[Simh] Bug in pdp11 simulator?
Bob Supnik
bob at supnik.org
Fri May 4 20:37:14 EDT 2012
I don't see the problem on RSTS/E 10.1 with the released code. I've
omitted some blank lines for concision.
Note that I am using my Visual Studio debug version, which is totally
unoptimized, but I get <almost> the same results with gcc and MINGW and
the optimized make file (although the simulator is noticeably faster).
---
PDP-11 simulator V3.9-0
sim> set pclk enabled
sim> set pclk 60
sim> set clk 60
sim> sho pclk
PCLK, address=17772540-17772545, vector=104, 60 Hz
sim> sho clk
CLK, 60Hz, address=17777546-17777547, vector=100
sim> set cpu 2m
Disabling CR
Disabling RK
Disabling HK
Disabling TM
sim> att -e rq c:\temp\rsts101_rqtq.dsk
sim> sho clk
CLK, 60Hz, address=17777546-17777547, vector=100
sim> boot rq
RSTS V10.1-L RSTS (DU0) INIT V10.1-0L
Today's date? 12\21\4-MAY-12
Current time? 20:19
Start timesharing? <Yes>
04-May-12 08:19 PM
6 devices disabled
Proceed with system startup? <YES>
Beginning RSTS/E system startup...
04-May-12 08:19 PM Installing monitor overlays
04-May-12 08:19 PM Mounting disks
04-May-12 08:19 PM Assigning logical names
04-May-12 08:19 PM Starting error logging
04-May-12 08:19 PM Setting system characteristics
04-May-12 08:19 PM Installing run-time systems and libraries
04-May-12 08:19 PM Starting Operator/Message Services
>>>>>>>>>>>>>>> OMS V10.1-A 04-May-12 08:19 PM <<<<<<<<<<<<<<<
Message 4 from OMS, user [1,2], Detached, job 3
Starting Operator/Message Services
04-May-12 08:19 PM Setting terminal characteristics
04-May-12 08:19 PM Defining system commands
04-May-12 08:19 PM Setting printer characteristics
04-May-12 08:19 PM Starting spoolers
*** From [1,2] "[ 1, 2]" on KB0: at 08:19 PM 04-May-12
** RSTS/E is on the air...
>>>>>>>>>>>>>>> OMS V10.1-A 04-May-12 08:19 PM <<<<<<<<<<<<<<<
Message 5 from user [1,2] on _KB0:, job 2
The system startup is complete
Simulation stopped, PC: 104040 (SOB R1,104036)
sim> sho clk
CLK, 60Hz, address=17777546-17777547, vector=100
sim> c
RSTS V10.1-L 04-May-12 08:19 PM
User: 1,2
Password:
Jobs detached under this account:
Job What Size State Run-time RTS
1 ERRCPY 5K SR 0.0 ...RSX
3 OMS 9K SL 0.0 ...RSX
4 PBS... 19K SL 0.0 ...RSX
Job number to attach to? 0
Last interactive login on 04-May-12, 08:19 PM at KB0:
Last non-interactive login on 04-May-12, 08:19 PM
3 other users are logged in under this account
Your password has not been changed in over 365 days.
Please change it.
$
Simulation stopped, PC: 104040 (SOB R1,104036)
sim> sho clk
CLK, 60Hz, address=17777546-17777547, vector=100
sim> c
$ run shutup
SHUTUP V10.1-H RSTS V10.1-L RSTS/E V10.1
08:19 PM 04-May-12 ######## Set-up Dialogue Phase ########
Type '^' to any query to backup one question
Allow Print/Batch Services jobs to complete <YES>?
