ASUS P2B / Award BIOS APM problems resolved
Jason Barlow
management works correctly, read no further. For those who
are truly obsessed, as was I, about getting APM/ACPI to work
w/ their ASUS motherboard - read on.
I promise nothing but it may lead you to a resolution to your
problem. Here is my system configuration:
P2-400 w/ ASUS P2B Motherboard
Award BIOS 4.51PG / ACPI Revision 1005
SB AWE64 - ISA
3COM/USR 56K Modem - ISA (obviously)
Iomega PCI Jaz SCSI controller
Linksys EtherLink PCI adapter
Matrox Millenium G200 AGP
ADI 6P - 19" Monitor w/ USB support
PS/2 Intellimouse (no Intellimouse
Here's what I did (PMTSHOOT is a P.O.S. - don't bother):
1. Backup critical data files
2. Restored BIOS to Setup Defaults
3. !!! Disabled USB support in the BIOS - this can raise havoc
with DirectX 6. MS is pursuing a fix. It also seems to be at
the root of my power problems (HD not spinning up on resume,
monitor staying in a state of perpetual sleep, etc...)
4. Removed *all* cards except for the display adapter
5. Reinstalled Windows 98 using a standard VGA adapter and
PnP monitor drivers.
6. Tested StandBy - verified correct behavior
7. Switched to my Matrox drivers....and verified
8. Added another card, restarted...had Win98 reinstall the
drivers....verified StandBy worked
9. Repeated this process and determined that it was important
that my LAN card precede my SCSI adapter on the PCI bus. I
moved my LAN card from Slot 4 to Slot 1. Reordering the cards
on the bus seems to affect APM behavior.
10. Verified that my ConfigFlags in HKLM\Enum\Root\*PNP0C05\0000 was set to
0 and not some other value (soft power off will not
work without this - see my earlier post)
11. Verified that SuspendFlags was also 0 (a value of 0x0200 (512)
disables APM. Look for:
HKLM\System\CurrentControlSet\Services\VxD\VPOWERD
Notes
--------
I repeated this process using an ACPI installation of Win98 and
had successful results as well.
There are some KB articles about USB interfering with APM.
That would be the very first thing I would try.
Be advised there is still a known problem with PS/2 port mice
and resuming from suspend - sometimes you get the blinking
cursor in the corner and have to press Alt-Tab to get the
screen to come back online.
I had minor problems with the Win98 Monitor drivers that
came w/ my monitor. Using the canned PnP Monitor
driver that ships with Win98, seemed to rectify the situation.
ACPI seems to be a bit more thorough about sequentially
shutting everything down (where APM just kinda shuts
it all down at once, and brings it all online at once) - I did
notice that when resuming from an ACPI suspend, it
does take 15 seconds or so for each device to power up
again - no idea why it takes this long.
Anyway...I hope this information can help somebody. What
a crappy OS - hope NT 5 proves better. Supposedly
there are replacement files for CONFIGMG.VXD and
the USB*.VXD drivers, but as it was stated in an earlier
post, nobody knows where to get them....I called MS and
they were as confused as I was (???)
Jason Barlow
jas...@xmission.com
Bill Drake
immensely in tracing ACPI bugs.
Further things to check:
1. Load-order dependencies are not supposed to
occur with PCI devices. (Yes, I know all about
the difference between theory and reality... <g>)
Are you using the ASUS SCSI Host Adapter or
a third-party adapter such as an Adaptec?
The problem you mention may have some
implications for the BIOS design. I recommend
you send ASUS an email with all the particulars
of the dependency you discovered.
ASUS may have to patch the PnP Manager in
their BIOS to remove the load-order dependency.
2. Does the USB problem go away when you stay
with DirectX5? If so, then the problem is definitely
in DirectX5. If not, then the interaction is again with
PnP and may require BIOS updates to fix completely.
Send feedback to ASUS on this issue as well, so they
can test for the problem and see if there are
housekeeping changes they can make in the BIOS to
improve compatibility between USB and DX6.
3. Let ASUS know about the LANcard as well. There may
be an interaction there which ASUS must accommodate.
4. The slowness on resume also sounds like something
ASUS may want to investigate. It sounds like each
device must timeout before the next can begin
initializing -- which is a common indicator of BIOS bugs.
