* [PATCH v9] drm/amdgpu: add drm buddy support to amdgpu
@ 2022-03-01 20:38 ` Arunpravin
0 siblings, 0 replies; 11+ messages in thread
From: Arunpravin @ 2022-03-01 20:38 UTC (permalink / raw)
To: amd-gfx, dri-devel, intel-gfx
Cc: Arunpravin, matthew.auld, tzimmermann, alexander.deucher,
christian.koenig
- Remove drm_mm references and replace with drm buddy functionalities
- Add res cursor support for drm buddy
v2(Matthew Auld):
- replace spinlock with mutex as we call kmem_cache_zalloc
(..., GFP_KERNEL) in drm_buddy_alloc() function
- lock drm_buddy_block_trim() function as it calls
mark_free/mark_split are all globally visible
v3(Matthew Auld):
- remove trim method error handling as we address the failure case
at drm_buddy_block_trim() function
v4:
- fix warnings reported by kernel test robot <lkp@intel.com>
v5:
- fix merge conflict issue
v6:
- fix warnings reported by kernel test robot <lkp@intel.com>
v7:
- remove DRM_BUDDY_RANGE_ALLOCATION flag usage
v8:
- keep DRM_BUDDY_RANGE_ALLOCATION flag usage
- resolve conflicts created by drm/amdgpu: remove VRAM accounting v2
v9(Christian):
- rename label name as fallback
- move struct amdgpu_vram_mgr to amdgpu_vram_mgr.h
- remove unnecessary flags from struct amdgpu_vram_reservation
- rewrite block NULL check condition
- change else style as per coding standard
- rewrite the node max size
- add a helper function to fetch the first entry from the list
Signed-off-by: Arunpravin <Arunpravin.PaneerSelvam@amd.com>
---
drivers/gpu/drm/Kconfig | 1 +
.../gpu/drm/amd/amdgpu/amdgpu_res_cursor.h | 97 +++++--
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h | 10 +-
drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c | 249 ++++++++++--------
drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h | 68 +++++
5 files changed, 289 insertions(+), 136 deletions(-)
create mode 100644 drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h
diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig
index f1422bee3dcc..5133c3f028ab 100644
--- a/drivers/gpu/drm/Kconfig
+++ b/drivers/gpu/drm/Kconfig
@@ -280,6 +280,7 @@ config DRM_AMDGPU
select HWMON
select BACKLIGHT_CLASS_DEVICE
select INTERVAL_TREE
+ select DRM_BUDDY
help
Choose this option if you have a recent AMD Radeon graphics card.
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h
index acfa207cf970..864c609ba00b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h
@@ -30,12 +30,15 @@
#include <drm/ttm/ttm_resource.h>
#include <drm/ttm/ttm_range_manager.h>
+#include "amdgpu_vram_mgr.h"
+
/* state back for walking over vram_mgr and gtt_mgr allocations */
struct amdgpu_res_cursor {
uint64_t start;
uint64_t size;
uint64_t remaining;
- struct drm_mm_node *node;
+ void *node;
+ uint32_t mem_type;
};
/**
@@ -52,27 +55,63 @@ static inline void amdgpu_res_first(struct ttm_resource *res,
uint64_t start, uint64_t size,
struct amdgpu_res_cursor *cur)
{
+ struct drm_buddy_block *block;
+ struct list_head *head, *next;
struct drm_mm_node *node;
- if (!res || res->mem_type == TTM_PL_SYSTEM) {
- cur->start = start;
- cur->size = size;
- cur->remaining = size;
- cur->node = NULL;
- WARN_ON(res && start + size > res->num_pages << PAGE_SHIFT);
- return;
- }
+ if (!res)
+ goto fallback;
BUG_ON(start + size > res->num_pages << PAGE_SHIFT);
- node = to_ttm_range_mgr_node(res)->mm_nodes;
- while (start >= node->size << PAGE_SHIFT)
- start -= node++->size << PAGE_SHIFT;
+ cur->mem_type = res->mem_type;
+
+ switch (cur->mem_type) {
+ case TTM_PL_VRAM:
+ head = &to_amdgpu_vram_mgr_node(res)->blocks;
+
+ block = list_first_entry_or_null(head,
+ struct drm_buddy_block,
+ link);
+ if (!block)
+ goto fallback;
+
+ while (start >= amdgpu_node_size(block)) {
+ start -= amdgpu_node_size(block);
+
+ next = block->link.next;
+ if (next != head)
+ block = list_entry(next, struct drm_buddy_block, link);
+ }
+
+ cur->start = amdgpu_node_start(block) + start;
+ cur->size = min(amdgpu_node_size(block) - start, size);
+ cur->remaining = size;
+ cur->node = block;
+ break;
+ case TTM_PL_TT:
+ node = to_ttm_range_mgr_node(res)->mm_nodes;
+ while (start >= node->size << PAGE_SHIFT)
+ start -= node++->size << PAGE_SHIFT;
+
+ cur->start = (node->start << PAGE_SHIFT) + start;
+ cur->size = min((node->size << PAGE_SHIFT) - start, size);
+ cur->remaining = size;
+ cur->node = node;
+ break;
+ default:
+ goto fallback;
+ }
- cur->start = (node->start << PAGE_SHIFT) + start;
- cur->size = min((node->size << PAGE_SHIFT) - start, size);
+ return;
+
+fallback:
+ cur->start = start;
+ cur->size = size;
cur->remaining = size;
- cur->node = node;
+ cur->node = NULL;
+ WARN_ON(res && start + size > res->num_pages << PAGE_SHIFT);
+ return;
}
/**
@@ -85,7 +124,9 @@ static inline void amdgpu_res_first(struct ttm_resource *res,
*/
static inline void amdgpu_res_next(struct amdgpu_res_cursor *cur, uint64_t size)
{
- struct drm_mm_node *node = cur->node;
+ struct drm_buddy_block *block;
+ struct drm_mm_node *node;
+ struct list_head *next;
BUG_ON(size > cur->remaining);
@@ -99,9 +140,27 @@ static inline void amdgpu_res_next(struct amdgpu_res_cursor *cur, uint64_t size)
return;
}
- cur->node = ++node;
- cur->start = node->start << PAGE_SHIFT;
- cur->size = min(node->size << PAGE_SHIFT, cur->remaining);
+ switch (cur->mem_type) {
+ case TTM_PL_VRAM:
+ block = cur->node;
+
+ next = block->link.next;
+ block = list_entry(next, struct drm_buddy_block, link);
+
+ cur->node = block;
+ cur->start = amdgpu_node_start(block);
+ cur->size = min(amdgpu_node_size(block), cur->remaining);
+ break;
+ case TTM_PL_TT:
+ node = cur->node;
+
+ cur->node = ++node;
+ cur->start = node->start << PAGE_SHIFT;
+ cur->size = min(node->size << PAGE_SHIFT, cur->remaining);
+ break;
+ default:
+ return;
+ }
}
#endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
index 9120ae80ef52..6a70818039dd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
@@ -26,6 +26,7 @@
#include <linux/dma-direction.h>
#include <drm/gpu_scheduler.h>
+#include "amdgpu_vram_mgr.h"
#include "amdgpu.h"
#define AMDGPU_PL_GDS (TTM_PL_PRIV + 0)
@@ -38,15 +39,6 @@
#define AMDGPU_POISON 0xd0bed0be
-struct amdgpu_vram_mgr {
- struct ttm_resource_manager manager;
- struct drm_mm mm;
- spinlock_t lock;
- struct list_head reservations_pending;
- struct list_head reserved_pages;
- atomic64_t vis_usage;
-};
-
struct amdgpu_gtt_mgr {
struct ttm_resource_manager manager;
struct drm_mm mm;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
index 0a7611648573..cdf66802e6bc 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
@@ -32,8 +32,10 @@
#include "atom.h"
struct amdgpu_vram_reservation {
+ u64 start;
+ u64 size;
+ struct list_head block;
struct list_head node;
- struct drm_mm_node mm_node;
};
static inline struct amdgpu_vram_mgr *
@@ -194,10 +196,10 @@ const struct attribute_group amdgpu_vram_mgr_attr_group = {
* Calculate how many bytes of the MM node are inside visible VRAM
*/
static u64 amdgpu_vram_mgr_vis_size(struct amdgpu_device *adev,
- struct drm_mm_node *node)
+ struct drm_buddy_block *block)
{
- uint64_t start = node->start << PAGE_SHIFT;
- uint64_t end = (node->size + node->start) << PAGE_SHIFT;
+ u64 start = amdgpu_node_start(block);
+ u64 end = start + amdgpu_node_size(block);
if (start >= adev->gmc.visible_vram_size)
return 0;
@@ -218,9 +220,9 @@ u64 amdgpu_vram_mgr_bo_visible_size(struct amdgpu_bo *bo)
{
struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
struct ttm_resource *res = bo->tbo.resource;
- unsigned pages = res->num_pages;
- struct drm_mm_node *mm;
- u64 usage;
+ struct amdgpu_vram_mgr_node *node = to_amdgpu_vram_mgr_node(res);
+ struct drm_buddy_block *block;
+ u64 usage = 0;
if (amdgpu_gmc_vram_full_visible(&adev->gmc))
return amdgpu_bo_size(bo);
@@ -228,9 +230,8 @@ u64 amdgpu_vram_mgr_bo_visible_size(struct amdgpu_bo *bo)
if (res->start >= adev->gmc.visible_vram_size >> PAGE_SHIFT)
return 0;
- mm = &container_of(res, struct ttm_range_mgr_node, base)->mm_nodes[0];
- for (usage = 0; pages; pages -= mm->size, mm++)
- usage += amdgpu_vram_mgr_vis_size(adev, mm);
+ list_for_each_entry(block, &node->blocks, link)
+ usage += amdgpu_vram_mgr_vis_size(adev, block);
return usage;
}
@@ -240,21 +241,28 @@ static void amdgpu_vram_mgr_do_reserve(struct ttm_resource_manager *man)
{
struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
struct amdgpu_device *adev = to_amdgpu_device(mgr);
- struct drm_mm *mm = &mgr->mm;
+ struct drm_buddy *mm = &mgr->mm;
struct amdgpu_vram_reservation *rsv, *temp;
+ struct drm_buddy_block *block;
uint64_t vis_usage;
list_for_each_entry_safe(rsv, temp, &mgr->reservations_pending, node) {
- if (drm_mm_reserve_node(mm, &rsv->mm_node))
+ if (drm_buddy_alloc_blocks(mm, rsv->start, rsv->start + rsv->size,
+ rsv->size, mm->chunk_size, &rsv->block,
+ DRM_BUDDY_RANGE_ALLOCATION))
+ continue;
+
+ block = amdgpu_get_node(&rsv->block);
+ if (!block)
continue;
dev_dbg(adev->dev, "Reservation 0x%llx - %lld, Succeeded\n",
- rsv->mm_node.start, rsv->mm_node.size);
+ rsv->start, rsv->size);
- vis_usage = amdgpu_vram_mgr_vis_size(adev, &rsv->mm_node);
+ vis_usage = amdgpu_vram_mgr_vis_size(adev, block);
atomic64_add(vis_usage, &mgr->vis_usage);
spin_lock(&man->bdev->lru_lock);
- man->usage += rsv->mm_node.size << PAGE_SHIFT;
+ man->usage += rsv->size;
spin_unlock(&man->bdev->lru_lock);
list_move(&rsv->node, &mgr->reserved_pages);
}
@@ -279,13 +287,15 @@ int amdgpu_vram_mgr_reserve_range(struct amdgpu_vram_mgr *mgr,
return -ENOMEM;
INIT_LIST_HEAD(&rsv->node);
- rsv->mm_node.start = start >> PAGE_SHIFT;
- rsv->mm_node.size = size >> PAGE_SHIFT;
+ INIT_LIST_HEAD(&rsv->block);
+
+ rsv->start = start;
+ rsv->size = size;
- spin_lock(&mgr->lock);
+ mutex_lock(&mgr->lock);
list_add_tail(&rsv->node, &mgr->reservations_pending);
amdgpu_vram_mgr_do_reserve(&mgr->manager);
- spin_unlock(&mgr->lock);
+ mutex_unlock(&mgr->lock);
return 0;
}
@@ -307,19 +317,19 @@ int amdgpu_vram_mgr_query_page_status(struct amdgpu_vram_mgr *mgr,
struct amdgpu_vram_reservation *rsv;
int ret;
- spin_lock(&mgr->lock);
+ mutex_lock(&mgr->lock);
list_for_each_entry(rsv, &mgr->reservations_pending, node) {
- if ((rsv->mm_node.start <= start) &&
- (start < (rsv->mm_node.start + rsv->mm_node.size))) {
+ if (rsv->start <= start &&
+ (start < (rsv->start + rsv->size))) {
ret = -EBUSY;
goto out;
}
}
list_for_each_entry(rsv, &mgr->reserved_pages, node) {
- if ((rsv->mm_node.start <= start) &&
- (start < (rsv->mm_node.start + rsv->mm_node.size))) {
+ if (rsv->start <= start &&
+ (start < (rsv->start + rsv->size))) {
ret = 0;
goto out;
}
@@ -327,32 +337,10 @@ int amdgpu_vram_mgr_query_page_status(struct amdgpu_vram_mgr *mgr,
ret = -ENOENT;
out:
- spin_unlock(&mgr->lock);
+ mutex_unlock(&mgr->lock);
return ret;
}
-/**
- * amdgpu_vram_mgr_virt_start - update virtual start address
- *
- * @mem: ttm_resource to update
- * @node: just allocated node
- *
- * Calculate a virtual BO start address to easily check if everything is CPU
- * accessible.
- */
-static void amdgpu_vram_mgr_virt_start(struct ttm_resource *mem,
- struct drm_mm_node *node)
-{
- unsigned long start;
-
- start = node->start + node->size;
- if (start > mem->num_pages)
- start -= mem->num_pages;
- else
- start = 0;
- mem->start = max(mem->start, start);
-}
-
/**
* amdgpu_vram_mgr_new - allocate new ranges
*
@@ -368,13 +356,13 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
const struct ttm_place *place,
struct ttm_resource **res)
{
- unsigned long lpfn, num_nodes, pages_per_node, pages_left, pages;
+ unsigned long lpfn, pages_per_node, pages_left, pages;
+ u64 vis_usage = 0, mem_bytes, max_bytes, min_page_size;
struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
struct amdgpu_device *adev = to_amdgpu_device(mgr);
- uint64_t vis_usage = 0, mem_bytes, max_bytes;
- struct ttm_range_mgr_node *node;
- struct drm_mm *mm = &mgr->mm;
- enum drm_mm_insert_mode mode;
+ struct amdgpu_vram_mgr_node *node;
+ struct drm_buddy *mm = &mgr->mm;
+ struct drm_buddy_block *block;
unsigned i;
int r;
@@ -389,7 +377,6 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
mem_bytes = tbo->base.size;
if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
pages_per_node = ~0ul;
- num_nodes = 1;
} else {
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
pages_per_node = HPAGE_PMD_NR;
@@ -399,11 +386,9 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
#endif
pages_per_node = max_t(uint32_t, pages_per_node,
tbo->page_alignment);
- num_nodes = DIV_ROUND_UP_ULL(PFN_UP(mem_bytes), pages_per_node);
}
- node = kvmalloc(struct_size(node, mm_nodes, num_nodes),
- GFP_KERNEL | __GFP_ZERO);
+ node = kzalloc(sizeof(*node), GFP_KERNEL);
if (!node)
return -ENOMEM;
@@ -415,48 +400,83 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
goto error_fini;
}
- mode = DRM_MM_INSERT_BEST;
+ INIT_LIST_HEAD(&node->blocks);
+
if (place->flags & TTM_PL_FLAG_TOPDOWN)
- mode = DRM_MM_INSERT_HIGH;
+ node->flags |= DRM_BUDDY_TOPDOWN_ALLOCATION;
- pages_left = node->base.num_pages;
+ if (place->fpfn || lpfn != man->size >> PAGE_SHIFT)
+ /* Allocate blocks in desired range */
+ node->flags |= DRM_BUDDY_RANGE_ALLOCATION;
- /* Limit maximum size to 2GB due to SG table limitations */
- pages = min(pages_left, 2UL << (30 - PAGE_SHIFT));
+ BUG_ON(!node->base.num_pages);
+ pages_left = node->base.num_pages;
i = 0;
- spin_lock(&mgr->lock);
while (pages_left) {
- uint32_t alignment = tbo->page_alignment;
+ if (tbo->page_alignment)
+ min_page_size = tbo->page_alignment << PAGE_SHIFT;
+ else
+ min_page_size = mgr->default_page_size;
+
+ /* Limit maximum size to 2GB due to SG table limitations */
+ pages = min(pages_left, 2UL << (30 - PAGE_SHIFT));
if (pages >= pages_per_node)
- alignment = pages_per_node;
-
- r = drm_mm_insert_node_in_range(mm, &node->mm_nodes[i], pages,
- alignment, 0, place->fpfn,
- lpfn, mode);
- if (unlikely(r)) {
- if (pages > pages_per_node) {
- if (is_power_of_2(pages))
- pages = pages / 2;
- else
- pages = rounddown_pow_of_two(pages);
- continue;
- }
- goto error_free;
+ min_page_size = pages_per_node << PAGE_SHIFT;
+
+ if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
+ pages = roundup_pow_of_two(pages);
+ min_page_size = pages << PAGE_SHIFT;
+
+ if (pages > lpfn)
+ lpfn = pages;
}
- vis_usage += amdgpu_vram_mgr_vis_size(adev, &node->mm_nodes[i]);
- amdgpu_vram_mgr_virt_start(&node->base, &node->mm_nodes[i]);
- pages_left -= pages;
+ BUG_ON(min_page_size < mm->chunk_size);
+
+ mutex_lock(&mgr->lock);
+ r = drm_buddy_alloc_blocks(mm, (u64)place->fpfn << PAGE_SHIFT,
+ (u64)lpfn << PAGE_SHIFT,
+ (u64)pages << PAGE_SHIFT,
+ min_page_size,
+ &node->blocks,
+ node->flags);
+ mutex_unlock(&mgr->lock);
+ if (unlikely(r))
+ goto error_free_blocks;
+
++i;
if (pages > pages_left)
- pages = pages_left;
+ pages_left = 0;
+ else
+ pages_left -= pages;
}
- spin_unlock(&mgr->lock);
- if (i == 1)
+ /* Free unused pages for contiguous allocation */
+ if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
+ u64 actual_size = (u64)node->base.num_pages << PAGE_SHIFT;
+
+ mutex_lock(&mgr->lock);
+ drm_buddy_block_trim(mm,
+ actual_size,
+ &node->blocks);
+ mutex_unlock(&mgr->lock);
+ }
+
+ list_for_each_entry(block, &node->blocks, link)
+ vis_usage += amdgpu_vram_mgr_vis_size(adev, block);
+
+ block = amdgpu_get_node(&node->blocks);
+ if (!block) {
+ r = -EINVAL;
+ goto error_fini;
+ }
+
+ node->base.start = amdgpu_node_start(block) >> PAGE_SHIFT;
+
+ if (i == 1 && place->flags & TTM_PL_FLAG_CONTIGUOUS)
node->base.placement |= TTM_PL_FLAG_CONTIGUOUS;
if (adev->gmc.xgmi.connected_to_cpu)
@@ -468,13 +488,13 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
*res = &node->base;
return 0;
-error_free:
- while (i--)
- drm_mm_remove_node(&node->mm_nodes[i]);
- spin_unlock(&mgr->lock);
+error_free_blocks:
+ mutex_lock(&mgr->lock);
+ drm_buddy_free_list(mm, &node->blocks);
+ mutex_unlock(&mgr->lock);
error_fini:
ttm_resource_fini(man, &node->base);
- kvfree(node);
+ kfree(node);
return r;
}
@@ -490,27 +510,26 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
static void amdgpu_vram_mgr_del(struct ttm_resource_manager *man,
struct ttm_resource *res)
{
- struct ttm_range_mgr_node *node = to_ttm_range_mgr_node(res);
+ struct amdgpu_vram_mgr_node *node = to_amdgpu_vram_mgr_node(res);
struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
struct amdgpu_device *adev = to_amdgpu_device(mgr);
+ struct drm_buddy *mm = &mgr->mm;
+ struct drm_buddy_block *block;
uint64_t vis_usage = 0;
- unsigned i, pages;
- spin_lock(&mgr->lock);
- for (i = 0, pages = res->num_pages; pages;
- pages -= node->mm_nodes[i].size, ++i) {
- struct drm_mm_node *mm = &node->mm_nodes[i];
+ mutex_lock(&mgr->lock);
+ list_for_each_entry(block, &node->blocks, link)
+ vis_usage += amdgpu_vram_mgr_vis_size(adev, block);
- drm_mm_remove_node(mm);
- vis_usage += amdgpu_vram_mgr_vis_size(adev, mm);
- }
amdgpu_vram_mgr_do_reserve(man);
- spin_unlock(&mgr->lock);
+
+ drm_buddy_free_list(mm, &node->blocks);
+ mutex_unlock(&mgr->lock);
atomic64_sub(vis_usage, &mgr->vis_usage);
ttm_resource_fini(man, res);
- kvfree(node);
+ kfree(node);
}
/**
@@ -648,13 +667,22 @@ static void amdgpu_vram_mgr_debug(struct ttm_resource_manager *man,
struct drm_printer *printer)
{
struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
+ struct drm_buddy *mm = &mgr->mm;
+ struct drm_buddy_block *block;
drm_printf(printer, " vis usage:%llu\n",
amdgpu_vram_mgr_vis_usage(mgr));
- spin_lock(&mgr->lock);
- drm_mm_print(&mgr->mm, printer);
- spin_unlock(&mgr->lock);
+ mutex_lock(&mgr->lock);
+ drm_printf(printer, "default_page_size: %lluKiB\n",
+ mgr->default_page_size >> 10);
+
+ drm_buddy_print(mm, printer);
+
+ drm_printf(printer, "reserved:\n");
+ list_for_each_entry(block, &mgr->reserved_pages, link)
+ drm_buddy_block_print(mm, block, printer);
+ mutex_unlock(&mgr->lock);
}
static const struct ttm_resource_manager_func amdgpu_vram_mgr_func = {
@@ -674,16 +702,21 @@ int amdgpu_vram_mgr_init(struct amdgpu_device *adev)
{
struct amdgpu_vram_mgr *mgr = &adev->mman.vram_mgr;
struct ttm_resource_manager *man = &mgr->manager;
+ int err;
ttm_resource_manager_init(man, &adev->mman.bdev,
adev->gmc.real_vram_size);
man->func = &amdgpu_vram_mgr_func;
- drm_mm_init(&mgr->mm, 0, man->size >> PAGE_SHIFT);
- spin_lock_init(&mgr->lock);
+ err = drm_buddy_init(&mgr->mm, man->size, PAGE_SIZE);
+ if (err)
+ return err;
+
+ mutex_init(&mgr->lock);
INIT_LIST_HEAD(&mgr->reservations_pending);
INIT_LIST_HEAD(&mgr->reserved_pages);
+ mgr->default_page_size = PAGE_SIZE;
ttm_set_driver_manager(&adev->mman.bdev, TTM_PL_VRAM, &mgr->manager);
ttm_resource_manager_set_used(man, true);
@@ -711,16 +744,16 @@ void amdgpu_vram_mgr_fini(struct amdgpu_device *adev)
if (ret)
return;
- spin_lock(&mgr->lock);
+ mutex_lock(&mgr->lock);
list_for_each_entry_safe(rsv, temp, &mgr->reservations_pending, node)
kfree(rsv);
list_for_each_entry_safe(rsv, temp, &mgr->reserved_pages, node) {
- drm_mm_remove_node(&rsv->mm_node);
+ drm_buddy_free_list(&mgr->mm, &rsv->block);
kfree(rsv);
}
- drm_mm_takedown(&mgr->mm);
- spin_unlock(&mgr->lock);
+ drm_buddy_fini(&mgr->mm);
+ mutex_unlock(&mgr->lock);
ttm_resource_manager_cleanup(man);
ttm_set_driver_manager(&adev->mman.bdev, TTM_PL_VRAM, NULL);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h
new file mode 100644
index 000000000000..e7d6bbf40c33
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h
@@ -0,0 +1,68 @@
+/* SPDX-License-Identifier: MIT
+ * Copyright 2021 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef __AMDGPU_VRAM_MGR_H__
+#define __AMDGPU_VRAM_MGR_H__
+
+#include <drm/drm_buddy.h>
+
+struct amdgpu_vram_mgr {
+ struct ttm_resource_manager manager;
+ struct drm_buddy mm;
+ /* protects access to buffer objects */
+ struct mutex lock;
+ struct list_head reservations_pending;
+ struct list_head reserved_pages;
+ atomic64_t vis_usage;
+ u64 default_page_size;
+};
+
+struct amdgpu_vram_mgr_node {
+ struct ttm_resource base;
+ struct list_head blocks;
+ unsigned long flags;
+};
+
+static inline u64 amdgpu_node_start(struct drm_buddy_block *block)
+{
+ return drm_buddy_block_offset(block);
+}
+
+static inline u64 amdgpu_node_size(struct drm_buddy_block *block)
+{
+ return PAGE_SIZE << drm_buddy_block_order(block);
+}
+
+static inline struct drm_buddy_block *
+amdgpu_get_node(struct list_head *list)
+{
+ return list_first_entry_or_null(list, struct drm_buddy_block, link);
+}
+
+static inline struct amdgpu_vram_mgr_node *
+to_amdgpu_vram_mgr_node(struct ttm_resource *res)
+{
+ return container_of(res, struct amdgpu_vram_mgr_node, base);
+}
+
+#endif
--
2.25.1
^ permalink raw reply related [flat|nested] 11+ messages in thread
* [Intel-gfx] [PATCH v9] drm/amdgpu: add drm buddy support to amdgpu
@ 2022-03-01 20:38 ` Arunpravin
0 siblings, 0 replies; 11+ messages in thread
From: Arunpravin @ 2022-03-01 20:38 UTC (permalink / raw)
To: amd-gfx, dri-devel, intel-gfx
Cc: Arunpravin, matthew.auld, tzimmermann, alexander.deucher,
christian.koenig
- Remove drm_mm references and replace with drm buddy functionalities
- Add res cursor support for drm buddy
v2(Matthew Auld):
- replace spinlock with mutex as we call kmem_cache_zalloc
(..., GFP_KERNEL) in drm_buddy_alloc() function
- lock drm_buddy_block_trim() function as it calls
mark_free/mark_split are all globally visible
v3(Matthew Auld):
- remove trim method error handling as we address the failure case
at drm_buddy_block_trim() function
v4:
- fix warnings reported by kernel test robot <lkp@intel.com>
v5:
- fix merge conflict issue
v6:
- fix warnings reported by kernel test robot <lkp@intel.com>
v7:
- remove DRM_BUDDY_RANGE_ALLOCATION flag usage
v8:
- keep DRM_BUDDY_RANGE_ALLOCATION flag usage
- resolve conflicts created by drm/amdgpu: remove VRAM accounting v2
v9(Christian):
- rename label name as fallback
- move struct amdgpu_vram_mgr to amdgpu_vram_mgr.h
- remove unnecessary flags from struct amdgpu_vram_reservation
- rewrite block NULL check condition
- change else style as per coding standard
- rewrite the node max size
- add a helper function to fetch the first entry from the list
Signed-off-by: Arunpravin <Arunpravin.PaneerSelvam@amd.com>
---
drivers/gpu/drm/Kconfig | 1 +
.../gpu/drm/amd/amdgpu/amdgpu_res_cursor.h | 97 +++++--
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h | 10 +-
drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c | 249 ++++++++++--------
drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h | 68 +++++
5 files changed, 289 insertions(+), 136 deletions(-)
create mode 100644 drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h
diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig
index f1422bee3dcc..5133c3f028ab 100644
--- a/drivers/gpu/drm/Kconfig
+++ b/drivers/gpu/drm/Kconfig
@@ -280,6 +280,7 @@ config DRM_AMDGPU
select HWMON
select BACKLIGHT_CLASS_DEVICE
select INTERVAL_TREE
+ select DRM_BUDDY
help
Choose this option if you have a recent AMD Radeon graphics card.
