Closed gfr10598 closed 3 years ago
I've pushed it manually to sandbox ndt pods.
Changes Missing Coverage | Covered Lines | Changed/Added Lines | % | ||
---|---|---|---|---|---|
tracer/scamper.go | 66 | 68 | 97.06% | ||
<!-- | Total: | 66 | 68 | 97.06% | --> |
Files with Coverage Reduction | New Missed Lines | % | ||
---|---|---|---|---|
tracer/scamper.go | 4 | 93.51% | ||
<!-- | Total: | 4 | --> |
Totals | |
---|---|
Change from base Build 301: | -0.9% |
Covered Lines: | 579 |
Relevant Lines: | 591 |
tracer/scamper.go, line 83 at r1 (raw file):
// called binaries have a much smaller "blast radius". func (s *Scamper) Trace(conn connection.Connection, t time.Time) (out []byte, err error) { // XXX Is this still useful? Does shx.PipeJob we ever panic?
shx does not panic, so this recover() should never happen. But I suggest we change (correct) error/panic/recovery handling in a separate PR.
tracer/scamper.go, line 180 at r1 (raw file):
// connection. func (d *ScamperDaemon) Trace(conn connection.Connection, t time.Time) (out []byte, err error) { // XXX Is this still useful? Does shx.PipeJob we ever panic?
See my previous comment.
tracer/scamper.go, line 272 at r1 (raw file):
} return buff.Bytes(), err } else {
Please remove this unnecessary else.
This is not a strict refactoring, but is not intended to change behavior in any way.
Passes unit tests. We should probably also run it in sandbox manually.
This change is