drm/i915: Document our lackluster FSB frequency readout
authorVille Syrjälä <ville.syrjala@linux.intel.com>
Thu, 14 May 2020 12:38:37 +0000 (15:38 +0300)
committerVille Syrjälä <ville.syrjala@linux.intel.com>
Tue, 19 May 2020 09:08:20 +0000 (12:08 +0300)
Document the fact that we aren't reading out the actual FSB
frequency but rather just the state of the FSB straps.
Some BIOSen allow you to configure the two independently.
So if someone sets the two up in an inconsistent manner
we'll get the wrong answer here and thus will end up with
incorrect aux/pps clock dividers. Alas, proper docs are no
longer around so we can't do any better.

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200514123838.3017-2-ville.syrjala@linux.intel.com
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
drivers/gpu/drm/i915/display/intel_cdclk.c

index c17cf61..d57dfec 100644 (file)
@@ -2704,7 +2704,16 @@ static int g4x_hrawclk(struct drm_i915_private *dev_priv)
 {
        u32 clkcfg;
 
-       /* hrawclock is 1/4 the FSB frequency */
+       /*
+        * hrawclock is 1/4 the FSB frequency
+        *
+        * Note that this only reads the state of the FSB
+        * straps, not the actual FSB frequency. Some BIOSen
+        * let you configure each independently. Ideally we'd
+        * read out the actual FSB frequency but sadly we
+        * don't know which registers have that information,
+        * and all the relevant docs have gone to bit heaven :(
+        */
        clkcfg = intel_de_read(dev_priv, CLKCFG) & CLKCFG_FSB_MASK;
 
        /* ELK seems to redefine some of the values */