#1181: GCOV compilation option #1219
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…o build http_parser.c with "profile-use=/root/tempesta/http_parser.c.gcov" in #pragma GCC optimize. http_parser.c.gcov is generated by $ cd ~/tempesta $ ./scripts/gcov_gather.sh /tmp/gcov.tgz $ cd /tmp && tar -zxf gcov.tgz && cd - $ gcov -o /tmp/sys/kernel/debug/gcov/root/tempesta/tempesta_fw http_parser.c perf(1) also can be used for the profile generation, see -fauto-profile GCC command line option. PGO generates slightly better code layout, but the profile counts total samples and doesn't record sequence of called code. E.g. for a normal HTTP request the parser spends more time in URI processing than in method processing, so PGO moves URI before method processing which is completely wrong. The second drawback of PGO is that each code update requires regeneration of the profile. Moreover, in general it's tricky to correctly choose the right workload for the profile, so it looks more like workload programming than the parser programming. All in all, I leave the GCOV compilation option only for the test coverage and we need to use more flexible way to generate proper code layout.
i-rinat
approved these changes
Mar 24, 2019
scripts/gcov_gather.sh
Outdated
| #!/bin/bash | ||
| # | ||
| # Gcov gather script for Tempesta FW data. | ||
| # Mainly bothered from linux/Documentation/dev-tools/gcov.rst |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The option is useful for profile guided optimization (PGO). I tried to build
http_parser.c with "profile-use=/root/tempesta/http_parser.c.gcov" in #pragma
GCC optimize. http_parser.c.gcov is generated by
$ cd ~/tempesta
$ ./scripts/gcov_gather.sh /tmp/gcov.tgz
$ cd /tmp && tar -zxf gcov.tgz && cd -
$ gcov -o /tmp/sys/kernel/debug/gcov/root/tempesta/tempesta_fw http_parser.c
perf(1) also can be used for the profile generation, see -fauto-profile GCC
command line option.
PGO generates slightly better code layout, but the profile counts total
samples and doesn't record sequence of called code. E.g. for a normal HTTP
request the parser spends more time in URI processing than in method
processing, so PGO moves URI before method processing which is completely
wrong.
The second drawback of PGO is that each code update requires regeneration of
the profile. Moreover, in general it's tricky to correctly choose the right
workload for the profile, so it looks more like workload programming than the
parser programming.
All in all, I leave the GCOV compilation option only for the test coverage and
we need to use more flexible way to generate proper code layout.