Allow specifying a drm_exec object in TTMs operation context which is
used to lock objects during eviction.
This allows to handle deadlocks much more gracefully and with that
avoid returning -ENOMEM on heavily contended domains.
v2: rebased on top of Thomas work
TODO: This still doesn't handle BOs which are about to be torn down
correctly.
Signed-off-by: Christian König <christian.koe...@amd.com>
---
drivers/gpu/drm/ttm/ttm_bo_util.c | 45 +++++++++++++++++++++++++------
drivers/gpu/drm/ttm/ttm_bo_util.h | 2 ++
include/drm/ttm/ttm_bo.h | 3 +++
3 files changed, 42 insertions(+), 8 deletions(-)
diff --git a/drivers/gpu/drm/ttm/ttm_bo_util.c
b/drivers/gpu/drm/ttm/ttm_bo_util.c
index 7a4bc7e9950b..850e329ab5a5 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_util.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_util.c
@@ -36,6 +36,7 @@
#include <drm/ttm/ttm_tt.h>
#include <drm/drm_cache.h>
+#include <drm/drm_exec.h>
#include "ttm_bo_util.h"
@@ -776,15 +777,22 @@ static bool ttm_lru_walk_trylock(struct ttm_lru_walk
*walk,
{
struct ttm_operation_ctx *ctx = walk->ctx;
+ walk->needs_drop = false;
walk->needs_unlock = false;
- if (dma_resv_trylock(bo->base.resv)) {
- walk->needs_unlock = true;
+ if (bo->base.resv == ctx->resv && ctx->allow_res_evict) {
+ dma_resv_assert_held(bo->base.resv);
return true;
}
- if (bo->base.resv == ctx->resv && ctx->allow_res_evict) {
- dma_resv_assert_held(bo->base.resv);
+ if (walk->ctx->exec) {
+ if (drm_exec_trylock_obj(walk->ctx->exec, &bo->base)) {
+ walk->needs_drop = true;
+ return true;
+ }
+
+ } else if (dma_resv_trylock(bo->base.resv)) {
+ walk->needs_unlock = true;
return true;
}
@@ -797,7 +805,9 @@ static int ttm_lru_walk_ticketlock(struct ttm_lru_walk
*walk,
struct dma_resv *resv = bo->base.resv;
int ret;
- if (walk->ctx->interruptible)
+ if (walk->ctx->exec)
+ ret = drm_exec_lock_obj(walk->ctx->exec, &bo->base);
+ else if (walk->ctx->interruptible)
ret = dma_resv_lock_interruptible(resv, walk->ticket);
else
ret = dma_resv_lock(resv, walk->ticket);
@@ -811,7 +821,8 @@ static int ttm_lru_walk_ticketlock(struct ttm_lru_walk
*walk,
* trylocking for this walk.
*/
walk->ticket = NULL;
- } else if (ret == -EDEADLK) {
+
+ } else if (!walk->ctx->exec && ret == -EDEADLK) {
/* Caller needs to exit the ww transaction. */
ret = -ENOSPC;
}
@@ -822,7 +833,15 @@ static int ttm_lru_walk_ticketlock(struct ttm_lru_walk
*walk,
static void ttm_lru_walk_unlock(struct ttm_lru_walk *walk,
struct ttm_buffer_object *bo)
{
- if (walk->needs_unlock)
+ if (walk->needs_drop)
+ drm_exec_drop_trylocked_obj(walk->ctx->exec, &bo->base);
+
+ if (!walk->needs_unlock)
+ return;
+
+ if (walk->ctx->exec)
+ drm_exec_unlock_obj(walk->ctx->exec, &bo->base);
+ else
dma_resv_unlock(bo->base.resv);
}
@@ -891,8 +910,18 @@ s64 ttm_lru_walk_for_evict(struct ttm_lru_walk *walk,
struct ttm_device *bdev,
spin_unlock(&bdev->lru_lock);
lret = 0;
- if (!bo_locked)
+ if (!bo_locked) {
lret = ttm_lru_walk_ticketlock(walk, bo);
+ } else if (walk->ctx->exec && !bo->deleted) {
+ lret = drm_exec_keep_trylocked_obj(walk->ctx->exec,
+ &bo->base);
+ if (!lret) {
+ walk->needs_drop = false;
+ walk->needs_unlock = true;
+ }
+ } else {
+ lret = 0;
+ }
/*
* Note that in between the release of the lru lock and the
diff --git a/drivers/gpu/drm/ttm/ttm_bo_util.h
b/drivers/gpu/drm/ttm/ttm_bo_util.h
index c653e16ccb76..5e1bb156837f 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_util.h
+++ b/drivers/gpu/drm/ttm/ttm_bo_util.h
@@ -59,6 +59,8 @@ struct ttm_lru_walk {
struct ww_acquire_ctx *ticket;
/** @tryock_only: Only use trylock for locking. */
bool trylock_only;
+ /** @needs_drop: If the current BO needs a drm_exec trylock drop */
+ bool needs_drop;
/** @needs_unlock: If the current BO needs unlocking */
bool needs_unlock;
};
diff --git a/include/drm/ttm/ttm_bo.h b/include/drm/ttm/ttm_bo.h
index 5f7c967222a2..5bee917e01e2 100644
--- a/include/drm/ttm/ttm_bo.h
+++ b/include/drm/ttm/ttm_bo.h
@@ -180,6 +180,8 @@ struct ttm_bo_kmap_obj {
* faults. Should only be used by TTM internally.
* @resv: Reservation object to allow reserved evictions with.
* @bytes_moved: Statistics on how many bytes have been moved.
+ * @exec: optional drm_exec object to use for locking BOs and tracking which
are
+ * locked.
*
* Context for TTM operations like changing buffer placement or general memory
* allocation.
@@ -192,6 +194,7 @@ struct ttm_operation_ctx {
bool force_alloc;
struct dma_resv *resv;
uint64_t bytes_moved;
+ struct drm_exec *exec;
};
/**
--
2.34.1
