From: Pranjal Shrivastava <praan@google•com>
To: Jason Gunthorpe <jgg@nvidia•com>
Cc: Alexandre Ghiti <alex@ghiti•fr>,
AngeloGioacchino Del Regno
<angelogioacchino.delregno@collabora•com>,
Albert Ou <aou@eecs•berkeley.edu>,
asahi@lists•linux.dev,
Baolin Wang <baolin.wang@linux•alibaba.com>,
iommu@lists•linux.dev, Janne Grunau <j@jannau•net>,
Jernej Skrabec <jernej.skrabec@gmail•com>,
Joerg Roedel <joro@8bytes•org>,
Jean-Philippe Brucker <jpb@kernel•org>,
linux-arm-kernel@lists•infradead.org,
linux-mediatek@lists•infradead.org,
linux-riscv@lists•infradead.org, linux-sunxi@lists•linux.dev,
Matthias Brugger <matthias.bgg@gmail•com>,
Neal Gompa <neal@gompa•dev>, Orson Zhai <orsonzhai@gmail•com>,
Palmer Dabbelt <palmer@dabbelt•com>,
Paul Walmsley <pjw@kernel•org>,
Samuel Holland <samuel@sholland•org>,
Sven Peter <sven@kernel•org>,
virtualization@lists•linux.dev, Chen-Yu Tsai <wens@kernel•org>,
Will Deacon <will@kernel•org>, Yong Wu <yong.wu@mediatek•com>,
Chunyan Zhang <zhang.lyra@gmail•com>,
Lu Baolu <baolu.lu@linux•intel.com>,
Janusz Krzysztofik <janusz.krzysztofik@linux•intel.com>,
Joerg Roedel <joerg.roedel@amd•com>,
Jon Hunter <jonathanh@nvidia•com>,
patches@lists•linux.dev, Robin Murphy <robin.murphy@arm•com>,
Samiullah Khawaja <skhawaja@google•com>,
stable@vger•kernel.org, Vasant Hegde <vasant.hegde@amd•com>
Subject: Re: [PATCH] iommu: Always fill in gather when unmapping
Date: Wed, 1 Apr 2026 11:23:23 +0000 [thread overview]
Message-ID: <ac0AKyvHMYHlqL5i@google.com> (raw)
In-Reply-To: <0-v1-664d3acaabb9+78b-iommu_gather_always_jgg@nvidia.com>
On Tue, Mar 31, 2026 at 04:56:22PM -0300, Jason Gunthorpe wrote:
> The fixed commit assumed that the gather would always be populated if
> an iotlb_sync was required.
>
> arm-smmu-v3, amd, VT-d, riscv, s390, mtk all use information from the
> gather during their iotlb_sync() and this approach works for them.
>
> However, arm-smmu, qcom_iommu, ipmmu-vmsa, sun50i, sprd, virtio,
> apple-dart all ignore the gather during their iotlb_sync(). They
> mostly issue a full flush.
>
> Unfortunately the latter set of drivers often don't bother to add
> anything to the gather since they don't intend on using it. Since the
> core code now blocks gathers that were never filled, this caused those
> drivers to stop getting their iotlb_sync() calls and breaks them.
>
> Since it is impossible to tell the difference between gathers that are
> empty because there is nothing to do and gathers that are empty
> because they are not used, fill in the gathers for the missing cases.
>
I believe the problem is a fundamental disagreement between the core
layer and these drivers. The core assumes an empty gather means there
is no work to do, while these drivers expect a sync regardless. With
this, it seems we're forcing the drivers to lie to the core by
populating a gather they don't actually use just to trigger the sync.
I was wondering if, as a longer-term direction, having an explicit flag
for these drivers to indicate they always require a sync would be a
cleaner way to handle this than the trivial population?
Just a thought, not a hard disagreement with the current approach..
> io-pgtable might have intended to allow the driver to choose between
> gather or immediate flush because it passed gather to
> ops->tlb_add_page(), however no driver does anything with it.
>
> mtk uses io-pgtable-arm-v7s but added the range to the gather in the
> unmap callback. Move this into the io-pgtable-arm unmap itself. That
> will fix all the armv7 using drivers (arm-smmu, qcom_iommu,
> ipmmu-vmsa).
