[Vm-dev] SystemTracing: SystemTracing-dtl.23.mcz

David T. Lewis lewis at mail.msen.com
Fri Dec 9 22:30:16 UTC 2011


Thanks Stef,

I'll try to look at it when I get a chance. I did a quick check with
Pharo 1.4, but I just did not have enough time to follow up on the issues.
I'll have a look at PharoTaskForces because I'm sure it's addressed there.

Dave

On Fri, Dec 09, 2011 at 10:52:44PM +0100, stephane ducasse wrote:
> 
> hi david
> 
> the systemTracing package for pharo is available in PharoTaskForces
> I do not know what changed compared with the one of squeak.
> 
> Stef
> 
> On Dec 8, 2011, at 6:49 PM, David T. Lewis wrote:
> 
> > 
> > This is an update to the system tracer that makes it possible to
> > trace current Squeak images to 64-bit image format. Prior to this
> > update, only pre-closure images could be converted to 64-bit object
> > format.
> > 
> > Thanks to Eliot for the tip that pointed me to the solution, which
> > is that instances of BlockClosure need to have the value of their
> > startpc instance variable adjusted to match byte offsets in the
> > enclosing compiled method when converting from 32-bit to 64-bit
> > object format.
> > 
> > This works on Squeak, not yet tested on Pharo.
> > 
> > Dave
> > 
> > 
> > On Thu, Dec 08, 2011 at 05:17:49AM +0000, squeak-dev-noreply at lists.squeakfoundation.org wrote:
> >> Dave Lewis uploaded a new version of SystemTracing to project SystemTracing:
> >> http://www.squeaksource.com/SystemTracing/SystemTracing-dtl.23.mcz
> >> 
> >> ==================== Summary ====================
> >> 
> >> Name: SystemTracing-dtl.23
> >> Author: dtl
> >> Time: 7 December 2011, 11:17:47 am
> >> UUID: 9f6136af-8117-416c-bf54-baf2a32bb571
> >> Ancestors: SystemTracing-dtl.22
> >> 
> >> Support tracing closure-enabled images to 64-bit object format.
> >> 
> >> When tracing to a different word size, e.g. trace to 64-bit image from a 32-bit image, adjust the startpc of all BlockClosure instances to account for different word size in the enclosing compiled method.


More information about the Vm-dev mailing list