[PATCH] x86/mm, mm/hwpoison: Fix the unmap kernel 1:1 pages check condition

[luofei <luofei@unicloud.com>]

[ Upstream commit fd0e786d9d09024f67bd71ec094b110237dc3840 ]

This commit solves the problem of unmap kernel 1:1 pages
unconditionally, it appears in Linus's tree 4.16 and later
versions, and is backported to 4.14.x and 4.15.x stable branches.

But the backported patch has its logic reversed when calling
memory_failure() to determine whether it needs to unmap the
kernel page. Only when memory_failure() returns successfully,
the kernel page can be unmapped.

Signed-off-by: luofei <luofei@unicloud.com>
Cc: stable@vger.kernel.org #v4.14.x
Cc: stable@vger.kernel.org #v4.15.x
---
 arch/x86/kernel/cpu/mcheck/mce.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c
index 95c09db1bba2..d8399a689165 100644
--- a/arch/x86/kernel/cpu/mcheck/mce.c
+++ b/arch/x86/kernel/cpu/mcheck/mce.c
@@ -589,7 +589,7 @@ static int srao_decode_notifier(struct notifier_block *nb, unsigned long val,
 
 	if (mce_usable_address(mce) && (mce->severity == MCE_AO_SEVERITY)) {
 		pfn = mce->addr >> PAGE_SHIFT;
-		if (memory_failure(pfn, MCE_VECTOR, 0))
+		if (!memory_failure(pfn, MCE_VECTOR, 0))
 			mce_unmap_kpfn(pfn);
 	}
 
-- 
2.27.0