| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 1927.1 | Works Under OpenVMS... | XDELTA::HOFFMAN | Steve, OpenVMS Engineering | Wed Apr 16 1997 10:26 | 10 |
| : I have a cust with a 255 and 3 hds...Powerstorm 3d30's. He also has
: some with single hds. They are running Unix 3.2G and Open 3D v 3.4.
You'll have to ask the UNIX folks.
Three PowerStorm 3D30 graphics controllers do work under OpenVMS on an
AlphaStation 255. (Several folks locally run dual-headed AlphaStation
255 systems, and there is at least one local triple-headed system.)
NB: I do not know if this configuration is officially supported by OpenVMS.
|
| 1927.2 | Not officially supported yet! | WRKSYS::BROOKS | | Wed Apr 16 1997 10:35 | 132 |
| Three heads are not formally supported. If you want this contact
Connie Doyle the Product Manager. Worksystems Support Engineering did
some tests a while ago and this is what we found on some experiments in
the lab:
From: WRKSYS::SCANTLEBURY "Pete Scantlebury DTN:223-3629"
[3HD-UNIX.TXT]
13-DEC-96
This is a brief overview of what we have done with this 3 head systems
with
this IMPT case.
At this time the LX3+(233 MHZ) with UNIX V3.2g and UNIX V3.4 only seems
to
support two TGA2(8 pl). Any attempt to add the 3rd head results in
machine
check - hardware error(crash). TGA2 with VGA enabled for the console
seems
to prefer PCI slot 11(bottom), which is the shared interrupt slot.
In parallel we have another LX3+(300 MHZ) with UNIX V4.0A and no
OPEN3D. All
three heads seem to operate properly. Here again the TGA2 with VGA
enabled
for the console seems to prefer PCI slot 11(bottom), which is the
shared
interrupt slot.
Next we will remove OPEN3D from the UNIX V3.2g system. We will also,
try
moving the console TGA2 around again.
Stay tuned.
13-DEC-96 PM update
-------------------
I had problems with trying to generate a new kernel on the LX3+(233
MHZ) with
UNIX V3.2g after I unloaded all the OPEN3D V3.4 code. I ended up
rebuilding
the system. Now, however it seems ALL THREE HEADS are working under
UNIX
V3.2g and no OPEN3D. I have not seen any hint of the infamous 'white'
screen.
I have been trying during the afternoon to ftp a copy of DECvet from
'clipr',
but there still seems to be network problems. Also, saw problems with
MPC
much the same as on the LX3+(300 MHZ). Finally pulled over DECvet from
CHUMP.
screen 0: vet(72 hrs.){errorlog at /screen0/decvet_err.log}
screen 1: vet(71 hrs.){errorlog at /screen1/decvet_err.log}
screen 2: vet(70 hrs.){errorlog at /screen3/decvet_err.log}
The LX3+(300 MHZ) with UNIX V4.0A and no OPEN3D is still working. There
have
been some problems running MPC on all three screens. This might be due
to a
system hardware problem. System is running X11PERF on screen 0.
---------------------------------------------------------------------------
LX3+ - 233 MHZ
==============
Initial configuration -
---------------------
PCI slot 13(top): TGA2(8 pl){ZLXp2}
PCI slot 12(mid):
PCI slot 11(bot):
ISA slot(top):
ISA slot(bot):
Installed UNIX V3.2g - ver. 62 via CD.
Installed OPEN3D V3.4 with initial configuration of 1 TGA2 in PCI slot
13.
Ran pst_cube(3D) to verify OPEN3D operation on monitor
Installed 2nd TGA2(8 pl) in PCI slot 12(middle).
Booted genvmunix.
Second monitor comes up white. I moved the mouse from the screen
0(console)
into the white screen 1. The mouse pointer image could be seen but when
I
returned to screen 0, the screen was frozen. I could rlogin from
another
system.
Installed 3rd TGA2(8 pl) in PCI slot 11(top).
Booted genvmunix -->machine panic - hardware error(crash)
Remove 3rd TGA2.
Move console TGA2(w/VGA enabled) to PCI slot 11(bot).
Second configuration:
--------------------
PCI slot 13(top): TGA2(8 pl){ZLXp2}
PCI slot 12(mid): TGA2(8 pl){ZLXp2}
PCI slot 11(bot): TGA2(8 pl){ZLXp2}[w/VGA enabled]
ISA slot(top):
ISA slot(bot):
Booted genvmunix
Run pst_cube(3D) in screen 0, OK.
