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
|
<!-- table 5.3 -->
<informaltable frame="all">
<tgroup cols="2" align="left" colsep="1" rowsep="1">
<colspec colname="col1" colnum="1" />
<thead>
<row>
<entry>cl_mem_flags</entry>
<entry>Description</entry>
</row>
</thead>
<tbody>
<row>
<entry><constant>CL_MEM_READ_WRITE</constant></entry>
<entry>
This flag specifies that the memory object will be read and written by a kernel. This
is the default.
</entry>
</row>
<row>
<entry><constant>CL_MEM_WRITE_ONLY</constant></entry>
<entry>
<para>
This flags specifies that the memory object will be written but not read by
a kernel.
</para>
<para>
Reading from a buffer or image object created with
<constant>CL_MEM_WRITE_ONLY</constant> inside a kernel is undefined.
</para>
<para>
<constant>CL_MEM_READ_WRITE</constant> and <constant>CL_MEM_WRITE_ONLY</constant>
are mutually exclusive.
</para>
</entry>
</row>
<row>
<entry><constant>CL_MEM_READ_ONLY</constant></entry>
<entry>
<para>
This flag specifies that the memory object is a read-only memory object when
used inside a kernel.
</para>
<para>
Writing to a buffer or image object created with
<constant>CL_MEM_READ_ONLY</constant> inside a kernel is undefined.
</para>
<para>
<constant>CL_MEM_READ_WRITE</constant> or <constant>CL_MEM_WRITE_ONLY</constant>
and <constant>CL_MEM_READ_ONLY</constant> are mutually exclusive.
</para>
</entry>
</row>
<row>
<entry><constant>CL_MEM_USE_HOST_PTR</constant></entry>
<entry>
<para>
This flag is valid only if <varname>host_ptr</varname> is not NULL. If
specified, it indicates that the application wants the OpenCL implementation
to use memory referenced by <varname>host_ptr</varname> as the storage bits
for the memory object.
</para>
<para>
OpenCL implementations are allowed to cache the buffer contents pointed to by
<varname>host_ptr</varname> in device memory. This cached copy can be used when
kernels are executed on a device.
</para>
<para>
The result of OpenCL commands that operate on multiple buffer objects created with
the same <varname>host_ptr</varname> or overlapping host regions is considered
to be undefined.
</para>
<para>
Refer to the <citerefentry href="dataTypes"> <refentrytitle>description
of the alignment rules</refentrytitle> </citerefentry> for
<varname>host_ptr</varname> for memory objects (buffer and images) created
using <constant>CL_MEM_USE_HOST_PTR</constant>.
</para>
</entry>
</row>
<row>
<entry><constant>CL_MEM_ALLOC_HOST_PTR</constant></entry>
<entry>
<para>
This flag specifies that the application wants the OpenCL implementation to
allocate memory from host accessible memory.
</para>
<para>
<constant>CL_MEM_ALLOC_HOST_PTR</constant> and
<constant>CL_MEM_USE_HOST_PTR</constant> are mutually exclusive.
</para>
</entry>
</row>
<row>
<entry><constant>CL_MEM_COPY_HOST_PTR</constant></entry>
<entry>
<para>
This flag is valid only if <varname>host_ptr</varname> is not NULL. If specified,
it indicates that the application wants the OpenCL implementation to allocate
memory for the memory object and copy the data from memory referenced by
<varname>host_ptr</varname>.
</para>
<para>
<constant>CL_MEM_COPY_HOST_PTR</constant> and
<constant>CL_MEM_USE_HOST_PTR</constant> are mutually exclusive.
</para>
<para>
<constant>CL_MEM_COPY_HOST_PTR</constant> can be used with
<constant>CL_MEM_ALLOC_HOST_PTR</constant> to initialize the contents of the
<type>cl_mem</type> object allocated using host-accessible (e.g. PCIe) memory.
</para>
</entry>
</row>
<row>
<entry><constant>CL_MEM_HOST_WRITE_ONLY</constant></entry>
<entry>
<para>
This flag specifies that the host will only write to the memory object (using
OpenCL APIs that enqueue a write or a map for write). This can be used to optimize
write access from the host (e.g. enable write combined allocations for memory
objects for devices that communicate with the host over a system bus such as PCIe).
</para>
</entry>
</row>
<row>
<entry><constant>CL_MEM_HOST_READ_ONLY</constant></entry>
<entry>
<para>
This flag specifies that the host will only read the memory object (using OpenCL
APIs that enqueue a read or a map for read).
</para>
<para>
<constant>CL_MEM_HOST_WRITE_ONLY</constant> and
<constant>CL_MEM_HOST_READ_ONLY</constant> are mutually exclusive.
</para>
</entry>
</row>
<row>
<entry><constant>CL_MEM_HOST_NO_ACCESS</constant></entry>
<entry>
<para>
This flag specifies that the host will not read or write the memory object.
</para>
<para>
<constant>CL_MEM_HOST_WRITE_ONLY</constant> or
<constant>CL_MEM_HOST_READ_ONLY</constant> and
<constant>CL_MEM_HOST_NO_ACCESS</constant> are mutually exclusive.
</para>
</entry>
</row>
</tbody>
</tgroup>
</informaltable>
<!-- 21-Oct-2011 -->
|
