On Tue, Nov 17, 2026 at 9:02 PM <[email protected]> wrote: > > Thank you for your contribution! Sashiko AI review found 1 potential > issue to consider: > - [Medium] drm/tyr: page table desync and GpuVa leak on pt_map() error > in sm_step_remap() > -- > > commit 890e6d1a0c182be186e5d6f90855d935bae30eda > Author: Sami Tolvanen <[email protected]> > > drm/tyr: vm: expand sub-block unmaps to 2MB boundaries on remap > > This commit expands sub-block unmaps in sm_step_remap() to 2MB boundaries when > the surviving fragments are physically contiguous and remaps the surviving > head and tail fragments as 4KB pages. > > > diff --git a/drivers/gpu/drm/tyr/vm.rs b/drivers/gpu/drm/tyr/vm.rs > > index 1dd11b2b12532..bcc864b174284 100644 > > --- a/drivers/gpu/drm/tyr/vm.rs > > +++ b/drivers/gpu/drm/tyr/vm.rs > > [ ... ] > > > @@ -749,10 +780,50 @@ fn sm_step_remap<'op>( > > op.va_to_unmap().addr() + op.va_to_unmap().length() > > }; > > > > - let unmap_length = unmap_end - unmap_start; > > + // The surviving fragments inherit the protection of the mapping > > being > > + // split. > > + let prot = op.va_to_unmap().data_ref().prot; > > + > > + let block = Alignment::SZ_2M; > > + let aligned_start = unmap_start.align_down(block); > > + let aligned_end = unmap_end.align_up(block).ok_or(EINVAL)?; > > + > > + // An end may only be expanded if the fragment the expansion > > sweeps and > > + // rebuilds is fully mapped and physically contiguous, because the > > + // rebuild maps it linearly from a single recovered address. A > > genuine > > + // 2MB block is always contiguous, so this never misses one and the > > + // partial-block unmap can never reach arm-lpae. A fragmented > > region is > > + // never expanded, so its sub-range unmap stays correct. > > + let head_paddr = match op.prev() { > > + Some(prev) if aligned_start < unmap_start && prev.addr() <= > > aligned_start => { > > + contiguous_phys(context.pt, aligned_start..unmap_start) > > + } > > + _ => None, > > + }; > > + let tail_paddr = match op.next() { > > + Some(next) if aligned_end > unmap_end && next.addr() + > > next.length() >= aligned_end => { > > + contiguous_phys(context.pt, unmap_end..aligned_end) > > + } > > + _ => None, > > + }; > > + > > + let region_start = if head_paddr.is_some() { > > + aligned_start > > + } else { > > + unmap_start > > + }; > > + let region_end = if tail_paddr.is_some() { > > + aligned_end > > + } else { > > + unmap_end > > + }; > > + > > + let prev_va = context.preallocated_gpuva()?; > > + let next_va = context.preallocated_gpuva()?; > > > > - if unmap_length > 0 { > > - let region = unmap_start..(unmap_start + unmap_length); > > + if region_end > region_start { > > + let region = region_start..region_end; > > + context.extend_lock(region.clone())?; > > pt_unmap(context.dev, context.pt, > > region.clone()).inspect_err(|e| { > > dev_err!( > > context.dev, > > @@ -764,14 +835,29 @@ fn sm_step_remap<'op>( > > })?; > > } > > > > - let prev_va = context.preallocated_gpuva()?; > > - let next_va = context.preallocated_gpuva()?; > > + if let Some(head_paddr) = head_paddr { > > + pt_map( > > + context.dev, > > + context.pt, > > + aligned_start, > > + head_paddr, > > + unmap_start - aligned_start, > > + prot, > > + )?; > > + } > > + if let Some(tail_paddr) = tail_paddr { > > + pt_map( > > + context.dev, > > + context.pt, > > + unmap_end, > > + tail_paddr, > > + aligned_end - unmap_end, > > + prot, > > + )?; > > + } > > > > - let (op_remapped, remap_ret) = op.remap( > > - [prev_va, next_va], > > - pin_init::init_zeroed(), > > - pin_init::init_zeroed(), > > - ); > > + let (op_remapped, remap_ret) = > > + op.remap([prev_va, next_va], GpuVaData { prot }, GpuVaData { > > prot }); > > [Severity: Medium] > Can a pt_map() failure here leave context.pt out of sync with GpuVm and leak > the GpuVa? > > pt_unmap() has already torn down the expanded 2MB range in context.pt before > pt_map() allocates L3 page tables for the surviving 4KB fragments. If either > pt_map() fails with ENOMEM, sm_step_remap() returns without restoring the > unmapped range in context.pt or updating the GpuVm tree via op.remap(). > > When Vm::kill() or Vm::unmap_range() later calls sm_step_unmap() on the > unsplit GpuVa, pt_unmap() hits the already-unmapped PTEs, triggers a WARN_ON > in __arm_lpae_unmap(), and returns Err(EINVAL) before op.remove() can unlink > and free the GpuVa and its GEM BO references.
This is valid, but allocation failure in pt_map() is a pre-existing problem. It will go away once these page table entries are pre-allocated. alice