Run pst_cube(3D) in screen 1, OK.
Run pst_cube(3D) in screen 0 & 1, OK.
Installed 3rd TGA2.
Booted genvmunix -->machine panic - hardware error(crash)
---------------------------------------------------------------------------
LX3+ - 300 MHZ
==============
Initial configuration -
---------------------
PCI slot 13(top): TGA2(8 pl){ZLXp2}[w/VGA enabled]
PCI slot 12(mid): TGA2(8 pl){ZLXp2}
PCI slot 11(bot): TGA2(8 pl){ZLXp2}
ISA slot(top):
ISA slot(bot):
Installed UNIX V4.0A - ver. 464 via RIS system
System was built with 3 TGA2. System crashes during kernel build.
TGA2 in PCI slot 11(bot.) removed. TGA2 from PCI slot 13(top) moved to
replace this PCI slot.
System was built with 2 TGA2. OK.
Add 3rd TGA2. OK
|
| 1927.3 | Need fix or official confirmation | KAOFS::R_SIEMENS | | Mon Apr 21 1997 14:36 | 26 |
| So, it appears that 3 hds are definitly NOT supported under Unix
3.2g...Is that correct?
We found that in the Open3D v.4.3 release notes that it states that
"the problem with the bottom(shared)slot has been fixed in this
release."
So I would like to confirm with someone that this is true:
3 hds are in no way supported under Unix 3.2G
3 hds are supported under Unix 4 with Open3D v4.3.
Is it possible to get this in writing(official) from engineering.
The problem is that this customer has already purchased 20 AS255's
based on the SOC. Because of their 3rd party software, upgrading to
Unix 4 is out of the question.
Unless we can get a fix for this, we will be eating 20 stations and
replacing them with AS500's, that's approx. $5k each!!! Not a very
desirable option.
The other problem is that 4 of these systems are to be shipped to NATO
within the next week or two.
Thank you
Randy Siemens
DTN 635-4300 x527
|
| 1927.4 | 3 heads in AS255 are supported | WRKSYS::BCLARK | I have dyslexia of the fingers | Tue Apr 22 1997 14:03 | 37 |
| I reviewed the history of a recent IPMT case our team had helped to
resolve and found it was closed once a few configuration details were
provided. If anyone from O3d group can offer other additional or more
recent information, please do so.
The solution to the problems with 3 heads is an easy one. Please try
the following 2 steps and either contact me directly at iav8::bclark
(exchange is too slooooow) or phone me at DTN 223-5733.
I have cut and pasted the field engineer's response to his own IPMT
case! here:
FIELD UPDATE
============
"This case is solved. I used the recommended config -> putting
the console in PCI slot 11 with VGA enabled -> booted genvmunix
->run doconfig -> OK!"
STEP 1:
=======
>>>Ensure that you set VGA ENabled jumper on the 3D30 module which is
>>>inserted into Slot 11 of the PCI. (refer to the Owner's or
Installation Manual)
>>>Ensure that the VGA enable jumper is DISabled for 2 of the 3
>>>3D30 modules which should be installed into Slots 12 and 13 of the
>>>PCI bus.
STEP 2:
=======
>>>DO NOT INSTALL Open3D ! I am not sure if removing Open3D will allow
>>>the configuration to work or not. The person who did this work has
>>>left our organization.
Good luck!
|
| 1927.5 | | WRKSYS::BCLARK | I have dyslexia of the fingers | Wed Apr 23 1997 08:33 | 13 |
| I recall that once O3D folks became involved with the IPMT case
mentioned back one, that all 3 heads were able to work properly if
certain subsets were removed from the installation. I recall that at
least 2 subsets should not be installed:
1. Button/dial box
2. Freedom series
I may be wrong, since these things aren't documented as one might
expect to find if the release notes.....but I am certain that if O3d is
not installed, the 3 heads work.
bc
|
| 1927.6 | 3head problem resolved with new tga.o | WRKSYS::BCLARK | I have dyslexia of the fingers | Thu May 15 1997 09:15 | 18 |
| Just to follow up in case others are having a similar problem...
In reviewing the console log files of the system failing, the following
message was found, which immediately preceeded many CAM errors (timeout
on disconnect request):
login: shared_intr_dispatch: stray interrupt = 0x5fe0460
The problem reported in .0 has been resolved with a new tga.o driver
provided 5/8/97. Keep an eye out for TIMA kits containing a new tga.o
driver.
best regards,
Bob
|