From: Andrew Cooper <andrew.cooper3@citrix.com>
To: Xen-devel <xen-devel@lists.xen.org>
Cc: Andrew Cooper <andrew.cooper3@citrix.com>,
Jan Beulich <JBeulich@suse.com>
Subject: [PATCH 5/6] x86/domctl: Fix migration of guests which are not using xsave
Date: Mon, 12 Sep 2016 10:51:39 +0100 [thread overview]
Message-ID: <1473673900-8585-6-git-send-email-andrew.cooper3@citrix.com> (raw)
In-Reply-To: <1473673900-8585-1-git-send-email-andrew.cooper3@citrix.com>
c/s da62246e "x86/xsaves: enable xsaves/xrstors/xsavec in xen" broke migration
of PV guests which were not using xsave.
In such a case, compress_xsave_states() gets passed a zero length buffer. The
first thing it tries to do is ASSERT() on user-provided data, if it hadn't
already wandered off the end of the buffer to do so.
Perform more verification of the input buffer before passing it to
compress_xsave_states(). This involves making xsave_area_compressed() public.
Similar problems exist on the HVM side, so make equivielent adjustments there.
This doesn't manifest in general, as hvm_save_cpu_xsave_states() elides the
entire record if xsave isn't used, but is a problem if a caller were to
construct an xsave record manually.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
---
CC: Jan Beulich <JBeulich@suse.com>
---
xen/arch/x86/domctl.c | 12 +++++++++---
xen/arch/x86/hvm/hvm.c | 20 ++++++++++++++++++--
xen/arch/x86/xstate.c | 6 ------
xen/include/asm-x86/xstate.h | 6 ++++++
4 files changed, 33 insertions(+), 11 deletions(-)
diff --git a/xen/arch/x86/domctl.c b/xen/arch/x86/domctl.c
index 5aa9f3a..2a2fe04 100644
--- a/xen/arch/x86/domctl.c
+++ b/xen/arch/x86/domctl.c
@@ -1158,7 +1158,15 @@ long arch_do_domctl(
goto vcpuextstate_out;
}
- if ( evc->size <= PV_XSAVE_SIZE(_xcr0_accum) )
+ if ( evc->size == PV_XSAVE_HDR_SIZE )
+ ; /* Nothing to restore. */
+ else if ( evc->size < PV_XSAVE_HDR_SIZE + XSTATE_AREA_MIN_SIZE )
+ ret = -EINVAL; /* Can't be legitimate data. */
+ else if ( xsave_area_compressed(_xsave_area) )
+ ret = -EOPNOTSUPP; /* Don't support compressed data. */
+ else if ( evc->size != PV_XSAVE_SIZE(_xcr0_accum) )
+ ret = -EINVAL; /* Not legitimate data. */
+ else
{
vcpu_pause(v);
v->arch.xcr0 = _xcr0;
@@ -1169,8 +1177,6 @@ long arch_do_domctl(
evc->size - PV_XSAVE_HDR_SIZE);
vcpu_unpause(v);
}
- else
- ret = -EINVAL;
xfree(receive_buf);
}
diff --git a/xen/arch/x86/hvm/hvm.c b/xen/arch/x86/hvm/hvm.c
index ca96643..a6c6621 100644
--- a/xen/arch/x86/hvm/hvm.c
+++ b/xen/arch/x86/hvm/hvm.c
@@ -1345,6 +1345,23 @@ static int hvm_load_cpu_xsave_states(struct domain *d, hvm_domain_context_t *h)
printk(XENLOG_G_WARNING
"HVM%d.%u restore mismatch: xsave length %#x > %#x\n",
d->domain_id, vcpuid, desc->length, size);
+ /* Rewind desc->length to ignore the extraneous zeros. */
+ desc->length = size;
+ }
+
+ if ( xsave_area_compressed((const void *)&ctxt->save_area) )
+ {
+ printk(XENLOG_G_WARNING
+ "HVM%d.%u restore: compressed xsave state not supported\n",
+ d->domain_id, vcpuid);
+ return -EOPNOTSUPP;
+ }
+ else if ( desc->length != size )
+ {
+ printk(XENLOG_G_WARNING
+ "HVM%d.%u restore mismatch: xsave length %#x != %#x\n",
+ d->domain_id, vcpuid, desc->length, size);
+ return -EINVAL;
}
/* Checking finished */
