summaryrefslogtreecommitdiffstats
path: root/drivers/gpu/drm/radeon/radeon_gart.c
diff options
context:
space:
mode:
authorJens Axboe <jens.axboe@oracle.com>2009-12-03 13:49:39 +0100
committerJens Axboe <jens.axboe@oracle.com>2009-12-03 13:49:39 +0100
commit220d0b1dbf78c6417a658c96e571415552d3abac (patch)
tree70cd3862540c38ea490e7a27c3c7acc35b680234 /drivers/gpu/drm/radeon/radeon_gart.c
parent474b18ccc264c472abeec50f48469b6477202699 (diff)
parent22763c5cf3690a681551162c15d34d935308c8d7 (diff)
Merge branch 'master' into for-2.6.33
Diffstat (limited to 'drivers/gpu/drm/radeon/radeon_gart.c')
-rw-r--r--drivers/gpu/drm/radeon/radeon_gart.c20
1 files changed, 10 insertions, 10 deletions
diff --git a/drivers/gpu/drm/radeon/radeon_gart.c b/drivers/gpu/drm/radeon/radeon_gart.c
index a931af065dd..a68d7566178 100644
--- a/drivers/gpu/drm/radeon/radeon_gart.c
+++ b/drivers/gpu/drm/radeon/radeon_gart.c
@@ -140,15 +140,15 @@ void radeon_gart_unbind(struct radeon_device *rdev, unsigned offset,
WARN(1, "trying to unbind memory to unitialized GART !\n");
return;
}
- t = offset / 4096;
- p = t / (PAGE_SIZE / 4096);
+ t = offset / RADEON_GPU_PAGE_SIZE;
+ p = t / (PAGE_SIZE / RADEON_GPU_PAGE_SIZE);
for (i = 0; i < pages; i++, p++) {
if (rdev->gart.pages[p]) {
pci_unmap_page(rdev->pdev, rdev->gart.pages_addr[p],
PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
rdev->gart.pages[p] = NULL;
rdev->gart.pages_addr[p] = 0;
- for (j = 0; j < (PAGE_SIZE / 4096); j++, t++) {
+ for (j = 0; j < (PAGE_SIZE / RADEON_GPU_PAGE_SIZE); j++, t++) {
radeon_gart_set_page(rdev, t, 0);
}
}
@@ -169,8 +169,8 @@ int radeon_gart_bind(struct radeon_device *rdev, unsigned offset,
DRM_ERROR("trying to bind memory to unitialized GART !\n");
return -EINVAL;
}
- t = offset / 4096;
- p = t / (PAGE_SIZE / 4096);
+ t = offset / RADEON_GPU_PAGE_SIZE;
+ p = t / (PAGE_SIZE / RADEON_GPU_PAGE_SIZE);
for (i = 0; i < pages; i++, p++) {
/* we need to support large memory configurations */
@@ -185,9 +185,9 @@ int radeon_gart_bind(struct radeon_device *rdev, unsigned offset,
}
rdev->gart.pages[p] = pagelist[i];
page_base = rdev->gart.pages_addr[p];
- for (j = 0; j < (PAGE_SIZE / 4096); j++, t++) {
+ for (j = 0; j < (PAGE_SIZE / RADEON_GPU_PAGE_SIZE); j++, t++) {
radeon_gart_set_page(rdev, t, page_base);
- page_base += 4096;
+ page_base += RADEON_GPU_PAGE_SIZE;
}
}
mb();
@@ -200,14 +200,14 @@ int radeon_gart_init(struct radeon_device *rdev)
if (rdev->gart.pages) {
return 0;
}
- /* We need PAGE_SIZE >= 4096 */
- if (PAGE_SIZE < 4096) {
+ /* We need PAGE_SIZE >= RADEON_GPU_PAGE_SIZE */
+ if (PAGE_SIZE < RADEON_GPU_PAGE_SIZE) {
DRM_ERROR("Page size is smaller than GPU page size!\n");
return -EINVAL;
}
/* Compute table size */
rdev->gart.num_cpu_pages = rdev->mc.gtt_size / PAGE_SIZE;
- rdev->gart.num_gpu_pages = rdev->mc.gtt_size / 4096;
+ rdev->gart.num_gpu_pages = rdev->mc.gtt_size / RADEON_GPU_PAGE_SIZE;
DRM_INFO("GART: num cpu pages %u, num gpu pages %u\n",
rdev->gart.num_cpu_pages, rdev->gart.num_gpu_pages);
/* Allocate pages table */