On 19/01/2022 11:37, Arunpravin wrote:
Implemented a function which walk through the order list,
compares the offset and returns the maximum offset block,
this method is unpredictable in obtaining the high range
address blocks which depends on allocation and deallocation.
for instance, if driver requests address at a low specific
range, allocator traverses from the root block and splits
the larger blocks until it reaches the specific block and
in the process of splitting, lower orders in the freelist
are occupied with low range address blocks and for the
subsequent TOPDOWN memory request we may return the low
range blocks.To overcome this issue, we may go with the
below approach.
The other approach, sorting each order list entries in
ascending order and compares the last entry of each
order list in the freelist and return the max block.
This creates sorting overhead on every drm_buddy_free()
request and split up of larger blocks for a single page
request.
v2:
- Fix alignment issues(Matthew Auld)
- Remove unnecessary list_empty check(Matthew Auld)
- merged the below patch to see the feature in action
- add top-down alloc support to i915 driver
Signed-off-by: Arunpravin <Arunpravin.PaneerSelvam@xxxxxxx>
---
drivers/gpu/drm/drm_buddy.c | 36 ++++++++++++++++---
drivers/gpu/drm/i915/i915_ttm_buddy_manager.c | 3 ++
include/drm/drm_buddy.h | 1 +
3 files changed, 35 insertions(+), 5 deletions(-)
diff --git a/drivers/gpu/drm/drm_buddy.c b/drivers/gpu/drm/drm_buddy.c
index 954e31962c74..6aa5c1ce25bf 100644
--- a/drivers/gpu/drm/drm_buddy.c
+++ b/drivers/gpu/drm/drm_buddy.c
@@ -371,6 +371,26 @@ alloc_range_bias(struct drm_buddy *mm,
return ERR_PTR(err);
}
+static struct drm_buddy_block *
+get_maxblock(struct list_head *head)
+{
+ struct drm_buddy_block *max_block = NULL, *node;
+
+ max_block = list_first_entry_or_null(head,
+ struct drm_buddy_block,
+ link);
+ if (!max_block)
+ return NULL;
+
+ list_for_each_entry(node, head, link) {
+ if (drm_buddy_block_offset(node) >
+ drm_buddy_block_offset(max_block))
+ max_block = node;
+ }
If we feed in the knowledge of the visible_size(or perhaps implement
that generically as "zones"), I think this can be done more efficiently.
It could also be useful to track directly in the allocator how much of
the visible_size is still available, rather than having to do that in
the upper levels by scanning the entire list. But hopefully in practice
this should be good enough for our needs,
Reviewed-by: Matthew Auld <matthew.auld@xxxxxxxxx>
+
+ return max_block;
+}
+
static struct drm_buddy_block *
alloc_from_freelist(struct drm_buddy *mm,
unsigned int order,
@@ -381,11 +401,17 @@ alloc_from_freelist(struct drm_buddy *mm,
int err;
for (i = order; i <= mm->max_order; ++i) {
- block = list_first_entry_or_null(&mm->free_list[i],
- struct drm_buddy_block,
- link);
- if (block)
- break;
+ if (flags & DRM_BUDDY_TOPDOWN_ALLOCATION) {
+ block = get_maxblock(&mm->free_list[i]);
+ if (block)
+ break;
+ } else {
+ block = list_first_entry_or_null(&mm->free_list[i],
+ struct drm_buddy_block,
+ link);
+ if (block)
+ break;
+ }
}
if (!block)
diff --git a/drivers/gpu/drm/i915/i915_ttm_buddy_manager.c b/drivers/gpu/drm/i915/i915_ttm_buddy_manager.c
index 1411f4cf1f21..3662434b64bb 100644
--- a/drivers/gpu/drm/i915/i915_ttm_buddy_manager.c
+++ b/drivers/gpu/drm/i915/i915_ttm_buddy_manager.c
@@ -53,6 +53,9 @@ static int i915_ttm_buddy_man_alloc(struct ttm_resource_manager *man,
INIT_LIST_HEAD(&bman_res->blocks);
bman_res->mm = mm;
+ if (place->flags & TTM_PL_FLAG_TOPDOWN)
+ bman_res->flags |= DRM_BUDDY_TOPDOWN_ALLOCATION;
+
if (place->fpfn || lpfn != man->size)
bman_res->flags |= DRM_BUDDY_RANGE_ALLOCATION;
diff --git a/include/drm/drm_buddy.h b/include/drm/drm_buddy.h
index 865664b90a8a..424fc443115e 100644
--- a/include/drm/drm_buddy.h
+++ b/include/drm/drm_buddy.h
@@ -28,6 +28,7 @@
})
#define DRM_BUDDY_RANGE_ALLOCATION (1 << 0)
+#define DRM_BUDDY_TOPDOWN_ALLOCATION (1 << 1)
struct drm_buddy_block {
#define DRM_BUDDY_HEADER_OFFSET GENMASK_ULL(63, 12)
