Re: any performance profiling tools ??

> [Do any of these address the original question about profiling at a low
> enough level to count cache misses and the like? -John]


The Digital continuous profiling tools will help if you have Alpha
machines running Digital Unix (NT to be available soon.) I have
included an abstract from a paper describing this system that will
appear in the 16th SOSP next week:


--- begin abstract ---
This paper describes the DIGITAL Continuous Profiling Infrastructure,
a sampling-based profiling system designed to run continuously on
production systems. The system supports multiprocessors, works on
unmodified executables, and collects profiles for entire systems,
including user programs, shared libraries, and the operating system
kernel. Samples are collected at a high rate (over 5200 samples/sec
per 333-MHz processor), yet with low overhead (1-3% slowdown for most
workloads).




Analysis tools supplied with the profiling system use the sample data
to produce a precise and accurate accounting, down to the level of
pipeline stalls incurred by individual instructions, of where time is
being spent. When instructions incur stalls, the tools identify
possible reasons, such as cache misses, branch mispredictions, and
functional unit contention. The fine-grained instruction-level
analysis guides users and automated optimizers to the causes of
performance problems and provides important insights for fixing them.
--- end abstract ---


The profiling tools can directly measure the number of cache misses,
etc. that occur down to a procedure level. Instruction-level stalls
are inferred by some clever analysis of the collected data.


See "http://www.research.digital.com/SRC/dcpi/" for more information,
as well as to download a copy of the tools.


-Sanjay Ghemawat (one of the many authors of the tools)
--

Return to the comp.compilers page.
Search the comp.compilers archives again.