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h
index acfa207cf970..864c609ba00b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h
@@ -30,12 +30,15 @@
#include <drm/ttm/ttm_resource.h>
#include <drm/ttm/ttm_range_manager.h>
+#include "amdgpu_vram_mgr.h"
+
/* state back for walking over vram_mgr and gtt_mgr allocations */
struct amdgpu_res_cursor {
uint64_t start;
uint64_t size;
uint64_t remaining;
- struct drm_mm_node *node;
+ void *node;
+ uint32_t mem_type;
};
/**
@@ -52,27 +55,63 @@ static inline void amdgpu_res_first(struct ttm_resource *res,
uint64_t start, uint64_t size,
struct amdgpu_res_cursor *cur)
{
+ struct drm_buddy_block *block;
+ struct list_head *head, *next;
struct drm_mm_node *node;
- if (!res || res->mem_type == TTM_PL_SYSTEM) {
- cur->start = start;
- cur->size = size;
- cur->remaining = size;
- cur->node = NULL;
- WARN_ON(res && start + size > res->num_pages << PAGE_SHIFT);
- return;
- }
+ if (!res)
+ goto fallback;
BUG_ON(start + size > res->num_pages << PAGE_SHIFT);
- node = to_ttm_range_mgr_node(res)->mm_nodes;
- while (start >= node->size << PAGE_SHIFT)
- start -= node++->size << PAGE_SHIFT;
+ cur->mem_type = res->mem_type;
+
+ switch (cur->mem_type) {
+ case TTM_PL_VRAM:
+ head = &to_amdgpu_vram_mgr_node(res)->blocks;
+
+ block = list_first_entry_or_null(head,
+ struct drm_buddy_block,
+ link);
+ if (!block)
+ goto fallback;
+
+ while (start >= amdgpu_node_size(block)) {
+ start -= amdgpu_node_size(block);
+
+ next = block->link.next;
+ if (next != head)
+ block = list_entry(next, struct drm_buddy_block, link);
+ }
+
+ cur->start = amdgpu_node_start(block) + start;
+ cur->size = min(amdgpu_node_size(block) - start, size);
+ cur->remaining = size;
+ cur->node = block;
+ break;
+ case TTM_PL_TT:
+ node = to_ttm_range_mgr_node(res)->mm_nodes;
+ while (start >= node->size << PAGE_SHIFT)
+ start -= node++->size << PAGE_SHIFT;
+
+ cur->start = (node->start << PAGE_SHIFT) + start;
+ cur->size = min((node->size << PAGE_SHIFT) - start, size);
+ cur->remaining = size;
+ cur->node = node;
+ break;
+ default:
+ goto fallback;
+ }
- cur->start = (node->start << PAGE_SHIFT) + start;
- cur->size = min((node->size << PAGE_SHIFT) - start, size);
+ return;
+
+fallback:
+ cur->start = start;
+ cur->size = size;
cur->remaining = size;
- cur->node = node;
+ cur->node = NULL;
+ WARN_ON(res && start + size > res->num_pages << PAGE_SHIFT);
+ return;
}
/**
@@ -85,7 +124,9 @@ static inline void amdgpu_res_first(struct ttm_resource *res,
*/
static inline void amdgpu_res_next(struct amdgpu_res_cursor *cur, uint64_t size)
{
- struct drm_mm_node *node = cur->node;
+ struct drm_buddy_block *block;
+ struct drm_mm_node *node;
+ struct list_head *next;
BUG_ON(size > cur->remaining);
@@ -99,9 +140,27 @@ static inline void amdgpu_res_next(struct amdgpu_res_cursor *cur, uint64_t size)
return;
}
- cur->node = ++node;
- cur->start = node->start << PAGE_SHIFT;
- cur->size = min(node->size << PAGE_SHIFT, cur->remaining);
+ switch (cur->mem_type) {
+ case TTM_PL_VRAM:
+ block = cur->node;
+
+ next = block->link.next;
+ block = list_entry(next, struct drm_buddy_block, link);
+
+ cur->node = block;
+ cur->start = amdgpu_node_start(block);
+ cur->size = min(amdgpu_node_size(block), cur->remaining);
+ break;
+ case TTM_PL_TT:
+ node = cur->node;
+
+ cur->node = ++node;
+ cur->start = node->start << PAGE_SHIFT;
+ cur->size = min(node->size << PAGE_SHIFT, cur->remaining);
+ break;
+ default:
+ return;
+ }
}
#endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
index 9120ae80ef52..6a70818039dd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
@@ -26,6 +26,7 @@
#include <linux/dma-direction.h>
#include <drm/gpu_scheduler.h>
+#include "amdgpu_vram_mgr.h"
#include "amdgpu.h"
#define AMDGPU_PL_GDS (TTM_PL_PRIV + 0)
@@ -38,15 +39,6 @@
#define AMDGPU_POISON 0xd0bed0be
-struct amdgpu_vram_mgr {
- struct ttm_resource_manager manager;
- struct drm_mm mm;
- spinlock_t lock;
- struct list_head reservations_pending;
- struct list_head reserved_pages;
- atomic64_t vis_usage;
-};
-
struct amdgpu_gtt_mgr {
struct ttm_resource_manager manager;
struct drm_mm mm;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
index 0a7611648573..cdf66802e6bc 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
@@ -32,8 +32,10 @@
#include "atom.h"
struct amdgpu_vram_reservation {
+ u64 start;
+ u64 size;
+ struct list_head block;
struct list_head node;
- struct drm_mm_node mm_node;
};
static inline struct amdgpu_vram_mgr *
@@ -194,10 +196,10 @@ const struct attribute_group amdgpu_vram_mgr_attr_group = {
* Calculate how many bytes of the MM node are inside visible VRAM
*/
static u64 amdgpu_vram_mgr_vis_size(struct amdgpu_device *adev,
- struct drm_mm_node *node)
+ struct drm_buddy_block *block)
{
- uint64_t start = node->start << PAGE_SHIFT;
- uint64_t end = (node->size + node->start) << PAGE_SHIFT;
+ u64 start = amdgpu_node_start(block);
+ u64 end = start + amdgpu_node_size(block);
if (start >= adev->gmc.visible_vram_size)
return 0;
@@ -218,9 +220,9 @@ u64 amdgpu_vram_mgr_bo_visible_size(struct amdgpu_bo *bo)
{
struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
struct ttm_resource *res = bo->tbo.resource;
- unsigned pages = res->num_pages;
- struct drm_mm_node *mm;
- u64 usage;
+ struct amdgpu_vram_mgr_node *node = to_amdgpu_vram_mgr_node(res);
+ struct drm_buddy_block *block;
+ u64 usage = 0;
if (amdgpu_gmc_vram_full_visible(&adev->gmc))
return amdgpu_bo_size(bo);
@@ -228,9 +230,8 @@ u64 amdgpu_vram_mgr_bo_visible_size(struct amdgpu_bo *bo)
if (res->start >= adev->gmc.visible_vram_size >> PAGE_SHIFT)
return 0;
- mm = &container_of(res, struct ttm_range_mgr_node, base)->mm_nodes[0];
- for (usage = 0; pages; pages -= mm->size, mm++)
- usage += amdgpu_vram_mgr_vis_size(adev, mm);
+ list_for_each_entry(block, &node->blocks, link)
+ usage += amdgpu_vram_mgr_vis_size(adev, block);
return usage;
}
@@ -240,21 +241,28 @@ static void amdgpu_vram_mgr_do_reserve(struct ttm_resource_manager *man)
{
struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
struct amdgpu_device *adev = to_amdgpu_device(mgr);
- struct drm_mm *mm = &mgr->mm;
+ struct drm_buddy *mm = &mgr->mm;
struct amdgpu_vram_reservation *rsv, *temp;
+ struct drm_buddy_block *block;
uint64_t vis_usage;
list_for_each_entry_safe(rsv, temp, &mgr->reservations_pending, node) {
- if (drm_mm_reserve_node(mm, &rsv->mm_node))
+ if (drm_buddy_alloc_blocks(mm, rsv->start, rsv->start + rsv->size,
+ rsv->size, mm->chunk_size, &rsv->block,
+ DRM_BUDDY_RANGE_ALLOCATION))
+ continue;
+
+ block = amdgpu_get_node(&rsv->block);
+ if (!block)
continue;
dev_dbg(adev->dev, "Reservation 0x%llx - %lld, Succeeded\n",
- rsv->mm_node.start, rsv->mm_node.size);
+ rsv->start, rsv->size);
- vis_usage = amdgpu_vram_mgr_vis_size(adev, &rsv->mm_node);
+ vis_usage = amdgpu_vram_mgr_vis_size(adev, block);
atomic64_add(vis_usage, &mgr->vis_usage);
spin_lock(&man->bdev->lru_lock);
- man->usage += rsv->mm_node.size << PAGE_SHIFT;
+ man->usage += rsv->size;
spin_unlock(&man->bdev->lru_lock);
list_move(&rsv->node, &mgr->reserved_pages);
}
@@ -279,13 +287,15 @@ int amdgpu_vram_mgr_reserve_range(struct amdgpu_vram_mgr *mgr,
return -ENOMEM;
INIT_LIST_HEAD(&rsv->node);
- rsv->mm_node.start = start >> PAGE_SHIFT;
- rsv->mm_node.size = size >> PAGE_SHIFT;
+ INIT_LIST_HEAD(&rsv->block);
+
+ rsv->start = start;
+ rsv->size = size;
- spin_lock(&mgr->lock);
+ mutex_lock(&mgr->lock);
list_add_tail(&rsv->node, &mgr->reservations_pending);
amdgpu_vram_mgr_do_reserve(&mgr->manager);
- spin_unlock(&mgr->lock);
+ mutex_unlock(&mgr->lock);
return 0;
}
@@ -307,19 +317,19 @@ int amdgpu_vram_mgr_query_page_status(struct amdgpu_vram_mgr *mgr,
struct amdgpu_vram_reservation *rsv;
int ret;
- spin_lock(&mgr->lock);
+ mutex_lock(&mgr->lock);
list_for_each_entry(rsv, &mgr->reservations_pending, node) {
- if ((rsv->mm_node.start <= start) &&
- (start < (rsv->mm_node.start + rsv->mm_node.size))) {
+ if (rsv->start <= start &&
+ (start < (rsv->start + rsv->size))) {
ret = -EBUSY;
goto out;
}
}
list_for_each_entry(rsv, &mgr->reserved_pages, node) {
- if ((rsv->mm_node.start <= start) &&
- (start < (rsv->mm_node.start + rsv->mm_node.size))) {
+ if (rsv->start <= start &&
+ (start < (rsv->start + rsv->size))) {
ret = 0;
goto out;
}
@@ -327,32 +337,10 @@ int amdgpu_vram_mgr_query_page_status(struct amdgpu_vram_mgr *mgr,
ret = -ENOENT;
out:
- spin_unlock(&mgr->lock);
+ mutex_unlock(&mgr->lock);
return ret;
}
-/**
- * amdgpu_vram_mgr_virt_start - update virtual start address
- *
- * @mem: ttm_resource to update
- * @node: just allocated node
- *
- * Calculate a virtual BO start address to easily check if everything is CPU
- * accessible.
- */
-static void amdgpu_vram_mgr_virt_start(struct ttm_resource *mem,
- struct drm_mm_node *node)
-{
- unsigned long start;
-
- start = node->start + node->size;
- if (start > mem->num_pages)
- start -= mem->num_pages;
- else
- start = 0;
- mem->start = max(mem->start, start);
-}
-
/**
* amdgpu_vram_mgr_new - allocate new ranges
*
@@ -368,13 +356,13 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
const struct ttm_place *place,
struct ttm_resource **res)
{
- unsigned long lpfn, num_nodes, pages_per_node, pages_left, pages;
+ unsigned long lpfn, pages_per_node, pages_left, pages;
+ u64 vis_usage = 0, mem_bytes, max_bytes, min_page_size;
struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
struct amdgpu_device *adev = to_amdgpu_device(mgr);
- uint64_t vis_usage = 0, mem_bytes, max_bytes;
- struct ttm_range_mgr_node *node;
- struct drm_mm *mm = &mgr->mm;
- enum drm_mm_insert_mode mode;
+ struct amdgpu_vram_mgr_node *node;
+ struct drm_buddy *mm = &mgr->mm;
+ struct drm_buddy_block *block;
unsigned i;
int r;
@@ -389,7 +377,6 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
mem_bytes = tbo->base.size;
if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
pages_per_node = ~0ul;
- num_nodes = 1;
} else {
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
pages_per_node = HPAGE_PMD_NR;
@@ -399,11 +386,9 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
#endif
pages_per_node = max_t(uint32_t, pages_per_node,
tbo->page_alignment);
- num_nodes = DIV_ROUND_UP_ULL(PFN_UP(mem_bytes), pages_per_node);
}
- node = kvmalloc(struct_size(node, mm_nodes, num_nodes),
- GFP_KERNEL | __GFP_ZERO);
+ node = kzalloc(sizeof(*node), GFP_KERNEL);
if (!node)
return -ENOMEM;
@@ -415,48 +400,83 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
goto error_fini;
}
- mode = DRM_MM_INSERT_BEST;
+ INIT_LIST_HEAD(&node->blocks);
+
if (place->flags & TTM_PL_FLAG_TOPDOWN)
- mode = DRM_MM_INSERT_HIGH;
+ node->flags |= DRM_BUDDY_TOPDOWN_ALLOCATION;
- pages_left = node->base.num_pages;
+ if (place->fpfn || lpfn != man->size >> PAGE_SHIFT)
+ /* Allocate blocks in desired range */
+ node->flags |= DRM_BUDDY_RANGE_ALLOCATION;
- /* Limit maximum size to 2GB due to SG table limitations */
- pages = min(pages_left, 2UL << (30 - PAGE_SHIFT));
+ BUG_ON(!node->base.num_pages);
+ pages_left = node->base.num_pages;
i = 0;
- spin_lock(&mgr->lock);
while (pages_left) {
- uint32_t alignment = tbo->page_alignment;
+ if (tbo->page_alignment)
+ min_page_size = tbo->page_alignment << PAGE_SHIFT;
+ else
+ min_page_size = mgr->default_page_size;
+
+ /* Limit maximum size to 2GB due to SG table limitations */
+ pages = min(pages_left, 2UL << (30 - PAGE_SHIFT));
if (pages >= pages_per_node)
- alignment = pages_per_node;
-
- r = drm_mm_insert_node_in_range(mm, &node->mm_nodes[i], pages,
- alignment, 0, place->fpfn,
- lpfn, mode);
- if (unlikely(r)) {
- if (pages > pages_per_node) {
- if (is_power_of_2(pages))
- pages = pages / 2;
- else
- pages = rounddown_pow_of_two(pages);
- continue;
- }
- goto error_free;
+ min_page_size = pages_per_node << PAGE_SHIFT;
+
+ if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
+ pages = roundup_pow_of_two(pages);
+ min_page_size = pages << PAGE_SHIFT;
+
+ if (pages > lpfn)
+ lpfn = pages;
}
- vis_usage += amdgpu_vram_mgr_vis_size(adev, &node->mm_nodes[i]);
- amdgpu_vram_mgr_virt_start(&node->base, &node->mm_nodes[i]);
- pages_left -= pages;
+ BUG_ON(min_page_size < mm->chunk_size);
+
+ mutex_lock(&mgr->lock);
+ r = drm_buddy_alloc_blocks(mm, (u64)place->fpfn << PAGE_SHIFT,
+ (u64)lpfn << PAGE_SHIFT,
+ (u64)pages << PAGE_SHIFT,
+ min_page_size,
+ &node->blocks,
+ node->flags);
+ mutex_unlock(&mgr->lock);
+ if (unlikely(r))
+ goto error_free_blocks;
+
++i;
if (pages > pages_left)
- pages = pages_left;
+ pages_left = 0;
+ else
+ pages_left -= pages;
}
- spin_unlock(&mgr->lock);
- if (i == 1)
+ /* Free unused pages for contiguous allocation */
+ if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
+ u64 actual_size = (u64)node->base.num_pages << PAGE_SHIFT;
+
+ mutex_lock(&mgr->lock);
+ drm_buddy_block_trim(mm,
+ actual_size,
+ &node->blocks);
+ mutex_unlock(&mgr->lock);
+ }
+
+ list_for_each_entry(block, &node->blocks, link)
+ vis_usage += amdgpu_vram_mgr_vis_size(adev, block);
+
+ block = amdgpu_get_node(&node->blocks);
+ if (!block) {
+ r = -EINVAL;
+ goto error_fini;
+ }
+
+ node->base.start = amdgpu_node_start(block) >> PAGE_SHIFT;
+
+ if (i == 1 && place->flags & TTM_PL_FLAG_CONTIGUOUS)
node->base.placement |= TTM_PL_FLAG_CONTIGUOUS;
if (adev->gmc.xgmi.connected_to_cpu)
@@ -468,13 +488,13 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
*res = &node->base;
return 0;
-error_free:
- while (i--)
- drm_mm_remove_node(&node->mm_nodes[i]);
- spin_unlock(&mgr->lock);
+error_free_blocks:
+ mutex_lock(&mgr->lock);
+ drm_buddy_free_list(mm, &node->blocks);
+ mutex_unlock(&mgr->lock);
error_fini:
ttm_resource_fini(man, &node->base);
- kvfree(node);
+ kfree(node);
return r;
}
@@ -490,27 +510,26 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
static void amdgpu_vram_mgr_del(struct ttm_resource_manager *man,
struct ttm_resource *res)
{
- struct ttm_range_mgr_node *node = to_ttm_range_mgr_node(res);
+ struct amdgpu_vram_mgr_node *node = to_amdgpu_vram_mgr_node(res);
struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
struct amdgpu_device *adev = to_amdgpu_device(mgr);
+ struct drm_buddy *mm = &mgr->mm;
+ struct drm_buddy_block *block;
uint64_t vis_usage = 0;
- unsigned i, pages;
- spin_lock(&mgr->lock);
- for (i = 0, pages = res->num_pages; pages;
- pages -= node->mm_nodes[i].size, ++i) {
- struct drm_mm_node *mm = &node->mm_nodes[i];
+ mutex_lock(&mgr->lock);
+ list_for_each_entry(block, &node->blocks, link)
+ vis_usage += amdgpu_vram_mgr_vis_size(adev, block);
- drm_mm_remove_node(mm);
- vis_usage += amdgpu_vram_mgr_vis_size(adev, mm);
- }
amdgpu_vram_mgr_do_reserve(man);
- spin_unlock(&mgr->lock);
+
+ drm_buddy_free_list(mm, &node->blocks);
+ mutex_unlock(&mgr->lock);
atomic64_sub(vis_usage, &mgr->vis_usage);
ttm_resource_fini(man, res);
- kvfree(node);
+ kfree(node);
}
/**
@@ -648,13 +667,22 @@ static void amdgpu_vram_mgr_debug(struct ttm_resource_manager *man,
struct drm_printer *printer)
{
struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
+ struct drm_buddy *mm = &mgr->mm;
+ struct drm_buddy_block *block;
drm_printf(printer, " vis usage:%llu\n",
amdgpu_vram_mgr_vis_usage(mgr));
- spin_lock(&mgr->lock);
- drm_mm_print(&mgr->mm, printer);
- spin_unlock(&mgr->lock);
+ mutex_lock(&mgr->lock);
+ drm_printf(printer, "default_page_size: %lluKiB\n",
+ mgr->default_page_size >> 10);
+
+ drm_buddy_print(mm, printer);
+
+ drm_printf(printer, "reserved:\n");
+ list_for_each_entry(block, &mgr->reserved_pages, link)
+ drm_buddy_block_print(mm, block, printer);
+ mutex_unlock(&mgr->lock);
}
static const struct ttm_resource_manager_func amdgpu_vram_mgr_func = {
@@ -674,16 +702,21 @@ int amdgpu_vram_mgr_init(struct amdgpu_device *adev)
{
struct amdgpu_vram_mgr *mgr = &adev->mman.vram_mgr;
struct ttm_resource_manager *man = &mgr->manager;
+ int err;
ttm_resource_manager_init(man, &adev->mman.bdev,
adev->gmc.real_vram_size);
man->func = &amdgpu_vram_mgr_func;
- drm_mm_init(&mgr->mm, 0, man->size >> PAGE_SHIFT);
- spin_lock_init(&mgr->lock);
+ err = drm_buddy_init(&mgr->mm, man->size, PAGE_SIZE);
+ if (err)
+ return err;
+
+ mutex_init(&mgr->lock);
INIT_LIST_HEAD(&mgr->reservations_pending);
INIT_LIST_HEAD(&mgr->reserved_pages);
+ mgr->default_page_size = PAGE_SIZE;
ttm_set_driver_manager(&adev->mman.bdev, TTM_PL_VRAM, &mgr->manager);
ttm_resource_manager_set_used(man, true);
@@ -711,16 +744,16 @@ void amdgpu_vram_mgr_fini(struct amdgpu_device *adev)
if (ret)
return;
- spin_lock(&mgr->lock);
+ mutex_lock(&mgr->lock);
list_for_each_entry_safe(rsv, temp, &mgr->reservations_pending, node)
kfree(rsv);
list_for_each_entry_safe(rsv, temp, &mgr->reserved_pages, node) {
- drm_mm_remove_node(&rsv->mm_node);
+ drm_buddy_free_list(&mgr->mm, &rsv->block);
kfree(rsv);
}
- drm_mm_takedown(&mgr->mm);
- spin_unlock(&mgr->lock);
+ drm_buddy_fini(&mgr->mm);
+ mutex_unlock(&mgr->lock);
ttm_resource_manager_cleanup(man);
ttm_set_driver_manager(&adev->mman.bdev, TTM_PL_VRAM, NULL);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h
new file mode 100644
index 000000000000..e7d6bbf40c33
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h
@@ -0,0 +1,68 @@
+/* SPDX-License-Identifier: MIT
+ * Copyright 2021 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef __AMDGPU_VRAM_MGR_H__
+#define __AMDGPU_VRAM_MGR_H__
+
+#include <drm/drm_buddy.h>
+
+struct amdgpu_vram_mgr {
+ struct ttm_resource_manager manager;
+ struct drm_buddy mm;
+ /* protects access to buffer objects */
+ struct mutex lock;
+ struct list_head reservations_pending;
+ struct list_head reserved_pages;
+ atomic64_t vis_usage;
+ u64 default_page_size;
+};
+
+struct amdgpu_vram_mgr_node {
+ struct ttm_resource base;
+ struct list_head blocks;
+ unsigned long flags;
+};
+
+static inline u64 amdgpu_node_start(struct drm_buddy_block *block)
+{
+ return drm_buddy_block_offset(block);
+}
+
+static inline u64 amdgpu_node_size(struct drm_buddy_block *block)
+{
+ return PAGE_SIZE << drm_buddy_block_order(block);
+}
+
+static inline struct drm_buddy_block *
+amdgpu_get_node(struct list_head *list)
+{
+ return list_first_entry_or_null(list, struct drm_buddy_block, link);
+}
+
+static inline struct amdgpu_vram_mgr_node *
+to_amdgpu_vram_mgr_node(struct ttm_resource *res)
+{
+ return container_of(res, struct amdgpu_vram_mgr_node, base);
+}
+
+#endif
--
2.25.1
^ permalink raw reply related [flat|nested] 11+ messages in thread
* [PATCH v9] drm/amdgpu: add drm buddy support to amdgpu
@ 2022-03-01 20:38 ` Arunpravin
0 siblings, 0 replies; 11+ messages in thread
From: Arunpravin @ 2022-03-01 20:38 UTC (permalink / raw)
To: amd-gfx, dri-devel, intel-gfx
Cc: daniel, Arunpravin, matthew.auld, tzimmermann, alexander.deucher,
christian.koenig
- Remove drm_mm references and replace with drm buddy functionalities
- Add res cursor support for drm buddy
v2(Matthew Auld):
- replace spinlock with mutex as we call kmem_cache_zalloc
(..., GFP_KERNEL) in drm_buddy_alloc() function
- lock drm_buddy_block_trim() function as it calls
mark_free/mark_split are all globally visible
v3(Matthew Auld):
- remove trim method error handling as we address the failure case
at drm_buddy_block_trim() function
v4:
- fix warnings reported by kernel test robot <lkp@intel.com>
v5:
- fix merge conflict issue
v6:
- fix warnings reported by kernel test robot <lkp@intel.com>
v7:
- remove DRM_BUDDY_RANGE_ALLOCATION flag usage
v8:
- keep DRM_BUDDY_RANGE_ALLOCATION flag usage
- resolve conflicts created by drm/amdgpu: remove VRAM accounting v2
v9(Christian):
- rename label name as fallback
- move struct amdgpu_vram_mgr to amdgpu_vram_mgr.h
- remove unnecessary flags from struct amdgpu_vram_reservation
- rewrite block NULL check condition
- change else style as per coding standard
- rewrite the node max size
- add a helper function to fetch the first entry from the list
Signed-off-by: Arunpravin <Arunpravin.PaneerSelvam@amd.com>
---
drivers/gpu/drm/Kconfig | 1 +
.../gpu/drm/amd/amdgpu/amdgpu_res_cursor.h | 97 +++++--
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h | 10 +-
drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c | 249 ++++++++++--------
drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h | 68 +++++
5 files changed, 289 insertions(+), 136 deletions(-)
create mode 100644 drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h
diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig
index f1422bee3dcc..5133c3f028ab 100644
--- a/drivers/gpu/drm/Kconfig
+++ b/drivers/gpu/drm/Kconfig
@@ -280,6 +280,7 @@ config DRM_AMDGPU
select HWMON
select BACKLIGHT_CLASS_DEVICE
select INTERVAL_TREE
+ select DRM_BUDDY
help
Choose this option if you have a recent AMD Radeon graphics card.
