Lines Matching full:profiling

22 // By default we support profiling upto 1024 perf events. Build
23 // targets can override this to increase the profiling buffer size
28 // By default we support profiling upto 1024 memory allocation events.
30 // the effect of increasing/decreasing the profiling buffer size.
34 // By default we support profiling only upto 16 allocators. If users
36 // compilation time. There will be an increase/decrease in the profiling
41 // By default we support only one profiling block. If users want to profile
92 This is what the layout of the profiling buffer looks like.
94 | Profiling header                    |
104 // offsets of the various sections in the profiling buffer
105 // Total size required for profiling buffer
122 // This starts the profiling of this event and returns a token
126 // End profiling event represented by token_id
141 // This class enables scope based profiling where needed. Profiling
221 // Convenience macros to begin and end profiling. These can be inserted