Lines Matching refs:be
42 * @return status Error status of the call, must be:
55 * not be supported for any reason, getSupportedOperations must return
59 * operations may only be fully supported if the vendor driver fully
63 * are to be verified by the driver.
64 * @return status Error status of the call, must be:
87 * Optionally, caching information may be provided for the driver to save
94 * The file descriptors must be opened with read and write permission. A
97 * to that file. The file descriptors may be closed by the client once the
118 * must be invoked with an error status of ErrorStatus::NONE and the
120 * the callback object must be invoked with the appropriate ErrorStatus
128 * prepareModel_1_3 can be called with an optional deadline. If the model
129 * is not able to be prepared before the provided deadline, the model
130 * preparation may be aborted, and either {@link
132 * ErrorStatus::MISSED_DEADLINE_PERSISTENT} may be returned. The error due
133 * to an abort must be sent the same way as other errors, described above.
146 * related to the cache may be invalid, or the driver may fail to save to
151 * The only information that may be unknown to the model at this stage is
152 * the shape of the tensors, which may only be known at execution time. As
154 * the prepared model may only be finished when it is paired with a set of
155 * inputs to the model. Note that the same prepared model object may be used
161 * @param model The model to be prepared for execution.
166 * @param deadline The time by which the model is expected to be prepared.
167 * If the model cannot be prepared by the deadline, the preparation may
168 * be aborted.
171 * the vector must either be 0 indicating that caching information is
173 * getNumberOfCacheFilesNeeded. The cache handles will be provided in
178 * vector must either be 0 indicating that caching information is not
180 * getNumberOfCacheFilesNeeded. The cache handles will be provided in
184 * identifying the prepared model. The same token will be provided when
186 * prepareModelFromCache_1_3. Tokens should be chosen to have a low rate of
191 * information is not provided, this token must be ignored.
195 * must be called exactly once, even if the model could not be prepared.
197 * in the background; must be:
204 * cannot be prepared by the deadline
223 * The file descriptors must be opened with read and write permission. A file may
226 * descriptors may be closed by the client once the asynchronous preparation has
249 * callback object must be invoked with an error status of ErrorStatus::NONE
251 * the model, the callback object must be invoked with the appropriate
254 * prepareModelFromCache_1_3 can be called with an optional deadline. If the
256 * preparation may be aborted, and either {@link
258 * or {@link ErrorStatus::MISSED_DEADLINE_PERSISTENT} may be returned. The
259 * error due to an abort must be sent the same way as other errors,
270 * The only information that may be unknown to the model at this stage is
271 * the shape of the tensors, which may only be known at execution time. As
273 * the prepared model may only be finished when it is paired with a set of
274 * inputs to the model. Note that the same prepared model object may be
278 * @param deadline The time by which the model is expected to be prepared.
279 * If the model cannot be prepared by the deadline, the preparation may
280 * be aborted.
284 * The cache handles will be provided in the same order as with prepareModel_1_3.
288 * The cache handles will be provided in the same order as with prepareModel_1_3.
291 * the cache files with prepareModel_1_3. Tokens should be chosen
298 * must be called exactly once, even if the model could not be prepared.
300 * in the background; must be:
307 * cannot be prepared by the deadline
327 * The buffer is allocated to an uninitialized state. An uninitialized buffer may only be used
330 * the buffer. An initialized buffer may be used according to all roles specified in inputRoles
334 * The dimensions of the buffer can be deduced from the buffer descriptor as well as the
336 * or rank of the buffer may be unknown at this stage. As such, some driver services may only
338 * same buffer may be used for different shapes of outputs on different executions. When the
339 * buffer is used as an input, the input shape must be the same as the output shape from the
350 * Each role specified in inputRoles and outputRoles must be unique. The corresponding
353 * descriptor must be compatible with each other. Two dimensions are incompatible if there
355 * @return status Error status of the buffer allocation. Must be:
361 * @return buffer The allocated IBuffer object. If the buffer was unable to be allocated
362 * due to an error, nullptr must be returned.
363 * @return token A positive token identifying the allocated buffer. The same token will be
366 * currently alive in the same driver service. If the buffer was unable to be allocated
367 * due to an error, the token must be 0.