diff options
author | Adrian Szyndela <adrian.s@samsung.com> | 2022-04-29 08:58:42 +0200 |
---|---|---|
committer | Adrian Szyndela <adrian.s@samsung.com> | 2022-04-29 08:58:42 +0200 |
commit | 2091493e996ce159cb5d92adcf10def691214881 (patch) | |
tree | 75437c52e57c19fb1f71d4bb8e4fcd97735f7714 /README.cmake | |
parent | dd366f2e664d660084fcbaaba30ae6c8dc8cdd4a (diff) | |
download | dbus-2091493e996ce159cb5d92adcf10def691214881.tar.gz dbus-2091493e996ce159cb5d92adcf10def691214881.tar.bz2 dbus-2091493e996ce159cb5d92adcf10def691214881.zip |
bus/policy: separate prefix rules in default contextsubmit/tizen/20220525.001052submit/tizen/20220502.052537accepted/tizen/unified/20220526.144439
To handle prefix rules stored with all other rules in the default context
we need to match each prefix of each name against policy rules.
That's because names are looked up in the hash tables, so we can
miss a prefix rule for a prefix of the name.
However, if prefix rules are separated from non-prefix rules, we
can simply check them all once for each name, and also check hash tables once
for each name.
This is what this commit changes. It separates prefix rules
from non-prefix rules, and handles them in sequence.
This gives a little boost, especially if there are no prefix rules.
Change-Id: Ifade906d35af96a973920ce9c2f6065f5b9b549e
Diffstat (limited to 'README.cmake')
0 files changed, 0 insertions, 0 deletions