Patched in this series of patches from bwidawsk to test RC6 state on ILK. http://lists.freedesktop.org/archives/intel-gfx/2013-July/030263.html Issue was seeing if it ever reached RC6 state, sysfs has no interface for this anymore. I poked around in debugFS and found i915_drpc_info. cat /sys/kernel/debug/dri/64/i915_drpc_info HD boost: yes Boost freq: 9 HW control enabled: no SW control enabled: yes Gated voltage change: no Starting frequency: P8 Max P-state: P0 Min P-state: P10 RS1 VID: 0 RS2 VID: 8 Render standby enabled: yes Current RS state: RC1 It also shows RS state: on if performing activity on the GPU. Tried polling repeatedly using: while [ 1 ] ; do grep "Current" /sys/kernel/debug/dri/0/i915_drpc_info >> out.txt ; sleep .01; done Still bounced between RC1 and on, never hitting RC6 However, if I use intel-gpu-tools and run: intel_reg_read 0x111b8 I see: 0x111B8 : 0x474C3000 (RS2/RC6) or 0x111B8 : 0x471C3000 (RS1) or 0x111B8 : 0x470C3000 (on) Though today I'm seeing just: 0x111B8 : 0x0 0x111B8 : 0x100000
Reassigning to the newer Jesse :)
Hmm. Just a guess, can you try this: diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c index a6f4cb5..b0347cc 100644 --- a/drivers/gpu/drm/i915/i915_debugfs.c +++ b/drivers/gpu/drm/i915/i915_debugfs.c @@ -1020,8 +1020,8 @@ static int ironlake_drpc_info(struct seq_file *m) if (ret) return ret; - rgvmodectl = I915_READ(MEMMODECTL); rstdbyctl = I915_READ(RSTDBYCTL); + rgvmodectl = I915_READ(MEMMODECTL); crstandvid = I915_READ16(CRSTANDVID); mutex_unlock(&dev->struct_mutex);
timed out on this one.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.