>
> arm-smmu uses both ARM_V7S and ARM LPAE formats. The LPAE formats
> already have the gather population because SMMUv3 requires it, so it
> becomes consistent.
>
> Add a trivial gather population to io-pgtable-dart.
>
> Add trivial populations to sprd, sun50i and virtio-iommu in their
> unmap functions.
>
> Fixes: 90c5def10bea ("iommu: Do not call drivers for empty gathers")
> Reported-by: Jon Hunter <jonathanh@nvidia•com>
> Closes: https://lore.kernel.org/r/8800a38b-8515-4bbe-af15-0dae81274bf7@nvidia.com
> Signed-off-by: Jason Gunthorpe <jgg@nvidia•com>
> ---
> drivers/iommu/io-pgtable-arm.c | 4 +++-
> drivers/iommu/io-pgtable-dart.c | 3 +++
> drivers/iommu/mtk_iommu.c | 1 -
> drivers/iommu/sprd-iommu.c | 1 +
> drivers/iommu/sun50i-iommu.c | 1 +
> drivers/iommu/virtio-iommu.c | 2 ++
> 6 files changed, 10 insertions(+), 2 deletions(-)
>
Acked-by: Pranjal Shrivastava <praan@google•com>
Thanks,
Praan
next prev parent reply other threads:[~2026-04-01 11:23 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-03-31 19:56 [PATCH] iommu: Always fill in gather when unmapping Jason Gunthorpe
2026-04-01 10:40 ` Jon Hunter
2026-04-01 11:23 ` Pranjal Shrivastava [this message]
2026-04-01 12:58 ` Jason Gunthorpe
2026-04-01 16:33 ` Robin Murphy
2026-04-01 17:36 ` Jason Gunthorpe
2026-04-02 18:11 ` Robin Murphy
2026-04-02 22:51 ` Jason Gunthorpe
2026-04-08 8:42 ` Jon Hunter
2026-04-08 13:34 ` Jon Hunter
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=ac0AKyvHMYHlqL5i@google.com \
--to=praan@google$(echo .)com \
--cc=alex@ghiti$(echo .)fr \
--cc=angelogioacchino.delregno@collabora$(echo .)com \
--cc=aou@eecs$(echo .)berkeley.edu \
--cc=asahi@lists$(echo .)linux.dev \
--cc=baolin.wang@linux$(echo .)alibaba.com \
--cc=baolu.lu@linux$(echo .)intel.com \
--cc=iommu@lists$(echo .)linux.dev \
--cc=j@jannau$(echo .)net \
--cc=janusz.krzysztofik@linux$(echo .)intel.com \
--cc=jernej.skrabec@gmail$(echo .)com \
--cc=jgg@nvidia$(echo .)com \
--cc=joerg.roedel@amd$(echo .)com \
--cc=jonathanh@nvidia$(echo .)com \
--cc=joro@8bytes$(echo .)org \
--cc=jpb@kernel$(echo .)org \
--cc=linux-arm-kernel@lists$(echo .)infradead.org \
--cc=linux-mediatek@lists$(echo .)infradead.org \
--cc=linux-riscv@lists$(echo .)infradead.org \
--cc=linux-sunxi@lists$(echo .)linux.dev \
--cc=matthias.bgg@gmail$(echo .)com \
--cc=neal@gompa$(echo .)dev \
--cc=orsonzhai@gmail$(echo .)com \
--cc=palmer@dabbelt$(echo .)com \
--cc=patches@lists$(echo .)linux.dev \
--cc=pjw@kernel$(echo .)org \
--cc=robin.murphy@arm$(echo .)com \
--cc=samuel@sholland$(echo .)org \
--cc=skhawaja@google$(echo .)com \
--cc=stable@vger$(echo .)kernel.org \
--cc=sven@kernel$(echo .)org \
--cc=vasant.hegde@amd$(echo .)com \
--cc=virtualization@lists$(echo .)linux.dev \
--cc=wens@kernel$(echo .)org \
--cc=will@kernel$(echo .)org \
--cc=yong.wu@mediatek$(echo .)com \
--cc=zhang.lyra@gmail$(echo .)com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox