mfm_emu messages
43 views
Skip to first unread message
d...@pdp8online.com
Jan 17, 2021, 7:10:22 PM1/17/21
to MFM Discuss
A user was having issues with the output messages causing the emulator to pause when the output backed up. That caused the computer using the emulator to give an error. They were using the physical serial port on the board which is slower than the USB port or Ethernet. I'm going to see about setting stdout to non blocking to prevent this.
During operation mfm_emu does generate a lot of messages. I assume its useful to have some messages but seems like during normal operation less messages may be better. Do users have any suggestions on what information they would like to see it print? Less or leave it as is?
The simplest would be S each time a seek is done or W if the current cylinder was written to. Anything else useful?
If I change I plan to add a new level to quiet so the messages can be turned back on for debugging.
Convergent MightyFrame
Jan 17, 2021, 7:34:56 PM1/17/21
to MFM Discuss
I am the user David speaks of here, but independently of that, I have the first suggestion for reducing messages.
Right now, a typical output looks like this:
Drive 0 Cyl 0->30 select 3, head 7 dirty 0
Waiting, seek time 7.2 ms max 7.2 min free buffers 75
Drive 0 Cyl 30->29 select 3, head 7 dirty 0
Waiting, seek time 1.8 ms max 7.2 min free buffers 75
Drive 0 Cyl 29->27 select 3, head 7 dirty 0
Waiting, seek time 0.3 ms max 7.2 min free buffers 75
Drive 0 Cyl 27->23 select 3, head 7 dirty 0
Waiting, seek time 1.4 ms max 7.2 min free buffers 75
Drive 0 Cyl 23->21 select 3, head 7 dirty 0
Waiting, seek time 2.5 ms max 7.2 min free buffers 75
Drive 0 Cyl 21->19 select 3, head 7 dirty 0
Waiting, seek time 0.4 ms max 7.2 min free buffers 75
Drive 0 Cyl 19->16 select 3, head 7 dirty 0
Waiting, seek time 2.9 ms max 7.2 min free buffers 75
Drive 0 Cyl 16->12 select 3, head 7 dirty 0
Waiting, seek time 1.1 ms max 7.2 min free buffers 75
Drive 0 Cyl 12->9 select 3, head 7 dirty 0
Waiting, seek time 2.8 ms max 7.2 min free buffers 75
But what if we remove the "Waiting, seek time..." messages, so it only looks like this?
Drive 0 Cyl 0->30 select 3, head 7 dirty 0
Drive 0 Cyl 30->29 select 3, head 7 dirty 0
Drive 0 Cyl 29->27 select 3, head 7 dirty 0
Drive 0 Cyl 27->23 select 3, head 7 dirty 0
Drive 0 Cyl 23->21 select 3, head 7 dirty 0
Drive 0 Cyl 21->19 select 3, head 7 dirty 0
Drive 0 Cyl 19->16 select 3, head 7 dirty 0
Drive 0 Cyl 16->12 select 3, head 7 dirty 0
Drive 0 Cyl 12->9 select 3, head 7 dirty 0
In some of my testing, I've done this after copying the console to text file, I wrote a text file modification basic program to remove all of these lines, so that I could easily scroll through the output and much more easily visually see the history of the Cylinders and heads selected as I scrolled.
Those are my first thoughts on reducing output in a meaningful way.
Best,
AJ
Jay Jaeger
Jan 18, 2021, 8:16:27 AM1/18/21
to mfm-d...@googlegroups.com
Perhaps have the log written to a (rotating) log file and have the console do a "tail -f" on that?
JRJ
--
You received this message because you are subscribed to the Google Groups "MFM Discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mfm-discuss...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/52caf130-c049-4049-818c-eaf75aba6a7dn%40googlegroups.com.
Joan Touzet
Jan 18, 2021, 12:09:49 PM1/18/21
to mfm-d...@googlegroups.com
On 18/01/2021 08:16, Jay Jaeger wrote:
> Perhaps have the log written to a (rotating) log file and have the
> console do a "tail -f" on that?
