From: Pan Xinhui <xinhuix....@intel.com>
It's safe and more reasonable to unlock memtype_lock right after
rbt_memtype_check_insert. It's not cool to call kfree, pr_info, etc with
this lock held. So move spin_unlock a little ahead.
memory_lock protects data stored in rb-tree, if *new* succeed to be
stored into the rb-tree, we might hit panic. Because we access *new* in
dprintk "cattr_name(new->type)". data stored in the rb-tree might be
freed at any possbile time. It's abviously wrong to access such data
without lock. So save new->type to actual_type, and use actual_type in
dprintk.
Signed-off-by: Pan Xinhui <xinhuix....@intel.com>
---
change from V1:
fix an access of *new* without memtype_lock held.
---
arch/x86/mm/pat.c | 10 ++++------
1 file changed, 4 insertions(+), 6 deletions(-)
diff --git a/arch/x86/mm/pat.c b/arch/x86/mm/pat.c
index 188e3e0..f3c49fa 100644
--- a/arch/x86/mm/pat.c
+++ b/arch/x86/mm/pat.c
@@ -538,22 +538,20 @@ int reserve_memtype(u64 start, u64 end, enum
page_cache_mode req_type,
new->type = actual_type;
spin_lock(&memtype_lock);
-
err = rbt_memtype_check_insert(new, new_type);
+ actual_type = new->type;
+ spin_unlock(&memtype_lock);
+
if (err) {
pr_info("x86/PAT: reserve_memtype failed [mem %#010Lx-%#010Lx],
track %s, req %s\n",
start, end - 1,
cattr_name(new->type), cattr_name(req_type));
kfree(new);
- spin_unlock(&memtype_lock);
-
return err;
}
- spin_unlock(&memtype_lock);
-
dprintk("reserve_memtype added [mem %#010Lx-%#010Lx], track %s, req %s,
ret %s\n",
- start, end - 1, cattr_name(new->type), cattr_name(req_type),
+ start, end - 1, cattr_name(actual_type), cattr_name(req_type),
new_type ? cattr_name(*new_type) : "-");
return err;
--
1.9.1
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
