From 985b8bb486e7dd924925898f86fffca546d698db Mon Sep 17 00:00:00 2001 From: Ville Syrjälä Date: Thu, 11 Jun 2015 16:31:15 +0300 Subject: drm/i915: Align DSPSURF to 128k on VLV/CHV VLV/CHV have problems with 4k aligned linear scanout buffers. The VLV docs got updated at some point to say that we need to align them to 128k, just like we do on gen4. So far I've seen the problem manifest when the stride is an odd multiple of 512 bytes, and the surface address meets the following pattern '(addr & 0xf000) == 0x1000' (also == 0x2000 is problematic on VLV). The result is a starcase effect (so some pages get dropped maybe?), with a few pages here and there clearly getting scannout out at the wrong position. I've not actually been able to reproduce this problem on gen4, so it's not clear of the issue is any way related to the 128k restrictions supposedly inherited from gen4. But let's hope the 128k alignment is sufficient to hide it all. Signed-off-by: Ville Syrjälä Reviewed-by: Clint Taylor Reviewed-by: Arun R Murthy Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'drivers/gpu/drm/i915/intel_display.c') diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 1d440e7e0344..e94a9a0a6d91 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2333,7 +2333,8 @@ static unsigned int intel_linear_alignment(struct drm_i915_private *dev_priv) { if (INTEL_INFO(dev_priv)->gen >= 9) return 256 * 1024; - else if (IS_BROADWATER(dev_priv) || IS_CRESTLINE(dev_priv)) + else if (IS_BROADWATER(dev_priv) || IS_CRESTLINE(dev_priv) || + IS_VALLEYVIEW(dev_priv)) return 128 * 1024; else if (INTEL_INFO(dev_priv)->gen >= 4) return 4 * 1024; -- cgit v1.2.3-55-g7522