This is a good option. Especially when emulation is being done headless
> Perhaps have the log written to a (rotating) log file and have the
> console do a "tail -f" on that?
(inside of an old machine), you may not care to see any output unless
you're debugging the setup.
Or, have the console only print messages that are at a higher level of
priority. Using standard UNIX syslog levels, I'd say the majority of the
output is at "info" level. Most of the time an end-user only cares about
"warning" level and up. when emulating or dumping.
The suggested suppression of the "Waiting, seek time..." messages could
be designated "debug" level, with the "Drive 0 Cyl 0->30 select..."
messages at "info" level. Or, any other priority level system you like.
-Joan
> <mailto:d...@pdp8online.com> <d...@pdp8online.com
> <mailto:d...@pdp8online.com>> wrote:
>
> A user was having issues with the output messages causing the
> emulator to pause when the output backed up. That caused the
> computer using the emulator to give an error. They were using the
> physical serial port on the board which is slower than the USB port
> or Ethernet. I'm going to see about setting stdout to non blocking
> to prevent this.
>
> During operation mfm_emu does generate a lot of messages. I assume
> its useful to have some messages but seems like during normal
> operation less messages may be better. Do users have any suggestions
> on what information they would like to see it print? Less or leave
> it as is?
>
> The simplest would be S each time a seek is done or W if the current
> cylinder was written to. Anything else useful?
>
> If I change I plan to add a new level to quiet so the messages can
> be turned back on for debugging.
>
> --
> You received this message because you are subscribed to the Google
> Groups "MFM Discuss" group.
> To unsubscribe from this group and stop receiving emails from it,
> send an email to mfm-discuss...@googlegroups.com
> <mailto:mfm-discuss...@googlegroups.com>.
>
> A user was having issues with the output messages causing the
> emulator to pause when the output backed up. That caused the
> computer using the emulator to give an error. They were using the
> physical serial port on the board which is slower than the USB port
> or Ethernet. I'm going to see about setting stdout to non blocking
> to prevent this.
>
> During operation mfm_emu does generate a lot of messages. I assume
> its useful to have some messages but seems like during normal
> operation less messages may be better. Do users have any suggestions
> on what information they would like to see it print? Less or leave
> it as is?
>
> The simplest would be S each time a seek is done or W if the current
> cylinder was written to. Anything else useful?
>
> If I change I plan to add a new level to quiet so the messages can
> be turned back on for debugging.
>
> --
> You received this message because you are subscribed to the Google
> Groups "MFM Discuss" group.
> To unsubscribe from this group and stop receiving emails from it,
> send an email to mfm-discuss...@googlegroups.com
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/mfm-discuss/52caf130-c049-4049-818c-eaf75aba6a7dn%40googlegroups.com
> <https://groups.google.com/d/msgid/mfm-discuss/52caf130-c049-4049-818c-eaf75aba6a7dn%40googlegroups.com?utm_medium=email&utm_source=footer>.
> https://groups.google.com/d/msgid/mfm-discuss/52caf130-c049-4049-818c-eaf75aba6a7dn%40googlegroups.com
>
> --
> You received this message because you are subscribed to the Google
> Groups "MFM Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to mfm-discuss...@googlegroups.com
> <mailto:mfm-discuss...@googlegroups.com>.
> --
> You received this message because you are subscribed to the Google
> Groups "MFM Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to mfm-discuss...@googlegroups.com
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/mfm-discuss/CAAXuDe_CFXiccnmfu8E6aEFJrHFVEYiUESDBw%2B8qVhFzBexy2Q%40mail.gmail.com
> <https://groups.google.com/d/msgid/mfm-discuss/CAAXuDe_CFXiccnmfu8E6aEFJrHFVEYiUESDBw%2B8qVhFzBexy2Q%40mail.gmail.com?utm_medium=email&utm_source=footer>.
Reply all
Reply to author
Forward
0 new messages