[Openmcl-devel] Shark profiling works with openmcl!

[Hamilton Link]

  No, that's what the call to PURIFY is for, it puts all the code 
vectors in a non-gced place so they won't move around.
h

On Oct 17, 2004, at 9:57 PM, Alan Ruttenberg wrote:

> Cool! I've been waiting for this for a while.
> Am I correct to presume that a gc in the middle of profiling will mess 
> it up?
> -Alan
> On Oct 17, 2004, at 4:38 PM, Hamilton Link wrote:
>
>> Attention all, good news! I have gone back and forth with the shark 
>> people at Apple, I sent them a shrink-wrapped bug that was preventing 
>> shark from reporting lisp symbol names for functions, and... they 
>> fixed the problem!
>>
>> Here is the file I use to demonstrate how to use shark to profile 
>> your code, a screen shot of what the result looks like, and the 
>> .spatch file that was generated in the process (for a syntactic 
>> example). Obviously you'll want to do things a bit differently, and 
>> if I come up with a swanky API for putting a set of packages and 
>> other symbols and named lambdas into a .spatch file before someone 
>> else does, I'll post that too.
>>
>> (btw, sharkloopfn in the image is just a function I defined as a test 
>> that was guaranteed to bubble to the top, it has nothing to do with 
>> interfacing to shark)
>>
>> This works as of 4.0.0b11 as far as the CHUD/Shark people say, and 
>> I've confirmed that it works with the GM release, 4.0.0.
>>
>> Happy profiling...
>>
>> h
>> <sharktest.lisp><dppccl_879.spatch>
>> <sharkworking.tiff>_______________________________________________
>> Openmcl-devel mailing list
>> Openmcl-devel at clozure.com
>> http://clozure.com/mailman/listinfo/openmcl-devel
>
> _______________________________________________
> Openmcl-devel mailing list
> Openmcl-devel at clozure.com
> http://clozure.com/mailman/listinfo/openmcl-devel
>