On Thu, Apr 04, 2024 at 09:02:17PM +0200, David Hildenbrand wrote:
On 04.04.24 01:34, Deepak Gupta wrote:} - } else if (!(vm_flags & VM_READ)) { + } else if (!(vm_flags & VM_READ) && !vma_is_shadow_stack(vm_flags)) { + /* reads allowed if its shadow stack vma */ if (!(gup_flags & FOLL_FORCE)) return -EFAULT; /*Unless I am missing something, this is not a simple cleanup. It should go into a separate patch with a clearly documented reason for that change.
I tried that here https://lore.kernel.org/linux-mm/CAKC1njTPBqtsAOn-CWhB+-8FaZ2KWkkz-vRZr7MZq=0yLUdjcQ@xxxxxxxxxxxxxx/T/ But at that time, VM_SHADOW_STACK for riscv meant only VM_WRITE. So I think there was obvious uneasiness with that part. Now we have VM_SHADOW_STACK pretty much same for all arches and only 64bit. I'll try it again as a separate patch.
-- Cheers, David / dhildenb
