aboutsummaryrefslogtreecommitdiff
path: root/tools/perf/scripts/python
diff options
context:
space:
mode:
authorAlexei Starovoitov <[email protected]>2019-04-01 21:27:42 -0700
committerDaniel Borkmann <[email protected]>2019-04-04 01:27:37 +0200
commit25af32dad8047d180e70e233c85b909dd6587cc5 (patch)
treef5a1e79c875c48395775bf94bfe217dc44effc17 /tools/perf/scripts/python
parent9f4686c41bdff051f557accb531af79dd1773687 (diff)
bpf: improve verification speed by not remarking live_read
With large verifier speed improvement brought by the previous patch mark_reg_read() becomes the hottest function during verification. On a typical program it consumes 40% of cpu. mark_reg_read() walks parentage chain of registers to mark parents as LIVE_READ. Once the register is marked there is no need to remark it again in the future. Hence stop walking the chain once first LIVE_READ is seen. This optimization drops mark_reg_read() time from 40% of cpu to <1% and overall 2x improvement of verification speed. For some programs the longest_mark_read_walk counter improves from ~500 to ~5 Signed-off-by: Alexei Starovoitov <[email protected]> Reviewed-by: Jakub Kicinski <[email protected]> Reviewed-by: Edward Cree <[email protected]> Signed-off-by: Daniel Borkmann <[email protected]>
Diffstat (limited to 'tools/perf/scripts/python')
0 files changed, 0 insertions, 0 deletions