Re: The demise of utracer.

[Chris Moller]

Adam Jackson wrote:

On Tue, 2008-06-24 at 22:49 -0400, Chris Moller wrote:

Comments, up to and including "What the hell are you smoking?" welcome.

I'd rather phrase this as "what are we not getting out of ptrace that we
really want".  Answering that will frame how we go about getting it.  (I
admit to not having looked closely at utracer yet, so I'm not sure how
much this was done there.)

utracer was intended to be very much a "debugger assist" thing that consolidated normal ptrace capabilities with stuff like asynchronous event notification (all the various utrace report_* stuff) and much better thread support--e.g., it could be set to automatically attach to vfork()ed tasks, notify the app that that that had been done, and allow the app to control both parent and child tasks independently. I plan to implement all this stuff in son-of-utracer--actually, a lot of it can probably be lifted straight from the old code.

Two features I would like to see, more for debugger performance than

anything else, are direct API support for breakpoints

There was a bit of discussion a couple of months ago about breakpoint support in utrace, but I don't know if it's gone anywhere. If it gets real, it would certainly be high on the list of stuff I'd want to support.

and watchpoints.
The former could be implemented as single-stepping entirely within the
kernel and that'd be just fine, although running until illegal

instruction trap would be even better.

The discussion I'm thinking of (the SSOL stuff) seems to imply the latter, but I haven't looked that closely.

 Watchpoint support might require
the usual page protection tricks, which is admittedly non-trivial.  Both
of these seem like they slot in reasonably in the existing ptrace
interface, but that's very much a non-expert's opinion.

In terms of correctness it seems like at least gdb still has real
trouble handling multiple threads and processes sanely, but that could
easily be more about gdb's failings than those of ptrace.

- ajax

--
Chris Moller

 I know that you believe you understand what you think I said, but
 I'm not sure you realize that what you heard is not what I meant.
     -- Robert McCloskey

signature.asc
Description: OpenPGP digital signature