在 2022/4/6 17:22, Borislav Petkov 写道:
On Wed, Apr 06, 2022 at 09:13:05AM +0000, Tong Tiangen wrote:
The follow patch
There's no concept of "follow patch" in git - you need to explain this
differently.
will add copy_mc_to_user to include/linux/uaccess.h, X86
must declare Implemented to avoid compile error.
I don't know what that means. Try again pls.
This description is not good, will redescribe it in next version.
Here I describe the reasons for this:
Patch 3/7 in patchset[1] introduce copy_mc_to_user() in
include/linux/uaccess.h and the prototype is:
static inline unsigned long __must_check
copy_mc_to_user(void *dst, const void *src, size_t cnt)
The prototype in x86 is:
unsigned long __must_check
copy_mc_to_user(void *to, const void *from, unsigned len);
This two are a little different, so I added the follow code to x86 to
avoid prototype conflict compile error.
#define copy_mc_to_user copy_mc_to_user
In addition, I think this #define should be added here.
[1]https://patchwork.kernel.org/project/linux-mm/cover/20220406091311.3354723-1-tongtiangen@xxxxxxxxxx/
Thanks.
Tong.
