pstore/ram: Use memcpy_toio instead of memcpy
authorFurquan Shaikh <furquan@google.com>
Mon, 15 Feb 2016 08:19:48 +0000 (09:19 +0100)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Fri, 28 Oct 2016 07:01:27 +0000 (03:01 -0400)
commit 7e75678d23167c2527e655658a8ef36a36c8b4d9 upstream.

persistent_ram_update uses vmap / iomap based on whether the buffer is in
memory region or reserved region. However, both map it as non-cacheable
memory. For armv8 specifically, non-cacheable mapping requests use a
memory type that has to be accessed aligned to the request size. memcpy()
doesn't guarantee that.

Signed-off-by: Furquan Shaikh <furquan@google.com>
Signed-off-by: Enric Balletbo Serra <enric.balletbo@collabora.com>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Olof Johansson <olofj@chromium.org>
Tested-by: Furquan Shaikh <furquan@chromium.org>
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
fs/pstore/ram_core.c

index 9cc6efe0b78568a94296ea1e3381d030a1ba4148..cc83231d91689ba6c5370d637081ce24a1209540 100644 (file)
@@ -263,7 +263,7 @@ static void notrace persistent_ram_update(struct persistent_ram_zone *prz,
        const void *s, unsigned int start, unsigned int count)
 {
        struct persistent_ram_buffer *buffer = prz->buffer;
-       memcpy(buffer->data + start, s, count);
+       memcpy_toio(buffer->data + start, s, count);
        persistent_ram_update_ecc(prz, start, count);
 }