[Qemu-devel] [PATCH] os: don't corrupt pre-existing memory-backend data with prealloc

[Daniel P. Berrange]

When using a memory-backend object with prealloc turned on, QEMU
will memset() the first byte in every memory page to zero. While
this might have been acceptable for memory backends associated
with RAM, this corrupts application data for NVDIMMs.

Instead of setting every page to zero, read the current byte
value and then just write that same value back, so we are not
corrupting the original data. Directly write the value instead
of memset()ing it, since there's no benefit to memset for a
single byte write.

Signed-off-by: Daniel P. Berrange <address@hidden>
---

NB, I have not tested performance of this, so no idea if this
makes it better/worse/no-change. Would appreciate if Jitendra
could repeat tests to see if it impacts scalability at all.

 util/oslib-posix.c | 15 ++++++++++++++-
 1 file changed, 14 insertions(+), 1 deletion(-)

diff --git a/util/oslib-posix.c b/util/oslib-posix.c
index 35012b9..2a5bb93 100644
--- a/util/oslib-posix.c
+++ b/util/oslib-posix.c
@@ -355,7 +355,20 @@ void os_mem_prealloc(int fd, char *area, size_t memory, 
Error **errp)
 
         /* MAP_POPULATE silently ignores failures */
         for (i = 0; i < numpages; i++) {
-            memset(area + (hpagesize * i), 0, 1);
+            /*
+             * Read & write back the same value, so we don't
+             * corrupt existinng user/app data that might be
+             * stored.
+             *
+             * 'volatile' to stop compiler optimizing this away
+             * to a no-op
+             *
+             * TODO: get a better solution from kernel so we
+             * don't need to write at all so we don't cause
+             * wear on the storage backing the region...
+             */
+            volatile char val = *(area + (hpagesize * i));
+            *(area + (hpagesize * i)) = val;
         }
     }
 
-- 
2.9.3