All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Christian König" <deathsimple@vodafone.de>
To: dri-devel@lists.freedesktop.org
Subject: [PATCH 5/6] drm/radeon: implement UVD hw workarounds for R6xx v3
Date: Sun, 24 Aug 2014 15:14:29 +0200	[thread overview]
Message-ID: <1408886070-31983-6-git-send-email-deathsimple@vodafone.de> (raw)
In-Reply-To: <1408886070-31983-1-git-send-email-deathsimple@vodafone.de>

From: Christian König <christian.koenig@amd.com>

Only the essentials, cause this hw generation is really buggy.

v2: start supporting RV670,RV620 and RV635 as well
v3: activate more workarounds

Signed-off-by: Christian König <christian.koenig@amd.com>
---
 drivers/gpu/drm/radeon/r600d.h    |  3 +++
 drivers/gpu/drm/radeon/uvd_v1_0.c | 26 +++++++++++++++++++++++++-
 2 files changed, 28 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/radeon/r600d.h b/drivers/gpu/drm/radeon/r600d.h
index ab200c4..e80153b 100644
--- a/drivers/gpu/drm/radeon/r600d.h
+++ b/drivers/gpu/drm/radeon/r600d.h
@@ -330,6 +330,7 @@
 #define	HDP_TILING_CONFIG				0x2F3C
 #define HDP_DEBUG1                                      0x2F34
 
+#define MC_CONFIG					0x2000
 #define MC_VM_AGP_TOP					0x2184
 #define MC_VM_AGP_BOT					0x2188
 #define	MC_VM_AGP_BASE					0x218C
@@ -375,6 +376,8 @@
 #define MC_VM_SYSTEM_APERTURE_HIGH_ADDR			0x2194
 #define MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR		0x2198
 
+#define RS_DQ_RD_RET_CONF				0x2348
+
 #define	PA_CL_ENHANCE					0x8A14
 #define		CLIP_VTX_REORDER_ENA				(1 << 0)
 #define		NUM_CLIP_SEQ(x)					((x) << 1)
diff --git a/drivers/gpu/drm/radeon/uvd_v1_0.c b/drivers/gpu/drm/radeon/uvd_v1_0.c
index 4d95a7d..e251624 100644
--- a/drivers/gpu/drm/radeon/uvd_v1_0.c
+++ b/drivers/gpu/drm/radeon/uvd_v1_0.c
@@ -207,8 +207,32 @@ done:
 	/* lower clocks again */
 	radeon_set_uvd_clocks(rdev, 0, 0);
 
-	if (!r)
+	if (!r) {
+		switch (rdev->family) {
+		case CHIP_RV610:
+		case CHIP_RV630:
+		case CHIP_RV620:
+			/* 64byte granularity workaround */
+			WREG32(MC_CONFIG, 0);
+			WREG32(MC_CONFIG, 1 << 4);
+			WREG32(RS_DQ_RD_RET_CONF, 0x3f);
+			WREG32(MC_CONFIG, 0x1f);
+
+			/* fall through */
+		case CHIP_RV670:
+		case CHIP_RV635:
+
+			/* write clean workaround */
+			WREG32_P(UVD_VCPU_CNTL, 0x10, ~0x10);
+			break;
+
+		default:
+			/* TODO: Do we need more? */
+			break;
+		}
+
 		DRM_INFO("UVD initialized successfully.\n");
+	}
 
 	return r;
 }
-- 
1.9.1

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

  parent reply	other threads:[~2014-08-24 13:15 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <UVD support for older ASICs>
2014-08-24 13:14 ` (unknown), Christian König
2014-08-24 13:14   ` [PATCH 1/6] drm/radeon: force UVD buffers into VRAM on RS[78]80 v2 Christian König
2014-08-25 17:14     ` Alex Deucher
2014-08-24 13:14   ` [PATCH 2/6] drm/radeon: properly init UVD MC bits on R600 Christian König
2014-08-24 13:14   ` [PATCH 3/6] drm/radeon: add set_uvd_clocks callback for r6xx v4 Christian König
2014-08-24 13:14   ` [PATCH 4/6] drm/radeon: add UVD support for older asics v4 Christian König
2014-08-24 13:14   ` Christian König [this message]
2014-08-24 13:14   ` [PATCH 6/6] drm/radeon: add UVD fw names for older asic Christian König
2014-08-24 13:34   ` Mike Lothian
2014-08-25  9:10     ` Re: Christian König
2014-08-26 14:14       ` (UVD RS780 first impression) Stefan Lange
2014-08-27  3:40         ` Michel Dänzer
2014-08-27  7:00           ` Christian König
2014-08-28 19:25             ` Stefan Lange
2014-08-30 15:58             ` Stefan Lange
2014-09-07 13:24       ` Markus Trippelsdorf
2014-09-08  3:47         ` Re: Alex Deucher
2014-09-08  7:13           ` Re: Markus Trippelsdorf

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1408886070-31983-6-git-send-email-deathsimple@vodafone.de \
    --to=deathsimple@vodafone.de \
    --cc=dri-devel@lists.freedesktop.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body. 
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.