MicroVAX3900 aborts on Widows due to sim_idle() - waiting too long

[wpappas]

When running VMS in the MicroVAX3900 simulator under Windows 10 with idle enabled, it will eventually crash with the a message similar to the following:
sim_idle() - waiting too long: w_ms=1253 usecs, w_idle=973 usecs, sim_interval=6265, rtc->currd=6435
sim_idle() - waiting too long - aborting from C:...\sim_timer.c:1792

It seem be triggered by heavy IO activity and most recently happened while running a DEC Windows session on the host PC.

Show Version:
MicroVAX 3900 simulator V4.0-0 Current
Simulator Framework Capabilities:
64b data
64b addresses
Threaded Ethernet Packet transports:PCAP:NAT:UDP
Idle/Throttling support is available
Virtual Hard Disk (VHD) support
RAW disk and CD/DVD ROM support
Asynchronous I/O support (Lock free asynchronous event queue)
Asynchronous Clock support
FrontPanel API Version 15
Host Platform:
Compiler: Microsoft Visual C++ 19.29.30152.00
Simulator Compiled as C arch: x86 (Release Build) on Oct 12 2023 at 19:34:29
Build Tool: simh-Visual-Studio-Project
Memory Access: Little Endian
Memory Pointer Size: 32 bits
Large File (>2GB) support
SDL Video support: SDL Version 2.0.20, PNG Version 1.6.37
PCRE RegEx (Version 8.43 2019-02-23) support for EXPECT commands
OS clock resolution: 1ms
Time taken by msleep(1): 1ms
WinEditLine Version: 2.208
Ethernet packet info: NAT, UDP, PCAP: Npcap version 1.71, based on libpcap version 1.10.2-PRE-GIT
OS: Microsoft Windows [Version 10.0.19045.3570]
Architecture: x86 on AMD64, Processors: 4
Processor Id: Intel64 Family 6 Model 42 Stepping 7, GenuineIntel, Level: 6, Revision: 2a07
Processor Name: Intel(R) Core(TM) i5-2520M CPU @ 2.50GHz
tar tool: bsdtar 3.5.2 - libarchive 3.5.2 zlib/1.2.5.f-ipp
curl tool: curl 8.0.1 (Windows) libcurl/8.0.1 Schannel WinIDN
git commit id: 985137c+uncommitted-changes
git commit time: 2023-10-06T13:40:11-10:00

• how you built the simulator or that you're using prebuilt binaries

Standard, no debug build using Visual Studio 2019 and sources recently pulled from the git repository

• the simulator configuration file (or commands) which were used when the problem occurred.

set cr disable
set dz disable
set vh disable
set rl disable
set rqb disable
set rqc disable
set rqd disable
set ts disable
set xqb disable
attach nvr 3900.nvr
set cpu 64m
set cpu idle=vms:90
set rq0 ra92
set rq1 ra92
set rq2 ra92
set rq3 cdrom
set rq3 locked
attach rq0 d0.dsk
attach rq1 d1.dsk
attach rq2 d2.dsk
set tq tk70
set tq2 disabled
set tq3 disabled
attach xq nat:tcp=23:10.0.2.4:23,tcp=22:10.0.2.4:22,udp=177:10.0.2.4:177
attach lpt LPA0.txt
dep bdr 80
set cpu conhalt
;set cpu simhalt
set console brk=10
boot cpu

• the expected behavior and the actual behavior

sim_idle() - waiting too long: w_ms=1253 usecs, w_idle=973 usecs, sim_interval=6265, rtc->currd=6435
sim_idle() - waiting too long - aborting from C:...\sim_timer.c:1792

• you may also need to provide specific pointers to data files that may be necessary to demonstrate the problem

[markpizz]

Hi there Bill,

Couple of comments/questions about your config:

• You've got: set cpu 64m. Why so little RAM. You certainly should be able to run with set cpu 256m.

• You've got: set cpu idle:90. Why did you pick the :90? How long does the simulator need to be running before you see the "sim_idle() - waiting too long" message?

• You've got: dep bdr 80. This is equivalent to set cpu noautoboot which is the default and thus not needed, however set cpu noautoboot makes your intention more clear to the casual observer.

• The next time you see this problem, please return the full "waiting too long" message output and then enter the following commands and return their output as well:

    sim> EXAMINE IDLE_CYC_MS
    sim> SHOW CLOCK

The version you're running is: git commit id: 985137c+uncommitted-changes, to be sure that any uncommitted changes are irrelevant, please try to reproduce this with the latest prebuild windows binaries at https://github.com/simh/Development-Binaries

Thanks for this report.

• Mark

[wpappas]

Hi Mark, Thanks for the speedy response. My comments below: - ​64M works just fine for what I've been running, but I can bump it up to see if it matters - ​the IDLE=VMS:90 is a leftover from seeing if it makes a difference when booting. It doesn't and I can produce the issue without setting to delay to anything - The dep BDR 80 seems to be necessary to enable the console break to boot rom functionality even though set noautoboot seems to deposit 80 in BDR. Can you reproduce that behavior? - I can't do anything with SimH after this happens as it pops up a windows message saying the programs has terminated abnormally and do I want to run the debugger. At this point the SimH window becomes unresponsive. I've done a debug build to see if I can get more info next time it happens. What I posted is the full extent of SimH's console output when it crashes - I am not sure what the uncommitted changes are all about as this was a "clean" git pull and build. It could be due to me having to stash my project files to keep them from getting clobbered. I'll run with the builds provided to see if it makes any difference Finally I am pretty sure I can get this to happen pretty consistently with a simple two node cluster setup. If so I can package up the setup and send to you to see if you can reproduce it. Thanks, Bill

…

________________________________ From: Mark Pizzolato ***@***.***> Sent: Saturday, October 28, 2023 1:09 AM To: simh/simh ***@***.***> Cc: Bill Pappas ***@***.***>; Author ***@***.***> Subject: Re: [simh/simh] MicroVAX3900 aborts on Widows due to sim_idle() - waiting too long (Issue #1187) Hi there Bill, Couple of comments/questions about your config: * You've got: set cpu 64m. Why so little RAM. You certainly should be able to run with set cpu 256m. * You've got: set cpu idle:90. Why did you pick the :90? How long does the simulator need to be running before you see the "sim_idle() - waiting too long" message? * You've got: dep bdr 80. This is equivalent to set cpu noautoboot which is the default and thus not needed, however set cpu noautoboot makes your intention more clear to the casual observer. * The next time you see this problem, please return the full "waiting too long" message output and then enter the following commands and return their output as well: sim> EXAMINE IDLE_CYC_MS sim> SHOW CLOCK The version you're running is: git commit id: 985137c+uncommitted-changes, to be sure that any uncommitted changes are irrelevant, please try to reproduce this with the latest prebuild windows binaries at https://github.com/simh/Development-Binaries Thanks for this report. * Mark — Reply to this email directly, view it on GitHub<#1187 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AKLCEWGU7D3EN3KGENRU53LYBSHPXAVCNFSM6AAAAAA6TS6YP6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOBTG4YDINBRGU>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[markpizz]

Hi Bill,

I tried to take this offline a few days back and sent email to an eMail I had for you at outlook.com. Do you still use/check that email? If not please contact me directly at mark@infocomm.com.

Thanks,

• Mark