[PATCH] drm/radeon/nouveau: fix build regression on alpha due to Xen changes.

[PATCH] drm/radeon/nouveau: fix build regression on alpha due to Xen changes.

[Dave Airlie]

From: Dave Airlie <airlied@redhat.com>

The Xen changes were using DMA_ERROR_CODE which isn't defined on a few
platforms, however we reverted the Xen patch that caused use to try and
use this code path earlier in 2.6.39 cycle, so for now lets just force
the code to never take this path and allow it to build again on alpha.

The proper long term answer is probably to store if the dma_addr has
been assigned to alongside the dma_addr in the higher level code,
though I think Thomas wanted to rewrite most of this anyways properly.

Signed-off-by: Dave Airlie <airlied@redhat.com>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
---
 drivers/gpu/drm/nouveau/nouveau_sgdma.c |    3 ++-
 drivers/gpu/drm/radeon/radeon_gart.c    |    6 +++---
 2 files changed, 5 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nouveau_sgdma.c b/drivers/gpu/drm/nouveau/nouveau_sgdma.c
index 4bce801..c77111e 100644
--- a/drivers/gpu/drm/nouveau/nouveau_sgdma.c
+++ b/drivers/gpu/drm/nouveau/nouveau_sgdma.c
@@ -42,7 +42,8 @@ nouveau_sgdma_populate(struct ttm_backend *be, unsigned long num_pages,
 
 	nvbe->nr_pages = 0;
 	while (num_pages--) {
-		if (dma_addrs[nvbe->nr_pages] != DMA_ERROR_CODE) {
+		/* this code path isn't called and is incorrect anyways */
+		if (0) { /*dma_addrs[nvbe->nr_pages] != DMA_ERROR_CODE)*/
 			nvbe->pages[nvbe->nr_pages] =
 					dma_addrs[nvbe->nr_pages];
 		 	nvbe->ttm_alloced[nvbe->nr_pages] = true;
diff --git a/drivers/gpu/drm/radeon/radeon_gart.c b/drivers/gpu/drm/radeon/radeon_gart.c
index 8a955bb..a533f52 100644
--- a/drivers/gpu/drm/radeon/radeon_gart.c
+++ b/drivers/gpu/drm/radeon/radeon_gart.c
@@ -181,9 +181,9 @@ int radeon_gart_bind(struct radeon_device *rdev, unsigned offset,
 	p = t / (PAGE_SIZE / RADEON_GPU_PAGE_SIZE);
 
 	for (i = 0; i < pages; i++, p++) {
-		/* On TTM path, we only use the DMA API if TTM_PAGE_FLAG_DMA32
-		 * is requested. */
-		if (dma_addr[i] != DMA_ERROR_CODE) {
+		/* we reverted the patch using dma_addr in TTM for now but this
+		 * code stops building on alpha so just comment it out for now */
+		if (0) { /*dma_addr[i] != DMA_ERROR_CODE) */
 			rdev->gart.ttm_alloced[p] = true;
 			rdev->gart.pages_addr[p] = dma_addr[i];
 		} else {
-- 
1.7.1

Re: [PATCH] drm/radeon/nouveau: fix build regression on alpha due to Xen changes.

[Konrad Rzeszutek Wilk]

On Mon, May 09, 2011 at 12:24:04PM +1000, Dave Airlie wrote:
> From: Dave Airlie <airlied@redhat.com>
> 
> The Xen changes were using DMA_ERROR_CODE which isn't defined on a few
> platforms, however we reverted the Xen patch that caused use to try and
> use this code path earlier in 2.6.39 cycle, so for now lets just force
> the code to never take this path and allow it to build again on alpha.
> 
> The proper long term answer is probably to store if the dma_addr has
> been assigned to alongside the dma_addr in the higher level code,
> though I think Thomas wanted to rewrite most of this anyways properly.

<nods> Yes, just need to find the time :-)
> 
> Signed-off-by: Dave Airlie <airlied@redhat.com>
> Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>

Acked-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>

Thanks for sending this patch out!

> ---
>  drivers/gpu/drm/nouveau/nouveau_sgdma.c |    3 ++-
>  drivers/gpu/drm/radeon/radeon_gart.c    |    6 +++---
>  2 files changed, 5 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/gpu/drm/nouveau/nouveau_sgdma.c b/drivers/gpu/drm/nouveau/nouveau_sgdma.c
> index 4bce801..c77111e 100644
> --- a/drivers/gpu/drm/nouveau/nouveau_sgdma.c
> +++ b/drivers/gpu/drm/nouveau/nouveau_sgdma.c
> @@ -42,7 +42,8 @@ nouveau_sgdma_populate(struct ttm_backend *be, unsigned long num_pages,
>  
>  	nvbe->nr_pages = 0;
>  	while (num_pages--) {
> -		if (dma_addrs[nvbe->nr_pages] != DMA_ERROR_CODE) {
> +		/* this code path isn't called and is incorrect anyways */
> +		if (0) { /*dma_addrs[nvbe->nr_pages] != DMA_ERROR_CODE)*/
>  			nvbe->pages[nvbe->nr_pages] =
>  					dma_addrs[nvbe->nr_pages];
>  		 	nvbe->ttm_alloced[nvbe->nr_pages] = true;
> diff --git a/drivers/gpu/drm/radeon/radeon_gart.c b/drivers/gpu/drm/radeon/radeon_gart.c
> index 8a955bb..a533f52 100644
> --- a/drivers/gpu/drm/radeon/radeon_gart.c
> +++ b/drivers/gpu/drm/radeon/radeon_gart.c
> @@ -181,9 +181,9 @@ int radeon_gart_bind(struct radeon_device *rdev, unsigned offset,
>  	p = t / (PAGE_SIZE / RADEON_GPU_PAGE_SIZE);
>  
>  	for (i = 0; i < pages; i++, p++) {
> -		/* On TTM path, we only use the DMA API if TTM_PAGE_FLAG_DMA32
> -		 * is requested. */
> -		if (dma_addr[i] != DMA_ERROR_CODE) {
> +		/* we reverted the patch using dma_addr in TTM for now but this
> +		 * code stops building on alpha so just comment it out for now */
> +		if (0) { /*dma_addr[i] != DMA_ERROR_CODE) */
>  			rdev->gart.ttm_alloced[p] = true;
>  			rdev->gart.pages_addr[p] = dma_addr[i];
>  		} else {
> -- 
> 1.7.1