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h
index acfa207cf970..864c609ba00b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h
@@ -30,12 +30,15 @@
#include <drm/ttm/ttm_resource.h>
#include <drm/ttm/ttm_range_manager.h>
+#include "amdgpu_vram_mgr.h"
+
/* state back for walking over vram_mgr and gtt_mgr allocations */
struct amdgpu_res_cursor {
uint64_t start;
uint64_t size;
uint64_t remaining;
- struct drm_mm_node *node;
+ void *node;
+ uint32_t mem_type;
};
/**
@@ -52,27 +55,63 @@ static inline void amdgpu_res_first(struct ttm_resource *res,
uint64_t start, uint64_t size,
struct amdgpu_res_cursor *cur)
{
+ struct drm_buddy_block *block;
+ struct list_head *head, *next;
struct drm_mm_node *node;
- if (!res || res->mem_type == TTM_PL_SYSTEM) {
- cur->start = start;
- cur->size = size;
- cur->remaining = size;
- cur->node = NULL;
- WARN_ON(res && start + size > res->num_pages << PAGE_SHIFT);
- return;
- }
+ if (!res)
+ goto fallback;
BUG_ON(start + size > res->num_pages << PAGE_SHIFT);
- node = to_ttm_range_mgr_node(res)->mm_nodes;
- while (start >= node->size << PAGE_SHIFT)
- start -= node++->size << PAGE_SHIFT;
+ cur->mem_type = res->mem_type;
+
+ switch (cur->mem_type) {
+ case TTM_PL_VRAM:
+ head = &to_amdgpu_vram_mgr_node(res)->blocks;
+
+ block = list_first_entry_or_null(head,
+ struct drm_buddy_block,
+ link);
+ if (!block)
+ goto fallback;
+
+ while (start >= amdgpu_node_size(block)) {
+ start -= amdgpu_node_size(block);
+
+ next = block->link.next;
+ if (next != head)
+ block = list_entry(next, struct drm_buddy_block, link);
+ }
+
+ cur->start = amdgpu_node_start(block) + start;
+ cur->size = min(amdgpu_node_size(block) - start, size);
+ cur->remaining = size;
+ cur->node = block;
+ break;
+ case TTM_PL_TT:
+ node = to_ttm_range_mgr_node(res)->mm_nodes;
+ while (start >= node->size << PAGE_SHIFT)
+ start -= node++->size << PAGE_SHIFT;
+
+ cur->start = (node->start << PAGE_SHIFT) + start;
+ cur->size = min((node->size << PAGE_SHIFT) - start, size);
+ cur->remaining = size;
+ cur->node = node;
+ break;
+ default:
+ goto fallback;
+ }
- cur->start = (node->start << PAGE_SHIFT) + start;
- cur->size = min((node->size << PAGE_SHIFT) - start, size);
+ return;
+
+fallback:
+ cur->start = start;
+ cur->size = size;
cur->remaining = size;
- cur->node = node;
+ cur->node = NULL;
+ WARN_ON(res && start + size > res->num_pages << PAGE_SHIFT);
+ return;
}
/**
@@ -85,7 +124,9 @@ static inline void amdgpu_res_first(struct ttm_resource *res,
*/
static inline void amdgpu_res_next(struct amdgpu_res_cursor *cur, uint64_t size)
{
- struct drm_mm_node *node = cur->node;
+ struct drm_buddy_block *block;
+ struct drm_mm_node *node;
+ struct list_head *next;
BUG_ON(size > cur->remaining);
@@ -99,9 +140,27 @@ static inline void amdgpu_res_next(struct amdgpu_res_cursor *cur, uint64_t size)
return;
}
- cur->node = ++node;
- cur->start = node->start << PAGE_SHIFT;
- cur->size = min(node->size << PAGE_SHIFT, cur->remaining);
+ switch (cur->mem_type) {
+ case TTM_PL_VRAM:
+ block = cur->node;
+
+ next = block->link.next;
+ block = list_entry(next, struct drm_buddy_block, link);
+
+ cur->node = block;
+ cur->start = amdgpu_node_start(block);
+ cur->size = min(amdgpu_node_size(block), cur->remaining);
+ break;
+ case TTM_PL_TT:
+ node = cur->node;
+
+ cur->node = ++node;
+ cur->start = node->start << PAGE_SHIFT;
+ cur->size = min(node->size << PAGE_SHIFT, cur->remaining);
+ break;
+ default:
+ return;
+ }
}
#endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
index 9120ae80ef52..6a70818039dd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
@@ -26,6 +26,7 @@
#include <linux/dma-direction.h>
#include <drm/gpu_scheduler.h>
+#include "amdgpu_vram_mgr.h"
#include "amdgpu.h"
#define AMDGPU_PL_GDS (TTM_PL_PRIV + 0)
@@ -38,15 +39,6 @@
#define AMDGPU_POISON 0xd0bed0be
-struct amdgpu_vram_mgr {
- struct ttm_resource_manager manager;
- struct drm_mm mm;
- spinlock_t lock;
- struct list_head reservations_pending;
- struct list_head reserved_pages;
- atomic64_t vis_usage;
-};
-
struct amdgpu_gtt_mgr {
struct ttm_resource_manager manager;
struct drm_mm mm;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
index 0a7611648573..cdf66802e6bc 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
@@ -32,8 +32,10 @@
#include "atom.h"
struct amdgpu_vram_reservation {
+ u64 start;
+ u64 size;
+ struct list_head block;
struct list_head node;
- struct drm_mm_node mm_node;
};
static inline struct amdgpu_vram_mgr *
@@ -194,10 +196,10 @@ const struct attribute_group amdgpu_vram_mgr_attr_group = {
* Calculate how many bytes of the MM node are inside visible VRAM
*/
static u64 amdgpu_vram_mgr_vis_size(struct amdgpu_device *adev,
- struct drm_mm_node *node)
+ struct drm_buddy_block *block)
{
- uint64_t start = node->start << PAGE_SHIFT;
- uint64_t end = (node->size + node->start) << PAGE_SHIFT;
+ u64 start = amdgpu_node_start(block);
+ u64 end = start + amdgpu_node_size(block);
if (start >= adev->gmc.visible_vram_size)
return 0;
@@ -218,9 +220,9 @@ u64 amdgpu_vram_mgr_bo_visible_size(struct amdgpu_bo *bo)
{
struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
struct ttm_resource *res = bo->tbo.resource;
- unsigned pages = res->num_pages;
- struct drm_mm_node *mm;
- u64 usage;
+ struct amdgpu_vram_mgr_node *node = to_amdgpu_vram_mgr_node(res);
+ struct drm_buddy_block *block;
+ u64 usage = 0;
if (amdgpu_gmc_vram_full_visible(&adev->gmc))
return amdgpu_bo_size(bo);
@@ -228,9 +230,8 @@ u64 amdgpu_vram_mgr_bo_visible_size(struct amdgpu_bo *bo)
if (res->start >= adev->gmc.visible_vram_size >> PAGE_SHIFT)
return 0;
- mm = &container_of(res, struct ttm_range_mgr_node, base)->mm_nodes[0];
- for (usage = 0; pages; pages -= mm->size, mm++)
- usage += amdgpu_vram_mgr_vis_size(adev, mm);
+ list_for_each_entry(block, &node->blocks, link)
+ usage += amdgpu_vram_mgr_vis_size(adev, block);
return usage;
}
@@ -240,21 +241,28 @@ static void amdgpu_vram_mgr_do_reserve(struct ttm_resource_manager *man)
{
struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
struct amdgpu_device *adev = to_amdgpu_device(mgr);
- struct drm_mm *mm = &mgr->mm;
+ struct drm_buddy *mm = &mgr->mm;
struct amdgpu_vram_reservation *rsv, *temp;
+ struct drm_buddy_block *block;
uint64_t vis_usage;
list_for_each_entry_safe(rsv, temp, &mgr->reservations_pending, node) {
- if (drm_mm_reserve_node(mm, &rsv->mm_node))
+ if (drm_buddy_alloc_blocks(mm, rsv->start, rsv->start + rsv->size,
+ rsv->size, mm->chunk_size, &rsv->block,
+ DRM_BUDDY_RANGE_ALLOCATION))
+ continue;
+
+ block = amdgpu_get_node(&rsv->block);
+ if (!block)
continue;
dev_dbg(adev->dev, "Reservation 0x%llx - %lld, Succeeded\n",
- rsv->mm_node.start, rsv->mm_node.size);
+ rsv->start, rsv->size);
- vis_usage = amdgpu_vram_mgr_vis_size(adev, &rsv->mm_node);
+ vis_usage = amdgpu_vram_mgr_vis_size(adev, block);
atomic64_add(vis_usage, &mgr->vis_usage);
spin_lock(&man->bdev->lru_lock);
- man->usage += rsv->mm_node.size << PAGE_SHIFT;
+ man->usage += rsv->size;
spin_unlock(&man->bdev->lru_lock);
list_move(&rsv->node, &mgr->reserved_pages);
}
@@ -279,13 +287,15 @@ int amdgpu_vram_mgr_reserve_range(struct amdgpu_vram_mgr *mgr,
return -ENOMEM;
INIT_LIST_HEAD(&rsv->node);
- rsv->mm_node.start = start >> PAGE_SHIFT;
- rsv->mm_node.size = size >> PAGE_SHIFT;
+ INIT_LIST_HEAD(&rsv->block);
+
+ rsv->start = start;
+ rsv->size = size;
- spin_lock(&mgr->lock);
+ mutex_lock(&mgr->lock);
list_add_tail(&rsv->node, &mgr->reservations_pending);
amdgpu_vram_mgr_do_reserve(&mgr->manager);
- spin_unlock(&mgr->lock);
+ mutex_unlock(&mgr->lock);
return 0;
}
@@ -307,19 +317,19 @@ int amdgpu_vram_mgr_query_page_status(struct amdgpu_vram_mgr *mgr,
struct amdgpu_vram_reservation *rsv;
int ret;
- spin_lock(&mgr->lock);
+ mutex_lock(&mgr->lock);
list_for_each_entry(rsv, &mgr->reservations_pending, node) {
- if ((rsv->mm_node.start <= start) &&
- (start < (rsv->mm_node.start + rsv->mm_node.size))) {
+ if (rsv->start <= start &&
+ (start < (rsv->start + rsv->size))) {
ret = -EBUSY;
goto out;
}
}
list_for_each_entry(rsv, &mgr->reserved_pages, node) {
- if ((rsv->mm_node.start <= start) &&
- (start < (rsv->mm_node.start + rsv->mm_node.size))) {
+ if (rsv->start <= start &&
+ (start < (rsv->start + rsv->size))) {
ret = 0;
goto out;
}
@@ -327,32 +337,10 @@ int amdgpu_vram_mgr_query_page_status(struct amdgpu_vram_mgr *mgr,
ret = -ENOENT;
out:
- spin_unlock(&mgr->lock);
+ mutex_unlock(&mgr->lock);
return ret;
}
-/**
- * amdgpu_vram_mgr_virt_start - update virtual start address
- *
- * @mem: ttm_resource to update
- * @node: just allocated node
- *
- * Calculate a virtual BO start address to easily check if everything is CPU
- * accessible.
- */
-static void amdgpu_vram_mgr_virt_start(struct ttm_resource *mem,
- struct drm_mm_node *node)
-{
- unsigned long start;
-
- start = node->start + node->size;
- if (start > mem->num_pages)
- start -= mem->num_pages;
- else
- start = 0;
- mem->start = max(mem->start, start);
-}
-
/**
* amdgpu_vram_mgr_new - allocate new ranges
*
@@ -368,13 +356,13 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
const struct ttm_place *place,
struct ttm_resource **res)
{
- unsigned long lpfn, num_nodes, pages_per_node, pages_left, pages;
+ unsigned long lpfn, pages_per_node, pages_left, pages;
+ u64 vis_usage = 0, mem_bytes, max_bytes, min_page_size;
struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
struct amdgpu_device *adev = to_amdgpu_device(mgr);
- uint64_t vis_usage = 0, mem_bytes, max_bytes;
- struct ttm_range_mgr_node *node;
- struct drm_mm *mm = &mgr->mm;
- enum drm_mm_insert_mode mode;
+ struct amdgpu_vram_mgr_node *node;
+ struct drm_buddy *mm = &mgr->mm;
+ struct drm_buddy_block *block;
unsigned i;
int r;
@@ -389,7 +377,6 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
mem_bytes = tbo->base.size;
if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
pages_per_node = ~0ul;
- num_nodes = 1;
} else {
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
pages_per_node = HPAGE_PMD_NR;
@@ -399,11 +386,9 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
#endif
pages_per_node = max_t(uint32_t, pages_per_node,
tbo->page_alignment);
- num_nodes = DIV_ROUND_UP_ULL(PFN_UP(mem_bytes), pages_per_node);
}
- node = kvmalloc(struct_size(node, mm_nodes, num_nodes),
- GFP_KERNEL | __GFP_ZERO);
+ node = kzalloc(sizeof(*node), GFP_KERNEL);
if (!node)
return -ENOMEM;
@@ -415,48 +400,83 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
goto error_fini;
}
- mode = DRM_MM_INSERT_BEST;
+ INIT_LIST_HEAD(&node->blocks);
+
if (place->flags & TTM_PL_FLAG_TOPDOWN)
- mode = DRM_MM_INSERT_HIGH;
+ node->flags |= DRM_BUDDY_TOPDOWN_ALLOCATION;
- pages_left = node->base.num_pages;
+ if (place->fpfn || lpfn != man->size >> PAGE_SHIFT)
+ /* Allocate blocks in desired range */
+ node->flags |= DRM_BUDDY_RANGE_ALLOCATION;
- /* Limit maximum size to 2GB due to SG table limitations */
- pages = min(pages_left, 2UL << (30 - PAGE_SHIFT));
+ BUG_ON(!node->base.num_pages);
+ pages_left = node->base.num_pages;
i = 0;
- spin_lock(&mgr->lock);
while (pages_left) {
- uint32_t alignment = tbo->page_alignment;
+ if (tbo->page_alignment)
+ min_page_size = tbo->page_alignment << PAGE_SHIFT;
+ else
+ min_page_size = mgr->default_page_size;
+
+ /* Limit maximum size to 2GB due to SG table limitations */
+ pages = min(pages_left, 2UL << (30 - PAGE_SHIFT));
if (pages >= pages_per_node)
- alignment = pages_per_node;
-
- r = drm_mm_insert_node_in_range(mm, &node->mm_nodes[i], pages,
- alignment, 0, place->fpfn,
- lpfn, mode);
- if (unlikely(r)) {
- if (pages > pages_per_node) {
- if (is_power_of_2(pages))
- pages = pages / 2;
- else
- pages = rounddown_pow_of_two(pages);
- continue;
- }
- goto error_free;
+ min_page_size = pages_per_node << PAGE_SHIFT;
+
+ if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
+ pages = roundup_pow_of_two(pages);
+ min_page_size = pages << PAGE_SHIFT;
+
+ if (pages > lpfn)
+ lpfn = pages;
}
- vis_usage += amdgpu_vram_mgr_vis_size(adev, &node->mm_nodes[i]);
- amdgpu_vram_mgr_virt_start(&node->base, &node->mm_nodes[i]);
- pages_left -= pages;
+ BUG_ON(min_page_size < mm->chunk_size);
+
+ mutex_lock(&mgr->lock);
+ r = drm_buddy_alloc_blocks(mm, (u64)place->fpfn << PAGE_SHIFT,
+ (u64)lpfn << PAGE_SHIFT,
+ (u64)pages << PAGE_SHIFT,
+ min_page_size,
+ &node->blocks,
+ node->flags);
+ mutex_unlock(&mgr->lock);
+ if (unlikely(r))
+ goto error_free_blocks;
+
++i;
if (pages > pages_left)
- pages = pages_left;
+ pages_left = 0;
+ else
+ pages_left -= pages;
}
- spin_unlock(&mgr->lock);
- if (i == 1)
+ /* Free unused pages for contiguous allocation */
+ if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
+ u64 actual_size = (u64)node->base.num_pages << PAGE_SHIFT;
+
+ mutex_lock(&mgr->lock);
+ drm_buddy_block_trim(mm,
+ actual_size,
+ &node->blocks);
+ mutex_unlock(&mgr->lock);
+ }
+
+ list_for_each_entry(block, &node->blocks, link)
+ vis_usage += amdgpu_vram_mgr_vis_size(adev, block);
+
+ block = amdgpu_get_node(&node->blocks);
+ if (!block) {
+ r = -EINVAL;
+ goto error_fini;
+ }
+
+ node->base.start = amdgpu_node_start(block) >> PAGE_SHIFT;
+
+ if (i == 1 && place->flags & TTM_PL_FLAG_CONTIGUOUS)
node->base.placement |= TTM_PL_FLAG_CONTIGUOUS;
if (adev->gmc.xgmi.connected_to_cpu)
@@ -468,13 +488,13 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
*res = &node->base;
return 0;
-error_free:
- while (i--)
- drm_mm_remove_node(&node->mm_nodes[i]);
- spin_unlock(&mgr->lock);
+error_free_blocks:
+ mutex_lock(&mgr->lock);
+ drm_buddy_free_list(mm, &node->blocks);
+ mutex_unlock(&mgr->lock);
error_fini:
ttm_resource_fini(man, &node->base);
- kvfree(node);
+ kfree(node);
return r;
}
@@ -490,27 +510,26 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
static void amdgpu_vram_mgr_del(struct ttm_resource_manager *man,
struct ttm_resource *res)
{
- struct ttm_range_mgr_node *node = to_ttm_range_mgr_node(res);
+ struct amdgpu_vram_mgr_node *node = to_amdgpu_vram_mgr_node(res);
struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
struct amdgpu_device *adev = to_amdgpu_device(mgr);
+ struct drm_buddy *mm = &mgr->mm;
+ struct drm_buddy_block *block;
uint64_t vis_usage = 0;
- unsigned i, pages;
- spin_lock(&mgr->lock);
- for (i = 0, pages = res->num_pages; pages;
- pages -= node->mm_nodes[i].size, ++i) {
- struct drm_mm_node *mm = &node->mm_nodes[i];
+ mutex_lock(&mgr->lock);
+ list_for_each_entry(block, &node->blocks, link)
+ vis_usage += amdgpu_vram_mgr_vis_size(adev, block);
- drm_mm_remove_node(mm);
- vis_usage += amdgpu_vram_mgr_vis_size(adev, mm);
- }
amdgpu_vram_mgr_do_reserve(man);
- spin_unlock(&mgr->lock);
+
+ drm_buddy_free_list(mm, &node->blocks);
+ mutex_unlock(&mgr->lock);
atomic64_sub(vis_usage, &mgr->vis_usage);
ttm_resource_fini(man, res);
- kvfree(node);
+ kfree(node);
}
/**
@@ -648,13 +667,22 @@ static void amdgpu_vram_mgr_debug(struct ttm_resource_manager *man,
struct drm_printer *printer)
{
struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
+ struct drm_buddy *mm = &mgr->mm;
+ struct drm_buddy_block *block;
drm_printf(printer, " vis usage:%llu\n",
amdgpu_vram_mgr_vis_usage(mgr));
- spin_lock(&mgr->lock);
- drm_mm_print(&mgr->mm, printer);
- spin_unlock(&mgr->lock);
+ mutex_lock(&mgr->lock);
+ drm_printf(printer, "default_page_size: %lluKiB\n",
+ mgr->default_page_size >> 10);
+
+ drm_buddy_print(mm, printer);
+
+ drm_printf(printer, "reserved:\n");
+ list_for_each_entry(block, &mgr->reserved_pages, link)
+ drm_buddy_block_print(mm, block, printer);
+ mutex_unlock(&mgr->lock);
}
static const struct ttm_resource_manager_func amdgpu_vram_mgr_func = {
@@ -674,16 +702,21 @@ int amdgpu_vram_mgr_init(struct amdgpu_device *adev)
{
struct amdgpu_vram_mgr *mgr = &adev->mman.vram_mgr;
struct ttm_resource_manager *man = &mgr->manager;
+ int err;
ttm_resource_manager_init(man, &adev->mman.bdev,
adev->gmc.real_vram_size);
man->func = &amdgpu_vram_mgr_func;
- drm_mm_init(&mgr->mm, 0, man->size >> PAGE_SHIFT);
- spin_lock_init(&mgr->lock);
+ err = drm_buddy_init(&mgr->mm, man->size, PAGE_SIZE);
+ if (err)
+ return err;
+
+ mutex_init(&mgr->lock);
INIT_LIST_HEAD(&mgr->reservations_pending);
INIT_LIST_HEAD(&mgr->reserved_pages);
+ mgr->default_page_size = PAGE_SIZE;
ttm_set_driver_manager(&adev->mman.bdev, TTM_PL_VRAM, &mgr->manager);
ttm_resource_manager_set_used(man, true);
@@ -711,16 +744,16 @@ void amdgpu_vram_mgr_fini(struct amdgpu_device *adev)
if (ret)
return;
- spin_lock(&mgr->lock);
+ mutex_lock(&mgr->lock);
list_for_each_entry_safe(rsv, temp, &mgr->reservations_pending, node)
kfree(rsv);
list_for_each_entry_safe(rsv, temp, &mgr->reserved_pages, node) {
- drm_mm_remove_node(&rsv->mm_node);
+ drm_buddy_free_list(&mgr->mm, &rsv->block);
kfree(rsv);
}
- drm_mm_takedown(&mgr->mm);
- spin_unlock(&mgr->lock);
+ drm_buddy_fini(&mgr->mm);
+ mutex_unlock(&mgr->lock);
ttm_resource_manager_cleanup(man);
ttm_set_driver_manager(&adev->mman.bdev, TTM_PL_VRAM, NULL);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h
new file mode 100644
index 000000000000..e7d6bbf40c33
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h
@@ -0,0 +1,68 @@
+/* SPDX-License-Identifier: MIT
+ * Copyright 2021 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef __AMDGPU_VRAM_MGR_H__
+#define __AMDGPU_VRAM_MGR_H__
+
+#include <drm/drm_buddy.h>
+
+struct amdgpu_vram_mgr {
+ struct ttm_resource_manager manager;
+ struct drm_buddy mm;
+ /* protects access to buffer objects */
+ struct mutex lock;
+ struct list_head reservations_pending;
+ struct list_head reserved_pages;
+ atomic64_t vis_usage;
+ u64 default_page_size;
+};
+
+struct amdgpu_vram_mgr_node {
+ struct ttm_resource base;
+ struct list_head blocks;
+ unsigned long flags;
+};
+
+static inline u64 amdgpu_node_start(struct drm_buddy_block *block)
+{
+ return drm_buddy_block_offset(block);
+}
+
+static inline u64 amdgpu_node_size(struct drm_buddy_block *block)
+{
+ return PAGE_SIZE << drm_buddy_block_order(block);
+}
+
+static inline struct drm_buddy_block *
+amdgpu_get_node(struct list_head *list)
+{
+ return list_first_entry_or_null(list, struct drm_buddy_block, link);
+}
+
+static inline struct amdgpu_vram_mgr_node *
+to_amdgpu_vram_mgr_node(struct ttm_resource *res)
+{
+ return container_of(res, struct amdgpu_vram_mgr_node, base);
+}
+
+#endif
--
2.25.1
^ permalink raw reply related [flat|nested] 11+ messages in thread
* [Intel-gfx] ✗ Fi.CI.CHECKPATCH: warning for drm/amdgpu: add drm buddy support to amdgpu
2022-03-01 20:38 ` [Intel-gfx] " Arunpravin
(?)
(?)
@ 2022-03-01 23:05 ` Patchwork
-1 siblings, 0 replies; 11+ messages in thread
From: Patchwork @ 2022-03-01 23:05 UTC (permalink / raw)
To: Arunpravin; +Cc: intel-gfx
== Series Details ==
Series: drm/amdgpu: add drm buddy support to amdgpu
URL : https://patchwork.freedesktop.org/series/100908/
State : warning
== Summary ==
$ dim checkpatch origin/drm-tip
8e830a55eacb drm/amdgpu: add drm buddy support to amdgpu
-:76: CHECK:PREFER_KERNEL_TYPES: Prefer kernel type 'u32' over 'uint32_t'
#76: FILE: drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h:41:
+ uint32_t mem_type;
-:454: WARNING:AVOID_BUG: Avoid crashing the kernel - try using WARN_ON & recovery code rather than BUG() or BUG_ON()
#454: FILE: drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c:412:
+ BUG_ON(!node->base.num_pages);
-:497: WARNING:AVOID_BUG: Avoid crashing the kernel - try using WARN_ON & recovery code rather than BUG() or BUG_ON()
#497: FILE: drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c:436:
+ BUG_ON(min_page_size < mm->chunk_size);
-:676: WARNING:FILE_PATH_CHANGES: added, moved or deleted file(s), does MAINTAINERS need updating?
#676:
new file mode 100644
total: 0 errors, 3 warnings, 1 checks, 656 lines checked
^ permalink raw reply [flat|nested] 11+ messages in thread
* [Intel-gfx] ✓ Fi.CI.BAT: success for drm/amdgpu: add drm buddy support to amdgpu
2022-03-01 20:38 ` [Intel-gfx] " Arunpravin
` (2 preceding siblings ...)
(?)
@ 2022-03-01 23:38 ` Patchwork
-1 siblings, 0 replies; 11+ messages in thread
From: Patchwork @ 2022-03-01 23:38 UTC (permalink / raw)
To: Arunpravin; +Cc: intel-gfx
[-- Attachment #1: Type: text/plain, Size: 7284 bytes --]
== Series Details ==
Series: drm/amdgpu: add drm buddy support to amdgpu
URL : https://patchwork.freedesktop.org/series/100908/
State : success
== Summary ==
CI Bug Log - changes from CI_DRM_11306 -> Patchwork_22453
====================================================
Summary
-------
**SUCCESS**
No regressions found.
