From: Andrew Cooper <andrew.cooper3@citrix.com>
To: Xen-devel <xen-devel@lists.xen.org>
Cc: Andrew Cooper <andrew.cooper3@citrix.com>,
Wei Liu <wei.liu2@citrix.com>, Jan Beulich <JBeulich@suse.com>
Subject: [PATCH 5/7] x86/mm: Carve steal_linear_address() out of replace_grant_host_mapping()
Date: Tue, 12 Sep 2017 13:14:44 +0100 [thread overview]
Message-ID: <1505218486-4416-6-git-send-email-andrew.cooper3@citrix.com> (raw)
In-Reply-To: <1505218486-4416-1-git-send-email-andrew.cooper3@citrix.com>
Document its curious semantics.
No functional change.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
---
CC: Jan Beulich <JBeulich@suse.com>
CC: Wei Liu <wei.liu2@citrix.com>
---
xen/arch/x86/mm.c | 105 ++++++++++++++++++++++++++++++------------------------
1 file changed, 58 insertions(+), 47 deletions(-)
diff --git a/xen/arch/x86/mm.c b/xen/arch/x86/mm.c
index e08b4a8..afd98d9 100644
--- a/xen/arch/x86/mm.c
+++ b/xen/arch/x86/mm.c
@@ -4075,14 +4075,68 @@ int create_grant_pv_mapping(uint64_t addr, unsigned long frame,
return rc;
}
-int replace_grant_pv_mapping(uint64_t addr, unsigned long frame,
- uint64_t new_addr, unsigned int flags)
+/*
+ * This exists soley for implementing GNTABOP_unmap_and_replace, the ABI of
+ * which is bizare. This GNTTABOP isn't used any more, but was used by
+ * classic-xen kernels and PVOps Linux before the M2P_OVERRIDE infrastructure
+ * was replaced with something which actually worked.
+ *
+ * Look up the L1e mapping linear, and zap it. Return the L1e via *out.
+ * Returns a boolean indicating success. If success, the caller is
+ * responsible for calling put_page_from_l1e().
+ */
+static bool steal_linear_address(unsigned long linear, l1_pgentry_t *out)
{
struct vcpu *curr = current;
struct domain *currd = curr->domain;
l1_pgentry_t *pl1e, ol1e;
+ struct page_info *page;
mfn_t gl1mfn;
- struct page_info *l1pg;
+ bool okay = false;
+
+ ASSERT(is_pv_domain(currd));
+
+ pl1e = map_guest_l1e(linear, &gl1mfn);
+ if ( !pl1e )
+ {
+ gdprintk(XENLOG_WARNING,
+ "Could not find L1 PTE for linear %"PRIx64"\n", linear);
+ goto out;
+ }
+
+ if ( !get_page_from_mfn(gl1mfn, currd) )
+ goto out_unmap;
+
+ page = mfn_to_page(gl1mfn);
+ if ( !page_lock(page) )
+ goto out_put;
+
+ if ( (page->u.inuse.type_info & PGT_type_mask) != PGT_l1_page_table )
+ goto out_unlock;
+
+ ol1e = *pl1e;
+ okay = UPDATE_ENTRY(l1, pl1e, ol1e, l1e_empty(), mfn_x(gl1mfn), curr, 0);
+
+ out_unlock:
+ page_unlock(page);
+ out_put:
+ put_page(page);
+ out_unmap:
+ unmap_domain_page(pl1e);
+
+ if ( okay )
+ *out = ol1e;
+
+ out:
+ return okay;
+}
+
+int replace_grant_pv_mapping(uint64_t addr, unsigned long frame,
+ uint64_t new_addr, unsigned int flags)
+{
+ struct vcpu *curr = current;
+ struct domain *currd = curr->domain;
+ l1_pgentry_t ol1e;
int rc;
unsigned int grant_pte_flags = grant_to_pte_flags(flags, 0);
@@ -4108,51 +4162,8 @@ int replace_grant_pv_mapping(uint64_t addr, unsigned long frame,
if ( !new_addr )
