[PATCH] KVM: arm64: Return on kvm_gmem_get_pfn() failure in gmem_abort()

[Suzuki K Poulose]

On 22/10/2026 13:30, Fuad Tabba wrote:
> The kvm-arm64/mmu-fixes-7.4 merge dropped the return from the
> kvm_gmem_get_pfn() error path in gmem_abort(). Both parents of the
> merge keep that return. On a failed kvm_gmem_get_pfn(), gmem_abort()
> then carries on with a zero pfn, and on a nested stage-2 it uses the
> mapping it has just freed.
> 
> Restore the return.
> 
> Fixes: 65f95b2874a15 ("Merge branch kvm-arm64/mmu-fixes-7.4 into kvmarm-master/next")
> Signed-off-by: Fuad Tabba <fuad.tabba at linux.dev>
> ---
> 
> Notes:
>      I went through the rest of the merge's conflict resolution with
>      git show --remerge-diff, and this is the only line it dropped.
> 
>   arch/arm64/kvm/mmu.c | 1 +
>   1 file changed, 1 insertion(+)
> 
> diff --git a/arch/arm64/kvm/mmu.c b/arch/arm64/kvm/mmu.c
> index a48f3f531e827..ba4fe58222ad5 100644
> --- a/arch/arm64/kvm/mmu.c
> +++ b/arch/arm64/kvm/mmu.c
> @@ -1802,6 +1802,7 @@ static int gmem_abort(const struct kvm_s2_fault_desc *s2fd,
>   						      s2vi.vma_pagesize, write_fault, exec_fault,
>   						      false);
>   		kfree(mapping);
> +		return ret;
>   	}
>   
>   	if (!(s2fd->memslot->flags & KVM_MEM_READONLY))

Thanks for fixing this ! Looks correct to me.

Reviewed-by: Suzuki K Poulose <suzuki.poulose at arm.com>