From: Jesse Barnes <jbarnes@virtuousgeek.org>
To: Daniel Vetter <daniel@ffwll.ch>
Cc: Ben Widawsky <ben@bwidawsk.net>, intel-gfx@lists.freedesktop.org
Subject: Re: [PATCH 20/26] drm/i915: ValleyView has limited cacheability
Date: Wed, 28 Mar 2012 10:43:02 -0700 [thread overview]
Message-ID: <20120328104302.45203e6a@jbarnes-desktop> (raw)
In-Reply-To: <20120326184945.GZ4014@phenom.ffwll.local>
[-- Attachment #1.1: Type: text/plain, Size: 1947 bytes --]
On Mon, 26 Mar 2012 20:49:45 +0200
Daniel Vetter <daniel@ffwll.ch> wrote:
> On Mon, Mar 26, 2012 at 11:34:21AM -0700, Ben Widawsky wrote:
> > On Thu, Mar 22, 2012 at 02:39:02PM -0700, Jesse Barnes wrote:
> > > The GT can snoop CPU writes, but doesn't snoop into the CPU cache when
> > > it does writes, so we can't use the cache bits the same way.
> >
> > I found this commit message to be confusing. Is it simply saying CPU
> > writes are snooped by the GT, but GT writes are not snooped bv the CPU?
> >
> > >
> > > So map the status and pipe control pages as uncached on ValleyView, and
> > > only set the pages to cached if we're on a supported platform.
> >
> > I'd like to see in the commit message why the pipe control page needs to
> > be uncached. The only workarounds on the top of my head don't care about
> > the coherency.
>
> Afaik we've cleared this up in our mtg yesterday:
> - Full coherent gtt mappings work, they simply moved the bit around (we
> need to set bit2 instead of bit1 like on snb/ivb).
> - It sounds like all the gpu functions can handle coherent memory, like
> on snb/ivb. But because there's no shared cache between the gpu and the
> cpu you don't gain anything, but only lose due to the required snoop
> traffic.
> - Because there's no last level cache it also means that when the gpu does
> a write and snoops the cpu cache, it essentially means the cpu
> completely drops it's cacheline and has to go back to main memory
> (instead of l3 like it does on llc platforms). Gpu reads snoop the cpu
> cache corectly.
>
> I hope this clears up the confusion around coherency on vlv. Let Jesse
> only needs to check with a real piece of silicon whether that's true ;-)
> -Daniel
Right, this patch is bogus; VLV should be just as capable here, so we
can drop this one and I'll verify the cache bits.
--
Jesse Barnes, Intel Open Source Technology Center
[-- Attachment #1.2: signature.asc --]
[-- Type: application/pgp-signature, Size: 836 bytes --]
[-- Attachment #2: Type: text/plain, Size: 159 bytes --]
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
next prev parent reply other threads:[~2012-03-28 17:43 UTC|newest]
Thread overview: 56+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-03-22 21:38 [RFCv2] ValleyView support Jesse Barnes
2012-03-22 21:38 ` [PATCH 01/26] drm/i915: move NEEDS_FORCE_WAKE to i915_drv.c Jesse Barnes
2012-03-22 22:20 ` Ben Widawsky
2012-03-23 22:46 ` Daniel Vetter
2012-03-22 21:38 ` [PATCH 02/26] drm/i915: re-order GT IIR bit definitions Jesse Barnes
2012-03-22 22:25 ` Ben Widawsky
2012-03-23 22:46 ` Daniel Vetter
2012-03-22 21:38 ` [PATCH 03/26] drm/i915: add ValleyView driver structs and IS_VALLEYVIEW macro Jesse Barnes