return destroy_grant_va_mapping(addr, frame, grant_pte_flags, curr);
- pl1e = map_guest_l1e(new_addr, &gl1mfn);
- if ( !pl1e )
- {
- gdprintk(XENLOG_WARNING,
- "Could not find L1 PTE for address %"PRIx64"\n", new_addr);
- return GNTST_general_error;
- }
-
- if ( !get_page_from_mfn(gl1mfn, currd) )
- {
- unmap_domain_page(pl1e);
- return GNTST_general_error;
- }
-
- l1pg = mfn_to_page(gl1mfn);
- if ( !page_lock(l1pg) )
- {
- put_page(l1pg);
- unmap_domain_page(pl1e);
- return GNTST_general_error;
- }
-
- if ( (l1pg->u.inuse.type_info & PGT_type_mask) != PGT_l1_page_table )
- {
- page_unlock(l1pg);
- put_page(l1pg);
- unmap_domain_page(pl1e);
- return GNTST_general_error;
- }
-
- ol1e = *pl1e;
-
- if ( unlikely(!UPDATE_ENTRY(l1, pl1e, ol1e, l1e_empty(),
- mfn_x(gl1mfn), curr, 0)) )
- {
- page_unlock(l1pg);
- put_page(l1pg);
- gdprintk(XENLOG_WARNING, "Cannot delete PTE entry at %p\n", pl1e);
- unmap_domain_page(pl1e);
+ if ( !steal_linear_address(new_addr, &ol1e) )
return GNTST_general_error;
- }
-
- page_unlock(l1pg);
- put_page(l1pg);
- unmap_domain_page(pl1e);
rc = replace_grant_va_mapping(addr, frame, grant_pte_flags, ol1e, curr);
if ( rc )
--
2.1.4
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
next prev parent reply other threads:[~2017-09-12 12:14 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-09-12 12:14 [PATCH 0/7] x86/mm: Post XSA-234 cleanup Andrew Cooper
2017-09-12 12:14 ` [PATCH 1/7] x86/mm: Improvements to PV l1e mapping helpers Andrew Cooper
2017-09-12 12:29 ` Wei Liu
2017-09-12 15:22 ` Jan Beulich
2017-09-12 12:14 ` [PATCH 2/7] x86/mm: Factor out the grant flags to pte flags conversion logic Andrew Cooper
2017-09-12 13:28 ` Wei Liu
2017-09-12 15:25 ` Jan Beulich
2017-09-12 12:14 ` [PATCH 3/7] x86/mm: Misc cleanup to {create, replace}_grant_host_mapping() Andrew Cooper
2017-09-12 13:40 ` Wei Liu
2017-09-12 15:25 ` Jan Beulich
2017-09-12 12:14 ` [PATCH 4/7] x86/mm: Combine create_grant_{pte, va}_mapping() Andrew Cooper
2017-09-12 14:04 ` Wei Liu
2017-09-12 15:31 ` Jan Beulich
2017-09-12 12:14 ` Andrew Cooper [this message]
2017-09-12 14:19 ` [PATCH 5/7] x86/mm: Carve steal_linear_address() out of replace_grant_host_mapping() Wei Liu
2017-09-12 15:41 ` Jan Beulich
2017-09-12 12:14 ` [PATCH 6/7] x86/mm: Combine {destroy, replace}_grant_{pte, va}_mapping() Andrew Cooper
2017-09-12 14:58 ` Wei Liu
2017-09-12 16:30 ` Andrew Cooper
2017-09-12 16:32 ` Wei Liu
2017-09-12 16:36 ` Andrew Cooper
2017-09-12 16:37 ` Wei Liu
2017-09-12 15:46 ` Jan Beulich
2017-09-12 12:14 ` [PATCH 7/7] x86/mm: Prevent 32bit PV guests using out-of-range linear addresses Andrew Cooper
2017-09-12 15:50 ` Jan Beulich
2017-09-12 16:04 ` Andrew Cooper
2017-09-13 8:28 ` 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=1505218486-4416-6-git-send-email-andrew.cooper3@citrix.com \
--to=andrew.cooper3@citrix.com \
--cc=JBeulich@suse.com \
--cc=wei.liu2@citrix.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).