[Intel-gfx] Framebuffer compression with 2.6.32-rc
jbarnes at virtuousgeek.org
Wed Oct 14 15:24:37 PDT 2009
On Thu, 15 Oct 2009 00:01:59 +0200
Florian Mickler <florian at mickler.org> wrote:
> On Tue, 13 Oct 2009 11:38:07 -0700
> Eric Anholt <eric at anholt.net> wrote:
> > On Tue, 2009-10-13 at 14:28 -0400, Matt Turner wrote:
> > > On Tue, Oct 13, 2009 at 1:58 PM, Jesse Barnes
> > > <jbarnes at virtuousgeek.org> wrote:
> > > > On Tue, 13 Oct 2009 19:23:05 +0200
> > > > Tino Keitel <tino.keitel+xorg at tikei.de> wrote:
> > > >
> > > >> Hi folks,
> > > >>
> > > >> I use Xserver 1.6.4, xorg-intel 2.9.0 and kernel 2.6.32-rc4.
> > > >> How to I see if framebuffer compression is enabled or not?
> > > >> Before KMS, I could see in in Xorg.0.log.
> > > >>
> > > >> The hardware is a ThinkPad X61s with i965 graphics.
> > > >
> > > > You have to do a register dump and check FBC regs or enable DRM
> > > > debug output and look for the "fbc enabled" messages.
> > >
> > > Do we want it to be _that_ transparent?
> > It may very well be that we want to build an intel_gpu_powertop to
> > track the runtime status of our power savings efforts, or maybe
> > integrate it into upstream powertop. But yes, this is all
> > debugging stuff that should be entirely transparent.
> what about smth like this? it is not overly useful, but...
> From 823dc6cf318c05019d9fe4ca2bf3c787a382da2f Mon Sep 17 00:00:00 2001
> From: Florian Mickler <florian at mickler.org>
> Date: Wed, 14 Oct 2009 17:59:03 +0200
> Subject: [PATCH] report framebuffer-compression status via debugfs
> This patch adds a debugfs-knob (i915_fbc) to query the fbc-status
> Signed-off-by: Florian Mickler <florian at mickler.org>
> drivers/gpu/drm/i915/i915_debugfs.c | 15 +++++++++++++++
> 1 files changed, 15 insertions(+), 0 deletions(-)
> diff --git a/drivers/gpu/drm/i915/i915_debugfs.c
> b/drivers/gpu/drm/i915/i915_debugfs.c index f8ce9a3..b634512 100644
> --- a/drivers/gpu/drm/i915/i915_debugfs.c
> +++ b/drivers/gpu/drm/i915/i915_debugfs.c
> @@ -348,6 +348,20 @@ static int i915_ringbuffer_info(struct seq_file
> *m, void *data) return 0;
> +static int i915_fbc_info(struct seq_file *m, void *data)
> + struct drm_info_node *node = (struct drm_info_node *)
> + struct drm_device *dev = node->minor->dev;
> + drm_i915_private_t *dev_priv = dev->dev_private;
> + unsigned int enabled = I915_READ(FBC_STATUS) &
> FBC_STAT_COMPRESSING; +
> + seq_printf(m, "Framebuffer compression enabled : %s\n",
> + enabled ? "yes" : "no");
> + return 0;
It should be checking the enable bit instead, the compressing bit will
be turning on & off a lot.
But we have a register dumper for this, so I don't think we need a
Jesse Barnes, Intel Open Source Technology Center
More information about the Intel-gfx