External URL: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/index.html
Participating hosts (52 -> 43)
------------------------------
Additional (1): fi-icl-u2
Missing (10): fi-kbl-soraka fi-cml-u2 shard-tglu fi-hsw-4200u fi-bsw-cyan bat-adlp-4 fi-ctg-p8600 shard-rkl shard-dg1 fi-bdw-samus
Known issues
------------
Here are the changes found in Patchwork_22453 that come from known issues:
### IGT changes ###
#### Issues hit ####
* igt@amdgpu/amd_cs_nop@fork-gfx0:
- fi-icl-u2: NOTRUN -> [SKIP][1] ([fdo#109315]) +17 similar issues
[1]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/fi-icl-u2/igt@amdgpu/amd_cs_nop@fork-gfx0.html
* igt@amdgpu/amd_cs_nop@nop-compute0:
- fi-pnv-d510: NOTRUN -> [SKIP][2] ([fdo#109271]) +17 similar issues
[2]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/fi-pnv-d510/igt@amdgpu/amd_cs_nop@nop-compute0.html
* igt@amdgpu/amd_cs_nop@sync-fork-compute0:
- fi-skl-6600u: NOTRUN -> [SKIP][3] ([fdo#109271]) +17 similar issues
[3]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/fi-skl-6600u/igt@amdgpu/amd_cs_nop@sync-fork-compute0.html
* igt@gem_huc_copy@huc-copy:
- fi-icl-u2: NOTRUN -> [SKIP][4] ([i915#2190])
[4]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/fi-icl-u2/igt@gem_huc_copy@huc-copy.html
* igt@gem_lmem_swapping@parallel-random-engines:
- fi-icl-u2: NOTRUN -> [SKIP][5] ([i915#4613]) +3 similar issues
[5]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/fi-icl-u2/igt@gem_lmem_swapping@parallel-random-engines.html
* igt@gem_lmem_swapping@random-engines:
- fi-ivb-3770: NOTRUN -> [SKIP][6] ([fdo#109271]) +35 similar issues
[6]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/fi-ivb-3770/igt@gem_lmem_swapping@random-engines.html
* igt@i915_selftest@live@hangcheck:
- bat-dg1-5: [PASS][7] -> [DMESG-FAIL][8] ([i915#4494] / [i915#4957])
[7]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/bat-dg1-5/igt@i915_selftest@live@hangcheck.html
[8]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/bat-dg1-5/igt@i915_selftest@live@hangcheck.html
* igt@kms_chamelium@dp-hpd-fast:
- fi-ivb-3770: NOTRUN -> [SKIP][9] ([fdo#109271] / [fdo#111827]) +8 similar issues
[9]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/fi-ivb-3770/igt@kms_chamelium@dp-hpd-fast.html
* igt@kms_chamelium@hdmi-hpd-fast:
- fi-icl-u2: NOTRUN -> [SKIP][10] ([fdo#111827]) +8 similar issues
[10]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/fi-icl-u2/igt@kms_chamelium@hdmi-hpd-fast.html
* igt@kms_cursor_legacy@basic-busy-flip-before-cursor-atomic:
- fi-icl-u2: NOTRUN -> [SKIP][11] ([fdo#109278]) +2 similar issues
[11]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/fi-icl-u2/igt@kms_cursor_legacy@basic-busy-flip-before-cursor-atomic.html
* igt@kms_force_connector_basic@force-load-detect:
- fi-icl-u2: NOTRUN -> [SKIP][12] ([fdo#109285])
[12]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/fi-icl-u2/igt@kms_force_connector_basic@force-load-detect.html
* igt@prime_vgem@basic-userptr:
- fi-icl-u2: NOTRUN -> [SKIP][13] ([i915#3301])
[13]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/fi-icl-u2/igt@prime_vgem@basic-userptr.html
* igt@runner@aborted:
- fi-bdw-5557u: NOTRUN -> [FAIL][14] ([i915#2426] / [i915#4312])
[14]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/fi-bdw-5557u/igt@runner@aborted.html
#### Possible fixes ####
* igt@core_hotunplug@unbind-rebind:
- fi-blb-e6850: [FAIL][15] ([i915#3194]) -> [PASS][16]
[15]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/fi-blb-e6850/igt@core_hotunplug@unbind-rebind.html
[16]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/fi-blb-e6850/igt@core_hotunplug@unbind-rebind.html
* igt@i915_pm_rps@basic-api:
- bat-dg1-5: [FAIL][17] ([i915#4032]) -> [PASS][18]
[17]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/bat-dg1-5/igt@i915_pm_rps@basic-api.html
[18]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/bat-dg1-5/igt@i915_pm_rps@basic-api.html
* igt@i915_selftest@live@gt_pm:
- {fi-jsl-1}: [DMESG-FAIL][19] ([i915#1886]) -> [PASS][20]
[19]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/fi-jsl-1/igt@i915_selftest@live@gt_pm.html
[20]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/fi-jsl-1/igt@i915_selftest@live@gt_pm.html
* igt@i915_selftest@live@requests:
- fi-pnv-d510: [DMESG-FAIL][21] ([i915#2927] / [i915#4528]) -> [PASS][22]
[21]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/fi-pnv-d510/igt@i915_selftest@live@requests.html
[22]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/fi-pnv-d510/igt@i915_selftest@live@requests.html
{name}: This element is suppressed. This means it is ignored when computing
the status of the difference (SUCCESS, WARNING, or FAILURE).
[fdo#109271]: https://bugs.freedesktop.org/show_bug.cgi?id=109271
[fdo#109278]: https://bugs.freedesktop.org/show_bug.cgi?id=109278
[fdo#109285]: https://bugs.freedesktop.org/show_bug.cgi?id=109285
[fdo#109315]: https://bugs.freedesktop.org/show_bug.cgi?id=109315
[fdo#111827]: https://bugs.freedesktop.org/show_bug.cgi?id=111827
[i915#1886]: https://gitlab.freedesktop.org/drm/intel/issues/1886
[i915#2190]: https://gitlab.freedesktop.org/drm/intel/issues/2190
[i915#2426]: https://gitlab.freedesktop.org/drm/intel/issues/2426
[i915#2927]: https://gitlab.freedesktop.org/drm/intel/issues/2927
[i915#3194]: https://gitlab.freedesktop.org/drm/intel/issues/3194
[i915#3301]: https://gitlab.freedesktop.org/drm/intel/issues/3301
[i915#3576]: https://gitlab.freedesktop.org/drm/intel/issues/3576
[i915#4032]: https://gitlab.freedesktop.org/drm/intel/issues/4032
[i915#4312]: https://gitlab.freedesktop.org/drm/intel/issues/4312
[i915#4494]: https://gitlab.freedesktop.org/drm/intel/issues/4494
[i915#4528]: https://gitlab.freedesktop.org/drm/intel/issues/4528
[i915#4613]: https://gitlab.freedesktop.org/drm/intel/issues/4613
[i915#4957]: https://gitlab.freedesktop.org/drm/intel/issues/4957
Build changes
-------------
* Linux: CI_DRM_11306 -> Patchwork_22453
CI-20190529: 20190529
CI_DRM_11306: 0eb492df610222f39eb2ad5a903626dd3ad9aea2 @ git://anongit.freedesktop.org/gfx-ci/linux
IGT_6361: 2372a4beb6a33c5f0799a4a8ccbb93794f52dbca @ https://gitlab.freedesktop.org/drm/igt-gpu-tools.git
Patchwork_22453: 8e830a55eacbeaf5fd3723f8f5e002bf2a786844 @ git://anongit.freedesktop.org/gfx-ci/linux
== Linux commits ==
8e830a55eacb drm/amdgpu: add drm buddy support to amdgpu
== Logs ==
For more details see: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/index.html
[-- Attachment #2: Type: text/html, Size: 8518 bytes --]
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [PATCH v9] drm/amdgpu: add drm buddy support to amdgpu
2022-03-01 20:38 ` [Intel-gfx] " Arunpravin
@ 2022-03-02 2:51 ` kernel test robot
-1 siblings, 0 replies; 11+ messages in thread
From: kernel test robot @ 2022-03-02 2:51 UTC (permalink / raw)
To: Arunpravin, amd-gfx, dri-devel, intel-gfx
Cc: kbuild-all, Arunpravin, matthew.auld, tzimmermann,
alexander.deucher, christian.koenig
Hi Arunpravin,
Thank you for the patch! Perhaps something to improve:
[auto build test WARNING on drm/drm-next]
[also build test WARNING on drm-tip/drm-tip]
[cannot apply to drm-intel/for-linux-next v5.17-rc6 next-20220301]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]
url: https://github.com/0day-ci/linux/commits/Arunpravin/drm-amdgpu-add-drm-buddy-support-to-amdgpu/20220302-043936
base: git://anongit.freedesktop.org/drm/drm drm-next
config: ia64-randconfig-r011-20220301 (https://download.01.org/0day-ci/archive/20220302/202203020700.yfCUTgQZ-lkp@intel.com/config)
compiler: ia64-linux-gcc (GCC) 11.2.0
reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
chmod +x ~/bin/make.cross
# https://github.com/0day-ci/linux/commit/22a06757ae067e29c09a9d95eaf2b9053833740f
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Arunpravin/drm-amdgpu-add-drm-buddy-support-to-amdgpu/20220302-043936
git checkout 22a06757ae067e29c09a9d95eaf2b9053833740f
# save the config file to linux build tree
mkdir build_dir
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-11.2.0 make.cross O=build_dir ARCH=ia64 SHELL=/bin/bash drivers/gpu/drm/amd/amdgpu/
If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>
All warnings (new ones prefixed by >>):
drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c: In function 'amdgpu_vram_mgr_new':
>> drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c:360:28: warning: variable 'mem_bytes' set but not used [-Wunused-but-set-variable]
360 | u64 vis_usage = 0, mem_bytes, max_bytes, min_page_size;
| ^~~~~~~~~
vim +/mem_bytes +360 drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
343
344 /**
345 * amdgpu_vram_mgr_new - allocate new ranges
346 *
347 * @man: TTM memory type manager
348 * @tbo: TTM BO we need this range for
349 * @place: placement flags and restrictions
350 * @res: the resulting mem object
351 *
352 * Allocate VRAM for the given BO.
353 */
354 static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
355 struct ttm_buffer_object *tbo,
356 const struct ttm_place *place,
357 struct ttm_resource **res)
358 {
359 unsigned long lpfn, pages_per_node, pages_left, pages;
> 360 u64 vis_usage = 0, mem_bytes, max_bytes, min_page_size;
361 struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
362 struct amdgpu_device *adev = to_amdgpu_device(mgr);
363 struct amdgpu_vram_mgr_node *node;
364 struct drm_buddy *mm = &mgr->mm;
365 struct drm_buddy_block *block;
366 unsigned i;
367 int r;
368
369 lpfn = place->lpfn;
370 if (!lpfn)
371 lpfn = man->size >> PAGE_SHIFT;
372
373 max_bytes = adev->gmc.mc_vram_size;
374 if (tbo->type != ttm_bo_type_kernel)
375 max_bytes -= AMDGPU_VM_RESERVED_VRAM;
376
377 mem_bytes = tbo->base.size;
378 if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
379 pages_per_node = ~0ul;
380 } else {
381 #ifdef CONFIG_TRANSPARENT_HUGEPAGE
382 pages_per_node = HPAGE_PMD_NR;
383 #else
384 /* default to 2MB */
385 pages_per_node = 2UL << (20UL - PAGE_SHIFT);
386 #endif
387 pages_per_node = max_t(uint32_t, pages_per_node,
388 tbo->page_alignment);
389 }
390
391 node = kzalloc(sizeof(*node), GFP_KERNEL);
392 if (!node)
393 return -ENOMEM;
394
395 ttm_resource_init(tbo, place, &node->base);
396
397 /* bail out quickly if there's likely not enough VRAM for this BO */
398 if (ttm_resource_manager_usage(man) > max_bytes) {
399 r = -ENOSPC;
400 goto error_fini;
401 }
402
403 INIT_LIST_HEAD(&node->blocks);
404
405 if (place->flags & TTM_PL_FLAG_TOPDOWN)
406 node->flags |= DRM_BUDDY_TOPDOWN_ALLOCATION;
407
408 if (place->fpfn || lpfn != man->size >> PAGE_SHIFT)
409 /* Allocate blocks in desired range */
410 node->flags |= DRM_BUDDY_RANGE_ALLOCATION;
411
412 BUG_ON(!node->base.num_pages);
413 pages_left = node->base.num_pages;
414
415 i = 0;
416 while (pages_left) {
417 if (tbo->page_alignment)
418 min_page_size = tbo->page_alignment << PAGE_SHIFT;
419 else
420 min_page_size = mgr->default_page_size;
421
422 /* Limit maximum size to 2GB due to SG table limitations */
423 pages = min(pages_left, 2UL << (30 - PAGE_SHIFT));
424
425 if (pages >= pages_per_node)
426 min_page_size = pages_per_node << PAGE_SHIFT;
427
428 if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
429 pages = roundup_pow_of_two(pages);
430 min_page_size = pages << PAGE_SHIFT;
431
432 if (pages > lpfn)
433 lpfn = pages;
434 }
435
436 BUG_ON(min_page_size < mm->chunk_size);
437
438 mutex_lock(&mgr->lock);
439 r = drm_buddy_alloc_blocks(mm, (u64)place->fpfn << PAGE_SHIFT,
440 (u64)lpfn << PAGE_SHIFT,
441 (u64)pages << PAGE_SHIFT,
442 min_page_size,
443 &node->blocks,
444 node->flags);
445 mutex_unlock(&mgr->lock);
446 if (unlikely(r))
447 goto error_free_blocks;
448
449 ++i;
450
451 if (pages > pages_left)
452 pages_left = 0;
453 else
454 pages_left -= pages;
455 }
456
457 /* Free unused pages for contiguous allocation */
458 if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
459 u64 actual_size = (u64)node->base.num_pages << PAGE_SHIFT;
460
461 mutex_lock(&mgr->lock);
462 drm_buddy_block_trim(mm,
463 actual_size,
464 &node->blocks);
465 mutex_unlock(&mgr->lock);
466 }
467
468 list_for_each_entry(block, &node->blocks, link)
469 vis_usage += amdgpu_vram_mgr_vis_size(adev, block);
470
471 block = amdgpu_get_node(&node->blocks);
472 if (!block) {
473 r = -EINVAL;
474 goto error_fini;
475 }
476
477 node->base.start = amdgpu_node_start(block) >> PAGE_SHIFT;
478
479 if (i == 1 && place->flags & TTM_PL_FLAG_CONTIGUOUS)
480 node->base.placement |= TTM_PL_FLAG_CONTIGUOUS;
481
482 if (adev->gmc.xgmi.connected_to_cpu)
483 node->base.bus.caching = ttm_cached;
484 else
485 node->base.bus.caching = ttm_write_combined;
486
487 atomic64_add(vis_usage, &mgr->vis_usage);
488 *res = &node->base;
489 return 0;
490
491 error_free_blocks:
492 mutex_lock(&mgr->lock);
493 drm_buddy_free_list(mm, &node->blocks);
494 mutex_unlock(&mgr->lock);
495 error_fini:
496 ttm_resource_fini(man, &node->base);
497 kfree(node);
498
499 return r;
500 }
501
---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [Intel-gfx] [PATCH v9] drm/amdgpu: add drm buddy support to amdgpu
@ 2022-03-02 2:51 ` kernel test robot
0 siblings, 0 replies; 11+ messages in thread
From: kernel test robot @ 2022-03-02 2:51 UTC (permalink / raw)
To: Arunpravin, amd-gfx, dri-devel, intel-gfx
Cc: kbuild-all, Arunpravin, matthew.auld, tzimmermann,
alexander.deucher, christian.koenig
Hi Arunpravin,
Thank you for the patch! Perhaps something to improve:
[auto build test WARNING on drm/drm-next]
[also build test WARNING on drm-tip/drm-tip]
[cannot apply to drm-intel/for-linux-next v5.17-rc6 next-20220301]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]
url: https://github.com/0day-ci/linux/commits/Arunpravin/drm-amdgpu-add-drm-buddy-support-to-amdgpu/20220302-043936
base: git://anongit.freedesktop.org/drm/drm drm-next
config: ia64-randconfig-r011-20220301 (https://download.01.org/0day-ci/archive/20220302/202203020700.yfCUTgQZ-lkp@intel.com/config)
compiler: ia64-linux-gcc (GCC) 11.2.0
reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
chmod +x ~/bin/make.cross
# https://github.com/0day-ci/linux/commit/22a06757ae067e29c09a9d95eaf2b9053833740f
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Arunpravin/drm-amdgpu-add-drm-buddy-support-to-amdgpu/20220302-043936
git checkout 22a06757ae067e29c09a9d95eaf2b9053833740f
# save the config file to linux build tree
mkdir build_dir
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-11.2.0 make.cross O=build_dir ARCH=ia64 SHELL=/bin/bash drivers/gpu/drm/amd/amdgpu/
If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>
All warnings (new ones prefixed by >>):
drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c: In function 'amdgpu_vram_mgr_new':
>> drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c:360:28: warning: variable 'mem_bytes' set but not used [-Wunused-but-set-variable]
360 | u64 vis_usage = 0, mem_bytes, max_bytes, min_page_size;
| ^~~~~~~~~
vim +/mem_bytes +360 drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
343
344 /**
345 * amdgpu_vram_mgr_new - allocate new ranges
346 *
347 * @man: TTM memory type manager
348 * @tbo: TTM BO we need this range for
349 * @place: placement flags and restrictions
350 * @res: the resulting mem object
351 *
352 * Allocate VRAM for the given BO.
353 */
354 static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
355 struct ttm_buffer_object *tbo,
356 const struct ttm_place *place,
357 struct ttm_resource **res)
358 {
359 unsigned long lpfn, pages_per_node, pages_left, pages;
> 360 u64 vis_usage = 0, mem_bytes, max_bytes, min_page_size;
361 struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
362 struct amdgpu_device *adev = to_amdgpu_device(mgr);
363 struct amdgpu_vram_mgr_node *node;
364 struct drm_buddy *mm = &mgr->mm;
365 struct drm_buddy_block *block;
366 unsigned i;
367 int r;
368
369 lpfn = place->lpfn;
370 if (!lpfn)
371 lpfn = man->size >> PAGE_SHIFT;
372
373 max_bytes = adev->gmc.mc_vram_size;
374 if (tbo->type != ttm_bo_type_kernel)
375 max_bytes -= AMDGPU_VM_RESERVED_VRAM;
376
377 mem_bytes = tbo->base.size;
378 if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
379 pages_per_node = ~0ul;
380 } else {
381 #ifdef CONFIG_TRANSPARENT_HUGEPAGE
382 pages_per_node = HPAGE_PMD_NR;
383 #else
384 /* default to 2MB */
385 pages_per_node = 2UL << (20UL - PAGE_SHIFT);
386 #endif
387 pages_per_node = max_t(uint32_t, pages_per_node,
388 tbo->page_alignment);
389 }
390
391 node = kzalloc(sizeof(*node), GFP_KERNEL);
392 if (!node)
393 return -ENOMEM;
394
395 ttm_resource_init(tbo, place, &node->base);
396
397 /* bail out quickly if there's likely not enough VRAM for this BO */
398 if (ttm_resource_manager_usage(man) > max_bytes) {
399 r = -ENOSPC;
400 goto error_fini;
401 }
402
403 INIT_LIST_HEAD(&node->blocks);
404
405 if (place->flags & TTM_PL_FLAG_TOPDOWN)
406 node->flags |= DRM_BUDDY_TOPDOWN_ALLOCATION;
407
408 if (place->fpfn || lpfn != man->size >> PAGE_SHIFT)
409 /* Allocate blocks in desired range */
410 node->flags |= DRM_BUDDY_RANGE_ALLOCATION;
411
412 BUG_ON(!node->base.num_pages);
413 pages_left = node->base.num_pages;
414
415 i = 0;
416 while (pages_left) {
417 if (tbo->page_alignment)
418 min_page_size = tbo->page_alignment << PAGE_SHIFT;
419 else
420 min_page_size = mgr->default_page_size;
421
422 /* Limit maximum size to 2GB due to SG table limitations */
423 pages = min(pages_left, 2UL << (30 - PAGE_SHIFT));
424
425 if (pages >= pages_per_node)
426 min_page_size = pages_per_node << PAGE_SHIFT;
427
428 if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
429 pages = roundup_pow_of_two(pages);
430 min_page_size = pages << PAGE_SHIFT;
431
432 if (pages > lpfn)
433 lpfn = pages;
434 }
435
436 BUG_ON(min_page_size < mm->chunk_size);
437
438 mutex_lock(&mgr->lock);
439 r = drm_buddy_alloc_blocks(mm, (u64)place->fpfn << PAGE_SHIFT,
440 (u64)lpfn << PAGE_SHIFT,
441 (u64)pages << PAGE_SHIFT,
442 min_page_size,
443 &node->blocks,
444 node->flags);
445 mutex_unlock(&mgr->lock);
446 if (unlikely(r))
447 goto error_free_blocks;
448
449 ++i;
450
451 if (pages > pages_left)
452 pages_left = 0;
453 else
454 pages_left -= pages;
455 }
456
457 /* Free unused pages for contiguous allocation */
458 if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
459 u64 actual_size = (u64)node->base.num_pages << PAGE_SHIFT;
460
461 mutex_lock(&mgr->lock);
462 drm_buddy_block_trim(mm,
463 actual_size,
464 &node->blocks);
465 mutex_unlock(&mgr->lock);
466 }
467
468 list_for_each_entry(block, &node->blocks, link)
469 vis_usage += amdgpu_vram_mgr_vis_size(adev, block);
470
471 block = amdgpu_get_node(&node->blocks);
472 if (!block) {
473 r = -EINVAL;
474 goto error_fini;
475 }
476
477 node->base.start = amdgpu_node_start(block) >> PAGE_SHIFT;
478
479 if (i == 1 && place->flags & TTM_PL_FLAG_CONTIGUOUS)
480 node->base.placement |= TTM_PL_FLAG_CONTIGUOUS;
481
482 if (adev->gmc.xgmi.connected_to_cpu)
483 node->base.bus.caching = ttm_cached;
484 else
485 node->base.bus.caching = ttm_write_combined;
486
487 atomic64_add(vis_usage, &mgr->vis_usage);
488 *res = &node->base;
489 return 0;
490
491 error_free_blocks:
492 mutex_lock(&mgr->lock);
493 drm_buddy_free_list(mm, &node->blocks);
494 mutex_unlock(&mgr->lock);
495 error_fini:
496 ttm_resource_fini(man, &node->base);
497 kfree(node);
498
499 return r;
500 }
501
---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org
^ permalink raw reply [flat|nested] 11+ messages in thread
* [Intel-gfx] ✓ Fi.CI.IGT: success for drm/amdgpu: add drm buddy support to amdgpu
2022-03-01 20:38 ` [Intel-gfx] " Arunpravin
` (4 preceding siblings ...)
(?)
@ 2022-03-02 4:21 ` Patchwork
-1 siblings, 0 replies; 11+ messages in thread
From: Patchwork @ 2022-03-02 4:21 UTC (permalink / raw)
To: Arunpravin; +Cc: intel-gfx
[-- Attachment #1: Type: text/plain, Size: 30267 bytes --]
== Series Details ==
Series: drm/amdgpu: add drm buddy support to amdgpu
URL : https://patchwork.freedesktop.org/series/100908/
State : success
== Summary ==
CI Bug Log - changes from CI_DRM_11306_full -> Patchwork_22453_full
====================================================
Summary
-------
**SUCCESS**
No regressions found.
Participating hosts (13 -> 13)
------------------------------
No changes in participating hosts
Possible new issues
-------------------
Here are the unknown changes that may have been introduced in Patchwork_22453_full:
### IGT changes ###
#### Suppressed ####
The following results come from untrusted machines, tests, or statuses.
They do not affect the overall result.
