I've just uploaded the latest diffs and binaries to www.opensolaris.org/os/community/dtrace/shells/.
So what changed?
There was a bug in that the
command-entry probe fired in both parent and child shell. This was a simple oversight that I should not have missed. I originally had the probe before sh forked, then moved it such that it fired after we knew we were able to fork (basically I didn't want it firing if we were not able to actually fork and run the command). Unfortunately, I forgot to specify that it was only to fire in the parent shell. Simple fix. My bad.
I haven't tested the diffs, but I did the same massaging to them that I did for the last lot, so they should be ok to use with gpatch.
Looks like things are starting to settle down now so I'll be able to think about progressing this one and starting to look at some others, using these probe names as some kind of standardisation.
There have been suggestions for extras in this provider, but the feeling that I'm getting is let's get a basic useful provider done as a v1.0 and look at things like stacks and other probes in a later update.