#PHAR LAP ERROR 10034
18 messages in this thread
help is all I can say! I'm in the middle of a drawing…have rendered to screen
about 15 times and may have passed the 13 megs available (Have 16 megs of ram
and may have left Smrtdrv on it)… anyway!.. it's in the middle of rendering
and it just stops and goes back to the dos prompt: c:\3ds2\maps and shows:
"Phar lap error 1034 386| vmm error reading exp file or dated file mapped in
virtual memory…" I have to reboot the computer since the screen is at a low
resolution and no graphics program like QEMM or 3DS will appear..
If this is a conflict as someone mentioned a month ago, why is it so infrequent
then it hits me twice today.,. of course all the changes to the project are
lost since I didnt save.. have to start up and it does it again.,.Sooo.. does
anyone have a good suggestion or even the phone number of the famous Phar lap?
whatever they or that is?(someone suggested a conflict between my Adaptec 1742A
controller and the rest of my software?
I know all you big pro's are still coming home from SIGGRAPH but
help..
Robert,
Consult the Installation and Performance Guide's memory usage section to find
out EXACTLY how much RAM you are going to need to finish your rendering. Then
make absolutely SURE that your swap drive can support it. You might de-frag
your hard drive too, as this helps the VMM do it's stuff. You really shouldn't
have SmartDrive running with 3DS, especially if you start to swap. Disable it
and start with as clean a configuration as possible.
Let us know how you do…
-Alan Iglesias
Robert:
>> of course all the changes to the project are lost since I didn't save
Always save before rendering an animation! Even a power outage in the
middle of the night will do this to you, and a frustrated animator is the
result!
Too bad 3D Studio doesnt have a feature like Word Perfect has, where you can
arrange that the "save function" happens at whatever interval you want.. every
few minutes or whatever.. But to return to the problem of that PHar lap
notice.. why is it that when I render to the screen a .tga drawing i'm working
on, it goes up by 3 megs each time until it gets to the max ram I have of 16
meg (available says 14,000 available) 2 then 4 then 7 then 9 etc.. till 14 then
it starts to page to disk.. that I understand is basically a lack of Ram …
But why if I go into keyframer, give it say 30 frames (.tga) and render to disk
they go up to about 9 megs and stay there…Why does one chew all my ram so
fast and the other stabilize after a few megs?
Robert:
>> Why does one chew all my ram so fast and the other stabilize after a few
>> megs?
I haven't a clue! This one is for the Yost Group! Gary, Jack?
Greg Pyros
Robert, once you _start_ paging to disk, Phar Lap automatically increases the
size of the swap file. If you haven't started paging, and you have a
reproducible situation in which RAM usage increases with subsequent renders,
please send me a test case.
– G
Robert, once you _start_ paging to disk, Phar Lap automatically increases the
size of the swap file. If you haven't started paging, and you have a
reproducible situation in which RAM usage increases with subsequent renders,
please send me a test case.
Sorry again for being slow on the uptake but the paging to disk I understand
as part of the setup ONCE all the RAM is used & the only way to avoid that
problem is to buy more RAM (too costly right now..)
For the project I'm on, the RAM usage increases substantially after each
render to screen..say I adjust a light and want to see the effect so I
render, but after checking RAM used, it might be 3 megs gone…The Project file
started paging to disk after 6 renders to screen but when I went into
keyframer and rendered to disk 30 frames with a camera pan to show movement
the memory used stayed at 9 megs..(I had not reset the program so I assume
if I had done so the memory used would have been even less..
.tga image with defaults = before first render: 14,936 available
used 2572 (upon loading it)
after one render to screen: 7916 used
after 2nd render " " : 8926 "
" 3rd " " " : 9796 "
" 4th " " " : 9652 "
" 5th " " " : 14,296 "
" 6th " " " : 14896 and it was SWap file : 532
and Page faults : 728
So If you want to look at it ..
1) where do I send the file
2) do you want the PROJECT file to look at?
3) what conference do I send it into or should it be sent by E-mail?
Thanks for any advice
I'd be interested in seeing the file to see if it's a problem with your
hardware/operating environment. Please zip up the prj file with a text file
explaining the problem (ie, the previous email) and email it as a binary file
to my ID# via Easyplex. Note that you're going to have to zip it with whatever
maps it needs to render. Make sure that I've got EVERYTHING necessary.
Thanks.
– G
Robert,
Intersting problem. What video driver are you using? I know older versions of
Vibrants DL-Xpress would do something similair. You could render an image like
5 times or something then it would start acting up. Just a thought.
