At least for now, continue to assume mask_user_address() is safe on AMD when combined with STAC/CLAC -- as get_user(), put_user() and masked_user_access_begin() already do today.
v2: - Separate copy_to_user() and clear_user() changes out into separate patches - Add masking to __get_user() and __put_user() v1: https://lore.kernel.org/b626840e55d4aa86b4b9b377a4cc2cda7038d33d.1728706156.git.jpoim...@kernel.org Josh Poimboeuf (6): x86/uaccess: Avoid barrier_nospec() in copy_from_user() x86/uaccess: Avoid barrier_nospec() in __get_user() x86/uaccess: Rearrange putuser.S x86/uaccess: Add user pointer masking to __put_user() x86/uaccess: Add user pointer masking to copy_to_user() x86/uaccess: Add user pointer masking to clear_user() arch/powerpc/include/asm/uaccess.h | 2 + arch/x86/include/asm/uaccess_64.h | 10 ++-- arch/x86/lib/getuser.S | 27 +++++++-- arch/x86/lib/putuser.S | 92 ++++++++++++++++++------------ include/linux/uaccess.h | 6 -- 5 files changed, 86 insertions(+), 51 deletions(-) -- 2.47.0
