1/* -*- c++ -*- */
2/*
3 * Copyright © 2020 Intel Corporation
4 *
5 * Permission is hereby granted, free of charge, to any person obtaining a
6 * copy of this software and associated documentation files (the "Software"),
7 * to deal in the Software without restriction, including without limitation
8 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
9 * and/or sell copies of the Software, and to permit persons to whom the
10 * Software is furnished to do so, subject to the following conditions:
11 *
12 * The above copyright notice and this permission notice (including the next
13 * paragraph) shall be included in all copies or substantial portions of the
14 * Software.
15 *
16 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
17 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
18 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
19 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
20 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
21 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
22 * IN THE SOFTWARE.
23 */
24
25#ifndef BRW_IR_PERFORMANCE_H
26#define BRW_IR_PERFORMANCE_H
27
28class fs_visitor;
29
30namespace brw {
31   class vec4_visitor;
32
33   /**
34    * Various estimates of the performance of a shader based on static
35    * analysis.
36    */
37   struct performance {
38      performance(const fs_visitor *v);
39      performance(const vec4_visitor *v);
40      ~performance();
41
42      analysis_dependency_class
43      dependency_class() const
44      {
45         return (DEPENDENCY_INSTRUCTIONS |
46                 DEPENDENCY_BLOCKS);
47      }
48
49      bool
50      validate(const backend_shader *) const
51      {
52         return true;
53      }
54
55      /**
56       * Array containing estimates of the runtime of each basic block of the
57       * program in cycle units.
58       */
59      unsigned *block_latency;
60
61      /**
62       * Estimate of the runtime of the whole program in cycle units assuming
63       * uncontended execution.
64       */
65      unsigned latency;
66
67      /**
68       * Estimate of the throughput of the whole program in
69       * invocations-per-cycle units.
70       *
71       * Note that this might be lower than the ratio between the dispatch
72       * width of the program and its latency estimate in cases where
73       * performance doesn't scale without limits as a function of its thread
74       * parallelism, e.g. due to the existence of a bottleneck in a shared
75       * function.
76       */
77      float throughput;
78
79   private:
80      performance(const performance &perf);
81      performance &
82      operator=(performance u);
83   };
84}
85
86#endif
87