* igt@i915_selftest@live@hangcheck:
- {shard-rkl}: NOTRUN -> [INCOMPLETE][1]
[1]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-rkl-5/igt@i915_selftest@live@hangcheck.html
* {igt@kms_plane_scaling@planes-unity-scaling-downscale-factor-0-25@pipe-a-hdmi-a-1-planes-upscale-downscale}:
- {shard-tglu}: NOTRUN -> [SKIP][2] +3 similar issues
[2]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-tglu-2/igt@kms_plane_scaling@planes-unity-scaling-downscale-factor-0-25@pipe-a-hdmi-a-1-planes-upscale-downscale.html
* {igt@kms_plane_scaling@scaler-with-rotation-unity-scaling@pipe-b-edp-1-scaler-with-rotation}:
- {shard-rkl}: NOTRUN -> [SKIP][3] +14 similar issues
[3]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-rkl-6/igt@kms_plane_scaling@scaler-with-rotation-unity-scaling@pipe-b-edp-1-scaler-with-rotation.html
* {igt@kms_plane_scaling@upscale-with-rotation-factor-0-25@pipe-b-hdmi-a-1-upscale-with-rotation}:
- {shard-dg1}: NOTRUN -> [SKIP][4] +15 similar issues
[4]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-dg1-19/igt@kms_plane_scaling@upscale-with-rotation-factor-0-25@pipe-b-hdmi-a-1-upscale-with-rotation.html
New tests
---------
New tests have been introduced between CI_DRM_11306_full and Patchwork_22453_full:
### New IGT tests (1) ###
* igt@kms_plane_scaling@planes-unity-scaling-downscale-factor-0-75@pipe-d-edp-1-planes-upscale-downscale:
- Statuses : 1 pass(s)
- Exec time: [1.28] s
Known issues
------------
Here are the changes found in Patchwork_22453_full that come from known issues:
### IGT changes ###
#### Issues hit ####
* igt@gem_create@create-massive:
- shard-kbl: NOTRUN -> [DMESG-WARN][5] ([i915#4991])
[5]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-kbl4/igt@gem_create@create-massive.html
* igt@gem_eio@in-flight-contexts-1us:
- shard-tglb: [PASS][6] -> [TIMEOUT][7] ([i915#3063])
[6]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-tglb2/igt@gem_eio@in-flight-contexts-1us.html
[7]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-tglb6/igt@gem_eio@in-flight-contexts-1us.html
* igt@gem_exec_balancer@parallel-out-fence:
- shard-kbl: NOTRUN -> [DMESG-WARN][8] ([i915#5076])
[8]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-kbl3/igt@gem_exec_balancer@parallel-out-fence.html
* igt@gem_exec_capture@pi@rcs0:
- shard-skl: [PASS][9] -> [INCOMPLETE][10] ([i915#4547])
[9]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-skl4/igt@gem_exec_capture@pi@rcs0.html
[10]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl3/igt@gem_exec_capture@pi@rcs0.html
* igt@gem_exec_fair@basic-none-share@rcs0:
- shard-iclb: [PASS][11] -> [FAIL][12] ([i915#2842])
[11]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-iclb4/igt@gem_exec_fair@basic-none-share@rcs0.html
[12]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb8/igt@gem_exec_fair@basic-none-share@rcs0.html
* igt@gem_exec_fair@basic-none-solo@rcs0:
- shard-kbl: NOTRUN -> [FAIL][13] ([i915#2842])
[13]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-kbl3/igt@gem_exec_fair@basic-none-solo@rcs0.html
* igt@gem_exec_fair@basic-none@vcs0:
- shard-apl: [PASS][14] -> [FAIL][15] ([i915#2842])
[14]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-apl3/igt@gem_exec_fair@basic-none@vcs0.html
[15]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-apl2/igt@gem_exec_fair@basic-none@vcs0.html
* igt@gem_exec_fair@basic-pace@vcs1:
- shard-iclb: NOTRUN -> [FAIL][16] ([i915#2842])
[16]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb2/igt@gem_exec_fair@basic-pace@vcs1.html
* igt@gem_exec_suspend@basic-s0@smem:
- shard-snb: [PASS][17] -> [SKIP][18] ([fdo#109271])
[17]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-snb7/igt@gem_exec_suspend@basic-s0@smem.html
[18]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-snb5/igt@gem_exec_suspend@basic-s0@smem.html
* igt@gem_huc_copy@huc-copy:
- shard-skl: NOTRUN -> [SKIP][19] ([fdo#109271] / [i915#2190])
[19]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl8/igt@gem_huc_copy@huc-copy.html
* igt@gem_lmem_swapping@heavy-multi:
- shard-apl: NOTRUN -> [SKIP][20] ([fdo#109271] / [i915#4613]) +1 similar issue
[20]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-apl2/igt@gem_lmem_swapping@heavy-multi.html
* igt@gem_lmem_swapping@heavy-verify-multi:
- shard-skl: NOTRUN -> [SKIP][21] ([fdo#109271] / [i915#4613]) +2 similar issues
[21]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl7/igt@gem_lmem_swapping@heavy-verify-multi.html
* igt@gem_lmem_swapping@heavy-verify-random:
- shard-iclb: NOTRUN -> [SKIP][22] ([i915#4613])
[22]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb6/igt@gem_lmem_swapping@heavy-verify-random.html
* igt@gem_pwrite@basic-exhaustion:
- shard-skl: NOTRUN -> [WARN][23] ([i915#2658])
[23]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl10/igt@gem_pwrite@basic-exhaustion.html
* igt@gem_render_copy@x-tiled-to-vebox-yf-tiled:
- shard-kbl: NOTRUN -> [SKIP][24] ([fdo#109271]) +80 similar issues
[24]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-kbl3/igt@gem_render_copy@x-tiled-to-vebox-yf-tiled.html
* igt@gem_render_copy@y-tiled-mc-ccs-to-yf-tiled-ccs:
- shard-iclb: NOTRUN -> [SKIP][25] ([i915#768])
[25]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb6/igt@gem_render_copy@y-tiled-mc-ccs-to-yf-tiled-ccs.html
* igt@gem_spin_batch@spin-each:
- shard-skl: NOTRUN -> [FAIL][26] ([i915#2898])
[26]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl8/igt@gem_spin_batch@spin-each.html
* igt@gen3_render_tiledy_blits:
- shard-iclb: NOTRUN -> [SKIP][27] ([fdo#109289])
[27]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb6/igt@gen3_render_tiledy_blits.html
* igt@gen9_exec_parse@shadow-peek:
- shard-tglb: NOTRUN -> [SKIP][28] ([i915#2527] / [i915#2856])
[28]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-tglb3/igt@gen9_exec_parse@shadow-peek.html
* igt@i915_pm_dc@dc6-psr:
- shard-skl: NOTRUN -> [FAIL][29] ([i915#454])
[29]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl4/igt@i915_pm_dc@dc6-psr.html
* igt@i915_pm_lpsp@kms-lpsp@kms-lpsp-dp:
- shard-kbl: NOTRUN -> [SKIP][30] ([fdo#109271] / [i915#1937])
[30]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-kbl4/igt@i915_pm_lpsp@kms-lpsp@kms-lpsp-dp.html
* igt@i915_pm_sseu@full-enable:
- shard-iclb: NOTRUN -> [SKIP][31] ([i915#4387])
[31]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb6/igt@i915_pm_sseu@full-enable.html
* igt@kms_async_flips@alternate-sync-async-flip:
- shard-skl: [PASS][32] -> [FAIL][33] ([i915#2521])
[32]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-skl7/igt@kms_async_flips@alternate-sync-async-flip.html
[33]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl6/igt@kms_async_flips@alternate-sync-async-flip.html
* igt@kms_big_fb@x-tiled-32bpp-rotate-180:
- shard-glk: [PASS][34] -> [DMESG-WARN][35] ([i915#118])
[34]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-glk9/igt@kms_big_fb@x-tiled-32bpp-rotate-180.html
[35]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-glk2/igt@kms_big_fb@x-tiled-32bpp-rotate-180.html
* igt@kms_big_fb@x-tiled-max-hw-stride-64bpp-rotate-180-async-flip:
- shard-skl: NOTRUN -> [FAIL][36] ([i915#3743]) +4 similar issues
[36]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl10/igt@kms_big_fb@x-tiled-max-hw-stride-64bpp-rotate-180-async-flip.html
* igt@kms_big_fb@y-tiled-max-hw-stride-64bpp-rotate-0-hflip:
- shard-skl: NOTRUN -> [SKIP][37] ([fdo#109271] / [i915#3777]) +3 similar issues
[37]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl7/igt@kms_big_fb@y-tiled-max-hw-stride-64bpp-rotate-0-hflip.html
* igt@kms_big_fb@y-tiled-max-hw-stride-64bpp-rotate-180-async-flip:
- shard-skl: NOTRUN -> [FAIL][38] ([i915#3763])
[38]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl4/igt@kms_big_fb@y-tiled-max-hw-stride-64bpp-rotate-180-async-flip.html
* igt@kms_big_fb@yf-tiled-max-hw-stride-32bpp-rotate-0-hflip-async-flip:
- shard-apl: NOTRUN -> [SKIP][39] ([fdo#109271] / [i915#3777])
[39]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-apl2/igt@kms_big_fb@yf-tiled-max-hw-stride-32bpp-rotate-0-hflip-async-flip.html
* igt@kms_ccs@pipe-a-crc-primary-rotation-180-y_tiled_gen12_rc_ccs_cc:
- shard-kbl: NOTRUN -> [SKIP][40] ([fdo#109271] / [i915#3886]) +3 similar issues
[40]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-kbl3/igt@kms_ccs@pipe-a-crc-primary-rotation-180-y_tiled_gen12_rc_ccs_cc.html
* igt@kms_ccs@pipe-b-missing-ccs-buffer-y_tiled_gen12_rc_ccs_cc:
- shard-apl: NOTRUN -> [SKIP][41] ([fdo#109271] / [i915#3886])
[41]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-apl2/igt@kms_ccs@pipe-b-missing-ccs-buffer-y_tiled_gen12_rc_ccs_cc.html
* igt@kms_ccs@pipe-b-random-ccs-data-y_tiled_gen12_rc_ccs_cc:
- shard-iclb: NOTRUN -> [SKIP][42] ([fdo#109278] / [i915#3886]) +1 similar issue
[42]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb6/igt@kms_ccs@pipe-b-random-ccs-data-y_tiled_gen12_rc_ccs_cc.html
* igt@kms_ccs@pipe-c-bad-rotation-90-y_tiled_gen12_rc_ccs_cc:
- shard-skl: NOTRUN -> [SKIP][43] ([fdo#109271] / [i915#3886]) +18 similar issues
[43]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl4/igt@kms_ccs@pipe-c-bad-rotation-90-y_tiled_gen12_rc_ccs_cc.html
* igt@kms_ccs@pipe-c-crc-primary-basic-y_tiled_ccs:
- shard-apl: NOTRUN -> [SKIP][44] ([fdo#109271]) +47 similar issues
[44]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-apl2/igt@kms_ccs@pipe-c-crc-primary-basic-y_tiled_ccs.html
* igt@kms_chamelium@dp-crc-multiple:
- shard-skl: NOTRUN -> [SKIP][45] ([fdo#109271] / [fdo#111827]) +33 similar issues
[45]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl4/igt@kms_chamelium@dp-crc-multiple.html
* igt@kms_chamelium@hdmi-hpd-for-each-pipe:
- shard-kbl: NOTRUN -> [SKIP][46] ([fdo#109271] / [fdo#111827]) +5 similar issues
[46]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-kbl3/igt@kms_chamelium@hdmi-hpd-for-each-pipe.html
* igt@kms_color_chamelium@pipe-b-ctm-limited-range:
- shard-iclb: NOTRUN -> [SKIP][47] ([fdo#109284] / [fdo#111827]) +1 similar issue
[47]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb6/igt@kms_color_chamelium@pipe-b-ctm-limited-range.html
* igt@kms_color_chamelium@pipe-b-ctm-red-to-blue:
- shard-apl: NOTRUN -> [SKIP][48] ([fdo#109271] / [fdo#111827]) +2 similar issues
[48]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-apl6/igt@kms_color_chamelium@pipe-b-ctm-red-to-blue.html
* igt@kms_content_protection@type1:
- shard-iclb: NOTRUN -> [SKIP][49] ([fdo#109300] / [fdo#111066])
[49]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb6/igt@kms_content_protection@type1.html
* igt@kms_content_protection@uevent:
- shard-kbl: NOTRUN -> [FAIL][50] ([i915#2105])
[50]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-kbl3/igt@kms_content_protection@uevent.html
* igt@kms_cursor_crc@pipe-a-cursor-max-size-sliding:
- shard-iclb: NOTRUN -> [SKIP][51] ([fdo#109278]) +4 similar issues
[51]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb6/igt@kms_cursor_crc@pipe-a-cursor-max-size-sliding.html
* igt@kms_cursor_edge_walk@pipe-a-128x128-left-edge:
- shard-glk: [PASS][52] -> [DMESG-FAIL][53] ([i915#118])
[52]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-glk7/igt@kms_cursor_edge_walk@pipe-a-128x128-left-edge.html
[53]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-glk8/igt@kms_cursor_edge_walk@pipe-a-128x128-left-edge.html
* igt@kms_cursor_legacy@cursorb-vs-flipa-legacy:
- shard-iclb: NOTRUN -> [SKIP][54] ([fdo#109274] / [fdo#109278]) +1 similar issue
[54]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb6/igt@kms_cursor_legacy@cursorb-vs-flipa-legacy.html
* igt@kms_cursor_legacy@pipe-d-single-bo:
- shard-kbl: NOTRUN -> [SKIP][55] ([fdo#109271] / [i915#533])
[55]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-kbl4/igt@kms_cursor_legacy@pipe-d-single-bo.html
* igt@kms_flip@2x-plain-flip-fb-recreate:
- shard-iclb: NOTRUN -> [SKIP][56] ([fdo#109274])
[56]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb6/igt@kms_flip@2x-plain-flip-fb-recreate.html
* igt@kms_flip@flip-vs-suspend-interruptible@c-dp1:
- shard-apl: [PASS][57] -> [DMESG-WARN][58] ([i915#180]) +3 similar issues
[57]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-apl7/igt@kms_flip@flip-vs-suspend-interruptible@c-dp1.html
[58]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-apl6/igt@kms_flip@flip-vs-suspend-interruptible@c-dp1.html
* igt@kms_flip@plain-flip-fb-recreate@a-hdmi-a1:
- shard-glk: [PASS][59] -> [FAIL][60] ([i915#2122])
[59]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-glk4/igt@kms_flip@plain-flip-fb-recreate@a-hdmi-a1.html
[60]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-glk1/igt@kms_flip@plain-flip-fb-recreate@a-hdmi-a1.html
* igt@kms_flip@plain-flip-ts-check-interruptible@a-edp1:
- shard-skl: [PASS][61] -> [FAIL][62] ([i915#2122])
[61]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-skl10/igt@kms_flip@plain-flip-ts-check-interruptible@a-edp1.html
[62]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl10/igt@kms_flip@plain-flip-ts-check-interruptible@a-edp1.html
* igt@kms_flip_scaled_crc@flip-32bpp-ytile-to-32bpp-ytileccs-upscaling:
- shard-glk: [PASS][63] -> [FAIL][64] ([i915#4911])
[63]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-glk2/igt@kms_flip_scaled_crc@flip-32bpp-ytile-to-32bpp-ytileccs-upscaling.html
[64]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-glk8/igt@kms_flip_scaled_crc@flip-32bpp-ytile-to-32bpp-ytileccs-upscaling.html
* igt@kms_flip_scaled_crc@flip-64bpp-ytile-to-32bpp-ytile-downscaling:
- shard-iclb: [PASS][65] -> [SKIP][66] ([i915#3701])
[65]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-iclb8/igt@kms_flip_scaled_crc@flip-64bpp-ytile-to-32bpp-ytile-downscaling.html
[66]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb2/igt@kms_flip_scaled_crc@flip-64bpp-ytile-to-32bpp-ytile-downscaling.html
* igt@kms_frontbuffer_tracking@fbc-2p-scndscrn-pri-indfb-draw-blt:
- shard-skl: NOTRUN -> [SKIP][67] ([fdo#109271]) +390 similar issues
[67]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl1/igt@kms_frontbuffer_tracking@fbc-2p-scndscrn-pri-indfb-draw-blt.html
* igt@kms_frontbuffer_tracking@fbc-suspend:
- shard-kbl: [PASS][68] -> [INCOMPLETE][69] ([i915#2828] / [i915#3614])
[68]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-kbl1/igt@kms_frontbuffer_tracking@fbc-suspend.html
[69]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-kbl3/igt@kms_frontbuffer_tracking@fbc-suspend.html
* igt@kms_frontbuffer_tracking@fbcpsr-2p-primscrn-pri-shrfb-draw-mmap-gtt:
- shard-iclb: NOTRUN -> [SKIP][70] ([fdo#109280]) +4 similar issues
[70]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb6/igt@kms_frontbuffer_tracking@fbcpsr-2p-primscrn-pri-shrfb-draw-mmap-gtt.html
* igt@kms_hdr@bpc-switch-suspend:
- shard-kbl: [PASS][71] -> [DMESG-WARN][72] ([i915#180]) +2 similar issues
[71]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-kbl3/igt@kms_hdr@bpc-switch-suspend.html
[72]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-kbl7/igt@kms_hdr@bpc-switch-suspend.html
- shard-skl: [PASS][73] -> [FAIL][74] ([i915#1188])
[73]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-skl4/igt@kms_hdr@bpc-switch-suspend.html
[74]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl9/igt@kms_hdr@bpc-switch-suspend.html
* igt@kms_pipe_crc_basic@suspend-read-crc-pipe-d:
- shard-skl: NOTRUN -> [SKIP][75] ([fdo#109271] / [i915#533]) +3 similar issues
[75]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl8/igt@kms_pipe_crc_basic@suspend-read-crc-pipe-d.html
* igt@kms_plane_alpha_blend@pipe-a-alpha-7efc:
- shard-skl: NOTRUN -> [FAIL][76] ([fdo#108145] / [i915#265]) +5 similar issues
[76]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl8/igt@kms_plane_alpha_blend@pipe-a-alpha-7efc.html
- shard-kbl: NOTRUN -> [FAIL][77] ([fdo#108145] / [i915#265]) +2 similar issues
[77]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-kbl3/igt@kms_plane_alpha_blend@pipe-a-alpha-7efc.html
* igt@kms_plane_alpha_blend@pipe-c-alpha-7efc:
- shard-apl: NOTRUN -> [FAIL][78] ([fdo#108145] / [i915#265])
[78]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-apl6/igt@kms_plane_alpha_blend@pipe-c-alpha-7efc.html
* igt@kms_plane_alpha_blend@pipe-c-alpha-transparent-fb:
- shard-skl: NOTRUN -> [FAIL][79] ([i915#265])
[79]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl4/igt@kms_plane_alpha_blend@pipe-c-alpha-transparent-fb.html
* igt@kms_plane_lowres@pipe-a-tiling-x:
- shard-iclb: NOTRUN -> [SKIP][80] ([i915#3536])
[80]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb6/igt@kms_plane_lowres@pipe-a-tiling-x.html
* igt@kms_psr2_su@frontbuffer-xrgb8888:
- shard-skl: NOTRUN -> [SKIP][81] ([fdo#109271] / [i915#658]) +2 similar issues
[81]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl4/igt@kms_psr2_su@frontbuffer-xrgb8888.html
- shard-iclb: [PASS][82] -> [SKIP][83] ([fdo#109642] / [fdo#111068] / [i915#658])
[82]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-iclb2/igt@kms_psr2_su@frontbuffer-xrgb8888.html
[83]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb1/igt@kms_psr2_su@frontbuffer-xrgb8888.html
* igt@kms_psr@psr2_no_drrs:
- shard-iclb: NOTRUN -> [SKIP][84] ([fdo#109441])
[84]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb6/igt@kms_psr@psr2_no_drrs.html
* igt@kms_tv_load_detect@load-detect:
- shard-iclb: NOTRUN -> [SKIP][85] ([fdo#109309])
[85]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb6/igt@kms_tv_load_detect@load-detect.html
* igt@kms_vrr@flip-suspend:
- shard-iclb: NOTRUN -> [SKIP][86] ([fdo#109502])
[86]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb6/igt@kms_vrr@flip-suspend.html
* igt@kms_writeback@writeback-invalid-parameters:
- shard-skl: NOTRUN -> [SKIP][87] ([fdo#109271] / [i915#2437]) +1 similar issue
[87]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl8/igt@kms_writeback@writeback-invalid-parameters.html
* igt@perf@blocking:
- shard-skl: [PASS][88] -> [FAIL][89] ([i915#1542])
[88]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-skl6/igt@perf@blocking.html
[89]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl1/igt@perf@blocking.html
* igt@perf@polling-small-buf:
- shard-skl: [PASS][90] -> [FAIL][91] ([i915#1722])
[90]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-skl7/igt@perf@polling-small-buf.html
[91]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl6/igt@perf@polling-small-buf.html
* igt@prime_vgem@fence-flip-hang:
- shard-iclb: NOTRUN -> [SKIP][92] ([fdo#109295])
[92]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb6/igt@prime_vgem@fence-flip-hang.html
* igt@syncobj_timeline@invalid-transfer-non-existent-point:
- shard-skl: NOTRUN -> [DMESG-WARN][93] ([i915#5098])
[93]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl1/igt@syncobj_timeline@invalid-transfer-non-existent-point.html
* igt@sysfs_clients@busy:
- shard-apl: NOTRUN -> [SKIP][94] ([fdo#109271] / [i915#2994])
[94]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-apl2/igt@sysfs_clients@busy.html
* igt@sysfs_clients@create:
- shard-skl: NOTRUN -> [SKIP][95] ([fdo#109271] / [i915#2994]) +3 similar issues
[95]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-skl4/igt@sysfs_clients@create.html
* igt@sysfs_clients@sema-50:
- shard-kbl: NOTRUN -> [SKIP][96] ([fdo#109271] / [i915#2994])
[96]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-kbl3/igt@sysfs_clients@sema-50.html
* igt@sysfs_clients@split-10:
- shard-iclb: NOTRUN -> [SKIP][97] ([i915#2994])
[97]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb6/igt@sysfs_clients@split-10.html
#### Possible fixes ####
* igt@fbdev@pan:
- {shard-rkl}: [SKIP][98] ([i915#2582]) -> [PASS][99]
[98]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-rkl-1/igt@fbdev@pan.html
[99]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-rkl-6/igt@fbdev@pan.html
* igt@feature_discovery@psr2:
- shard-iclb: [SKIP][100] ([i915#658]) -> [PASS][101]
[100]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-iclb8/igt@feature_discovery@psr2.html
[101]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb2/igt@feature_discovery@psr2.html
* igt@gem_eio@kms:
- shard-tglb: [FAIL][102] ([i915#232]) -> [PASS][103]
[102]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-tglb5/igt@gem_eio@kms.html
[103]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-tglb7/igt@gem_eio@kms.html
* igt@gem_eio@unwedge-stress:
- shard-iclb: [TIMEOUT][104] ([i915#2481] / [i915#3070]) -> [PASS][105]
[104]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-iclb8/igt@gem_eio@unwedge-stress.html
[105]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-iclb2/igt@gem_eio@unwedge-stress.html
* igt@gem_exec_capture@pi@rcs0:
- {shard-tglu}: [INCOMPLETE][106] ([i915#3371]) -> [PASS][107]
[106]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-tglu-3/igt@gem_exec_capture@pi@rcs0.html
[107]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-tglu-8/igt@gem_exec_capture@pi@rcs0.html
* igt@gem_exec_fair@basic-deadline:
- shard-glk: [FAIL][108] ([i915#2846]) -> [PASS][109]
[108]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-glk9/igt@gem_exec_fair@basic-deadline.html
[109]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-glk2/igt@gem_exec_fair@basic-deadline.html
* igt@gem_exec_fair@basic-none-solo@rcs0:
- {shard-rkl}: [FAIL][110] ([i915#2842]) -> [PASS][111]
[110]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-rkl-5/igt@gem_exec_fair@basic-none-solo@rcs0.html
[111]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-rkl-2/igt@gem_exec_fair@basic-none-solo@rcs0.html
* igt@gem_exec_fair@basic-pace-share@rcs0:
- shard-glk: [FAIL][112] ([i915#2842]) -> [PASS][113]
[112]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-glk9/igt@gem_exec_fair@basic-pace-share@rcs0.html
[113]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-glk2/igt@gem_exec_fair@basic-pace-share@rcs0.html
* igt@gem_exec_fair@basic-pace@bcs0:
- shard-tglb: [FAIL][114] ([i915#2842]) -> [PASS][115]
[114]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-tglb3/igt@gem_exec_fair@basic-pace@bcs0.html
[115]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-tglb6/igt@gem_exec_fair@basic-pace@bcs0.html
* igt@gem_huc_copy@huc-copy:
- shard-tglb: [SKIP][116] ([i915#2190]) -> [PASS][117]
[116]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-tglb7/igt@gem_huc_copy@huc-copy.html
[117]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-tglb8/igt@gem_huc_copy@huc-copy.html
* igt@gem_ppgtt@flink-and-close-vma-leak:
- shard-glk: [FAIL][118] ([i915#644]) -> [PASS][119]
[118]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-glk8/igt@gem_ppgtt@flink-and-close-vma-leak.html
[119]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-glk8/igt@gem_ppgtt@flink-and-close-vma-leak.html
* igt@gem_workarounds@suspend-resume-fd:
- shard-kbl: [DMESG-WARN][120] ([i915#180]) -> [PASS][121] +2 similar issues
[120]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-kbl6/igt@gem_workarounds@suspend-resume-fd.html
[121]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-kbl3/igt@gem_workarounds@suspend-resume-fd.html
* igt@i915_pm_backlight@basic-brightness:
- {shard-rkl}: [SKIP][122] ([i915#3012]) -> [PASS][123]
[122]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-rkl-1/igt@i915_pm_backlight@basic-brightness.html
[123]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-rkl-6/igt@i915_pm_backlight@basic-brightness.html
* igt@i915_pm_rpm@dpms-mode-unset-lpsp:
- {shard-rkl}: [SKIP][124] ([i915#1397]) -> [PASS][125]
[124]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-rkl-1/igt@i915_pm_rpm@dpms-mode-unset-lpsp.html
[125]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-rkl-6/igt@i915_pm_rpm@dpms-mode-unset-lpsp.html
* igt@i915_pm_rpm@modeset-non-lpsp-stress:
- {shard-dg1}: [SKIP][126] ([i915#1397]) -> [PASS][127] +2 similar issues
[126]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-dg1-12/igt@i915_pm_rpm@modeset-non-lpsp-stress.html
[127]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-dg1-18/igt@i915_pm_rpm@modeset-non-lpsp-stress.html
* igt@kms_atomic@atomic_plane_damage:
- {shard-rkl}: [SKIP][128] ([i915#4098]) -> [PASS][129] +1 similar issue
[128]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-rkl-1/igt@kms_atomic@atomic_plane_damage.html
[129]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-rkl-6/igt@kms_atomic@atomic_plane_damage.html
* igt@kms_atomic@plane-primary-legacy:
- {shard-rkl}: [SKIP][130] ([i915#1845]) -> [PASS][131] +19 similar issues
[130]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-rkl-1/igt@kms_atomic@plane-primary-legacy.html
[131]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-rkl-6/igt@kms_atomic@plane-primary-legacy.html
* igt@kms_big_fb@y-tiled-16bpp-rotate-0:
- shard-glk: [DMESG-WARN][132] ([i915#118]) -> [PASS][133]
[132]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-glk4/igt@kms_big_fb@y-tiled-16bpp-rotate-0.html
[133]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-glk1/igt@kms_big_fb@y-tiled-16bpp-rotate-0.html
* igt@kms_ccs@pipe-a-bad-pixel-format-y_tiled_gen12_rc_ccs:
- {shard-rkl}: [SKIP][134] ([i915#1845] / [i915#4098]) -> [PASS][135] +2 similar issues
[134]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-rkl-5/igt@kms_ccs@pipe-a-bad-pixel-format-y_tiled_gen12_rc_ccs.html
[135]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-rkl-6/igt@kms_ccs@pipe-a-bad-pixel-format-y_tiled_gen12_rc_ccs.html
* igt@kms_color@pipe-a-ctm-red-to-blue:
- {shard-rkl}: [SKIP][136] ([i915#1149] / [i915#1849]) -> [PASS][137]
[136]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-rkl-5/igt@kms_color@pipe-a-ctm-red-to-blue.html
[137]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/shard-rkl-6/igt@kms_color@pipe-a-ctm-red-to-blue.html
* igt@kms_color@pipe-b-gamma:
- {shard-rkl}: ([SKIP][138], [SKIP][139]) ([i915#1149] / [i915#1849] / [i915#4070] / [i915#4098]) -> [PASS][140]
[138]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_11306/shard-rkl-1/igt@kms_color@pipe-b-gamma.html
[139]: https://intel-gfx-ci.01.org/
== Logs ==
For more details see: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_22453/index.html
[-- Attachment #2: Type: text/html, Size: 33588 bytes --]
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [PATCH v9] drm/amdgpu: add drm buddy support to amdgpu
2022-03-01 20:38 ` [Intel-gfx] " Arunpravin
(?)
@ 2022-03-03 12:04 ` Christian König
-1 siblings, 0 replies; 11+ messages in thread
From: Christian König @ 2022-03-03 12:04 UTC (permalink / raw)
To: Arunpravin, amd-gfx, dri-devel, intel-gfx
Cc: alexander.deucher, tzimmermann, matthew.auld, christian.koenig
Am 01.03.22 um 21:38 schrieb Arunpravin:
> - Remove drm_mm references and replace with drm buddy functionalities
> - Add res cursor support for drm buddy
>
> v2(Matthew Auld):
> - replace spinlock with mutex as we call kmem_cache_zalloc
> (..., GFP_KERNEL) in drm_buddy_alloc() function
>
> - lock drm_buddy_block_trim() function as it calls
> mark_free/mark_split are all globally visible
>
> v3(Matthew Auld):
> - remove trim method error handling as we address the failure case
> at drm_buddy_block_trim() function
>
> v4:
> - fix warnings reported by kernel test robot <lkp@intel.com>
>
> v5:
> - fix merge conflict issue
>
> v6:
> - fix warnings reported by kernel test robot <lkp@intel.com>
>
> v7:
> - remove DRM_BUDDY_RANGE_ALLOCATION flag usage
>
> v8:
> - keep DRM_BUDDY_RANGE_ALLOCATION flag usage
> - resolve conflicts created by drm/amdgpu: remove VRAM accounting v2
>
> v9(Christian):
> - rename label name as fallback
> - move struct amdgpu_vram_mgr to amdgpu_vram_mgr.h
> - remove unnecessary flags from struct amdgpu_vram_reservation
> - rewrite block NULL check condition
> - change else style as per coding standard
> - rewrite the node max size
> - add a helper function to fetch the first entry from the list
>
> Signed-off-by: Arunpravin <Arunpravin.PaneerSelvam@amd.com>
> ---
> drivers/gpu/drm/Kconfig | 1 +
> .../gpu/drm/amd/amdgpu/amdgpu_res_cursor.h | 97 +++++--
> drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h | 10 +-
> drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c | 249 ++++++++++--------
> drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h | 68 +++++
> 5 files changed, 289 insertions(+), 136 deletions(-)
> create mode 100644 drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h
>
> diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig
> index f1422bee3dcc..5133c3f028ab 100644
> --- a/drivers/gpu/drm/Kconfig
> +++ b/drivers/gpu/drm/Kconfig
> @@ -280,6 +280,7 @@ config DRM_AMDGPU
> select HWMON
> select BACKLIGHT_CLASS_DEVICE
> select INTERVAL_TREE
> + select DRM_BUDDY
> help
> Choose this option if you have a recent AMD Radeon graphics card.