Minutes until system shutdown (0-99) <5>? 0
Shutdown with automatic RESTART <NO>? no
08:19 PM 04-May-12 ######## Warning Message Phase ########
Further LOGINs are now disabled
08:19 PM 04-May-12 ######## Initial Job Killing Phase ########
08:19 PM 04-May-12 ######## Print/Batch Shutdown Phase ########
Print/Batch Services shutdown complete at 08:19 PM
08:19 PM 04-May-12 ######## OMS Shutdown Phase ########
>>>>>>>>>>>>>>> OMS V10.1-A 04-May-12 08:19 PM <<<<<<<<<<<<<<<
Message 6 from SHUTUP, operator [1,2] on _KB0:, job 2
Operator/Message Services shutting down
Operator/Message Services shutdown complete at 08:19 PM
08:19 PM 04-May-12 ######## EVTLOG Shutdown Phase ########
08:19 PM 04-May-12 ######## ERRCPY Shutdown Phase ########
08:19 PM 04-May-12 ######## Remove RTS/RES LIB Phase ########
08:19 PM 04-May-12 ######## Log File Closing Phase ########
08:19 PM 04-May-12 ######## SWAP File Removal Phase ########
08:19 PM 04-May-12 ######## Disk DISMOUNT Phase ########
08:19 PM 04-May-12 ######## Final Shutdown Phase ########
Please wait for system to re-boot itself
RSTS V10.1-L RSTS (DU0) INIT V10.1-0L
04-May-12 08:19 PM
Start timesharing? <Yes>
Simulation stopped, PC: 005264 (BCC 5310)
sim> sho clk
CLK, 60Hz, address=17777546-17777547, vector=100
sim>
-------
There is one difference. On Visual Studio, idling works reasonably well.
On MingW, it doesn't work at all. However, if PCLK is left turned off
(as it should be), the idling works correctly on both Visual Studio and
MingW.
So...
1) How was this version compiled?
2) Was the CPU model number changed from its default of 11/73?
3) Which release of RSTS/E was being used?
/Bob
On 5/4/2012 12:27 PM, Mark Pizzolato - Info Comm wrote:
> Bob,
>
> Jordi's idle issue is with the released v3.9-0 version.
>
> I'm not a PDP11 user, and I know next to nothing about PDP11 OSes.
>
> Therefore, I really can't address this issue with authority, however I looked at the running system when this happens.
>
> His ini file does the following:
>
> set pclk enable
> set clk 60
> set pclk 60
>
> Before he boots the clocks look like this:
> sim> sh clk
> CLK, 60Hz, address=17777546-17777547, vector=100
> sim> sh pclk
> PCLK, address=17772540-17772545, vector=104, 60 Hz
> sim> b rp0
>
> RSTS/E requires a clock, enabling KW11-L at 60Hz.
> KW11-L doesn't work.
>
> Thinking this looks strange, I hit Ctrl-E
>
> Simulation stopped, PC: 062454 (CMP R5,(R0))
> sim> sh clk
> CLK, 50Hz, address=17777546-17777547, vector=100
> sim> sh pclk
> PCLK, address=17772540-17772545, vector=104, 60 Hz
>
> I don't know what the message means, but the clock rate seems to have changed from 60hz to 50hz. It would seem that this could cause these two clocks to have strange event timing interactions which could cause the logic in sim_idle to not actually idle....
>
> I have no idea where to go from here.
>
> - Mark
>
> On Friday, May 04, 2012, at 3:25 AM, Jordi Guillaumes i Pons wrote:
>> Hi everyone.
>>
>> Congratulations to Bob, Mark and everyone involved in this new release.
>> The networking new features are really outstanding.
>>
>> Unfortunately, I've jumped in whats seems to be a bug, present in the final
>> release but, weird enough, not apreciable in the 3.9-0-rc1 branch.
>> When I try to boot RSTS/E V10 using a pdp11 emulator built from the git
>> repository the simulator hangs and I have to kill the process. When i use a
>> binary built from the rc1 branch, it works (but it ignores the idle detection in
>> RSTS, works in RSX...).
>>
>> I'm compiling SIMH under linux amd64. If this is a real bug and not a stupid
>> mistake from my part please tell me what can I do to help to fix it.
>>
>> My config file, for reference, is attached to this email (I hope the list
>> processor won't strip it).
>>
>>
>
>
More information about the Simh
mailing list