xen-devel.lists.xenproject.org archive mirror
 help / color / mirror / Atom feed
From: Stefano Stabellini <sstabellini@kernel.org>
To: julien.grall@arm.com
Cc: xen-devel@lists.xenproject.org, sstabellini@kernel.org
Subject: [PATCH v2 4/4] The locking order is: first rank lock, then vgic lock. The order is respected everywhere, except for gic_update_one_lr.
Date: Wed, 21 Dec 2016 18:15:13 -0800	[thread overview]
Message-ID: <1482372913-18366-4-git-send-email-sstabellini@kernel.org> (raw)
In-Reply-To: <1482372913-18366-1-git-send-email-sstabellini@kernel.org>

gic_update_one_lr is called with the vgic lock held, but it calls
vgic_get_target_vcpu, which tries to obtain the rank lock. This can
cause deadlocks.

We already have a version of vgic_get_target_vcpu that doesn't take the
rank lock: __vgic_get_target_vcpu.

Solve the lock inversion problem, by not taking the rank lock in
gic_update_one_lr (calling __vgic_get_target_vcpu instead of
vgic_get_target_vcpu).  This is safe, because vcpu target modifications
are protected by the same vgic vcpu lock.

Signed-off-by: Stefano Stabellini <sstabellini@kernel.org>
---
 xen/arch/arm/gic.c         | 2 +-
 xen/arch/arm/vgic.c        | 4 +---
 xen/include/asm-arm/vgic.h | 1 +
 3 files changed, 3 insertions(+), 4 deletions(-)

diff --git a/xen/arch/arm/gic.c b/xen/arch/arm/gic.c
index 51148b4..28ab2f9 100644
--- a/xen/arch/arm/gic.c
+++ b/xen/arch/arm/gic.c
@@ -509,7 +509,7 @@ static void gic_update_one_lr(struct vcpu *v, int i)
             list_del_init(&p->inflight);
             if ( test_and_clear_bit(GIC_IRQ_GUEST_MIGRATING, &p->status) )
             {
-                struct vcpu *v_target = vgic_get_target_vcpu(v, irq);
+                struct vcpu *v_target = __vgic_get_target_vcpu(v, irq);
                 irq_set_affinity(p->desc, cpumask_of(v_target->processor));
             }
             /* 
diff --git a/xen/arch/arm/vgic.c b/xen/arch/arm/vgic.c
index cceac24..c89b85f 100644
--- a/xen/arch/arm/vgic.c
+++ b/xen/arch/arm/vgic.c
@@ -221,12 +221,10 @@ int vcpu_vgic_free(struct vcpu *v)
 }
 
 /* The function should be called by rank lock taken. */
-static struct vcpu *__vgic_get_target_vcpu(struct vcpu *v, unsigned int virq)
+struct vcpu *__vgic_get_target_vcpu(struct vcpu *v, unsigned int virq)
 {
     struct vgic_irq_rank *rank = vgic_rank_irq(v, virq);
 
-    ASSERT(spin_is_locked(&rank->lock));
-
     return v->domain->vcpu[rank->vcpu[virq & INTERRUPT_RANK_MASK]];
 }
 
diff --git a/xen/include/asm-arm/vgic.h b/xen/include/asm-arm/vgic.h
index dce2f84..26594b0 100644
--- a/xen/include/asm-arm/vgic.h
+++ b/xen/include/asm-arm/vgic.h
@@ -295,6 +295,7 @@ extern int domain_vgic_init(struct domain *d, unsigned int nr_spis);
 extern void domain_vgic_free(struct domain *d);
 extern int vcpu_vgic_init(struct vcpu *v);
 extern struct vcpu *vgic_get_target_vcpu(struct vcpu *v, unsigned int virq);
+extern struct vcpu *__vgic_get_target_vcpu(struct vcpu *v, unsigned int virq);
 extern void vgic_vcpu_inject_irq(struct vcpu *v, unsigned int virq);
 extern void vgic_vcpu_inject_spi(struct domain *d, unsigned int virq);
 extern void vgic_clear_pending_irqs(struct vcpu *v);
-- 
1.9.1


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

  parent reply	other threads:[~2016-12-22  2:15 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-12-22  2:14 [PATCH v2 0/4] xen/arm: fix rank/vgic lock inversion bug Stefano Stabellini
2016-12-22  2:15 ` [PATCH v2 1/4] xen/arm: fix GIC_INVALID_LR Stefano Stabellini
2016-12-22  2:15   ` [PATCH v2 2/4] arm: store vcpu id in struct irq_pending Stefano Stabellini
2016-12-22 11:52     ` Andrew Cooper
2016-12-22  2:15   ` [PATCH v2 3/4] arm, vgic_migrate_irq: take the right vgic lock Stefano Stabellini
2016-12-28 16:42     ` Julien Grall
2017-01-03 23:30       ` Stefano Stabellini
2017-01-16 16:31         ` Julien Grall
2016-12-22  2:15   ` Stefano Stabellini [this message]
2016-12-28 16:55     ` [PATCH v2 4/4] The locking order is: first rank lock, then vgic lock. The order is respected everywhere, except for gic_update_one_lr Julien Grall
2017-01-03 22:51       ` Stefano Stabellini
2017-01-16 16:55         ` Julien Grall
2017-01-16 19:10           ` Stefano Stabellini
2017-01-19 12:51             ` Julien Grall
2016-12-28 17:30   ` [PATCH v2 1/4] xen/arm: fix GIC_INVALID_LR Julien Grall
2017-01-03 22:52     ` Stefano Stabellini

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=1482372913-18366-4-git-send-email-sstabellini@kernel.org \
    --to=sstabellini@kernel.org \
    --cc=julien.grall@arm.com \
    --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).