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h
> index acfa207cf970..864c609ba00b 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h
> @@ -30,12 +30,15 @@
> #include <drm/ttm/ttm_resource.h>
> #include <drm/ttm/ttm_range_manager.h>
>
> +#include "amdgpu_vram_mgr.h"
> +
> /* state back for walking over vram_mgr and gtt_mgr allocations */
> struct amdgpu_res_cursor {
> uint64_t start;
> uint64_t size;
> uint64_t remaining;
> - struct drm_mm_node *node;
> + void *node;
> + uint32_t mem_type;
> };
>
> /**
> @@ -52,27 +55,63 @@ static inline void amdgpu_res_first(struct ttm_resource *res,
> uint64_t start, uint64_t size,
> struct amdgpu_res_cursor *cur)
> {
> + struct drm_buddy_block *block;
> + struct list_head *head, *next;
> struct drm_mm_node *node;
>
> - if (!res || res->mem_type == TTM_PL_SYSTEM) {
> - cur->start = start;
> - cur->size = size;
> - cur->remaining = size;
> - cur->node = NULL;
> - WARN_ON(res && start + size > res->num_pages << PAGE_SHIFT);
> - return;
> - }
> + if (!res)
> + goto fallback;
>
> BUG_ON(start + size > res->num_pages << PAGE_SHIFT);
>
> - node = to_ttm_range_mgr_node(res)->mm_nodes;
> - while (start >= node->size << PAGE_SHIFT)
> - start -= node++->size << PAGE_SHIFT;
> + cur->mem_type = res->mem_type;
> +
> + switch (cur->mem_type) {
> + case TTM_PL_VRAM:
> + head = &to_amdgpu_vram_mgr_node(res)->blocks;
> +
> + block = list_first_entry_or_null(head,
> + struct drm_buddy_block,
> + link);
> + if (!block)
> + goto fallback;
> +
> + while (start >= amdgpu_node_size(block)) {
> + start -= amdgpu_node_size(block);
> +
> + next = block->link.next;
> + if (next != head)
> + block = list_entry(next, struct drm_buddy_block, link);
> + }
> +
> + cur->start = amdgpu_node_start(block) + start;
> + cur->size = min(amdgpu_node_size(block) - start, size);
> + cur->remaining = size;
> + cur->node = block;
> + break;
> + case TTM_PL_TT:
> + node = to_ttm_range_mgr_node(res)->mm_nodes;
> + while (start >= node->size << PAGE_SHIFT)
> + start -= node++->size << PAGE_SHIFT;
> +
> + cur->start = (node->start << PAGE_SHIFT) + start;
> + cur->size = min((node->size << PAGE_SHIFT) - start, size);
> + cur->remaining = size;
> + cur->node = node;
> + break;
> + default:
> + goto fallback;
> + }
>
> - cur->start = (node->start << PAGE_SHIFT) + start;
> - cur->size = min((node->size << PAGE_SHIFT) - start, size);
> + return;
> +
> +fallback:
> + cur->start = start;
> + cur->size = size;
> cur->remaining = size;
> - cur->node = node;
> + cur->node = NULL;
> + WARN_ON(res && start + size > res->num_pages << PAGE_SHIFT);
> + return;
> }
>
> /**
> @@ -85,7 +124,9 @@ static inline void amdgpu_res_first(struct ttm_resource *res,
> */
> static inline void amdgpu_res_next(struct amdgpu_res_cursor *cur, uint64_t size)
> {
> - struct drm_mm_node *node = cur->node;
> + struct drm_buddy_block *block;
> + struct drm_mm_node *node;
> + struct list_head *next;
>
> BUG_ON(size > cur->remaining);
>
> @@ -99,9 +140,27 @@ static inline void amdgpu_res_next(struct amdgpu_res_cursor *cur, uint64_t size)
> return;
> }
>
> - cur->node = ++node;
> - cur->start = node->start << PAGE_SHIFT;
> - cur->size = min(node->size << PAGE_SHIFT, cur->remaining);
> + switch (cur->mem_type) {
> + case TTM_PL_VRAM:
> + block = cur->node;
> +
> + next = block->link.next;
> + block = list_entry(next, struct drm_buddy_block, link);
> +
> + cur->node = block;
> + cur->start = amdgpu_node_start(block);
> + cur->size = min(amdgpu_node_size(block), cur->remaining);
> + break;
> + case TTM_PL_TT:
> + node = cur->node;
> +
> + cur->node = ++node;
> + cur->start = node->start << PAGE_SHIFT;
> + cur->size = min(node->size << PAGE_SHIFT, cur->remaining);
> + break;
> + default:
> + return;
> + }
> }
>
> #endif
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
> index 9120ae80ef52..6a70818039dd 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
> @@ -26,6 +26,7 @@
>
> #include <linux/dma-direction.h>
> #include <drm/gpu_scheduler.h>
> +#include "amdgpu_vram_mgr.h"
> #include "amdgpu.h"
>
> #define AMDGPU_PL_GDS (TTM_PL_PRIV + 0)
> @@ -38,15 +39,6 @@
>
> #define AMDGPU_POISON 0xd0bed0be
>
> -struct amdgpu_vram_mgr {
> - struct ttm_resource_manager manager;
> - struct drm_mm mm;
> - spinlock_t lock;
> - struct list_head reservations_pending;
> - struct list_head reserved_pages;
> - atomic64_t vis_usage;
> -};
> -
> struct amdgpu_gtt_mgr {
> struct ttm_resource_manager manager;
> struct drm_mm mm;
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
> index 0a7611648573..cdf66802e6bc 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
> @@ -32,8 +32,10 @@
> #include "atom.h"
>
> struct amdgpu_vram_reservation {
> + u64 start;
> + u64 size;
> + struct list_head block;
> struct list_head node;
> - struct drm_mm_node mm_node;
> };
>
> static inline struct amdgpu_vram_mgr *
> @@ -194,10 +196,10 @@ const struct attribute_group amdgpu_vram_mgr_attr_group = {
> * Calculate how many bytes of the MM node are inside visible VRAM
> */
> static u64 amdgpu_vram_mgr_vis_size(struct amdgpu_device *adev,
> - struct drm_mm_node *node)
> + struct drm_buddy_block *block)
> {
> - uint64_t start = node->start << PAGE_SHIFT;
> - uint64_t end = (node->size + node->start) << PAGE_SHIFT;
> + u64 start = amdgpu_node_start(block);
> + u64 end = start + amdgpu_node_size(block);
>
> if (start >= adev->gmc.visible_vram_size)
> return 0;
> @@ -218,9 +220,9 @@ u64 amdgpu_vram_mgr_bo_visible_size(struct amdgpu_bo *bo)
> {
> struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
> struct ttm_resource *res = bo->tbo.resource;
> - unsigned pages = res->num_pages;
> - struct drm_mm_node *mm;
> - u64 usage;
> + struct amdgpu_vram_mgr_node *node = to_amdgpu_vram_mgr_node(res);
> + struct drm_buddy_block *block;
> + u64 usage = 0;
>
> if (amdgpu_gmc_vram_full_visible(&adev->gmc))
> return amdgpu_bo_size(bo);
> @@ -228,9 +230,8 @@ u64 amdgpu_vram_mgr_bo_visible_size(struct amdgpu_bo *bo)
> if (res->start >= adev->gmc.visible_vram_size >> PAGE_SHIFT)
> return 0;
>
> - mm = &container_of(res, struct ttm_range_mgr_node, base)->mm_nodes[0];
> - for (usage = 0; pages; pages -= mm->size, mm++)
> - usage += amdgpu_vram_mgr_vis_size(adev, mm);
> + list_for_each_entry(block, &node->blocks, link)
> + usage += amdgpu_vram_mgr_vis_size(adev, block);
>
> return usage;
> }
> @@ -240,21 +241,28 @@ static void amdgpu_vram_mgr_do_reserve(struct ttm_resource_manager *man)
> {
> struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
> struct amdgpu_device *adev = to_amdgpu_device(mgr);
> - struct drm_mm *mm = &mgr->mm;
> + struct drm_buddy *mm = &mgr->mm;
> struct amdgpu_vram_reservation *rsv, *temp;
> + struct drm_buddy_block *block;
> uint64_t vis_usage;
>
> list_for_each_entry_safe(rsv, temp, &mgr->reservations_pending, node) {
> - if (drm_mm_reserve_node(mm, &rsv->mm_node))
> + if (drm_buddy_alloc_blocks(mm, rsv->start, rsv->start + rsv->size,
> + rsv->size, mm->chunk_size, &rsv->block,
> + DRM_BUDDY_RANGE_ALLOCATION))
> + continue;
> +
> + block = amdgpu_get_node(&rsv->block);
> + if (!block)
> continue;
>
> dev_dbg(adev->dev, "Reservation 0x%llx - %lld, Succeeded\n",
> - rsv->mm_node.start, rsv->mm_node.size);
> + rsv->start, rsv->size);
>
> - vis_usage = amdgpu_vram_mgr_vis_size(adev, &rsv->mm_node);
> + vis_usage = amdgpu_vram_mgr_vis_size(adev, block);
> atomic64_add(vis_usage, &mgr->vis_usage);
> spin_lock(&man->bdev->lru_lock);
> - man->usage += rsv->mm_node.size << PAGE_SHIFT;
> + man->usage += rsv->size;
> spin_unlock(&man->bdev->lru_lock);
> list_move(&rsv->node, &mgr->reserved_pages);
> }
> @@ -279,13 +287,15 @@ int amdgpu_vram_mgr_reserve_range(struct amdgpu_vram_mgr *mgr,
> return -ENOMEM;
>
> INIT_LIST_HEAD(&rsv->node);
> - rsv->mm_node.start = start >> PAGE_SHIFT;
> - rsv->mm_node.size = size >> PAGE_SHIFT;
> + INIT_LIST_HEAD(&rsv->block);
> +
> + rsv->start = start;
> + rsv->size = size;
>
> - spin_lock(&mgr->lock);
> + mutex_lock(&mgr->lock);
> list_add_tail(&rsv->node, &mgr->reservations_pending);
> amdgpu_vram_mgr_do_reserve(&mgr->manager);
> - spin_unlock(&mgr->lock);
> + mutex_unlock(&mgr->lock);
>
> return 0;
> }
> @@ -307,19 +317,19 @@ int amdgpu_vram_mgr_query_page_status(struct amdgpu_vram_mgr *mgr,
> struct amdgpu_vram_reservation *rsv;
> int ret;
>
> - spin_lock(&mgr->lock);
> + mutex_lock(&mgr->lock);
>
> list_for_each_entry(rsv, &mgr->reservations_pending, node) {
> - if ((rsv->mm_node.start <= start) &&
> - (start < (rsv->mm_node.start + rsv->mm_node.size))) {
> + if (rsv->start <= start &&
> + (start < (rsv->start + rsv->size))) {
> ret = -EBUSY;
> goto out;
> }
> }
>
> list_for_each_entry(rsv, &mgr->reserved_pages, node) {
> - if ((rsv->mm_node.start <= start) &&
> - (start < (rsv->mm_node.start + rsv->mm_node.size))) {
> + if (rsv->start <= start &&
> + (start < (rsv->start + rsv->size))) {
> ret = 0;
> goto out;
> }
> @@ -327,32 +337,10 @@ int amdgpu_vram_mgr_query_page_status(struct amdgpu_vram_mgr *mgr,
>
> ret = -ENOENT;
> out:
> - spin_unlock(&mgr->lock);
> + mutex_unlock(&mgr->lock);
> return ret;
> }
>
> -/**
> - * amdgpu_vram_mgr_virt_start - update virtual start address
> - *
> - * @mem: ttm_resource to update
> - * @node: just allocated node
> - *
> - * Calculate a virtual BO start address to easily check if everything is CPU
> - * accessible.
> - */
> -static void amdgpu_vram_mgr_virt_start(struct ttm_resource *mem,
> - struct drm_mm_node *node)
> -{
> - unsigned long start;
> -
> - start = node->start + node->size;
> - if (start > mem->num_pages)
> - start -= mem->num_pages;
> - else
> - start = 0;
> - mem->start = max(mem->start, start);
> -}
> -
> /**
> * amdgpu_vram_mgr_new - allocate new ranges
> *
> @@ -368,13 +356,13 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
> const struct ttm_place *place,
> struct ttm_resource **res)
> {
> - unsigned long lpfn, num_nodes, pages_per_node, pages_left, pages;
> + unsigned long lpfn, pages_per_node, pages_left, pages;
> + u64 vis_usage = 0, mem_bytes, max_bytes, min_page_size;
> struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
> struct amdgpu_device *adev = to_amdgpu_device(mgr);
> - uint64_t vis_usage = 0, mem_bytes, max_bytes;
> - struct ttm_range_mgr_node *node;
> - struct drm_mm *mm = &mgr->mm;
> - enum drm_mm_insert_mode mode;
> + struct amdgpu_vram_mgr_node *node;
> + struct drm_buddy *mm = &mgr->mm;
> + struct drm_buddy_block *block;
> unsigned i;
> int r;
>
> @@ -389,7 +377,6 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
> mem_bytes = tbo->base.size;
> if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
> pages_per_node = ~0ul;
> - num_nodes = 1;
> } else {
> #ifdef CONFIG_TRANSPARENT_HUGEPAGE
> pages_per_node = HPAGE_PMD_NR;
> @@ -399,11 +386,9 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
> #endif
> pages_per_node = max_t(uint32_t, pages_per_node,
> tbo->page_alignment);
> - num_nodes = DIV_ROUND_UP_ULL(PFN_UP(mem_bytes), pages_per_node);
> }
>
> - node = kvmalloc(struct_size(node, mm_nodes, num_nodes),
> - GFP_KERNEL | __GFP_ZERO);
> + node = kzalloc(sizeof(*node), GFP_KERNEL);
> if (!node)
> return -ENOMEM;
>
> @@ -415,48 +400,83 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
> goto error_fini;
> }
>
> - mode = DRM_MM_INSERT_BEST;
> + INIT_LIST_HEAD(&node->blocks);
> +
> if (place->flags & TTM_PL_FLAG_TOPDOWN)
> - mode = DRM_MM_INSERT_HIGH;
> + node->flags |= DRM_BUDDY_TOPDOWN_ALLOCATION;
>
> - pages_left = node->base.num_pages;
> + if (place->fpfn || lpfn != man->size >> PAGE_SHIFT)
> + /* Allocate blocks in desired range */
> + node->flags |= DRM_BUDDY_RANGE_ALLOCATION;
>
> - /* Limit maximum size to 2GB due to SG table limitations */
> - pages = min(pages_left, 2UL << (30 - PAGE_SHIFT));
> + BUG_ON(!node->base.num_pages);
> + pages_left = node->base.num_pages;
>
> i = 0;
> - spin_lock(&mgr->lock);
> while (pages_left) {
> - uint32_t alignment = tbo->page_alignment;
> + if (tbo->page_alignment)
> + min_page_size = tbo->page_alignment << PAGE_SHIFT;
> + else
> + min_page_size = mgr->default_page_size;
> +
> + /* Limit maximum size to 2GB due to SG table limitations */
> + pages = min(pages_left, 2UL << (30 - PAGE_SHIFT));
>
> if (pages >= pages_per_node)
> - alignment = pages_per_node;
> -
> - r = drm_mm_insert_node_in_range(mm, &node->mm_nodes[i], pages,
> - alignment, 0, place->fpfn,
> - lpfn, mode);
> - if (unlikely(r)) {
> - if (pages > pages_per_node) {
> - if (is_power_of_2(pages))
> - pages = pages / 2;
> - else
> - pages = rounddown_pow_of_two(pages);
> - continue;
> - }
> - goto error_free;
> + min_page_size = pages_per_node << PAGE_SHIFT;
> +
> + if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
> + pages = roundup_pow_of_two(pages);
> + min_page_size = pages << PAGE_SHIFT;
> +
> + if (pages > lpfn)
> + lpfn = pages;
> }
>
> - vis_usage += amdgpu_vram_mgr_vis_size(adev, &node->mm_nodes[i]);
> - amdgpu_vram_mgr_virt_start(&node->base, &node->mm_nodes[i]);
> - pages_left -= pages;
> + BUG_ON(min_page_size < mm->chunk_size);
> +
> + mutex_lock(&mgr->lock);
> + r = drm_buddy_alloc_blocks(mm, (u64)place->fpfn << PAGE_SHIFT,
> + (u64)lpfn << PAGE_SHIFT,
> + (u64)pages << PAGE_SHIFT,
> + min_page_size,
> + &node->blocks,
> + node->flags);
> + mutex_unlock(&mgr->lock);
> + if (unlikely(r))
> + goto error_free_blocks;
> +
> ++i;
>
> if (pages > pages_left)
> - pages = pages_left;
> + pages_left = 0;
> + else
> + pages_left -= pages;
> }
> - spin_unlock(&mgr->lock);
>
> - if (i == 1)
> + /* Free unused pages for contiguous allocation */
> + if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
> + u64 actual_size = (u64)node->base.num_pages << PAGE_SHIFT;
> +
> + mutex_lock(&mgr->lock);
> + drm_buddy_block_trim(mm,
> + actual_size,
> + &node->blocks);
> + mutex_unlock(&mgr->lock);
> + }
> +
> + list_for_each_entry(block, &node->blocks, link)
> + vis_usage += amdgpu_vram_mgr_vis_size(adev, block);
> +
> + block = amdgpu_get_node(&node->blocks);
> + if (!block) {
> + r = -EINVAL;
> + goto error_fini;
> + }
> +
> + node->base.start = amdgpu_node_start(block) >> PAGE_SHIFT;
> +
> + if (i == 1 && place->flags & TTM_PL_FLAG_CONTIGUOUS)
> node->base.placement |= TTM_PL_FLAG_CONTIGUOUS;
>
> if (adev->gmc.xgmi.connected_to_cpu)
> @@ -468,13 +488,13 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
> *res = &node->base;
> return 0;
>
> -error_free:
> - while (i--)
> - drm_mm_remove_node(&node->mm_nodes[i]);
> - spin_unlock(&mgr->lock);
> +error_free_blocks:
> + mutex_lock(&mgr->lock);
> + drm_buddy_free_list(mm, &node->blocks);
> + mutex_unlock(&mgr->lock);
> error_fini:
> ttm_resource_fini(man, &node->base);
> - kvfree(node);
> + kfree(node);
>
> return r;
> }
> @@ -490,27 +510,26 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
> static void amdgpu_vram_mgr_del(struct ttm_resource_manager *man,
> struct ttm_resource *res)
> {
> - struct ttm_range_mgr_node *node = to_ttm_range_mgr_node(res);
> + struct amdgpu_vram_mgr_node *node = to_amdgpu_vram_mgr_node(res);
> struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
> struct amdgpu_device *adev = to_amdgpu_device(mgr);
> + struct drm_buddy *mm = &mgr->mm;
> + struct drm_buddy_block *block;
> uint64_t vis_usage = 0;
> - unsigned i, pages;
>
> - spin_lock(&mgr->lock);
> - for (i = 0, pages = res->num_pages; pages;
> - pages -= node->mm_nodes[i].size, ++i) {
> - struct drm_mm_node *mm = &node->mm_nodes[i];
> + mutex_lock(&mgr->lock);
> + list_for_each_entry(block, &node->blocks, link)
> + vis_usage += amdgpu_vram_mgr_vis_size(adev, block);
>
> - drm_mm_remove_node(mm);
> - vis_usage += amdgpu_vram_mgr_vis_size(adev, mm);
> - }
> amdgpu_vram_mgr_do_reserve(man);
> - spin_unlock(&mgr->lock);
> +
> + drm_buddy_free_list(mm, &node->blocks);
> + mutex_unlock(&mgr->lock);
>
> atomic64_sub(vis_usage, &mgr->vis_usage);
>
> ttm_resource_fini(man, res);
> - kvfree(node);
> + kfree(node);
> }
>
> /**
> @@ -648,13 +667,22 @@ static void amdgpu_vram_mgr_debug(struct ttm_resource_manager *man,
> struct drm_printer *printer)
> {
> struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
> + struct drm_buddy *mm = &mgr->mm;
> + struct drm_buddy_block *block;
>
> drm_printf(printer, " vis usage:%llu\n",
> amdgpu_vram_mgr_vis_usage(mgr));
>
> - spin_lock(&mgr->lock);
> - drm_mm_print(&mgr->mm, printer);
> - spin_unlock(&mgr->lock);
> + mutex_lock(&mgr->lock);
> + drm_printf(printer, "default_page_size: %lluKiB\n",
> + mgr->default_page_size >> 10);
> +
> + drm_buddy_print(mm, printer);
> +
> + drm_printf(printer, "reserved:\n");
> + list_for_each_entry(block, &mgr->reserved_pages, link)
> + drm_buddy_block_print(mm, block, printer);
> + mutex_unlock(&mgr->lock);
> }
>
> static const struct ttm_resource_manager_func amdgpu_vram_mgr_func = {
> @@ -674,16 +702,21 @@ int amdgpu_vram_mgr_init(struct amdgpu_device *adev)
> {
> struct amdgpu_vram_mgr *mgr = &adev->mman.vram_mgr;
> struct ttm_resource_manager *man = &mgr->manager;
> + int err;
>
> ttm_resource_manager_init(man, &adev->mman.bdev,
> adev->gmc.real_vram_size);
>
> man->func = &amdgpu_vram_mgr_func;
>
> - drm_mm_init(&mgr->mm, 0, man->size >> PAGE_SHIFT);
> - spin_lock_init(&mgr->lock);
> + err = drm_buddy_init(&mgr->mm, man->size, PAGE_SIZE);
> + if (err)
> + return err;
> +
> + mutex_init(&mgr->lock);
> INIT_LIST_HEAD(&mgr->reservations_pending);
> INIT_LIST_HEAD(&mgr->reserved_pages);
> + mgr->default_page_size = PAGE_SIZE;
>
> ttm_set_driver_manager(&adev->mman.bdev, TTM_PL_VRAM, &mgr->manager);
> ttm_resource_manager_set_used(man, true);
> @@ -711,16 +744,16 @@ void amdgpu_vram_mgr_fini(struct amdgpu_device *adev)
> if (ret)
> return;
>
> - spin_lock(&mgr->lock);
> + mutex_lock(&mgr->lock);
> list_for_each_entry_safe(rsv, temp, &mgr->reservations_pending, node)
> kfree(rsv);
>
> list_for_each_entry_safe(rsv, temp, &mgr->reserved_pages, node) {
> - drm_mm_remove_node(&rsv->mm_node);
> + drm_buddy_free_list(&mgr->mm, &rsv->block);
> kfree(rsv);
> }
> - drm_mm_takedown(&mgr->mm);
> - spin_unlock(&mgr->lock);
> + drm_buddy_fini(&mgr->mm);
> + mutex_unlock(&mgr->lock);
>
> ttm_resource_manager_cleanup(man);
> ttm_set_driver_manager(&adev->mman.bdev, TTM_PL_VRAM, NULL);
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h
> new file mode 100644
> index 000000000000..e7d6bbf40c33
> --- /dev/null
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h
> @@ -0,0 +1,68 @@
> +/* SPDX-License-Identifier: MIT
> + * Copyright 2021 Advanced Micro Devices, Inc.
> + *
> + * Permission is hereby granted, free of charge, to any person obtaining a
> + * copy of this software and associated documentation files (the "Software"),
> + * to deal in the Software without restriction, including without limitation
> + * the rights to use, copy, modify, merge, publish, distribute, sublicense,
> + * and/or sell copies of the Software, and to permit persons to whom the
> + * Software is furnished to do so, subject to the following conditions:
> + *
> + * The above copyright notice and this permission notice shall be included in
> + * all copies or substantial portions of the Software.
> + *
> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
> + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
> + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
> + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
> + * OTHER DEALINGS IN THE SOFTWARE.
> + *
> + */
> +
> +#ifndef __AMDGPU_VRAM_MGR_H__
> +#define __AMDGPU_VRAM_MGR_H__
> +
> +#include <drm/drm_buddy.h>
> +
> +struct amdgpu_vram_mgr {
> + struct ttm_resource_manager manager;
> + struct drm_buddy mm;
> + /* protects access to buffer objects */
> + struct mutex lock;
> + struct list_head reservations_pending;
> + struct list_head reserved_pages;
> + atomic64_t vis_usage;
> + u64 default_page_size;
> +};
> +
> +struct amdgpu_vram_mgr_node {
> + struct ttm_resource base;
> + struct list_head blocks;
> + unsigned long flags;
> +};
> +
> +static inline u64 amdgpu_node_start(struct drm_buddy_block *block)
> +{
> + return drm_buddy_block_offset(block);
> +}
> +
> +static inline u64 amdgpu_node_size(struct drm_buddy_block *block)
> +{
> + return PAGE_SIZE << drm_buddy_block_order(block);
> +}
> +
> +static inline struct drm_buddy_block *
> +amdgpu_get_node(struct list_head *list)
That function name is way to generic. Somebody could mix it up with
something completely different and _get_ in a name usually means that
you are grabbing a reference (which is not the case here).
I suggest to call it something like amdgpu_vram_mgr_first_blcok() and
move it from the headers into amdgpu_vram_mgr.c because code outside of
that file shouldn't mess with this.
Apart from that the looks really good now.
Regards,
Christian.
> +{
> + return list_first_entry_or_null(list, struct drm_buddy_block, link);
> +}
> +
> +static inline struct amdgpu_vram_mgr_node *
> +to_amdgpu_vram_mgr_node(struct ttm_resource *res)
> +{
> + return container_of(res, struct amdgpu_vram_mgr_node, base);
> +}
> +
> +#endif
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [PATCH v9] drm/amdgpu: add drm buddy support to amdgpu
@ 2022-03-03 12:04 ` Christian König
0 siblings, 0 replies; 11+ messages in thread
From: Christian König @ 2022-03-03 12:04 UTC (permalink / raw)
To: Arunpravin, amd-gfx, dri-devel, intel-gfx
Cc: alexander.deucher, tzimmermann, matthew.auld, daniel, christian.koenig
Am 01.03.22 um 21:38 schrieb Arunpravin:
> - Remove drm_mm references and replace with drm buddy functionalities
> - Add res cursor support for drm buddy
>
> v2(Matthew Auld):
> - replace spinlock with mutex as we call kmem_cache_zalloc
> (..., GFP_KERNEL) in drm_buddy_alloc() function
>
> - lock drm_buddy_block_trim() function as it calls
> mark_free/mark_split are all globally visible
>
> v3(Matthew Auld):
> - remove trim method error handling as we address the failure case
> at drm_buddy_block_trim() function
>
> v4:
> - fix warnings reported by kernel test robot <lkp@intel.com>
>
> v5:
> - fix merge conflict issue
>
> v6:
> - fix warnings reported by kernel test robot <lkp@intel.com>
>
> v7:
> - remove DRM_BUDDY_RANGE_ALLOCATION flag usage
>
> v8:
> - keep DRM_BUDDY_RANGE_ALLOCATION flag usage
> - resolve conflicts created by drm/amdgpu: remove VRAM accounting v2
>
> v9(Christian):
> - rename label name as fallback
> - move struct amdgpu_vram_mgr to amdgpu_vram_mgr.h
> - remove unnecessary flags from struct amdgpu_vram_reservation
> - rewrite block NULL check condition
> - change else style as per coding standard
> - rewrite the node max size
> - add a helper function to fetch the first entry from the list
>
> Signed-off-by: Arunpravin <Arunpravin.PaneerSelvam@amd.com>
> ---
> drivers/gpu/drm/Kconfig | 1 +
> .../gpu/drm/amd/amdgpu/amdgpu_res_cursor.h | 97 +++++--
> drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h | 10 +-
> drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c | 249 ++++++++++--------
> drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h | 68 +++++
> 5 files changed, 289 insertions(+), 136 deletions(-)
> create mode 100644 drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h
>
> diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig
> index f1422bee3dcc..5133c3f028ab 100644
> --- a/drivers/gpu/drm/Kconfig
> +++ b/drivers/gpu/drm/Kconfig
> @@ -280,6 +280,7 @@ config DRM_AMDGPU
> select HWMON
> select BACKLIGHT_CLASS_DEVICE
> select INTERVAL_TREE
> + select DRM_BUDDY
> help
> Choose this option if you have a recent AMD Radeon graphics card.