2012-03-22 22:31 ` Ben Widawsky
2012-03-22 21:38 ` [PATCH 04/26] drm/i915: ValleyView watermark support Jesse Barnes
2012-03-23 3:29 ` Ben Widawsky
2012-03-23 9:51 ` Daniel Vetter
2012-03-24 2:46 ` Ben Widawsky
2012-03-22 21:38 ` [PATCH 05/26] drm/i915: PLL defines for VLV Jesse Barnes
2012-03-23 3:35 ` Ben Widawsky
2012-03-22 21:38 ` [PATCH 06/26] drm/i915: interrupt bit definitions " Jesse Barnes
2012-03-22 21:38 ` [PATCH 07/26] drm/i915: add ValleyView clock gating init Jesse Barnes
2012-03-22 23:25 ` Ben Widawsky
2012-03-22 21:38 ` [PATCH 08/26] drm/i915: add DPIO read/write functions for ValleyView Jesse Barnes
2012-03-23 4:03 ` Ben Widawsky
2012-03-23 17:29 ` Eugeni Dodonov
2012-03-22 21:38 ` [PATCH 09/26] drm/i915: split PLL update code out of i9xx_crtc_mode_set Jesse Barnes
2012-03-23 4:16 ` Ben Widawsky
2012-03-23 23:00 ` Daniel Vetter
2012-03-22 21:38 ` [PATCH 10/26] drm/i915: split LVDS " Jesse Barnes
2012-03-23 23:04 ` Daniel Vetter
2012-03-22 21:38 ` [PATCH 11/26] drm/i915: ValleyView mode setting limits and PLL functions Jesse Barnes
2012-03-22 21:38 ` [PATCH 12/26] drm/i915: program drain latency regs on ValleyView Jesse Barnes
2012-03-26 1:50 ` Ben Widawsky
2012-03-22 21:38 ` [PATCH 13/26] drm/i915: Enable DP panel power sequencing for ValleyView Jesse Barnes
2012-03-22 21:38 ` [PATCH 14/26] drm/i915: Enable HDMI on ValleyView Jesse Barnes
2012-03-22 21:38 ` [PATCH 15/26] agp/intel: map more registers for use by the GTT code Jesse Barnes
2012-03-26 2:05 ` Ben Widawsky
2012-03-26 7:06 ` Daniel Vetter
2012-03-22 21:38 ` [PATCH 16/26] agp/intel: add ValleyView AGP driver Jesse Barnes
2012-03-26 2:16 ` Ben Widawsky
2012-03-26 7:08 ` Daniel Vetter
2012-03-26 18:17 ` Ben Widawsky
2012-03-22 21:38 ` [PATCH 17/26] agp/intel: bind " Jesse Barnes
2012-03-22 21:39 ` [PATCH 18/26] drm/i915: add ValleyView specific CRT detect function Jesse Barnes
2012-03-23 17:11 ` Eugeni Dodonov
2012-03-22 21:39 ` [PATCH 19/26] drm/i915: add ValleyView specific force wake get/put functions Jesse Barnes
2012-03-23 17:20 ` Eugeni Dodonov
2012-03-26 18:20 ` Ben Widawsky
2012-03-22 21:39 ` [PATCH 20/26] drm/i915: ValleyView has limited cacheability Jesse Barnes
2012-03-22 23:31 ` Jesse Barnes
2012-03-26 18:34 ` Ben Widawsky
2012-03-26 18:49 ` Daniel Vetter
2012-03-28 17:43 ` Jesse Barnes [this message]
2012-03-22 21:39 ` [PATCH 21/26] drm/i915: ValleyView IRQ support Jesse Barnes
2012-03-22 21:39 ` [PATCH 22/26] drm/i915: display regs are at 0x180000 on ValleyView Jesse Barnes
2012-03-22 21:39 ` [PATCH 23/26] drm/i915: check for disabled interrupts " Jesse Barnes
2012-03-22 21:39 ` [PATCH 24/26] drm/i915: add HDMI and DP port enumeration " Jesse Barnes
2012-03-22 21:39 ` [PATCH 25/26] drm/i915: disable turbo on ValleyView for now Jesse Barnes
2012-03-23 17:04 ` Eugeni Dodonov
2012-03-22 21:39 ` [PATCH 26/26] drm/i915: bind driver to ValleyView chipsets Jesse Barnes
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=20120328104302.45203e6a@jbarnes-desktop \
--to=jbarnes@virtuousgeek.org \
--cc=ben@bwidawsk.net \
--cc=daniel@ffwll.ch \
--cc=intel-gfx@lists.freedesktop.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