1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
|
From a001a34d8c17dcbd12ff418a7e79efe3879fb631 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@amd.com>
Date: Mon, 30 Jul 2018 16:16:01 +0200
Subject: [PATCH 5032/5725] drm/amdgpu: allocate the bo_list array after the
list
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
This avoids multiple allocations for the head and the array.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c | 114 +++++++++++-----------------
drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.h | 17 +++--
2 files changed, 57 insertions(+), 74 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c
index 3f96f38..b603249 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c
@@ -35,13 +35,15 @@
#define AMDGPU_BO_LIST_MAX_PRIORITY 32u
#define AMDGPU_BO_LIST_NUM_BUCKETS (AMDGPU_BO_LIST_MAX_PRIORITY + 1)
-static int amdgpu_bo_list_set(struct amdgpu_device *adev,
- struct drm_file *filp,
- struct amdgpu_bo_list *list,
- struct drm_amdgpu_bo_list_entry *info,
- unsigned num_entries);
+static void amdgpu_bo_list_free_rcu(struct rcu_head *rcu)
+{
+ struct amdgpu_bo_list *list = container_of(rcu, struct amdgpu_bo_list,
+ rhead);
+
+ kvfree(list);
+}
-static void amdgpu_bo_list_release_rcu(struct kref *ref)
+static void amdgpu_bo_list_free(struct kref *ref)
{
struct amdgpu_bo_list *list = container_of(ref, struct amdgpu_bo_list,
refcount);
@@ -50,67 +52,36 @@ static void amdgpu_bo_list_release_rcu(struct kref *ref)
amdgpu_bo_list_for_each_entry(e, list)
amdgpu_bo_unref(&e->robj);
- kvfree(list->array);
- kfree_rcu(list, rhead);
+ call_rcu(&list->rhead, amdgpu_bo_list_free_rcu);
}
-int amdgpu_bo_list_create(struct amdgpu_device *adev,
- struct drm_file *filp,
- struct drm_amdgpu_bo_list_entry *info,
- unsigned num_entries,
- struct amdgpu_bo_list **list_out)
+int amdgpu_bo_list_create(struct amdgpu_device *adev, struct drm_file *filp,
+ struct drm_amdgpu_bo_list_entry *info,
+ unsigned num_entries, struct amdgpu_bo_list **result)
{
+ unsigned last_entry = 0, first_userptr = num_entries;
+ struct amdgpu_bo_list_entry *array;
struct amdgpu_bo_list *list;
+ uint64_t total_size = 0;
+ size_t size;
+ unsigned i;
int r;
+ if (num_entries > SIZE_MAX / sizeof(struct amdgpu_bo_list_entry))
+ return -EINVAL;
- list = kzalloc(sizeof(struct amdgpu_bo_list), GFP_KERNEL);
+ size = sizeof(struct amdgpu_bo_list);
+ size += num_entries * sizeof(struct amdgpu_bo_list_entry);
+ list = kvmalloc(size, GFP_KERNEL);
if (!list)
return -ENOMEM;
- /* initialize bo list*/
kref_init(&list->refcount);
- r = amdgpu_bo_list_set(adev, filp, list, info, num_entries);
- if (r) {
- kfree(list);
- return r;
- }
-
- *list_out = list;
- return 0;
-}
-
-static void amdgpu_bo_list_destroy(struct amdgpu_fpriv *fpriv, int id)
-{
- struct amdgpu_bo_list *list;
-
- mutex_lock(&fpriv->bo_list_lock);
- list = idr_remove(&fpriv->bo_list_handles, id);
- mutex_unlock(&fpriv->bo_list_lock);
- if (list)
- kref_put(&list->refcount, amdgpu_bo_list_release_rcu);
-}
-
-static int amdgpu_bo_list_set(struct amdgpu_device *adev,
- struct drm_file *filp,
- struct amdgpu_bo_list *list,
- struct drm_amdgpu_bo_list_entry *info,
- unsigned num_entries)
-{
- struct amdgpu_bo_list_entry *array;
- struct amdgpu_bo *gds_obj = adev->gds.gds_gfx_bo;
- struct amdgpu_bo *gws_obj = adev->gds.gws_gfx_bo;
- struct amdgpu_bo *oa_obj = adev->gds.oa_gfx_bo;
-
- unsigned last_entry = 0, first_userptr = num_entries;