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h
> index acfa207cf970..864c609ba00b 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h
> @@ -30,12 +30,15 @@
> #include <drm/ttm/ttm_resource.h>
> #include <drm/ttm/ttm_range_manager.h>
>
> +#include "amdgpu_vram_mgr.h"
> +
> /* state back for walking over vram_mgr and gtt_mgr allocations */
> struct amdgpu_res_cursor {
> uint64_t start;
> uint64_t size;
> uint64_t remaining;
> - struct drm_mm_node *node;
> + void *node;
> + uint32_t mem_type;
> };
>
> /**
> @@ -52,27 +55,63 @@ static inline void amdgpu_res_first(struct ttm_resource *res,
> uint64_t start, uint64_t size,
> struct amdgpu_res_cursor *cur)
> {
> + struct drm_buddy_block *block;
> + struct list_head *head, *next;
> struct drm_mm_node *node;
>
> - if (!res || res->mem_type == TTM_PL_SYSTEM) {
> - cur->start = start;
> - cur->size = size;
> - cur->remaining = size;
> - cur->node = NULL;
> - WARN_ON(res && start + size > res->num_pages << PAGE_SHIFT);
> - return;
> - }
> + if (!res)
> + goto fallback;
>
> BUG_ON(start + size > res->num_pages << PAGE_SHIFT);
>
> - node = to_ttm_range_mgr_node(res)->mm_nodes;
> - while (start >= node->size << PAGE_SHIFT)
> - start -= node++->size << PAGE_SHIFT;
> + cur->mem_type = res->mem_type;
> +
> + switch (cur->mem_type) {
> + case TTM_PL_VRAM:
> + head = &to_amdgpu_vram_mgr_node(res)->blocks;
> +
> + block = list_first_entry_or_null(head,
> + struct drm_buddy_block,
> + link);
> + if (!block)
> + goto fallback;
> +
> + while (start >= amdgpu_node_size(block)) {
> + start -= amdgpu_node_size(block);
> +
> + next = block->link.next;
> + if (next != head)
> + block = list_entry(next, struct drm_buddy_block, link);
> + }
> +
> + cur->start = amdgpu_node_start(block) + start;
> + cur->size = min(amdgpu_node_size(block) - start, size);
> + cur->remaining = size;
> + cur->node = block;
> + break;
> + case TTM_PL_TT:
> + node = to_ttm_range_mgr_node(res)->mm_nodes;
> + while (start >= node->size << PAGE_SHIFT)
> + start -= node++->size << PAGE_SHIFT;
> +
> + cur->start = (node->start << PAGE_SHIFT) + start;
> + cur->size = min((node->size << PAGE_SHIFT) - start, size);
> + cur->remaining = size;
> + cur->node = node;
> + break;
> + default:
> + goto fallback;
> + }
>
> - cur->start = (node->start << PAGE_SHIFT) + start;
> - cur->size = min((node->size << PAGE_SHIFT) - start, size);
> + return;
> +
> +fallback:
> + cur->start = start;
> + cur->size = size;
> cur->remaining = size;
> - cur->node = node;
> + cur->node = NULL;
> + WARN_ON(res && start + size > res->num_pages << PAGE_SHIFT);
> + return;
> }
>
> /**
> @@ -85,7 +124,9 @@ static inline void amdgpu_res_first(struct ttm_resource *res,
> */
> static inline void amdgpu_res_next(struct amdgpu_res_cursor *cur, uint64_t size)
> {
> - struct drm_mm_node *node = cur->node;
> + struct drm_buddy_block *block;
> + struct drm_mm_node *node;
> + struct list_head *next;
>
> BUG_ON(size > cur->remaining);
>
> @@ -99,9 +140,27 @@ static inline void amdgpu_res_next(struct amdgpu_res_cursor *cur, uint64_t size)
> return;
> }
>
> - cur->node = ++node;
> - cur->start = node->start << PAGE_SHIFT;
> - cur->size = min(node->size << PAGE_SHIFT, cur->remaining);
> + switch (cur->mem_type) {
> + case TTM_PL_VRAM:
> + block = cur->node;
> +
> + next = block->link.next;
> + block = list_entry(next, struct drm_buddy_block, link);
> +
> + cur->node = block;
> + cur->start = amdgpu_node_start(block);
> + cur->size = min(amdgpu_node_size(block), cur->remaining);
> + break;
> + case TTM_PL_TT:
> + node = cur->node;
> +
> + cur->node = ++node;
> + cur->start = node->start << PAGE_SHIFT;
> + cur->size = min(node->size << PAGE_SHIFT, cur->remaining);
> + break;
> + default:
> + return;
> + }
> }
>
> #endif
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
> index 9120ae80ef52..6a70818039dd 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
> @@ -26,6 +26,7 @@
>
> #include <linux/dma-direction.h>
> #include <drm/gpu_scheduler.h>
> +#include "amdgpu_vram_mgr.h"
> #include "amdgpu.h"
>
> #define AMDGPU_PL_GDS (TTM_PL_PRIV + 0)
> @@ -38,15 +39,6 @@
>
> #define AMDGPU_POISON 0xd0bed0be
>
> -struct amdgpu_vram_mgr {
> - struct ttm_resource_manager manager;
> - struct drm_mm mm;
> - spinlock_t lock;
> - struct list_head reservations_pending;
> - struct list_head reserved_pages;
> - atomic64_t vis_usage;
> -};
> -
> struct amdgpu_gtt_mgr {
> struct ttm_resource_manager manager;
> struct drm_mm mm;
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
> index 0a7611648573..cdf66802e6bc 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
> @@ -32,8 +32,10 @@
> #include "atom.h"
>
> struct amdgpu_vram_reservation {
> + u64 start;
> + u64 size;
> + struct list_head block;
> struct list_head node;
> - struct drm_mm_node mm_node;
> };
>
> static inline struct amdgpu_vram_mgr *
> @@ -194,10 +196,10 @@ const struct attribute_group amdgpu_vram_mgr_attr_group = {
> * Calculate how many bytes of the MM node are inside visible VRAM
> */
> static u64 amdgpu_vram_mgr_vis_size(struct amdgpu_device *adev,
> - struct drm_mm_node *node)
> + struct drm_buddy_block *block)
> {
> - uint64_t start = node->start << PAGE_SHIFT;
> - uint64_t end = (node->size + node->start) << PAGE_SHIFT;
> + u64 start = amdgpu_node_start(block);
> + u64 end = start + amdgpu_node_size(block);
>
> if (start >= adev->gmc.visible_vram_size)
> return 0;
> @@ -218,9 +220,9 @@ u64 amdgpu_vram_mgr_bo_visible_size(struct amdgpu_bo *bo)
> {
> struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
> struct ttm_resource *res = bo->tbo.resource;
> - unsigned pages = res->num_pages;
> - struct drm_mm_node *mm;
> - u64 usage;
> + struct amdgpu_vram_mgr_node *node = to_amdgpu_vram_mgr_node(res);
> + struct drm_buddy_block *block;
> + u64 usage = 0;
>
> if (amdgpu_gmc_vram_full_visible(&adev->gmc))
> return amdgpu_bo_size(bo);
> @@ -228,9 +230,8 @@ u64 amdgpu_vram_mgr_bo_visible_size(struct amdgpu_bo *bo)
> if (res->start >= adev->gmc.visible_vram_size >> PAGE_SHIFT)
> return 0;
>
> - mm = &container_of(res, struct ttm_range_mgr_node, base)->mm_nodes[0];
> - for (usage = 0; pages; pages -= mm->size, mm++)
> - usage += amdgpu_vram_mgr_vis_size(adev, mm);
> + list_for_each_entry(block, &node->blocks, link)
> + usage += amdgpu_vram_mgr_vis_size(adev, block);
>
> return usage;
> }
> @@ -240,21 +241,28 @@ static void amdgpu_vram_mgr_do_reserve(struct ttm_resource_manager *man)
> {
> struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
> struct amdgpu_device *adev = to_amdgpu_device(mgr);
> - struct drm_mm *mm = &mgr->mm;
> + struct drm_buddy *mm = &mgr->mm;
> struct amdgpu_vram_reservation *rsv, *temp;
> + struct drm_buddy_block *block;
> uint64_t vis_usage;
>
> list_for_each_entry_safe(rsv, temp, &mgr->reservations_pending, node) {
> - if (drm_mm_reserve_node(mm, &rsv->mm_node))
> + if (drm_buddy_alloc_blocks(mm, rsv->start, rsv->start + rsv->size,
> + rsv->size, mm->chunk_size, &rsv->block,
> + DRM_BUDDY_RANGE_ALLOCATION))
> + continue;
> +
> + block = amdgpu_get_node(&rsv->block);
> + if (!block)
> continue;
>
> dev_dbg(adev->dev, "Reservation 0x%llx - %lld, Succeeded\n",
> - rsv->mm_node.start, rsv->mm_node.size);
> + rsv->start, rsv->size);
>
> - vis_usage = amdgpu_vram_mgr_vis_size(adev, &rsv->mm_node);
> + vis_usage = amdgpu_vram_mgr_vis_size(adev, block);
> atomic64_add(vis_usage, &mgr->vis_usage);
> spin_lock(&man->bdev->lru_lock);
> - man->usage += rsv->mm_node.size << PAGE_SHIFT;
> + man->usage += rsv->size;
> spin_unlock(&man->bdev->lru_lock);
> list_move(&rsv->node, &mgr->reserved_pages);
> }
> @@ -279,13 +287,15 @@ int amdgpu_vram_mgr_reserve_range(struct amdgpu_vram_mgr *mgr,
> return -ENOMEM;
>
> INIT_LIST_HEAD(&rsv->node);
> - rsv->mm_node.start = start >> PAGE_SHIFT;
> - rsv->mm_node.size = size >> PAGE_SHIFT;
> + INIT_LIST_HEAD(&rsv->block);
> +
> + rsv->start = start;
> + rsv->size = size;
>
> - spin_lock(&mgr->lock);
> + mutex_lock(&mgr->lock);
> list_add_tail(&rsv->node, &mgr->reservations_pending);
> amdgpu_vram_mgr_do_reserve(&mgr->manager);
> - spin_unlock(&mgr->lock);
> + mutex_unlock(&mgr->lock);
>
> return 0;
> }
> @@ -307,19 +317,19 @@ int amdgpu_vram_mgr_query_page_status(struct amdgpu_vram_mgr *mgr,
> struct amdgpu_vram_reservation *rsv;
> int ret;
>
> - spin_lock(&mgr->lock);
> + mutex_lock(&mgr->lock);
>
> list_for_each_entry(rsv, &mgr->reservations_pending, node) {
> - if ((rsv->mm_node.start <= start) &&
> - (start < (rsv->mm_node.start + rsv->mm_node.size))) {
> + if (rsv->start <= start &&
> + (start < (rsv->start + rsv->size))) {
> ret = -EBUSY;
> goto out;
> }
> }
>
> list_for_each_entry(rsv, &mgr->reserved_pages, node) {
> - if ((rsv->mm_node.start <= start) &&
> - (start < (rsv->mm_node.start + rsv->mm_node.size))) {
> + if (rsv->start <= start &&
> + (start < (rsv->start + rsv->size))) {
> ret = 0;
> goto out;
> }
> @@ -327,32 +337,10 @@ int amdgpu_vram_mgr_query_page_status(struct amdgpu_vram_mgr *mgr,
>
> ret = -ENOENT;
> out:
> - spin_unlock(&mgr->lock);
> + mutex_unlock(&mgr->lock);
> return ret;
> }
>
> -/**
> - * amdgpu_vram_mgr_virt_start - update virtual start address
> - *
> - * @mem: ttm_resource to update
> - * @node: just allocated node
> - *
> - * Calculate a virtual BO start address to easily check if everything is CPU
> - * accessible.
> - */
> -static void amdgpu_vram_mgr_virt_start(struct ttm_resource *mem,
> - struct drm_mm_node *node)
> -{
> - unsigned long start;
> -
> - start = node->start + node->size;
> - if (start > mem->num_pages)
> - start -= mem->num_pages;
> - else
> - start = 0;
> - mem->start = max(mem->start, start);
> -}
> -
> /**
> * amdgpu_vram_mgr_new - allocate new ranges
> *
> @@ -368,13 +356,13 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
> const struct ttm_place *place,
> struct ttm_resource **res)
> {
> - unsigned long lpfn, num_nodes, pages_per_node, pages_left, pages;
> + unsigned long lpfn, pages_per_node, pages_left, pages;
> + u64 vis_usage = 0, mem_bytes, max_bytes, min_page_size;
> struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
> struct amdgpu_device *adev = to_amdgpu_device(mgr);
> - uint64_t vis_usage = 0, mem_bytes, max_bytes;
> - struct ttm_range_mgr_node *node;
> - struct drm_mm *mm = &mgr->mm;
> - enum drm_mm_insert_mode mode;
> + struct amdgpu_vram_mgr_node *node;
> + struct drm_buddy *mm = &mgr->mm;
> + struct drm_buddy_block *block;
> unsigned i;
> int r;
>
> @@ -389,7 +377,6 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
> mem_bytes = tbo->base.size;
> if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
> pages_per_node = ~0ul;
> - num_nodes = 1;
> } else {
> #ifdef CONFIG_TRANSPARENT_HUGEPAGE
> pages_per_node = HPAGE_PMD_NR;
> @@ -399,11 +386,9 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
> #endif
> pages_per_node = max_t(uint32_t, pages_per_node,
> tbo->page_alignment);
> - num_nodes = DIV_ROUND_UP_ULL(PFN_UP(mem_bytes), pages_per_node);
> }
>
> - node = kvmalloc(struct_size(node, mm_nodes, num_nodes),
> - GFP_KERNEL | __GFP_ZERO);
> + node = kzalloc(sizeof(*node), GFP_KERNEL);
> if (!node)
> return -ENOMEM;
>
> @@ -415,48 +400,83 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
> goto error_fini;
> }
>
> - mode = DRM_MM_INSERT_BEST;
> + INIT_LIST_HEAD(&node->blocks);
> +
> if (place->flags & TTM_PL_FLAG_TOPDOWN)
> - mode = DRM_MM_INSERT_HIGH;
> + node->flags |= DRM_BUDDY_TOPDOWN_ALLOCATION;
>
> - pages_left = node->base.num_pages;
> + if (place->fpfn || lpfn != man->size >> PAGE_SHIFT)
> + /* Allocate blocks in desired range */
> + node->flags |= DRM_BUDDY_RANGE_ALLOCATION;
>
> - /* Limit maximum size to 2GB due to SG table limitations */
> - pages = min(pages_left, 2UL << (30 - PAGE_SHIFT));
> + BUG_ON(!node->base.num_pages);
> + pages_left = node->base.num_pages;
>
> i = 0;
> - spin_lock(&mgr->lock);
> while (pages_left) {
> - uint32_t alignment = tbo->page_alignment;
> + if (tbo->page_alignment)
> + min_page_size = tbo->page_alignment << PAGE_SHIFT;
> + else
> + min_page_size = mgr->default_page_size;
> +
> + /* Limit maximum size to 2GB due to SG table limitations */
> + pages = min(pages_left, 2UL << (30 - PAGE_SHIFT));
>
> if (pages >= pages_per_node)
> - alignment = pages_per_node;
> -
> - r = drm_mm_insert_node_in_range(mm, &node->mm_nodes[i], pages,
> - alignment, 0, place->fpfn,
> - lpfn, mode);
> - if (unlikely(r)) {
> - if (pages > pages_per_node) {
> - if (is_power_of_2(pages))
> - pages = pages / 2;
> - else
> - pages = rounddown_pow_of_two(pages);
> - continue;
> - }
> - goto error_free;
> + min_page_size = pages_per_node << PAGE_SHIFT;
> +
> + if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
> + pages = roundup_pow_of_two(pages);
> + min_page_size = pages << PAGE_SHIFT;
> +
> + if (pages > lpfn)
> + lpfn = pages;
> }
>
> - vis_usage += amdgpu_vram_mgr_vis_size(adev, &node->mm_nodes[i]);
> - amdgpu_vram_mgr_virt_start(&node->base, &node->mm_nodes[i]);
> - pages_left -= pages;
> + BUG_ON(min_page_size < mm->chunk_size);
> +
> + mutex_lock(&mgr->lock);
> + r = drm_buddy_alloc_blocks(mm, (u64)place->fpfn << PAGE_SHIFT,
> + (u64)lpfn << PAGE_SHIFT,
> + (u64)pages << PAGE_SHIFT,
> + min_page_size,
> + &node->blocks,
> + node->flags);
> + mutex_unlock(&mgr->lock);
> + if (unlikely(r))
> + goto error_free_blocks;
> +
> ++i;
>
> if (pages > pages_left)
> - pages = pages_left;
> + pages_left = 0;
> + else
> + pages_left -= pages;
> }
> - spin_unlock(&mgr->lock);
>
> - if (i == 1)
> + /* Free unused pages for contiguous allocation */
> + if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
> + u64 actual_size = (u64)node->base.num_pages << PAGE_SHIFT;
> +
> + mutex_lock(&mgr->lock);
> + drm_buddy_block_trim(mm,
> + actual_size,
> + &node->blocks);
> + mutex_unlock(&mgr->lock);
> + }
> +
> + list_for_each_entry(block, &node->blocks, link)
> + vis_usage += amdgpu_vram_mgr_vis_size(adev, block);
> +
> + block = amdgpu_get_node(&node->blocks);
> + if (!block) {
> + r = -EINVAL;
> + goto error_fini;
> + }
> +
> + node->base.start = amdgpu_node_start(block) >> PAGE_SHIFT;
> +
> + if (i == 1 && place->flags & TTM_PL_FLAG_CONTIGUOUS)
> node->base.placement |= TTM_PL_FLAG_CONTIGUOUS;
>
> if (adev->gmc.xgmi.connected_to_cpu)
> @@ -468,13 +488,13 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
> *res = &node->base;
> return 0;
>
> -error_free:
> - while (i--)
> - drm_mm_remove_node(&node->mm_nodes[i]);
> - spin_unlock(&mgr->lock);
> +error_free_blocks:
> + mutex_lock(&mgr->lock);
> + drm_buddy_free_list(mm, &node->blocks);
> + mutex_unlock(&mgr->lock);
> error_fini:
> ttm_resource_fini(man, &node->base);
> - kvfree(node);
> + kfree(node);
>
> return r;
> }
> @@ -490,27 +510,26 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
> static void amdgpu_vram_mgr_del(struct ttm_resource_manager *man,
> struct ttm_resource *res)
> {
> - struct ttm_range_mgr_node *node = to_ttm_range_mgr_node(res);
> + struct amdgpu_vram_mgr_node *node = to_amdgpu_vram_mgr_node(res);
> struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
> struct amdgpu_device *adev = to_amdgpu_device(mgr);
> + struct drm_buddy *mm = &mgr->mm;
> + struct drm_buddy_block *block;
> uint64_t vis_usage = 0;
> - unsigned i, pages;
>
> - spin_lock(&mgr->lock);
> - for (i = 0, pages = res->num_pages; pages;
> - pages -= node->mm_nodes[i].size, ++i) {
> - struct drm_mm_node *mm = &node->mm_nodes[i];
> + mutex_lock(&mgr->lock);
> + list_for_each_entry(block, &node->blocks, link)
> + vis_usage += amdgpu_vram_mgr_vis_size(adev, block);
>
> - drm_mm_remove_node(mm);
> - vis_usage += amdgpu_vram_mgr_vis_size(adev, mm);
> - }
> amdgpu_vram_mgr_do_reserve(man);
> - spin_unlock(&mgr->lock);
> +
> + drm_buddy_free_list(mm, &node->blocks);
> + mutex_unlock(&mgr->lock);
>
> atomic64_sub(vis_usage, &mgr->vis_usage);
>
> ttm_resource_fini(man, res);
> - kvfree(node);
> + kfree(node);
> }
>
> /**
> @@ -648,13 +667,22 @@ static void amdgpu_vram_mgr_debug(struct ttm_resource_manager *man,
> struct drm_printer *printer)
> {
> struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
> + struct drm_buddy *mm = &mgr->mm;
> + struct drm_buddy_block *block;
>
> drm_printf(printer, " vis usage:%llu\n",
> amdgpu_vram_mgr_vis_usage(mgr));
>
> - spin_lock(&mgr->lock);
> - drm_mm_print(&mgr->mm, printer);
> - spin_unlock(&mgr->lock);
> + mutex_lock(&mgr->lock);
> + drm_printf(printer, "default_page_size: %lluKiB\n",
> + mgr->default_page_size >> 10);
> +
> + drm_buddy_print(mm, printer);
> +
> + drm_printf(printer, "reserved:\n");
> + list_for_each_entry(block, &mgr->reserved_pages, link)
> + drm_buddy_block_print(mm, block, printer);
> + mutex_unlock(&mgr->lock);
> }
>
> static const struct ttm_resource_manager_func amdgpu_vram_mgr_func = {
> @@ -674,16 +702,21 @@ int amdgpu_vram_mgr_init(struct amdgpu_device *adev)
> {
> struct amdgpu_vram_mgr *mgr = &adev->mman.vram_mgr;
> struct ttm_resource_manager *man = &mgr->manager;
> + int err;
>
> ttm_resource_manager_init(man, &adev->mman.bdev,
> adev->gmc.real_vram_size);
>
> man->func = &amdgpu_vram_mgr_func;
>
> - drm_mm_init(&mgr->mm, 0, man->size >> PAGE_SHIFT);
> - spin_lock_init(&mgr->lock);
> + err = drm_buddy_init(&mgr->mm, man->size, PAGE_SIZE);
> + if (err)
> + return err;
> +
> + mutex_init(&mgr->lock);
> INIT_LIST_HEAD(&mgr->reservations_pending);
> INIT_LIST_HEAD(&mgr->reserved_pages);
> + mgr->default_page_size = PAGE_SIZE;
>
> ttm_set_driver_manager(&adev->mman.bdev, TTM_PL_VRAM, &mgr->manager);
> ttm_resource_manager_set_used(man, true);
> @@ -711,16 +744,16 @@ void amdgpu_vram_mgr_fini(struct amdgpu_device *adev)
> if (ret)
> return;
>
> - spin_lock(&mgr->lock);
> + mutex_lock(&mgr->lock);
> list_for_each_entry_safe(rsv, temp, &mgr->reservations_pending, node)
> kfree(rsv);
>
> list_for_each_entry_safe(rsv, temp, &mgr->reserved_pages, node) {
> - drm_mm_remove_node(&rsv->mm_node);
> + drm_buddy_free_list(&mgr->mm, &rsv->block);
> kfree(rsv);
> }
> - drm_mm_takedown(&mgr->mm);
> - spin_unlock(&mgr->lock);
> + drm_buddy_fini(&mgr->mm);
> + mutex_unlock(&mgr->lock);
>
> ttm_resource_manager_cleanup(man);
> ttm_set_driver_manager(&adev->mman.bdev, TTM_PL_VRAM, NULL);
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h
> new file mode 100644
> index 000000000000..e7d6bbf40c33
> --- /dev/null
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h
> @@ -0,0 +1,68 @@
> +/* SPDX-License-Identifier: MIT
> + * Copyright 2021 Advanced Micro Devices, Inc.
> + *
> + * Permission is hereby granted, free of charge, to any person obtaining a
> + * copy of this software and associated documentation files (the "Software"),
> + * to deal in the Software without restriction, including without limitation
> + * the rights to use, copy, modify, merge, publish, distribute, sublicense,
> + * and/or sell copies of the Software, and to permit persons to whom the
> + * Software is furnished to do so, subject to the following conditions:
> + *
> + * The above copyright notice and this permission notice shall be included in
> + * all copies or substantial portions of the Software.
> + *
> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
> + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
> + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
> + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
> + * OTHER DEALINGS IN THE SOFTWARE.
> + *
> + */
> +
> +#ifndef __AMDGPU_VRAM_MGR_H__
> +#define __AMDGPU_VRAM_MGR_H__
> +
> +#include <drm/drm_buddy.h>
> +
> +struct amdgpu_vram_mgr {
> + struct ttm_resource_manager manager;
> + struct drm_buddy mm;
> + /* protects access to buffer objects */
> + struct mutex lock;
> + struct list_head reservations_pending;
> + struct list_head reserved_pages;
> + atomic64_t vis_usage;
> + u64 default_page_size;
> +};
> +
> +struct amdgpu_vram_mgr_node {
> + struct ttm_resource base;
> + struct list_head blocks;
> + unsigned long flags;
> +};
> +
> +static inline u64 amdgpu_node_start(struct drm_buddy_block *block)
> +{
> + return drm_buddy_block_offset(block);
> +}
> +
> +static inline u64 amdgpu_node_size(struct drm_buddy_block *block)
> +{
> + return PAGE_SIZE << drm_buddy_block_order(block);
> +}
> +
> +static inline struct drm_buddy_block *
> +amdgpu_get_node(struct list_head *list)
That function name is way to generic. Somebody could mix it up with
something completely different and _get_ in a name usually means that
you are grabbing a reference (which is not the case here).
I suggest to call it something like amdgpu_vram_mgr_first_blcok() and
move it from the headers into amdgpu_vram_mgr.c because code outside of
that file shouldn't mess with this.
Apart from that the looks really good now.
Regards,
Christian.
> +{
> + return list_first_entry_or_null(list, struct drm_buddy_block, link);
> +}
> +
> +static inline struct amdgpu_vram_mgr_node *
> +to_amdgpu_vram_mgr_node(struct ttm_resource *res)
> +{
> + return container_of(res, struct amdgpu_vram_mgr_node, base);
> +}
> +
> +#endif
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [Intel-gfx] [PATCH v9] drm/amdgpu: add drm buddy support to amdgpu
@ 2022-03-03 12:04 ` Christian König
0 siblings, 0 replies; 11+ messages in thread
From: Christian König @ 2022-03-03 12:04 UTC (permalink / raw)
To: Arunpravin, amd-gfx, dri-devel, intel-gfx
Cc: alexander.deucher, tzimmermann, matthew.auld, christian.koenig
Am 01.03.22 um 21:38 schrieb Arunpravin:
> - Remove drm_mm references and replace with drm buddy functionalities
> - Add res cursor support for drm buddy
>
> v2(Matthew Auld):
> - replace spinlock with mutex as we call kmem_cache_zalloc
> (..., GFP_KERNEL) in drm_buddy_alloc() function
>
> - lock drm_buddy_block_trim() function as it calls
> mark_free/mark_split are all globally visible
>
> v3(Matthew Auld):
> - remove trim method error handling as we address the failure case
> at drm_buddy_block_trim() function
>
> v4:
> - fix warnings reported by kernel test robot <lkp@intel.com>
>
> v5:
> - fix merge conflict issue
>
> v6:
> - fix warnings reported by kernel test robot <lkp@intel.com>
>
> v7:
> - remove DRM_BUDDY_RANGE_ALLOCATION flag usage
>
> v8:
> - keep DRM_BUDDY_RANGE_ALLOCATION flag usage
> - resolve conflicts created by drm/amdgpu: remove VRAM accounting v2
>
> v9(Christian):
> - rename label name as fallback
> - move struct amdgpu_vram_mgr to amdgpu_vram_mgr.h
> - remove unnecessary flags from struct amdgpu_vram_reservation
> - rewrite block NULL check condition
> - change else style as per coding standard
> - rewrite the node max size
> - add a helper function to fetch the first entry from the list
>
> Signed-off-by: Arunpravin <Arunpravin.PaneerSelvam@amd.com>
> ---
> drivers/gpu/drm/Kconfig | 1 +
> .../gpu/drm/amd/amdgpu/amdgpu_res_cursor.h | 97 +++++--
> drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h | 10 +-
> drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c | 249 ++++++++++--------
> drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h | 68 +++++
> 5 files changed, 289 insertions(+), 136 deletions(-)
> create mode 100644 drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h
>
> diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig
> index f1422bee3dcc..5133c3f028ab 100644
> --- a/drivers/gpu/drm/Kconfig
> +++ b/drivers/gpu/drm/Kconfig
> @@ -280,6 +280,7 @@ config DRM_AMDGPU
> select HWMON
> select BACKLIGHT_CLASS_DEVICE
> select INTERVAL_TREE
> + select DRM_BUDDY
> help
> Choose this option if you have a recent AMD Radeon graphics card.
