2020-09-25 09:59:26

by John Garry

Subject: [PATCH 0/2] iommu/iova: Solve longterm IOVA issue

This series contains a patch to solve the longterm IOVA issue which
leizhen originally tried to address at [0].

I also included the small optimisation from Cong Wang, which never seems
to be have been accepted [1]. There was some debate of the other patches
in that series, but this one is quite straightforward.

@Cong Wang, Please resend your series if prefer I didn't upstream your
patch.

[0] https://lore.kernel.org/linux-iommu/[email protected]/
[1] https://lore.kernel.org/linux-iommu/[email protected]/

Cong Wang (1):
iommu: avoid taking iova_rbtree_lock twice

John Garry (1):
iommu/iova: Flush CPU rcache for when a depot fills

drivers/iommu/iova.c | 33 ++++++++++++++++++++++-----------
1 file changed, 22 insertions(+), 11 deletions(-)

--
2.26.2

2020-09-25 21:06:34

by Cong Wang

Subject: Re: [PATCH 0/2] iommu/iova: Solve longterm IOVA issue

On Fri, Sep 25, 2020 at 2:56 AM John Garry <[email protected]> wrote:
>
> This series contains a patch to solve the longterm IOVA issue which
> leizhen originally tried to address at [0].
>
> I also included the small optimisation from Cong Wang, which never seems
> to be have been accepted [1]. There was some debate of the other patches
> in that series, but this one is quite straightforward.
>
> @Cong Wang, Please resend your series if prefer I didn't upstream your
> patch.

Thanks for letting me know. But I still don't think it is worth any effort,
given it is hard to work with Robin. Users who care about latency here
should just disable IOMMU, it is really hard to optimize IOVA cache
performance to catch up with !IOMMU case.

So, please feel free to carry it on your own, I have no problem with it.

Thanks.