Re: [PATCH v9 3/7] iommu: Add verisilicon IOMMU driver

[Nicolas Dufresne <nicolas.dufresne@collabora.com>]

[-- Attachment #1: Type: text/plain, Size: 1623 bytes --]

Hi,

Le vendredi 12 septembre 2025 à 13:47 +0200, Jörg Rödel a écrit :
> On Thu, Sep 11, 2025 at 05:57:13PM +0200, Benjamin Gaignard wrote:
> > The Verisilicon IOMMU hardware block can be found in combination
> > with Verisilicon hardware video codecs (encoders or decoders) on
> > different SoCs.
> > Enable it will allow us to use non contiguous memory allocators
> > for Verisilicon video codecs.
> > 
> > Signed-off-by: Benjamin Gaignard <benjamin.gaignard@collabora.com>
> > ---
> >  drivers/iommu/Kconfig     |  11 +
> >  drivers/iommu/Makefile    |   1 +
> >  drivers/iommu/vsi-iommu.c | 808 ++++++++++++++++++++++++++++++++++++++
> >  include/linux/vsi-iommu.h |  21 +
> >  4 files changed, 841 insertions(+)
> >  create mode 100644 drivers/iommu/vsi-iommu.c
> >  create mode 100644 include/linux/vsi-iommu.h
> 
> This will not go in before Will Deacons comment about code duplication with the
> Rockchip driver is addressed.

So you'd like a new layer of ops to abstract the common parts of two drivers ? I
mean, I can see this happening for VSI/RK, but it seems rather pointless if in a
year from now Jason get his new framework ready. Re-doing the ops if they cause
too much duplicate seems like a better direction.

Benjamin already stated he'd be happy to port once there is enough example and
acceptability of the new code to do so. Why do mainline users have to suffer
this ?

To me this rejection isn't about Benjamin's driver, all iommu seems to look
alike, so anyone else that would have sent new driver would have face the same
issue.

Nicolas

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 228 bytes --]