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
|
//------------------------------------------------------------------------------
// GB_dev.h: definitions for code development
//------------------------------------------------------------------------------
// SuiteSparse:GraphBLAS, Timothy A. Davis, (c) 2017-2022, All Rights Reserved.
// SPDX-License-Identifier: Apache-2.0
//------------------------------------------------------------------------------
#ifndef GB_DEV_H
#define GB_DEV_H
//------------------------------------------------------------------------------
// code development settings: by default, all settings should be commented out
//------------------------------------------------------------------------------
// to turn on Debug for a single file of GraphBLAS, add '#define GB_DEBUG'
// just before the statement '#include "GB.h"'
// to turn on Debug for all of GraphBLAS, uncomment this line:
// (GraphBLAS will be exceedingly slow; this is for development only)
// #define GB_DEBUG
// to turn on a very verbose memory trace
// (GraphBLAS will be exceedingly slow; this is for development only)
// #define GB_MEMDUMP
// to turn on diagnostic timings. See GrB.timing in the @GrB interface.
// #define GB_TIMING
// By default, many internal temporary matrices use statically allocated
// headers to reduce the number of calls to malloc/free. This works fine for
// matrices on the CPU, but the static headers do not get automatically
// transfered to the GPU. Only dynamically allocated headers, allocated by
// rmm_wrap_malloc, get transfered. Set this to 1 to turn off static headers
// (required for CUDA; see GB_static_headers.h). Leave static headers
// enabled by default by leaving this commented out or setting GBNSTATIC to 0.
// #undef GBNSTATIC
// #define GBNSTATIC 1
#endif
|
