Re: [PATCH v2 0/6] arch, mm/execmem: resolve confusion about set_direct_map_valid_noflush()

On Thu, Aug 27, 2026 at 10:39:47PM -0700, Andrew Morton wrote:
> On Sun, 23 Aug 2026 14:46:11 +0300 "Mike Rapoport (Microsoft)" <[email protected]> wrote:
> 
> > Recent discussion about implementation of execmem's ROX caches on arm64
> 
> I'm assuming a "[1]" reference was intended here.

Yes :)
 
> > revealed a confusion about how set_direct_map_valid_noflush()
> > implemented on different architectures.
> > 
> > On arm64 it sets or clears the PTE_VALID bit marking a PTE as present or
> > not present.
> > 
> > On other architectures it's a range version of
> > set_direct_map_invalid_noflush() and set_direct_map_default_noflush()
> > 
> > Unlike arm64::set_direct_map_valid_noflush(),
> > set_direct_map_default_noflush() not only marks PTE as present, but also
> > sets its default protection mode.
> > 
> > Other than that, initial design of execmem ROX caches didn't rely on
> > restoration of large mappings that's now available on x86, but
> > completely removed the memory allocated for the ROX cache from the
> > direct map to ensure that large mappings are not split. This precluded
> > usage of VM_FLUSH_RESET_PERMS for the ROX cache allocations and required
> > execmem to implement manipulation of the direct map alias.
> > 
> > Current implementation of ROX caches does not remove the direct map
> > alias but simply calls set_memory_rox() that updates the permissions in
> > both vmalloc address space and the direct map and relies on
> > collapse_large_pages() in x86 CPA to keep large mappings.
> > 
> > This allow using VM_FLUSH_RESET_PERMS for execmem ROX cache allocations
> > with small adjustments to set_direct_map APIs and vmalloc::reset_perms()
> > behaviour: adding number of pages parameter to set_direct_map APIs and
> > making resetting of the direct map permissions in vmalloc VMAP_HUGE
> > friendly.
> > 
> > Implement these adjustments, make execmem always use
> > VM_FLUSH_RESET_PERMS and revert set_direct_map_valid_noflush() changes.
> > 
> > [1] https://lore.kernel.org/all/[email protected]
> 
> The patchset is perhaps a little green but I tossed it in there, thanks.
> 
> I hit a few rejects in arch/x86/mm/pat/set_memory.c, but minor stuff.
> 
> I trust you've seen Sashiko's incidental oh-btw comments.
> 
> 	https://sashiko.dev/#/patchset/[email protected]

Yeah, it did discover something fishy with the fallbacks handling.
I'm looking into it.
 

-- 
Sincerely yours,
Mike.