x86/mm: Teach pte_mkwrite() about stack memory
authorRick Edgecombe <rick.p.edgecombe@intel.com>
Tue, 13 Jun 2023 00:10:47 +0000 (17:10 -0700)
committerRick Edgecombe <rick.p.edgecombe@intel.com>
Tue, 11 Jul 2023 21:12:19 +0000 (14:12 -0700)
commitb497e52ddb2ab750b00d8cc78209613558fc503b
tree50a6375a6c6b7f0cf81c08081bec6831b1adf841
parent29f890d1050fc099fd578d9db844d6c0375902b6
x86/mm: Teach pte_mkwrite() about stack memory

If a VMA has the VM_SHADOW_STACK flag, it is shadow stack memory. So
when it is made writable with pte_mkwrite(), it should create shadow
stack memory, not conventionally writable memory. Now that all the places
where shadow stack memory might be created pass a VMA into pte_mkwrite(),
it can know when it should do this.

So make pte_mkwrite() create shadow stack memory when the VMA has the
VM_SHADOW_STACK flag. Do the same thing for pmd_mkwrite().

This requires referencing VM_SHADOW_STACK in these functions, which are
currently defined in pgtable.h, however mm.h (where VM_SHADOW_STACK is
located) can't be pulled in without causing problems for files that
reference pgtable.h. So also move pte/pmd_mkwrite() into pgtable.c, where
they can safely reference VM_SHADOW_STACK.

Signed-off-by: Rick Edgecombe <rick.p.edgecombe@intel.com>
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Borislav Petkov (AMD) <bp@alien8.de>
Reviewed-by: Kees Cook <keescook@chromium.org>
Acked-by: Mike Rapoport (IBM) <rppt@kernel.org>
Acked-by: Deepak Gupta <debug@rivosinc.com>
Tested-by: Pengfei Xu <pengfei.xu@intel.com>
Tested-by: John Allen <john.allen@amd.com>
Tested-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/all/20230613001108.3040476-22-rick.p.edgecombe%40intel.com
arch/x86/mm/pgtable.c