The LRO possible test was calling CURVNET_SET once for IPv4
and once for IPv6 for each packet in a chain. Only call it once per
chain instead.
Details
Details
- Reviewers
sbruno cem kmacy - Commits
- rS326577: Avoid calling CURVNET_[SET|RESTORE] for each packet
Check for improved performance
Diff Detail
Diff Detail
- Repository
- rS FreeBSD src repository - subversion
- Lint
Lint Passed - Unit
No Test Coverage - Build Status
Buildable 13333 Build 13566: arc lint + arc unit
Event Timeline
sys/net/iflib.c | ||
---|---|---|
2474 | Both conditions should be guarded with #ifdef INET and #ifdef INET6. |
Comment Actions
Integrate feedback from ae and cem
Rename functions and variables to better indicate what the represent
rather than what they're used for. Simplify logic, check INET and INET6
definitions.
Comment Actions
A static analyzer may reject passing uninitialized values to iflib_check_lro_possible in the case that one of INET or INET6 is disabled. But I don't think there is any real problem as the uninitialized value is unused.