 +998,9 @@ csched_vcpu_insert(const struct scheduler *ops, struct vcpu *vc)
 
     BUG_ON( is_idle_vcpu(vc) );
 
+    /* This is safe because vc isn't yet being scheduled */
+    vc->processor = csched_cpu_pick(ops, vc);
+
     lock = vcpu_schedule_lock_irq(vc);
 
     if ( !__vcpu_on_runq(svc) && vcpu_runnable(vc) && !vc->is_running )
diff --git a/xen/common/sched_credit2.c b/xen/common/sched_credit2.c
index 3d2716a..fb9dd98 100644
--- a/xen/common/sched_credit2.c
+++ b/xen/common/sched_credit2.c
@@ -2055,12 +2055,21 @@ csched2_vcpu_insert(const struct scheduler *ops, struct vcpu *vc)
     ASSERT(!is_idle_vcpu(vc));
     ASSERT(list_empty(&svc->runq_elem));
 
-    /* Add vcpu to runqueue of initial processor */
+    /* csched2_cpu_pick() expects the pcpu lock to be held */
     lock = vcpu_schedule_lock_irq(vc);
 
+    vc->processor = csched2_cpu_pick(ops, vc);
+
+    spin_unlock_irq(lock);
+
+    lock = vcpu_schedule_lock_irq(vc);
+
+    /* Add vcpu to runqueue of initial processor */
     runq_assign(ops, vc);
 
     vcpu_schedule_unlock_irq(lock, vc);
+    
+    local_irq_enable();
 
     sdom->nr_vcpus++;
 
diff --git a/xen/common/sched_rt.c b/xen/common/sched_rt.c
index bd3a2a0..41c61a7 100644
--- a/xen/common/sched_rt.c
+++ b/xen/common/sched_rt.c
@@ -874,6 +874,9 @@ rt_vcpu_insert(const struct scheduler *ops, struct vcpu *vc)
 
     BUG_ON( is_idle_vcpu(vc) );
 
+    /* This is safe because vc isn't yet being scheduled */
+    vc->processor = rt_cpu_pick(ops, vc);
+
     lock = vcpu_schedule_lock_irq(vc);
 
     now = NOW();
-- 
2.1.4


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

  reply	other threads:[~2016-07-25 17:26 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-07-25 15:28 [PATCH v2 1/3] xen: Some code motion to avoid having to do forward-declaration George Dunlap
2016-07-25 15:28 ` George Dunlap [this message]
2016-07-25 17:34   ` [PATCH v2 2/3] xen: Have schedulers revise initial placement Meng Xu
2016-07-26  9:30   ` Dario Faggioli
2016-07-26  9:35     ` George Dunlap
2016-07-25 15:28 ` [PATCH v2 3/3] xen: Remove buggy initial placement algorithm George Dunlap
2016-07-26  9:14   ` Dario Faggioli
2016-07-26  9:22   ` Andrew Cooper
2016-07-26  9:32 ` [PATCH v2 1/3] xen: Some code motion to avoid having to do forward-declaration Dario Faggioli

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1469460493-18842-2-git-send-email-george.dunlap@citrix.com \
    --to=george.dunlap@citrix.com \
    --cc=anshul.makkar@citrix.com \
    --cc=dario.faggioli@citrix.com \
    --cc=jbeulich@suse.com \
    --cc=mengxu@cis.upenn.edu \
    --cc=xen-devel@lists.xenproject.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).