Dave Mennenoh
Hi Dave, looked into my setup and noted the following concerning the Video
Driver..says ATI 28800 DL-XPRESS version 2.12 oct 6 1992 Do you know if there
is a more recent one that will work with my VGA Wonder XL ? I get a max of
32,768 colors at 640 x 480.. Not one of the best but seems ok for me…
Robert,
Hmmm.. that may be your problem. Contact Jack Bayt (he's fairly regular on
here) from Vibrant Graphics and you should be able to get an upgrade to the
latest and greatest. Though, this may not solve your problem. Maybe Jack could
be of further assistance on whether or not that version of the driver cause
memory problems. I know mine used to.
Dave Mennenoh
I had a similar problem like this when rendering tests and eventually running
out of ram and swapdir. It just keeps building when in your ram capacity?
right? I solved it by not using an RCPADI type driver for RDPADI set
statement. It was vibrants Softengine drivers for a Tiga board, in my case but
I've heard this happening to others with different drivers. In all cases they
corrected the problem by not setting RDPADI to an RCtype driver. This has
nothing to do with the pharlap problem of building the swap size, that is the
nature of the beast. Hope this will help
Thanks for the idea Don, but since some of this is still new to me can you
explain a little more what you mean when you say:
" I solved it by not using an RCPADI type driver for RDPADI set "
"they corrected the problem by not setting RDPADI to an RCtype driver." In my
setup the RCPADI goes to the ATI DLXpress driver while the RDPADI goes for the
RCPTARGP.EXP (DATED 4-1-92) (NOTE however that since the mentions on this
string I checked and found that it was an older driver: the RDPTARGP.exp driver
that was on the RDPADI line… Anyway, dont the two drivers I use seem to be
the ones that should be there? regards, Bob Smith
It sounds like DonL has an idea of what might be causing the problem… the ADI
driver. Please explore that with him before you send me the file. Thanks,
– G
Gary, I'll be trying today & tomorrow to recreate the problem now that I
updated the Targa driver…I will also call up tonite ATI in Toronto and
download the latest DLXpress driver ( i"m told by Vibrant that it's version
2.15) and see if that helps// But since I cant remember if the Phar Lap error
happened when I was rendering to a TV screen thru the TARGA or to the screen on
my VGA monitor I have some homework to do…This problem has been really
intermittent but:
1: Still cannot understand the memory usage that goes up so fast compared
to when I would render to Disk 30 frames or even more… Rendering 30 frames to
disk never caused it to swap or page to disk, it climbs but does eventually
stabilize.. But doing these test renders chews the RAM within up to 10 tests..
I'll let you know..
Bob Smith
Robert,
Actually the newest DL-Xpress is 2.16. Hope you can fix the prob.
Dave Mennenoh
>> RCPADI goes to the ATI DLXpress driver while the RDPADI goes for the
RCPTARGP.EXP (DATED 4-1-92)
This is what I thought you had incorrectly set up and has been what has caused
many weird problems including the one you mentioned. However, you now stated
the correct targa RDPTARGP.exp for the RDPADI line. This has been what usually
corrects the problem. If you still have it then maybe you need to suspect the
ATI driver. I can't help you there because I have no experience with that
particular driver. I do know that ATI drivers have been the subject of much
discussion regarding bugs on this forum.
Hi Don, havent been able to get new drivers and as hey say here's the story so
everyone can learn a lesson: 1rst my Nec 4FG monitor had a slight defect (a
horizontal line scrolled & there was an intermittent high pitched hum.) so
being 9 months old & under warranty I brought it to BULL systems who do
warranty work for NEC in Canada…they said Oh a day or two but when I went to
pick it up they said here's a new one.>>(Wrong…They had me back a
"refurbished" model that was older and also had a defect, so back it goes and I
have to get another one..That one now shows signs of a problem but I cannot
adjust it properly since no one at ATI has answered my questions about where is
the latest DLXpress version 2.16 driver & if my configuration drivers for the
ATI card are actually correct for my NEC monitor… I called the ATI BBS the
other nite and after downloading a file dated Jan 93 found out it was the same
driver package I already had (but dated Oct 92) So so smart fellow at ATI
changed the date on the file and people who want the upgrade download the wrong
package !! The drivers are not on the CIS library of ATI and two messages later
still no answer… GREAT SERVICE FOR A CANADIAN COMPANY GUYS !!!! So buyer
beware… On a positive note: When I contacted Vibrant Graphics yesterday and
mentioned the PHarLap error message I was given the following statement to
include in my batch file that starts 3D Studio.
SET DOSX=-SAVERGS Am told this helps to remove the Phar Lap
error message.. I'll try it out and see if it eliminates it.. I'll also let you
know how this story with ATI develops/ Regards, Bob Smith