@@ -1353,8 +1370,7 @@ static int hvm_load_cpu_xsave_states(struct domain *d, hvm_domain_context_t *h)
if ( ctxt->xcr0_accum & XSTATE_NONLAZY )
v->arch.nonlazy_xstate_used = 1;
compress_xsave_states(v, &ctxt->save_area,
- min(desc->length, size) -
- offsetof(struct hvm_hw_cpu_xsave,save_area));
+ size - offsetof(struct hvm_hw_cpu_xsave, save_area));
return 0;
}
diff --git a/xen/arch/x86/xstate.c b/xen/arch/x86/xstate.c
index 1973ba0..f6157f5 100644
--- a/xen/arch/x86/xstate.c
+++ b/xen/arch/x86/xstate.c
@@ -86,12 +86,6 @@ uint64_t get_msr_xss(void)
return this_cpu(xss);
}
-static bool_t xsave_area_compressed(const struct xsave_struct *xsave_area)
-{
- return xsave_area && (xsave_area->xsave_hdr.xcomp_bv
- & XSTATE_COMPACTION_ENABLED);
-}
-
static int setup_xstate_features(bool_t bsp)
{
unsigned int leaf, eax, ebx, ecx, edx;
diff --git a/xen/include/asm-x86/xstate.h b/xen/include/asm-x86/xstate.h
index 51a9ed4..a8b8751 100644
--- a/xen/include/asm-x86/xstate.h
+++ b/xen/include/asm-x86/xstate.h
@@ -131,4 +131,10 @@ static inline bool_t xstate_all(const struct vcpu *v)
(v->arch.xcr0_accum & XSTATE_LAZY & ~XSTATE_FP_SSE);
}
+static inline bool __nonnull(1)
+ xsave_area_compressed(const struct xsave_struct *xsave_area)
+{
+ return xsave_area->xsave_hdr.xcomp_bv & XSTATE_COMPACTION_ENABLED;
+}
+
#endif /* __ASM_XSTATE_H */
--
2.1.4
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
next prev parent reply other threads:[~2016-09-12 9:51 UTC|newest]
Thread overview: 29+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-09-12 9:51 [PATCH 0/6] Fix multiple issues with xsave state handling on migrate Andrew Cooper
2016-09-12 9:51 ` [PATCH 1/6] x86/domctl: Introduce PV_XSAVE_HDR_SIZE and remove its opencoding Andrew Cooper
2016-09-12 11:05 ` Jan Beulich
2016-09-12 9:51 ` [PATCH 2/6] x86/domctl: Fix TOCTOU race with the use of XEN_DOMCTL_getvcpuextstate Andrew Cooper
2016-09-12 11:17 ` Jan Beulich
2016-09-12 12:02 ` Andrew Cooper
2016-09-12 12:33 ` Jan Beulich
2016-09-12 13:09 ` Andrew Cooper
2016-09-12 13:35 ` Jan Beulich
2016-09-12 13:37 ` Jan Beulich
2016-09-12 9:51 ` [PATCH 3/6] x86/domctl: Simplfy XEN_DOMCTL_getvcpuextstate when xsave is not in use Andrew Cooper
2016-09-12 11:26 ` Jan Beulich
2016-09-12 9:51 ` [PATCH 4/6] x86/xstate: Fix latent bugs in expand_xsave_states() Andrew Cooper
2016-09-12 11:41 ` Jan Beulich
2016-09-12 12:29 ` Andrew Cooper
2016-09-12 12:41 ` Jan Beulich
2016-09-12 12:43 ` Jan Beulich
2016-09-12 13:57 ` Andrew Cooper
2016-09-12 14:13 ` Jan Beulich
2016-09-12 9:51 ` Andrew Cooper [this message]
2016-09-12 12:14 ` [PATCH 5/6] x86/domctl: Fix migration of guests which are not using xsave Jan Beulich
2016-09-12 12:46 ` Andrew Cooper
2016-09-12 13:41 ` Jan Beulich
2016-09-12 9:51 ` [PATCH 6/6] x86/xstate: Fix latent bugs in compress_xsave_states() Andrew Cooper
2016-09-12 12:27 ` Jan Beulich
2016-09-12 12:59 ` Andrew Cooper
2016-09-12 13:47 ` Jan Beulich
2016-09-12 15:28 ` Andrew Cooper
2016-09-12 16:10 ` Jan Beulich
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=1473673900-8585-6-git-send-email-andrew.cooper3@citrix.com \
--to=andrew.cooper3@citrix.com \
--cc=JBeulich@suse.com \
--cc=xen-devel@lists.xen.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).