On Fri, Jun 17, 2022 at 05:48:11AM +0000, Christophe Leroy wrote:
> Is it a good idea to duplicate vm_get_page_prot() in each architecture ?
It is a completely trivial array index. And I really like the idea
of not having the protection_map in common code - it really is an
implementation detail. But what we could do is something like
#define DECLARE_VM_GET_PAGE_PROT \
pgprot_t vm_get_page_prot(unsigned long vm_flags) \
{ \
return protection_map[vm_flags & \
(VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)]; \
} \
EXPORT_SYMBOL(vm_get_page_prot);
as a helper for the architectures.
> Maybe it is, but it will also mean changing common code like
> mm/debug_vm_pgtable.c which accesses protection_map[] directly as of today.
That's already gone thanks to the good work from Anshuman.
