On Wed, 20 Jul 2016 14:34:30 +1000
Alexey Kardashevskiy <aik@xxxxxxxxx> wrote:
> static long tce_iommu_register_pages(struct tce_container *container,
> @@ -128,10 +129,17 @@ static long tce_iommu_register_pages(struct
> tce_container *container, ((vaddr + size) < vaddr))
> return -EINVAL;
>
> - ret = mm_iommu_get(vaddr, entries, &mem);
> + if (!container->mm) {
> + if (!current->mm)
> + return -ESRCH; /* process exited */
This shouldn't happen if we're a userspace process.
> +
> + atomic_inc(¤t->mm->mm_count);
> + container->mm = current->mm;
> + }
> +
> + ret = mm_iommu_get(container->mm, vaddr, entries, &mem);
Is it possible for processes (different mm) to be using the same
container?
> @@ -354,6 +362,8 @@ static void tce_iommu_release(void *iommu_data)
> tce_iommu_free_table(tbl);
> }
>
> + if (container->mm)
> + mmdrop(container->mm);
> tce_iommu_disable(container);
> mutex_destroy(&container->lock);
I'm wondering why keep the mm around at all. There is a bit of
locked_vm accounting there (which maybe doesn't exactly do the
right thing if we're talking about current task's rlimit if the
mm does not belong to current anyway).
The interesting cases are only the ones where a thread does
something with container->mm when current->mm != container->mm
(either a different process or a kernel thread). In what
situations does that happen?
Thanks,
Nick
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
