All of lore.kernel.org
 help / color / mirror / Atom feed
From: Dave Airlie <airlied@gmail.com>
To: Benjamin Herrenschmidt <benh@kernel.crashing.org>,
	Jerome Glisse <j.glisse@gmail.com>,
	Oded Gabbay <oded.gabbay@gmail.com>,
	dri-devel <dri-devel@lists.freedesktop.org>
Subject: ongoing writecombine on ppc
Date: Fri, 2 Oct 2015 14:42:53 +1000	[thread overview]
Message-ID: <CAPM=9twpPBLBb6j1ee5_DDv5UqfTzCJh-63CHyDxRU4Xd0ny4g@mail.gmail.com> (raw)

[-- Attachment #1: Type: text/plain, Size: 281 bytes --]

I don't think we resolved this the last time we talked about it,

but radeon writecombine maps fail hard on ppc, I think all the fixes
either did something bad to AGP systems or weren't liked.

My patch attached just fixes radeon, which is where I'm still seeing the issue.

Dave.

[-- Attachment #2: 0001-radeon-don-t-attempt-WC-mappings-on-powerpc.patch --]
[-- Type: text/x-patch, Size: 2479 bytes --]

From 67d00d2fcb756a3d2b6061d0831242138889a81f Mon Sep 17 00:00:00 2001
From: Dave Airlie <airlied@redhat.com>
Date: Fri, 2 Oct 2015 12:22:08 +1000
Subject: [PATCH] radeon: don't attempt WC mappings on powerpc

---
 drivers/gpu/drm/radeon/radeon_object.c | 16 ++++++++++++----
 include/drm/drm_cache.h                |  9 +++++++++
 2 files changed, 21 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c
index d302488..4acda0e 100644
--- a/drivers/gpu/drm/radeon/radeon_object.c
+++ b/drivers/gpu/drm/radeon/radeon_object.c
@@ -33,6 +33,7 @@
 #include <linux/slab.h>
 #include <drm/drmP.h>
 #include <drm/radeon_drm.h>
+#include <drm/drm_cache.h>
 #include "radeon.h"
 #include "radeon_trace.h"
 
@@ -92,6 +93,15 @@ bool radeon_ttm_bo_is_radeon_bo(struct ttm_buffer_object *bo)
 	return false;
 }
 
+static inline bool radeon_placement_can_wc(struct radeon_bo *rbo)
+{
+	if ((rbo->flags & RADEON_GEM_GTT_WC) && drm_arch_can_wc_memory())
+		return true;
+	if (rbo->rdev->flags & RADEON_IS_AGP)
+		return true;
+	return false;
+}
+
 void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain)
 {
 	u32 c = 0, i;
@@ -123,8 +133,7 @@ void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain)
 			rbo->placements[c++].flags = TTM_PL_FLAG_UNCACHED |
 				TTM_PL_FLAG_TT;
 
-		} else if ((rbo->flags & RADEON_GEM_GTT_WC) ||
-			   (rbo->rdev->flags & RADEON_IS_AGP)) {
+		} else if (radeon_placement_can_wc(rbo)) {
 			rbo->placements[c].fpfn = 0;
 			rbo->placements[c++].flags = TTM_PL_FLAG_WC |
 				TTM_PL_FLAG_UNCACHED |
@@ -142,8 +151,7 @@ void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain)
 			rbo->placements[c++].flags = TTM_PL_FLAG_UNCACHED |
 				TTM_PL_FLAG_SYSTEM;
 
-		} else if ((rbo->flags & RADEON_GEM_GTT_WC) ||
-		    rbo->rdev->flags & RADEON_IS_AGP) {
+		} else if (radeon_placement_can_wc(rbo)) {
 			rbo->placements[c].fpfn = 0;
 			rbo->placements[c++].flags = TTM_PL_FLAG_WC |
 				TTM_PL_FLAG_UNCACHED |
diff --git a/include/drm/drm_cache.h b/include/drm/drm_cache.h
index 7bfb063..461a055 100644
--- a/include/drm/drm_cache.h
+++ b/include/drm/drm_cache.h
@@ -35,4 +35,13 @@
 
 void drm_clflush_pages(struct page *pages[], unsigned long num_pages);
 
+static inline bool drm_arch_can_wc_memory(void)
+{
+#if defined(CONFIG_PPC) && !defined(CONFIG_NOT_COHERENT_CACHE)
+	return false;
+#else
+	return true;
+#endif
+}
+
 #endif
-- 
2.4.3


[-- Attachment #3: Type: text/plain, Size: 159 bytes --]

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

             reply	other threads:[~2015-10-02  4:42 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-10-02  4:42 Dave Airlie [this message]
2015-10-02  4:45 ` ongoing writecombine on ppc Benjamin Herrenschmidt
2015-10-02  4:47   ` Benjamin Herrenschmidt
2015-10-02  4:50     ` Benjamin Herrenschmidt
2015-10-02  4:53   ` Dave Airlie
2015-10-02  4:56     ` Benjamin Herrenschmidt
2016-01-21 15:24       ` Oded Gabbay
2016-01-21 15:39         ` Oded Gabbay
2016-01-21 17:10           ` Alex Deucher
2016-01-22  2:32             ` Michel Dänzer
2016-01-28 14:30               ` Oded Gabbay

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='CAPM=9twpPBLBb6j1ee5_DDv5UqfTzCJh-63CHyDxRU4Xd0ny4g@mail.gmail.com' \
    --to=airlied@gmail.com \
    --cc=benh@kernel.crashing.org \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=j.glisse@gmail.com \
    --cc=oded.gabbay@gmail.com \
    /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.