From: Rob Clark <robdclark@chromium.org>
To: Christoph Hellwig <hch@lst.de>
Cc: Rob Clark <robdclark@gmail.com>,
dri-devel <dri-devel@lists.freedesktop.org>,
Catalin Marinas <catalin.marinas@arm.com>,
Will Deacon <will@kernel.org>,
Maarten Lankhorst <maarten.lankhorst@linux.intel.com>,
Maxime Ripard <maxime.ripard@bootlin.com>,
Sean Paul <sean@poorly.run>, David Airlie <airlied@linux.ie>,
Daniel Vetter <daniel@ffwll.ch>,
Allison Randal <allison@lohutok.net>,
Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
Thomas Gleixner <tglx@linutronix.de>,
linux-arm-kernel@lists.infradead.org,
LKML <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH 1/2] drm: add cache support for arm64
Date: Tue, 6 Aug 2019 09:23:51 -0700 [thread overview]
Message-ID: <CAJs_Fx6uztwDy2PqRy3Tc9p12k8r_ovS2tAcsMV6HqnAp=Ggug@mail.gmail.com> (raw)
In-Reply-To: <20190806155044.GC25050@lst.de>
On Tue, Aug 6, 2019 at 8:50 AM Christoph Hellwig <hch@lst.de> wrote:
>
> On Tue, Aug 06, 2019 at 07:11:41AM -0700, Rob Clark wrote:
> > Agreed that drm_cflush_* isn't a great API. In this particular case
> > (IIUC), I need wb+inv so that there aren't dirty cache lines that drop
> > out to memory later, and so that I don't get a cache hit on
> > uncached/wc mmap'ing.
>
> So what is the use case here? Allocate pages using the page allocator
> (or CMA for that matter), and then mmaping them to userspace and never
> touching them again from the kernel?
Currently, it is pages coming from tmpfs. Ideally we want pages that
are swappable when unpinned.
CPU mappings are *mostly* just mapping to userspace. There are a few
exceptions that are vmap'd (fbcon, and ringbuffer).
(Eventually I'd like to support pages passed in from userspace.. but
that is down the road.)
> > Tying it in w/ iommu seems a bit weird to me.. but maybe that is just
> > me, I'm certainly willing to consider proposals or to try things and
> > see how they work out.
>
> This was just my through as the fit seems easy. But maybe you'll
> need to explain your use case(s) a bit more so that we can figure out
> what a good high level API is.
Tying it to iommu_map/unmap would be awkward, as we could need to
setup cpu mmap before it ends up mapped to iommu. And the plan to
support per-process pagetables involved creating an iommu_domain per
userspace gl context.. some buffers would end up mapped into multiple
contexts/iommu_domains.
If the cache operation was detached from iommu_map/unmap, then it
would seem weird to be part of the iommu API.
I guess I'm not entirely sure what you had in mind, but this is why
iommu seemed to me like a bad fit.
> > Exposing the arch_sync_* API and using that directly (bypassing
> > drm_cflush_*) actually seems pretty reasonable and pragmatic. I did
> > have one doubt, as phys_to_virt() is only valid for kernel direct
> > mapped memory (AFAIU), what happens for pages that are not in kernel
> > linear map? Maybe it is ok to ignore those pages, since they won't
> > have an aliased mapping?
>
> They could have an aliased mapping in vmalloc/vmap space for example,
> if you created one. We have the flush_kernel_vmap_range /
> invalidate_kernel_vmap_range APIs for those, that are implement
> on architectures with VIVT caches.
If I only have to worry about aliased mappings that I create myself,
that doesn't seem too bad.. I could track that myself easily enough.
BR,
-R
next prev parent reply other threads:[~2019-08-06 16:24 UTC|newest]
Thread overview: 31+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-08-05 21:14 [PATCH 1/2] drm: add cache support for arm64 Rob Clark
2019-08-05 21:14 ` [PATCH 2/2] drm/msm: use drm_cache when available Rob Clark
2019-08-06 8:48 ` [PATCH 1/2] drm: add cache support for arm64 Christoph Hellwig
2019-08-06 9:38 ` Daniel Vetter
2019-08-06 11:38 ` Christoph Hellwig
2019-08-06 14:11 ` Rob Clark
2019-08-06 14:34 ` Mark Rutland
2019-08-06 16:31 ` Rob Clark
2019-08-07 12:38 ` Mark Rutland
2019-08-07 16:15 ` Rob Clark
2019-08-07 16:49 ` Mark Rutland
2019-08-07 17:30 ` Rob Clark
2019-08-08 7:59 ` Christoph Hellwig
2019-08-08 16:44 ` Rob Clark
2019-08-09 8:18 ` Christoph Hellwig
2019-08-08 7:58 ` Christoph Hellwig
2019-08-08 10:20 ` Mark Rutland
2019-08-08 10:24 ` Mark Rutland
2019-08-08 10:32 ` Will Deacon
2019-08-08 7:53 ` Christoph Hellwig
2019-08-06 15:50 ` Christoph Hellwig
2019-08-06 16:23 ` Rob Clark [this message]
2019-08-06 16:26 ` Rob Clark
2019-08-07 6:25 ` Christoph Hellwig
2019-08-07 8:48 ` Daniel Vetter
2019-08-08 9:55 ` Christoph Hellwig
2019-08-08 11:58 ` Daniel Vetter
2019-08-09 8:14 ` Christoph Hellwig
2019-08-07 16:09 ` Rob Clark
2019-08-08 10:00 ` Christoph Hellwig
2019-08-08 16:32 ` Rob Clark
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='CAJs_Fx6uztwDy2PqRy3Tc9p12k8r_ovS2tAcsMV6HqnAp=Ggug@mail.gmail.com' \
--to=robdclark@chromium.org \
--cc=airlied@linux.ie \
--cc=allison@lohutok.net \
--cc=catalin.marinas@arm.com \
--cc=daniel@ffwll.ch \
--cc=dri-devel@lists.freedesktop.org \
--cc=gregkh@linuxfoundation.org \
--cc=hch@lst.de \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=maarten.lankhorst@linux.intel.com \
--cc=maxime.ripard@bootlin.com \
--cc=robdclark@gmail.com \
--cc=sean@poorly.run \
--cc=tglx@linutronix.de \
--cc=will@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
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;
as well as URLs for NNTP newsgroup(s).