Q^6.1 Fitter dies

I too was seeing this on a dual core machine. I haven't been able to reproduce it in a while to figure out what is causing it.

Can you please email a brief system description of your system (processor, memory, and OS).

Shure.

It is a Dell Optiplex GX620 (it is a very silent but incredible fast system)

P4 HT 670 Prescott 3.80GHz FSB800

Family F Model 4 Stepping 3

L1 16KB L2 2MB Cache

MainBoard is a Dell Model 0F8096 with a i945G/GZ Rev A2 Chipset and a 82801GB ICH/ Southbridge

Memory is 3GB DDR2 (Dell Default)

Video is a ATI Radeon X600 Dual

All drivers are up2date.

The OS is a Windows XP with the lastest SP installed.

Regards.

Michael

just some more comments about it.

this behavioral of a dieing Q6.1 is not related to a specific design. it happens here will ALL designs. it happens every now and then. but always at 96%.

but i haven't seen it if i set the number of cores to 1.

we have another P4 server with HT or real dual core (not shure at the moment). today i will start some complete compiles there and see if it happens there too.

With MaxII designs, it happens that Q6.1 reports at 81% of the Fitter :

Internal Error: Sub-system: FMAIN, File: /quartus/fitter/fmain/fmain_location_constraint.cpp, Line: 3024

m_soft_locked_constraint.check_constraint( other_constraint->m_soft_locked_constraint)

Stack Trace:

0xA5CD : FITCC_CLOCK_REGION_CHECKER::set_oterm_alias + 0x26D6 (FITTER_FITCC)

0xC2DB : FITCC_CLOCK_REGION_CHECKER::clear_usage_for_all_oterms_on_atom + 0xE28 (FITTER_FITCC)

0x18DDA : FITCC_DESIGN::are_atom_location_constraints_mergeable + 0x132 (FITTER_FITCC)

0x1579A : fygr_execute + 0x1655 (fitter_fygr)

0x16FE2 : fygr_execute + 0x2E9D (fitter_fygr)

0xC57E1 : FYGR_TEMPLATE_REGISTER_PACKING_OP::find_shared_input_lonely_register_matings + 0x206 (fitter_fygr)

End-trace

Quartus II Version 6.1 Build 201 11/27/2006 SJ Full Version

Intresting that MaxII freezes at 81% .... if it freezes.

are there any new experiences regarding this topic - is there a way to avoid this fitter locking ?

(it just happened three times in a row now ...)

thanks

Klaus

Hi Stonie,

there is currently only one solution known to me ...

just set max number of processors allowed back to 1.

But most of the time the fitter runs fine with more than 1.

So if your fitter dies too, just set this setting back to 1 re-run the fitter and after a successfull job, set it back und run with all processors you have.

2 processors saves about 30% of the overall fitting time here.

I also have this issue, thanks for the workaround, my system specs.

Intel E660

2 Gig DDR2 6400

Gigabyte 965P-DS3

Nvidia 7600GT

windows xp sp2 with all updates.

Thanks,

Richard

This week i had asked mysupport if a fix for this problem is available bevor the next service pack. As i had noticed this beginning of december and thought that now 2 months later a fix could be available.

but this is the answer of mysupport :

"Unfortunately, we don’t have such a fix since we didn’t receive similar issue report before." http://forum.niosforum.com/work2/style_emoticons/<#EMO_DIR#>/blink.gif

Does this mean nobody has reported this to mysupport ? Well i havn&#39;t. But does mysupport not read this forum ?

I hope Altera is aware of this issue and has a fix that will be available soon.