- struct amdgpu_bo_list_entry *e;
- uint64_t total_size = 0;
- unsigned i;
- int r;
+ list->gds_obj = adev->gds.gds_gfx_bo;
+ list->gws_obj = adev->gds.gws_gfx_bo;
+ list->oa_obj = adev->gds.oa_gfx_bo;
- array = kvmalloc_array(num_entries, sizeof(struct amdgpu_bo_list_entry), GFP_KERNEL);
- if (!array)
- return -ENOMEM;
+ array = amdgpu_bo_list_array_entry(list, 0);
memset(array, 0, num_entries * sizeof(struct amdgpu_bo_list_entry));
for (i = 0; i < num_entries; ++i) {
@@ -147,36 +118,41 @@ static int amdgpu_bo_list_set(struct amdgpu_device *adev,
entry->tv.shared = !entry->robj->prime_shared_count;
if (entry->robj->preferred_domains == AMDGPU_GEM_DOMAIN_GDS)
- gds_obj = entry->robj;
+ list->gds_obj = entry->robj;
if (entry->robj->preferred_domains == AMDGPU_GEM_DOMAIN_GWS)
- gws_obj = entry->robj;
+ list->gws_obj = entry->robj;
if (entry->robj->preferred_domains == AMDGPU_GEM_DOMAIN_OA)
- oa_obj = entry->robj;
+ list->oa_obj = entry->robj;
total_size += amdgpu_bo_size(entry->robj);
trace_amdgpu_bo_list_set(list, entry->robj);
}
- amdgpu_bo_list_for_each_entry(e, list)
- amdgpu_bo_unref(&list->array[i].robj);
-
- kvfree(list->array);
-
- list->gds_obj = gds_obj;
- list->gws_obj = gws_obj;
- list->oa_obj = oa_obj;
list->first_userptr = first_userptr;
- list->array = array;
list->num_entries = num_entries;
trace_amdgpu_cs_bo_status(list->num_entries, total_size);
+
+ *result = list;
return 0;
error_free:
while (i--)
amdgpu_bo_unref(&array[i].robj);
- kvfree(array);
+ kvfree(list);
return r;
+
+}
+
+static void amdgpu_bo_list_destroy(struct amdgpu_fpriv *fpriv, int id)
+{
+ struct amdgpu_bo_list *list;
+
+ mutex_lock(&fpriv->bo_list_lock);
+ list = idr_remove(&fpriv->bo_list_handles, id);
+ mutex_unlock(&fpriv->bo_list_lock);
+ if (list)
+ kref_put(&list->refcount, amdgpu_bo_list_free);
}
int amdgpu_bo_list_get(struct amdgpu_fpriv *fpriv, int id,
@@ -229,7 +205,7 @@ void amdgpu_bo_list_get_list(struct amdgpu_bo_list *list,
void amdgpu_bo_list_put(struct amdgpu_bo_list *list)
{
- kref_put(&list->refcount, amdgpu_bo_list_release_rcu);
+ kref_put(&list->refcount, amdgpu_bo_list_free);
}
int amdgpu_bo_create_list_entry_array(struct drm_amdgpu_bo_list_in *in,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.h
index 3d77abf..61b0897 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.h
@@ -48,7 +48,6 @@ struct amdgpu_bo_list {
struct amdgpu_bo *oa_obj;
unsigned first_userptr;
unsigned num_entries;
- struct amdgpu_bo_list_entry *array;
};
int amdgpu_bo_list_get(struct amdgpu_fpriv *fpriv, int id,
@@ -65,14 +64,22 @@ int amdgpu_bo_list_create(struct amdgpu_device *adev,
unsigned num_entries,
struct amdgpu_bo_list **list);
+static inline struct amdgpu_bo_list_entry *
+amdgpu_bo_list_array_entry(struct amdgpu_bo_list *list, unsigned index)
+{
+ struct amdgpu_bo_list_entry *array = (void *)&list[1];
+
+ return &array[index];
+}
+
#define amdgpu_bo_list_for_each_entry(e, list) \
- for (e = &(list)->array[0]; \
- e != &(list)->array[(list)->num_entries]; \
+ for (e = amdgpu_bo_list_array_entry(list, 0); \
+ e != amdgpu_bo_list_array_entry(list, (list)->num_entries); \
++e)
#define amdgpu_bo_list_for_each_userptr_entry(e, list) \
- for (e = &(list)->array[(list)->first_userptr]; \
- e != &(list)->array[(list)->num_entries]; \
+ for (e = amdgpu_bo_list_array_entry(list, (list)->first_userptr); \
+ e != amdgpu_bo_list_array_entry(list, (list)->num_entries); \
++e)
#endif
--
2.7.4
|
