diff options
author | SZEDER Gábor <szeder.dev@gmail.com> | 2023-04-20 22:53:50 +0200 |
---|---|---|
committer | Junio C Hamano <gitster@pobox.com> | 2023-04-20 14:53:00 -0700 |
commit | 138ef8068c54d72c9bd1b09753408fde7750ec86 (patch) | |
tree | 57219d1469ee85975fda8c58b4892f53fa6915cb /builtin/commit-graph.c | |
parent | 667fcf4e15379790f0b609d6a83d578e69f20301 (diff) |
cocci: remove 'unused.cocci'
When 'unused.cocci' was added in 4f40f6cb73 (cocci: add and apply a
rule to find "unused" strbufs, 2022-07-05) it found three unused
strbufs, and when it was generalized in the next commit it managed to
find an unused string_list as well. That's four unused variables in
over 17 years, so apparently we rarely make this mistake.
Unfortunately, applying 'unused.cocci' is quite expensive, e.g. it
increases the from-scratch runtime of 'make coccicheck' by over 5:30
minutes or over 160%:
$ make -s cocciclean
$ time make -s coccicheck
* new spatch flags
real 8m56.201s
user 0m0.420s
sys 0m0.406s
$ rm contrib/coccinelle/unused.cocci contrib/coccinelle/tests/unused.*
$ make -s cocciclean
$ time make -s coccicheck
* new spatch flags
real 3m23.893s
user 0m0.228s
sys 0m0.247s
That's a lot of runtime spent for not much in return, and arguably an
unused struct instance sneaking in is not that big of a deal to
justify the significantly increased runtime.
Remove 'unused.cocci', because we are not getting our CPU cycles'
worth.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Diffstat (limited to 'builtin/commit-graph.c')
0 files changed, 0 insertions, 0 deletions