From: Marcelo Tosatti <marcelo-Bw31MaZKKs3YtjvyW6yDsg@public.gmane.org>
To: Avi Kivity <avi-atKUWr5tajBWk0Htik3J/w@public.gmane.org>
Cc: kvm-devel <kvm-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org>
Subject: [PATCH] Fix SMP shadow instantiation race
Date: Mon, 10 Dec 2007 11:19:07 -0500 [thread overview]
Message-ID: <20071210161907.GA13917@dmt> (raw)
There is a race where VCPU0 is shadowing a pagetable entry while VCPU1
is updating it, which results in a stale shadow copy.
Fix that by comparing the contents of the cached guest pte with the
current guest pte after write-protecting the guest pagetable.
Attached program kvm_shadow_race.c demonstrates the problem.
Signed-off-by: Marcelo Tosatti <mtosatti-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
diff --git a/drivers/kvm/paging_tmpl.h b/drivers/kvm/paging_tmpl.h
index 72d4816..4fece01 100644
--- a/drivers/kvm/paging_tmpl.h
+++ b/drivers/kvm/paging_tmpl.h
@@ -66,6 +66,7 @@ struct guest_walker {
int level;
gfn_t table_gfn[PT_MAX_FULL_LEVELS];
pt_element_t pte;
+ gpa_t pte_gpa;
unsigned pt_access;
unsigned pte_access;
gfn_t gfn;
@@ -212,6 +213,7 @@ walk:
}
walker->pte = pte;
+ walker->pte_gpa = pte_gpa;
walker->pt_access = pt_access;
walker->pte_access = pte_access;
pgprintk("%s: pte %llx pte_access %x pt_access %x\n",
@@ -267,6 +269,7 @@ static u64 *FNAME(fetch)(struct kvm_vcpu *vcpu, gva_t addr,
int level;
u64 *shadow_ent;
unsigned access = walker->pt_access;
+ pt_element_t curr_pte;
if (!is_present_pte(walker->pte))
return NULL;
@@ -316,6 +319,11 @@ static u64 *FNAME(fetch)(struct kvm_vcpu *vcpu, gva_t addr,
*shadow_ent = shadow_pte;
}
+ kvm_read_guest(vcpu->kvm, walker->pte_gpa, &curr_pte, sizeof(curr_pte));
+
+ if (curr_pte != walker->pte)
+ return 0;
+
mmu_set_spte(vcpu, shadow_ent, access, walker->pte_access & access,
user_fault, write_fault, walker->pte & PT_DIRTY_MASK,
ptwrite, walker->gfn);
@@ -382,10 +390,11 @@ static int FNAME(page_fault)(struct kvm_vcpu *vcpu, gva_t addr,
/*
* mmio: emulate if accessible, otherwise its a guest fault.
*/
- if (is_io_pte(*shadow_pte))
+ if (shadow_pte && is_io_pte(*shadow_pte))
return 1;
- ++vcpu->stat.pf_fixed;
+ if (shadow_pte)
+ ++vcpu->stat.pf_fixed;
kvm_mmu_audit(vcpu, "post page fault (fixed)");
return write_pt;
-------------------------------------------------------------------------
SF.Net email is sponsored by:
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
next reply other threads:[~2007-12-10 16:19 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-12-10 16:19 Marcelo Tosatti [this message]
2007-12-10 17:07 ` [PATCH] Fix SMP shadow instantiation race Avi Kivity
[not found] ` <475D726A.2040901-atKUWr5tajBWk0Htik3J/w@public.gmane.org>
2007-12-10 19:12 ` Marcelo Tosatti
2007-12-10 21:27 ` Avi Kivity
[not found] ` <475DAF51.8060804-atKUWr5tajBWk0Htik3J/w@public.gmane.org>
2007-12-10 22:22 ` Marcelo Tosatti
2007-12-12 0:12 ` Marcelo Tosatti
2007-12-13 8:37 ` Avi Kivity
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=20071210161907.GA13917@dmt \
--to=marcelo-bw31mazkks3ytjvyw6ydsg@public.gmane.org \
--cc=avi-atKUWr5tajBWk0Htik3J/w@public.gmane.org \
--cc=kvm-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.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