|
Post by foxpropgmr on Mar 28, 2019 16:00:40 GMT 1
Since Feb 15, I've had only 2 page faults (should have been 5 or 6) after including the SAVEREGS command. So something is working.
I'd be interested in using & testing the preliminary new version.
btw, are these page faults recorded in a error.log file somewhere in Windows?
Walt
|
|
|
Post by Jos on Mar 28, 2019 18:59:58 GMT 1
At this moment even the new version can have Page Faults errors, though less frequent. Since yesterday I have an automated test environment to enforce those errors in a somewhat consistent manner, within 15 minutes. First investigations indicate a rollover of the 12 offset bits of a virtual memory address under specific circumstances to blame. Will take some time to actually fix that in the involved memory access routines. Therefor also nothing to be tested for now.
The Page Faults errors are caught by vDos, it simply can’t translate a virtual to a linear memory address. Reports that and exits. These errors are out of the scoop of FoxProX or Windows.
Jos
|
|
|
Post by Jos on Apr 12, 2019 22:03:21 GMT 1
John and I finally found a method to consistently force a Page Fault error. With that the cause also became apparent. Not directly related to Page Fault errors, so also the main reason I couldn’t solve it until now. Due to more extensive CPU operational checking, a Page Fault was just trigged (or even not) as the FoxProX program already ran wild for some time.
Jos
|
|
|
Post by foxpropgmr on Apr 19, 2019 3:11:49 GMT 1
That's good news ... right?
A solution forthcoming soon?
|
|
|
Post by Jos on Apr 19, 2019 5:56:00 GMT 1
FoxProX was plagued by those illusive Page Fault errors since vDos supported virtual page mapping, so nice it’s finally fixed in the next 2019.05.01 version.
Jos
|
|
|
Post by acgeido on May 3, 2019 17:59:59 GMT 1
Hi Folks! We have good news!
The new version of May 1st. has solved our memory problems with FoxproX 2.5 / 2.6. Even some problems I was giving broken indexes, which were not true, since if they ran under the old DOS it worked correctly. I had to erase all the memory configuration that could previously be done in the file: config.txt (now warn me that this configuration is invalid) and I ran it without added. The configuration that I left is of course in the autoexec.txt file with respect to the disk mapping and the set that correspond to the environment.
|
|
|
Post by johngoebel on May 22, 2019 19:52:11 GMT 1
Jos, We have been using 2019.05.01 version and have had no page faults. Thank you very much for your hard work and patience. Regards, John
|
|
|
Post by foxpropgmr on Jun 1, 2019 4:46:58 GMT 1
Is there a way from vDos's command prompt to get it to reveal the version that's running? There's nothing unique in the registration nag. I have so many versions of it on my machine, I just want to be sure I'm referring to the right one; and I don't want to trust the file date.
Just wondering, Walt
|
|
|
Post by Jos on Jun 1, 2019 7:45:27 GMT 1
That would be the ABOUT command. But it’s the same as the registration nag, the version is displayed in the first line.
Jos
|
|
|
Post by foxpropgmr on Jun 1, 2019 8:23:40 GMT 1
vDos-Base?
The older versions have a date when I use the ABOUT command
|
|
|
Post by Jos on Jun 1, 2019 8:31:08 GMT 1
vDos-Base is static, so a date in About is not relevant.
Jos
|
|
|
Post by foxpropgmr on Jun 1, 2019 8:34:52 GMT 1
Does that mean this is the final version then? No more updates or improvements?
|
|
|
Post by Jos on Jun 1, 2019 8:38:26 GMT 1
As stated at the download page: “The Base edition is 100% opensource, basically vDos 2017.08.01. However w/o printing and executing Windows programs, no later improvements, and unsupported like other older vDos versions.”.
Jos
|
|
|
Post by foxpropgmr on Aug 1, 2019 1:52:03 GMT 1
The page faults have been greatly reduced, and I thank you. But I just had one the other day. (7/29/19 ... latest version of vDos)
|
|
|
Post by Jos on Aug 1, 2019 7:09:47 GMT 1
You’re the second to report FoxProX page faults are still around. But I have no idea what could still cause those.
Jos
|
|