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h
> index acfa207cf970..864c609ba00b 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h
> @@ -30,12 +30,15 @@
> #include <drm/ttm/ttm_resource.h>
> #include <drm/ttm/ttm_range_manager.h>
>
> +#include "amdgpu_vram_mgr.h"
> +
> /* state back for walking over vram_mgr and gtt_mgr allocations */
> struct amdgpu_res_cursor {
> uint64_t start;
> uint64_t size;
> uint64_t remaining;
> - struct drm_mm_node *node;
> + void *node;
> + uint32_t mem_type;
> };
>
> /**
> @@ -52,27 +55,63 @@ static inline void amdgpu_res_first(struct ttm_resource *res,
> uint64_t start, uint64_t size,
> struct amdgpu_res_cursor *cur)
> {
> + struct drm_buddy_block *block;
> + struct list_head *head, *next;
> struct drm_mm_node *node;
>
> - if (!res || res->mem_type == TTM_PL_SYSTEM) {
> - cur->start = start;
> - cur->size = size;
> - cur->remaining = size;
> - cur->node = NULL;
> - WARN_ON(res && start + size > res->num_pages << PAGE_SHIFT);
> - return;
> - }
> + if (!res)
> + goto fallback;
>
> BUG_ON(start + size > res->num_pages << PAGE_SHIFT);
>
> - node = to_ttm_range_mgr_node(res)->mm_nodes;
> - while (start >= node->size << PAGE_SHIFT)
> - start -= node++->size << PAGE_SHIFT;
> + cur->mem_type = res->mem_type;
> +
> + switch (cur->mem_type) {
> + case TTM_PL_VRAM:
> + head = &to_amdgpu_vram_mgr_node(res)->blocks;
> +
> + block = list_first_entry_or_null(head,
> + struct drm_buddy_block,
> + link);
> + if (!block)
> + goto fallback;
> +
> + while (start >= amdgpu_node_size(block)) {
> + start -= amdgpu_node_size(block);
> +
> + next = block->link.next;
> + if (next != head)
> + block = list_entry(next, struct drm_buddy_block, link);
> + }
> +
> + cur->start = amdgpu_node_start(block) + start;
> + cur->size = min(amdgpu_node_size(block) - start, size);
> + cur->remaining = size;
> + cur->node = block;
> + break;
> + case TTM_PL_TT:
> + node = to_ttm_range_mgr_node(res)->mm_nodes;
> + while (start >= node->size << PAGE_SHIFT)
> + start -= node++->size << PAGE_SHIFT;
> +
> + cur->start = (node->start << PAGE_SHIFT) + start;
> + cur->size = min((node->size << PAGE_SHIFT) - start, size);
> + cur->remaining = size;
> + cur->node = node;
> + break;
> + default:
> + goto fallback;
> + }
>
> - cur->start = (node->start << PAGE_SHIFT) + start;
> - cur->size = min((node->size << PAGE_SHIFT) - start, size);
> + return;
> +
> +fallback:
> + cur->start = start;
> + cur->size = size;
> cur->remaining = size;
> - cur->node = node;
> + cur->node = NULL;
> + WARN_ON(res && start + size > res->num_pages << PAGE_SHIFT);
> + return;
> }
>
> /**
> @@ -85,7 +124,9 @@ static inline void amdgpu_res_first(struct ttm_resource *res,
> */
> static inline void amdgpu_res_next(struct amdgpu_res_cursor *cur, uint64_t size)
> {
> - struct drm_mm_node *node = cur->node;
> + struct drm_buddy_block *block;
> + struct drm_mm_node *node;
> + struct list_head *next;
>
> BUG_ON(size > cur->remaining);
>
> @@ -99,9 +140,27 @@ static inline void amdgpu_res_next(struct amdgpu_res_cursor *cur, uint64_t size)
> return;
> }
>
> - cur->node = ++node;
> - cur->start = node->start << PAGE_SHIFT;
> - cur->size = min(node->size << PAGE_SHIFT, cur->remaining);
> + switch (cur->mem_type) {
> + case TTM_PL_VRAM:
> + block = cur->node;
> +
> + next = block->link.next;
> + block = list_entry(next, struct drm_buddy_block, link);
> +
> + cur->node = block;
> + cur->start = amdgpu_node_start(block);
> + cur->size = min(amdgpu_node_size(block), cur->remaining);
> + break;
> + case TTM_PL_TT:
> + node = cur->node;
> +
> + cur->node = ++node;
> + cur->start = node->start << PAGE_SHIFT;
> + cur->size = min(node->size << PAGE_SHIFT, cur->remaining);
> + break;
> + default:
> + return;
> + }
> }
>
> #endif
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
> index 9120ae80ef52..6a70818039dd 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
> @@ -26,6 +26,7 @@
>
> #include <linux/dma-direction.h>
> #include <drm/gpu_scheduler.h>
> +#include "amdgpu_vram_mgr.h"
> #include "amdgpu.h"
>
> #define AMDGPU_PL_GDS (TTM_PL_PRIV + 0)
> @@ -38,15 +39,6 @@
>
> #define AMDGPU_POISON 0xd0bed0be
>
> -struct amdgpu_vram_mgr {
> - struct ttm_resource_manager manager;
> - struct drm_mm mm;
> - spinlock_t lock;
> - struct list_head reservations_pending;
> - struct list_head reserved_pages;
> - atomic64_t vis_usage;
> -};
> -
> struct amdgpu_gtt_mgr {
> struct ttm_resource_manager manager;
> struct drm_mm mm;
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
> index 0a7611648573..cdf66802e6bc 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
> @@ -32,8 +32,10 @@
> #include "atom.h"
>
> struct amdgpu_vram_reservation {
> + u64 start;
> + u64 size;
> + struct list_head block;
> struct list_head node;
> - struct drm_mm_node mm_node;
> };
>
> static inline struct amdgpu_vram_mgr *
> @@ -194,10 +196,10 @@ const struct attribute_group amdgpu_vram_mgr_attr_group = {
> * Calculate how many bytes of the MM node are inside visible VRAM
> */
> static u64 amdgpu_vram_mgr_vis_size(struct amdgpu_device *adev,
> - struct drm_mm_node *node)
> + struct drm_buddy_block *block)
> {
> - uint64_t start = node->start << PAGE_SHIFT;
> - uint64_t end = (node->size + node->start) << PAGE_SHIFT;
> + u64 start = amdgpu_node_start(block);
> + u64 end = start + amdgpu_node_size(block);
>
> if (start >= adev->gmc.visible_vram_size)
> return 0;
> @@ -218,9 +220,9 @@ u64 amdgpu_vram_mgr_bo_visible_size(struct amdgpu_bo *bo)
> {
> struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
> struct ttm_resource *res = bo->tbo.resource;
> - unsigned pages = res->num_pages;
> - struct drm_mm_node *mm;
> - u64 usage;
> + struct amdgpu_vram_mgr_node *node = to_amdgpu_vram_mgr_node(res);
> + struct drm_buddy_block *block;
> + u64 usage = 0;
>
> if (amdgpu_gmc_vram_full_visible(&adev->gmc))
> return amdgpu_bo_size(bo);
> @@ -228,9 +230,8 @@ u64 amdgpu_vram_mgr_bo_visible_size(struct amdgpu_bo *bo)
> if (res->start >= adev->gmc.visible_vram_size >> PAGE_SHIFT)
> return 0;
>
> - mm = &container_of(res, struct ttm_range_mgr_node, base)->mm_nodes[0];
> - for (usage = 0; pages; pages -= mm->size, mm++)
> - usage += amdgpu_vram_mgr_vis_size(adev, mm);
> + list_for_each_entry(block, &node->blocks, link)
> + usage += amdgpu_vram_mgr_vis_size(adev, block);
>
> return usage;
> }
> @@ -240,21 +241,28 @@ static void amdgpu_vram_mgr_do_reserve(struct ttm_resource_manager *man)
> {
> struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
> struct amdgpu_device *adev = to_amdgpu_device(mgr);
> - struct drm_mm *mm = &mgr->mm;
> + struct drm_buddy *mm = &mgr->mm;
> struct amdgpu_vram_reservation *rsv, *temp;
> + struct drm_buddy_block *block;
> uint64_t vis_usage;
>
> list_for_each_entry_safe(rsv, temp, &mgr->reservations_pending, node) {
> - if (drm_mm_reserve_node(mm, &rsv->mm_node))
> + if (drm_buddy_alloc_blocks(mm, rsv->start, rsv->start + rsv->size,
> + rsv->size, mm->chunk_size, &rsv->block,
> + DRM_BUDDY_RANGE_ALLOCATION))
> + continue;
> +
> + block = amdgpu_get_node(&rsv->block);
> + if (!block)
> continue;
>
> dev_dbg(adev->dev, "Reservation 0x%llx - %lld, Succeeded\n",
> - rsv->mm_node.start, rsv->mm_node.size);
> + rsv->start, rsv->size);
>
> - vis_usage = amdgpu_vram_mgr_vis_size(adev, &rsv->mm_node);
> + vis_usage = amdgpu_vram_mgr_vis_size(adev, block);
> atomic64_add(vis_usage, &mgr->vis_usage);
> spin_lock(&man->bdev->lru_lock);
> - man->usage += rsv->mm_node.size << PAGE_SHIFT;
> + man->usage += rsv->size;
> spin_unlock(&man->bdev->lru_lock);
> list_move(&rsv->node, &mgr->reserved_pages);
> }
> @@ -279,13 +287,15 @@ int amdgpu_vram_mgr_reserve_range(struct amdgpu_vram_mgr *mgr,
> return -ENOMEM;
>
> INIT_LIST_HEAD(&rsv->node);
> - rsv->mm_node.start = start >> PAGE_SHIFT;
> - rsv->mm_node.size = size >> PAGE_SHIFT;
> + INIT_LIST_HEAD(&rsv->block);
> +
> + rsv->start = start;
> + rsv->size = size;
>
> - spin_lock(&mgr->lock);
> + mutex_lock(&mgr->lock);
> list_add_tail(&rsv->node, &mgr->reservations_pending);
> amdgpu_vram_mgr_do_reserve(&mgr->manager);
> - spin_unlock(&mgr->lock);
> + mutex_unlock(&mgr->lock);
>
> return 0;
> }
> @@ -307,19 +317,19 @@ int amdgpu_vram_mgr_query_page_status(struct amdgpu_vram_mgr *mgr,
> struct amdgpu_vram_reservation *rsv;
> int ret;
>
> - spin_lock(&mgr->lock);
> + mutex_lock(&mgr->lock);
>
> list_for_each_entry(rsv, &mgr->reservations_pending, node) {
> - if ((rsv->mm_node.start <= start) &&
> - (start < (rsv->mm_node.start + rsv->mm_node.size))) {
> + if (rsv->start <= start &&
> + (start < (rsv->start + rsv->size))) {
> ret = -EBUSY;
> goto out;
> }
> }
>
> list_for_each_entry(rsv, &mgr->reserved_pages, node) {
> - if ((rsv->mm_node.start <= start) &&
> - (start < (rsv->mm_node.start + rsv->mm_node.size))) {
> + if (rsv->start <= start &&
> + (start < (rsv->start + rsv->size))) {
> ret = 0;
> goto out;
> }
> @@ -327,32 +337,10 @@ int amdgpu_vram_mgr_query_page_status(struct amdgpu_vram_mgr *mgr,
>
> ret = -ENOENT;
> out:
> - spin_unlock(&mgr->lock);
> + mutex_unlock(&mgr->lock);
> return ret;
> }
>
> -/**
> - * amdgpu_vram_mgr_virt_start - update virtual start address
> - *
> - * @mem: ttm_resource to update
> - * @node: just allocated node
> - *
> - * Calculate a virtual BO start address to easily check if everything is CPU
> - * accessible.
> - */
> -static void amdgpu_vram_mgr_virt_start(struct ttm_resource *mem,
> - struct drm_mm_node *node)
> -{
> - unsigned long start;
> -
> - start = node->start + node->size;
> - if (start > mem->num_pages)
> - start -= mem->num_pages;
> - else
> - start = 0;
> - mem->start = max(mem->start, start);
> -}
> -
> /**
> * amdgpu_vram_mgr_new - allocate new ranges
> *
> @@ -368,13 +356,13 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
> const struct ttm_place *place,
> struct ttm_resource **res)
> {
> - unsigned long lpfn, num_nodes, pages_per_node, pages_left, pages;
> + unsigned long lpfn, pages_per_node, pages_left, pages;
> + u64 vis_usage = 0, mem_bytes, max_bytes, min_page_size;
> struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
> struct amdgpu_device *adev = to_amdgpu_device(mgr);
> - uint64_t vis_usage = 0, mem_bytes, max_bytes;
> - struct ttm_range_mgr_node *node;
> - struct drm_mm *mm = &mgr->mm;
> - enum drm_mm_insert_mode mode;
> + struct amdgpu_vram_mgr_node *node;
> + struct drm_buddy *mm = &mgr->mm;
> + struct drm_buddy_block *block;
> unsigned i;
> int r;
>
> @@ -389,7 +377,6 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
> mem_bytes = tbo->base.size;
> if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
> pages_per_node = ~0ul;
> - num_nodes = 1;
> } else {
> #ifdef CONFIG_TRANSPARENT_HUGEPAGE
> pages_per_node = HPAGE_PMD_NR;
> @@ -399,11 +386,9 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
> #endif
> pages_per_node = max_t(uint32_t, pages_per_node,
> tbo->page_alignment);
> - num_nodes = DIV_ROUND_UP_ULL(PFN_UP(mem_bytes), pages_per_node);
> }
>
> - node = kvmalloc(struct_size(node, mm_nodes, num_nodes),
> - GFP_KERNEL | __GFP_ZERO);
> + node = kzalloc(sizeof(*node), GFP_KERNEL);
> if (!node)
> return -ENOMEM;
>
> @@ -415,48 +400,83 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
> goto error_fini;
> }
>
> - mode = DRM_MM_INSERT_BEST;
> + INIT_LIST_HEAD(&node->blocks);
> +
> if (place->flags & TTM_PL_FLAG_TOPDOWN)
> - mode = DRM_MM_INSERT_HIGH;
> + node->flags |= DRM_BUDDY_TOPDOWN_ALLOCATION;
>
> - pages_left = node->base.num_pages;
> + if (place->fpfn || lpfn != man->size >> PAGE_SHIFT)
> + /* Allocate blocks in desired range */
> + node->flags |= DRM_BUDDY_RANGE_ALLOCATION;
>
> - /* Limit maximum size to 2GB due to SG table limitations */
> - pages = min(pages_left, 2UL << (30 - PAGE_SHIFT));
> + BUG_ON(!node->base.num_pages);
> + pages_left = node->base.num_pages;
>
> i = 0;
> - spin_lock(&mgr->lock);
> while (pages_left) {
> - uint32_t alignment = tbo->page_alignment;
> + if (tbo->page_alignment)
> + min_page_size = tbo->page_alignment << PAGE_SHIFT;
> + else
> + min_page_size = mgr->default_page_size;
> +
> + /* Limit maximum size to 2GB due to SG table limitations */
> + pages = min(pages_left, 2UL << (30 - PAGE_SHIFT));
>
> if (pages >= pages_per_node)
> - alignment = pages_per_node;
> -
> - r = drm_mm_insert_node_in_range(mm, &node->mm_nodes[i], pages,
> - alignment, 0, place->fpfn,
> - lpfn, mode);
> - if (unlikely(r)) {
> - if (pages > pages_per_node) {
> - if (is_power_of_2(pages))
> - pages = pages / 2;
> - else
> - pages = rounddown_pow_of_two(pages);
> - continue;
> - }
> - goto error_free;
> + min_page_size = pages_per_node << PAGE_SHIFT;
> +
> + if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
> + pages = roundup_pow_of_two(pages);
> + min_page_size = pages << PAGE_SHIFT;
> +
> + if (pages > lpfn)
> + lpfn = pages;
> }
>
> - vis_usage += amdgpu_vram_mgr_vis_size(adev, &node->mm_nodes[i]);
> - amdgpu_vram_mgr_virt_start(&node->base, &node->mm_nodes[i]);
> - pages_left -= pages;
> + BUG_ON(min_page_size < mm->chunk_size);
> +
> + mutex_lock(&mgr->lock);
> + r = drm_buddy_alloc_blocks(mm, (u64)place->fpfn << PAGE_SHIFT,
> + (u64)lpfn << PAGE_SHIFT,
> + (u64)pages << PAGE_SHIFT,
> + min_page_size,
> + &node->blocks,
> + node->flags);
> + mutex_unlock(&mgr->lock);
> + if (unlikely(r))
> + goto error_free_blocks;
> +
> ++i;
>
> if (pages > pages_left)
> - pages = pages_left;
> + pages_left = 0;
> + else
> + pages_left -= pages;
> }
> - spin_unlock(&mgr->lock);
>
> - if (i == 1)
> + /* Free unused pages for contiguous allocation */
> + if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
> + u64 actual_size = (u64)node->base.num_pages << PAGE_SHIFT;
> +
> + mutex_lock(&mgr->lock);
> + drm_buddy_block_trim(mm,
> + actual_size,
> + &node->blocks);
> + mutex_unlock(&mgr->lock);
> + }
> +
> + list_for_each_entry(block, &node->blocks, link)
> + vis_usage += amdgpu_vram_mgr_vis_size(adev, block);
> +
> + block = amdgpu_get_node(&node->blocks);
> + if (!block) {
> + r = -EINVAL;
> + goto error_fini;
> + }
> +
> + node->base.start = amdgpu_node_start(block) >> PAGE_SHIFT;
> +
> + if (i == 1 && place->flags & TTM_PL_FLAG_CONTIGUOUS)
> node->base.placement |= TTM_PL_FLAG_CONTIGUOUS;
>
> if (adev->gmc.xgmi.connected_to_cpu)
> @@ -468,13 +488,13 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
> *res = &node->base;
> return 0;
>
> -error_free:
> - while (i--)
> - drm_mm_remove_node(&node->mm_nodes[i]);
> - spin_unlock(&mgr->lock);
> +error_free_blocks:
> + mutex_lock(&mgr->lock);
> + drm_buddy_free_list(mm, &node->blocks);
> + mutex_unlock(&mgr->lock);
> error_fini:
> ttm_resource_fini(man, &node->base);
> - kvfree(node);
> + kfree(node);
>
> return r;
> }
> @@ -490,27 +510,26 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
> static void amdgpu_vram_mgr_del(struct ttm_resource_manager *man,
> struct ttm_resource *res)
> {
> - struct ttm_range_mgr_node *node = to_ttm_range_mgr_node(res);
> + struct amdgpu_vram_mgr_node *node = to_amdgpu_vram_mgr_node(res);
> struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
> struct amdgpu_device *adev = to_amdgpu_device(mgr);
> + struct drm_buddy *mm = &mgr->mm;
> + struct drm_buddy_block *block;
> uint64_t vis_usage = 0;
> - unsigned i, pages;
>
> - spin_lock(&mgr->lock);
> - for (i = 0, pages = res->num_pages; pages;
> - pages -= node->mm_nodes[i].size, ++i) {
> - struct drm_mm_node *mm = &node->mm_nodes[i];
> + mutex_lock(&mgr->lock);
> + list_for_each_entry(block, &node->blocks, link)
> + vis_usage += amdgpu_vram_mgr_vis_size(adev, block);
>
> - drm_mm_remove_node(mm);
> - vis_usage += amdgpu_vram_mgr_vis_size(adev, mm);
> - }
> amdgpu_vram_mgr_do_reserve(man);
> - spin_unlock(&mgr->lock);
> +
> + drm_buddy_free_list(mm, &node->blocks);
> + mutex_unlock(&mgr->lock);
>
> atomic64_sub(vis_usage, &mgr->vis_usage);
>
> ttm_resource_fini(man, res);
> - kvfree(node);
> + kfree(node);
> }
>
> /**
> @@ -648,13 +667,22 @@ static void amdgpu_vram_mgr_debug(struct ttm_resource_manager *man,
> struct drm_printer *printer)
> {
> struct amdgpu_vram_mgr *mgr = to_vram_mgr(man);
> + struct drm_buddy *mm = &mgr->mm;
> + struct drm_buddy_block *block;
>
> drm_printf(printer, " vis usage:%llu\n",
> amdgpu_vram_mgr_vis_usage(mgr));
>
> - spin_lock(&mgr->lock);
> - drm_mm_print(&mgr->mm, printer);
> - spin_unlock(&mgr->lock);
> + mutex_lock(&mgr->lock);
> + drm_printf(printer, "default_page_size: %lluKiB\n",
> + mgr->default_page_size >> 10);
> +
> + drm_buddy_print(mm, printer);
> +
> + drm_printf(printer, "reserved:\n");
> + list_for_each_entry(block, &mgr->reserved_pages, link)
> + drm_buddy_block_print(mm, block, printer);
> + mutex_unlock(&mgr->lock);
> }
>
> static const struct ttm_resource_manager_func amdgpu_vram_mgr_func = {
> @@ -674,16 +702,21 @@ int amdgpu_vram_mgr_init(struct amdgpu_device *adev)
> {
> struct amdgpu_vram_mgr *mgr = &adev->mman.vram_mgr;
> struct ttm_resource_manager *man = &mgr->manager;
> + int err;
>
> ttm_resource_manager_init(man, &adev->mman.bdev,
> adev->gmc.real_vram_size);
>
> man->func = &amdgpu_vram_mgr_func;
>
> - drm_mm_init(&mgr->mm, 0, man->size >> PAGE_SHIFT);
> - spin_lock_init(&mgr->lock);
> + err = drm_buddy_init(&mgr->mm, man->size, PAGE_SIZE);
> + if (err)
> + return err;
> +
> + mutex_init(&mgr->lock);
> INIT_LIST_HEAD(&mgr->reservations_pending);
> INIT_LIST_HEAD(&mgr->reserved_pages);
> + mgr->default_page_size = PAGE_SIZE;
>
> ttm_set_driver_manager(&adev->mman.bdev, TTM_PL_VRAM, &mgr->manager);
> ttm_resource_manager_set_used(man, true);
> @@ -711,16 +744,16 @@ void amdgpu_vram_mgr_fini(struct amdgpu_device *adev)
> if (ret)
> return;
>
> - spin_lock(&mgr->lock);
> + mutex_lock(&mgr->lock);
> list_for_each_entry_safe(rsv, temp, &mgr->reservations_pending, node)
> kfree(rsv);
>
> list_for_each_entry_safe(rsv, temp, &mgr->reserved_pages, node) {
> - drm_mm_remove_node(&rsv->mm_node);
> + drm_buddy_free_list(&mgr->mm, &rsv->block);
> kfree(rsv);
> }
> - drm_mm_takedown(&mgr->mm);
> - spin_unlock(&mgr->lock);
> + drm_buddy_fini(&mgr->mm);
> + mutex_unlock(&mgr->lock);
>
> ttm_resource_manager_cleanup(man);
> ttm_set_driver_manager(&adev->mman.bdev, TTM_PL_VRAM, NULL);
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h
> new file mode 100644
> index 000000000000..e7d6bbf40c33
> --- /dev/null
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h
> @@ -0,0 +1,68 @@
> +/* SPDX-License-Identifier: MIT
> + * Copyright 2021 Advanced Micro Devices, Inc.
> + *
> + * Permission is hereby granted, free of charge, to any person obtaining a
> + * copy of this software and associated documentation files (the "Software"),
> + * to deal in the Software without restriction, including without limitation
> + * the rights to use, copy, modify, merge, publish, distribute, sublicense,
> + * and/or sell copies of the Software, and to permit persons to whom the
> + * Software is furnished to do so, subject to the following conditions:
> + *
> + * The above copyright notice and this permission notice shall be included in
> + * all copies or substantial portions of the Software.
> + *
> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
> + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
> + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
> + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
> + * OTHER DEALINGS IN THE SOFTWARE.
> + *
> + */
> +
> +#ifndef __AMDGPU_VRAM_MGR_H__
> +#define __AMDGPU_VRAM_MGR_H__
> +
> +#include <drm/drm_buddy.h>
> +
> +struct amdgpu_vram_mgr {
> + struct ttm_resource_manager manager;
> + struct drm_buddy mm;
> + /* protects access to buffer objects */
> + struct mutex lock;
> + struct list_head reservations_pending;
> + struct list_head reserved_pages;
> + atomic64_t vis_usage;
> + u64 default_page_size;
> +};
> +
> +struct amdgpu_vram_mgr_node {
> + struct ttm_resource base;
> + struct list_head blocks;
> + unsigned long flags;
> +};
> +
> +static inline u64 amdgpu_node_start(struct drm_buddy_block *block)
> +{
> + return drm_buddy_block_offset(block);
> +}
> +
> +static inline u64 amdgpu_node_size(struct drm_buddy_block *block)
> +{
> + return PAGE_SIZE << drm_buddy_block_order(block);
> +}
> +
> +static inline struct drm_buddy_block *
> +amdgpu_get_node(struct list_head *list)
That function name is way to generic. Somebody could mix it up with
something completely different and _get_ in a name usually means that
you are grabbing a reference (which is not the case here).
I suggest to call it something like amdgpu_vram_mgr_first_blcok() and
move it from the headers into amdgpu_vram_mgr.c because code outside of
that file shouldn't mess with this.
Apart from that the looks really good now.
Regards,
Christian.
> +{
> + return list_first_entry_or_null(list, struct drm_buddy_block, link);
> +}
> +
> +static inline struct amdgpu_vram_mgr_node *
> +to_amdgpu_vram_mgr_node(struct ttm_resource *res)
> +{
> + return container_of(res, struct amdgpu_vram_mgr_node, base);
> +}
> +
> +#endif
^ permalink raw reply [flat|nested] 11+ messages in thread
end of thread, other threads:[~2022-03-03 18:07 UTC | newest]
Thread overview: 11+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-03-01 20:38 [PATCH v9] drm/amdgpu: add drm buddy support to amdgpu Arunpravin
2022-03-01 20:38 ` Arunpravin
2022-03-01 20:38 ` [Intel-gfx] " Arunpravin
2022-03-01 23:05 ` [Intel-gfx] ✗ Fi.CI.CHECKPATCH: warning for " Patchwork
2022-03-01 23:38 ` [Intel-gfx] ✓ Fi.CI.BAT: success " Patchwork
2022-03-02 2:51 ` [PATCH v9] " kernel test robot
2022-03-02 2:51 ` [Intel-gfx] " kernel test robot
2022-03-02 4:21 ` [Intel-gfx] ✓ Fi.CI.IGT: success for " Patchwork
2022-03-03 12:04 ` [PATCH v9] " Christian König
2022-03-03 12:04 ` [Intel-gfx] " Christian König
2022-03-03 12:04 ` Christian König
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.