Continue scheduler / queue optimization with pktin queue poll optimization. Remove special packet IO poll command and use queue instead. Optimize atomic pktin queue throughput by passing packets directly to application. Atomic queue packet rate : +18 ... 100% Parallel queue packet rate : + 7 ... 73% Especially, atomic queue packet rate is close to direct mode figures (<10% less than direct mode).
----------------github------------------------ /** Email created from pull request 504 (psavol:master-sched-optim-2) ** https://github.com/Linaro/odp/pull/504 ** Patch: https://github.com/Linaro/odp/pull/504.patch ** Base sha: 284f52d72ec19df3774c7409780f1f9eea33b8e6 ** Merge commit sha: 081cdd2f01d5a4e5b5b4d254f689e55d60a489cd **/ ----------------/github------------------------ ----------------checkpatch.pl------------------------ total: 0 errors, 0 warnings, 0 checks, 600 lines checked to_send-p-000.patch has no obvious style problems and is ready for submission. total: 0 errors, 0 warnings, 0 checks, 314 lines checked to_send-p-001.patch has no obvious style problems and is ready for submission. total: 0 errors, 0 warnings, 0 checks, 112 lines checked to_send-p-002.patch has no obvious style problems and is ready for submission. ----------------/checkpatch.pl------------------------
