Shouldn''t dtruss be using progenyof() instead of thread local variables set by fork for tracking child PIDs? I used the self->child method in a number of scripts I wrote before I discovered that I was missing the data from all but the first thread of child processes. Fortunately I came across progenyof(). Am I missing something here? -- This message posted from opensolaris.org
Brendan Gregg - Sun Microsystems
2007-Jun-30 02:15 UTC
[dtrace-discuss] DTraceToolkit Dtruss bug
G''Day Aaron, On Fri, Jun 29, 2007 at 06:59:01PM -0700, Aaron Gutman wrote:> Shouldn''t dtruss be using progenyof() instead of thread local variables set by fork for tracking child PIDs? I used the self->child method in a number of scripts I wrote before I discovered that I was missing the data from all but the first thread of child processes. Fortunately I came across progenyof().Yep. I wrote dtruss before I knew of progenyof(), so it probably should be using it. (I''ve added it to the todo list).> Am I missing something here?No, rather, you are the first not to miss picking up on this. :) cheers, Brendan -- Brendan [CA, USA]
Well...here was the lazy change I made =) 189c189 < (OPT_follow && progenyof(PID))/ ---> (self->child)/You use the self->child for printing so it wasn''t worth moving around that code... -- This message posted from opensolaris.org