dtrace: prevent forked child from running after an error condition

[vangyzen]

Summary

The pid/killonerror test uses an invalid probe specifier to verify that
the child process is killed. It occasionally fails because the "date"
command is allowed to run long enough to print the date. This is harmless
in this case, but is clearly not ideal.

When the dt_proc_control thread is about to exit, and the dtrace command
forked the child, do not make the child runnable.

Test Plan

I ran the test in a tight loop until failure, then repeated for a few
dozen loops. Before the change, every loop failed, and the iterations were
almost always less than 10, with a maximum of around 50. After the change,
the first-and-only loop passed over 5,000 iterations.

I'll also run the entire DTrace test suite.

[markj]

Seems reasonable.

cddl/contrib/opensolaris/lib/libdtrace/common/dt_proc.c

693

Maybe "an initialization error"? I'd have a hard time understanding what's meant by a command-line error if I were just reading the code.

[vangyzen]

cddl/contrib/opensolaris/lib/libdtrace/common/dt_proc.c

693

Yes, that sounds much better.

[vangyzen]

• wordsmith comment

[vangyzen]

The DTrace test suite had fewer failures with this change. They were not fixed by this change; they're just unreliable.