Forum Discussion

Altera_Forum's avatar
Altera_Forum
Icon for Honored Contributor rankHonored Contributor
17 years ago

Internal Error: SubSystem VPR20KMAIN

Hi All:

I've just recently ran into a problem about 65% though the fitter stage of my design:

Internal Error: Sub-system: VPR20KMAIN, File: /quartus/fitter/vpr20k/altera_arch_common/altera_arch_delay_cache.c, Line: 722

I've submitted a service request to Altera on this, But I was wondering if any of you have seen this error before, and know of a good resolution:

Running Quartus II 7.2 SP3 on Windows XP Pro. Licensed version.

It's an SOPC system design targeting a Cyclone III EP3C55U484C6.

Thanks

Pete

6 Replies

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    17 years ago

    Ok, I did the normal bug work arounds:

    Reboot the computer, and deleted the db directory.

    Now the error occurs at 45% in the Fitter stage, and the message has changed slightly.

    Internal Error: Sub-system: VPR20KMAIN, File: /quartus/fitter/vpr20k/altera_arch_common/altera_arch_place_bounding_box.c, Line: 2586

    Internal Error

    Stack Trace:

    End-trace

    Quartus II Version 7.2 Build 207 03/18/2008 SJ Full Version

    Service Pack Installed: 3

    Any ideas would be appreciated.

    Pete
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    17 years ago

    Hey Pete, I had a look at both of those errors and they're completely unrelated to one another, which is unusual (Internal Errors are almost always easily reproducible).

    Given that the error changes when you reboot your computer, I'm wondering if you might have some faulty RAM. Why don't you give the MS RAM checker (http://oca.microsoft.com/en/windiag.asp) a spin and report back here? If it doesn't turn up anything we'll have a closer look.

    Cheers, Adrian

    Altera Corp.
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    17 years ago

    Hi Adrian:

    Thanks for the reply. I did have a third message as well, related to the TimeQuest engine during fitter. There are two of us who have had this issue, and it seems it might be related to the Cisco VPN client we're using.

    I disabled the VPN during the fitter stage and it seem to complete OK.

    I'll run the ram checker over the weekend and report the results. I'll also let you know if I can find anything more definite with the issue.

    It's not completely random like I would expect from memory fault, but it's random enough to make me scratch my head.

    (I received the line 722 error at least 3 times now with reboots in-between)

    Pete
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    17 years ago

    Hi Pete, what's the TimeQuest error? The Fitter shouldn't be hitting the network (other than to check your license when it starts) so that seems kind of weird.

    Thanks, Adrian
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    17 years ago

    Hi Adrian:

    I can't give you that exact error but it was similar if not identical to:

    Internal Error: Sub-system: TDC, File: /quartus/tsm/tdc/tdc_accessories_analysis.cpp, Line: 1571

    m_edge_domain_data[iedge].m_relaxed_domain_slack_ratio < JUST_BELOW_HUGE_FLOAT

    Stack Trace:

    0x59E2 : _Dinkum_std::hash_compare<TDC_PATH_TIMING *,_Dinkum_std::less<TDC_PATH_TIMING *> >::operator() + 0x3B62 (TSM_TDC)

    End-trace

    This particular error came from a co-worker that was experiencing a similar issue.

    Like I said before it seems to only occur when the Cisco VPN was active. I have since upgraded the VPN client from 4.8 to 5.0 and now the problem seems to have gone away.

    I've told my co-worker to also upgrade his vpn client, in the hopes that this solve the issue.

    Thanks for the assistance!

    Pete
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    16 years ago

    Hello Searchers,

    I experienced a VERY similar problem with Quartus II 8.1, again an SOPC system targeting EP3C55U484C6, and it was also fixed by upgrading Cisco VPN from 4.8 to 5.0.

    I installed Quartus 8.1 on a second build machine, and on the 2nd machine my design always crashed. Pulling design files and scripts from revision control system so very confident that source files were the same. Spent time checking assignment_defaults.qdf, environment variables, etc. to no avail.

    The 2nd machine gave several different random errors, including VPR20K main subsystem, etc. After uninstall/reinstall several times, it began to consistently fail with:

    Internal Error: Sub-system TDC, File: /quartus/tsm/tdc/tdc_accessories_analysis.cpp

    sub_domain_slack_ratio < JUST_BELOW_HUGE_FLOAT.

    Cisco VPN 4.8 was the culprit. No idea what is going on that causes Quartus to interact with VPN client (license check ? something more mysterious ?).

    -- JMS