Regression in v4.5-rc1, bisected to commit 39bfcd5235e0 ("drm/i915: more virtual south bridge detection")

* Regression in v4.5-rc1, bisected to commit 39bfcd5235e0 ("drm/i915: more virtual south bridge detection")
@ 2016-01-25  2:52 Bjørn Mork
  2016-01-25  8:26 ` [PATCH] Revert "drm/i915: more virtual south bridge detection" Jani Nikula
                   ` (7 more replies)
  0 siblings, 8 replies; 10+ messages in thread
From: Bjørn Mork @ 2016-01-25  2:52 UTC (permalink / raw)
  To: Gerd Hoffmann; +Cc: Daniel Vetter, Jani Nikula, intel-gfx, linux-kernel

Hello,

my oldish Thinkpad X301 only wanted to show a blank screen in v4.5-rc1.
Bisecting resulted in:

HEAD is now at 39bfcd5235e0 drm/i915: more virtual south bridge detection
39bfcd5235e07e95ad3e70eab8e0b85db181de9e is the first bad commit
commit 39bfcd5235e07e95ad3e70eab8e0b85db181de9e
Author: Gerd Hoffmann <kraxel@redhat.com>
Date:   Thu Nov 26 12:03:51 2015 +0100

    drm/i915: more virtual south bridge detection
    
    Commit "30c964a drm/i915: Detect virtual south bridge" detects and
    handles the southbridge emulated by vmware esx.  Add the ich9 south
    bridge emulated by 'qemu -M q35'.
    
    Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
    Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

:040000 040000 b59ceb519d517a00e41e575346505b9ebde06288 825eb4e5684952de0931312183d1cf163c43219a M      drivers


I have confirmed tha reverting this commit on top of v4.5-rc1 fixes the
problem.  And looking into the details, I believe this makes sense.  The
patch does this (where id is a 0xff00 masked device ID):

diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index 649e20aaa3ff..90faa8e03fca 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -565,7 +565,8 @@ void intel_detect_pch(struct drm_device *dev)
                                DRM_DEBUG_KMS("Found SunrisePoint LP PCH\n");
                                WARN_ON(!IS_SKYLAKE(dev) &&
                                        !IS_KABYLAKE(dev));
-                       } else if (id == INTEL_PCH_P2X_DEVICE_ID_TYPE) {
+                       } else if ((id == INTEL_PCH_P2X_DEVICE_ID_TYPE) ||
+                                  (id == INTEL_PCH_QEMU_DEVICE_ID_TYPE)) {
                                dev_priv->pch_type = intel_virt_detect_pch(dev);
                        } else
                                continue;
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 3d8741eff7d3..439226d39437 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -2609,6 +2609,7 @@ struct drm_i915_cmd_table {
 #define INTEL_PCH_SPT_DEVICE_ID_TYPE           0xA100
 #define INTEL_PCH_SPT_LP_DEVICE_ID_TYPE                0x9D00
 #define INTEL_PCH_P2X_DEVICE_ID_TYPE           0x7100
+#define INTEL_PCH_QEMU_DEVICE_ID_TYPE          0x2900 /* qemu q35 has 2918 */
 
 #define INTEL_PCH_TYPE(dev) (__I915__(dev)->pch_type)
 #define HAS_PCH_SPT(dev) (INTEL_PCH_TYPE(dev) == PCH_SPT)



FWIW, the very real ISA bridge of my laptop matches the above:

bjorn@nemi:~$ lspci -vvvnns1f.0
00:1f.0 ISA bridge [0601]: Intel Corporation ICH9M-E LPC Interface Controller [8086:2917] (rev 03)
        Subsystem: Lenovo ICH9M-E LPC Interface Controller [17aa:20f5]
        Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
        Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
        Latency: 0
        Capabilities: <access denied>
        Kernel driver in use: lpc_ich
        Kernel modules: lpc_ich


This is a GM45 chipset.  I believe most chipsets of this generation use
29xx device IDs for the ISA bridge.  For example the P35:

 bjorn@canardo:~$ lspci -nns1f.0
 00:1f.0 ISA bridge [0601]: Intel Corporation 82801IB (ICH9) LPC Interface Controller [8086:2918] (rev 02)

You cannot assume all these are virtual systems. There are still crazy
people out there using the real thing :)



Bjørn

^ permalink raw reply related	[flat|nested] 10+ messages in thread