New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
vax/microvax3900 waits for an (practically) infinite duration when talking to anita (NetBSD) rather than a terminal. #478
Comments
Hi Utkarsh, While I'm looking at your suggested change to sim_timer.c, please try to run the latest code in the master branch (with and without your change) AND remove the "load -r /extra/vax/data/ka655x.bin" line from your configuration. Also, please provide the output of SHOW VERSION and what is produced when you build the vax simulator with make... Thanks.
|
Here's what I get when I build the emulator:
Here's the output for SHOW VERSION:
Here's what I output without the suggested changes:
Here's the output with the suggested changes:
The kernel now boots after removing the 'load -r' line but there are still errors in the system checks. Regards, |
I reformatted your last response to include the desired output inline in the text since downloading and examining separate files is tedious. I also reduced the make output to just the relevant 'make vax' rather than all the simulators... In any case, looking at your suggested change, I'm wondering how you got to that change? The original code subtracts 1 from the result of sim_activate_time() which always adds one to the activation time of specified unit if the unit is active (which it should be on the code paths that your change includes). I know there are problems with some of the timer code, but tracking these down is hard due:
Your test case happens very early in the simulator's execution, so it would seem to be an excellent case to track down at least one problem. Can you provide the pieces I would need to reproduce this? Thanks.
|
I got to that change just by looking at the output of :
Just clone this repo and run the it with the following command: |
Thanks. I see the failure. I'll let you know what I find. |
Hi Utkarsh, Please give the latest code a try and let me know how things look now. Thanks.
|
Sorry! I got busy. The latest build, as of now, fixes the issue. 😄 |
…patched As discussed in simh#478
Context
I am one of the contributors to anita, the automated test suite for NetBSD https://github.com/gson1703/anita . We are planning to add support for simh to it for testing the NetBSD/vax port. It works by screenscaping the output sent by the emulator and sends predefined input to it. Starting from the commit e36873f, we are having problem trying to communicate with the emulator via anita. It used to work prior to that. We are logging the output sent by the emulator.
the expected behavior
Here's an example log for any working commit (expected behaviour):
the actual behavior
And here's an example log for a commit that doesn't work:
As you can see, the emulator just keeps waiting for infinite duration after sending '[c'. Also, you might notice that for a working commit, there's a delay in between '[' and 'c' (look at the timestamps).
Do note, however, that running the emulator manually from the terminal works just fine.
the simulator configuration file (or commands) which were used when the problem occurred.
Here's the content of config file that we are using:
After that, at the '>>>' prompt we issue: boot dua3
I made the following changes which gets past '[c' but now the there's a problem in system tests and the netbsd kernel isn't starting. Here's the output after applying the fix:
I have attached the diff for this fix.
vax.txt
Any help with fixing the issue is appreciated.
Regards,
Utkarsh Anand
The text was updated successfully, but these errors were encountered: