Dtrace woes

[blkblt]

I have an older Pentium D940 CPU, and whenever I run dtrace, I get an error:

 

$ sudo dtrace -n 'syscall:::return { total[probename] = 1; }'

dtrace: invalid probe specifier syscall:::return { total[probename] = 1; }: "/usr/lib/dtrace/darwin.d", line 26: syntax error near "uthread_t"

 

Can someone else with that sort of processor give it a try and see if they get the same error?

 

Thanks,

 

/blkblt

[blagec]

I have an older Pentium D940 CPU, and whenever I run dtrace, I get an error:

Can someone else with that sort of processor give it a try and see if they get the same error?

 

Hi,

 

I have Pentium D 915, DTrace likes that CPU, and D 940 is probably OK!

 

Buuuuut... While I was on Kalyway 9a527 Leo beta (upgraded to 528d I think) DTrace/XRay worked great.

But, when I switched to ToH RC2 10.5.0 (9a581), and than to 10.5.1 (regular software update - very few .kexts are changed from 10.5.0 to 10.5.1 btw DTrace broke - got the same error message (line 26)

 

AFAIK, the cause of this behavior is missing file: mach_kernel.ctfsys in / dir (it should reside in / along with mach_kernel). I placed that file into the / folder, and voila! DTrace is working again. I don't know the exact purpose of .ctfsys file (but it is some kind of non-slimmed x86+PPC kernel image) so please someone explain ...

 

I even use cftsys from 9a528d installation in 10.5.1 (as I was lazy/couldn't find file on ToH disc?) and it works for me (correct me if I'm wrong).

 

My kernel is: Darwin aeroflot.local 9.1.0 Darwin Kernel Version 9.1.0: Fri Nov 16 14:50:43 SCT 2007; made by ToH:xnu-1228.0.2~1/BUILD/obj/RELEASE_I386 i386

 

Best regards,

blagec