5. The monitor INF problem sounds odd. There are many
things which can cause this. You may wish to mention
the problem to Matrox and see what they say.
Please note that W98 has two shutdown modes. The "fast"
shutdown mode simply abandons each driver at shutdown. The
"normal" shutdown mode forces driver deinitialization for each
driver before shutdown (the standard W95 methodology).
Many people have noted dependencies where certain programs
*require* proper deinitialization of drivers before a successful
shutdown can occur. WinFax Pro 8.0 is notorious for this.
Systems which require driver deinitialization will hang at the W98
shutdown screen if fast-shutdown is enabled (W98 default.) The
fix is to use MSCONFIG to disable fast shutdown.
Finally, as we all know there are a bunch of things wrong
with W98's implementation of APM/ACPI. Uncrossed T's
and undotted I's abound. Many of these problems are slated
for fixes in W98 SP1 -- supposedly due around the middle
of September.
I suspect we are all going to have to endure a bit of
"forth and back" between MS and the motherboard
manufacturers as ACPI matures. This is the usual
consequence of MS rushing their software out the door
before it is truly ready.
Best I can do for now. <tm>
Bill
Jason Barlow wrote in message ...
Jason Barlow
SCSI adapter for my Jaz drive.
I'll drop ASUS a line about the load order issue on the
PCI bus.
Regarding DX and USB - there is a problem with DINPUT
when it enumerates devices and encounters a USB
device. This is already being looked into by Micro$oft.
There seems to be a related problem with APM and
having USB support enabled.
Its entirely possible that both problems stem from poor USB compliance with
my monitor - it's still a pretty new technology
for most hardware vendors.
J
Bill Drake wrote in message ...
Barry Allard
I've got ACPI working properly with Win98 & ASUS P2B bios ver 1005 beta 15.
From Asus: P2B does not currently support ACPI Level=S4 (same as MS OnNow
Level=Hibernate)
Barry Allard
all...@ix.netcom.com
procedure:
1) Win98 OEM-ver full install
2) Add value:HKLM\Software\Microsoft\Windows\Current
Version\Detect\ACPIOption: DWORD =1
3) Add Hardware Wizard to redetect all
all done a couple of reboots later
VERSION=4.10.1998
LANGUAGE=English
OPERSYSVER=4.10.1998
OPERSYSTEM=Windows 95 / 98
CPUTYPE=6
COMPSPEED=400 MHz
MEMRAM=128Mb Hitachi 6ns
DRIVES=C: 3385Mb/8706Mb (FAT32) IBM 9ES Ultra2/LVD 9.1GB 68-pin
COMPMAKE=
COMPMODEL=
PRINTERS=
BIOS=AwardAward Modular BIOS v4.51PG07/17/98
DISPLAY=Matrox Millennium G200 AGP bios ver 1.2;
MULTIMEDIA=TBS Montego Sound Blaster Emulation;U.S. Robotics Voice Serial
Wave Device;TBS Montego PCI Audio;TBS Montego Gameport Interface;
CDROM=PLEXTOR CD-ROM PX-32TS bios ver 1.01
MODEM=U.S. Robotics 56K Voice Pro Ext;
NETCARD=Dial-Up Adapter;
NETCLIENT=Microsoft Family Logon;
NETSERVICES=
NETTRANS=TCP/IP;
MONITOR=Plug and Play Monitor;
KEYBOARD=FOCUS 2001 PS/2
PCMCIA=
DEVICESUSB=Intel 82371AB/EB PCI to USB Universal Host Controller;USB Root
Hub;
DEVICES1394=
INFRARED=
MOUSE=PS/2 Compatible Mouse Port (ms intellimouse pro);
SCSIADAPTER=Adaptec AHA-2940U2/AHA-2940U2W PCI SCSI Controller bios ver
2.00.0;
Jason Barlow wrote in message ...
>The SCSI Host Adapter I'm using is an Iomega PCI
>SCSI adapter for my Jaz drive.
>
>I'll drop ASUS a line about the load order issue on the
>PCI bus.
>
>Regarding DX and USB - there is a problem with DINPUT
>when it enumerates devices and encounters a USB
>device. This is already being looked into by Micro$oft.
>There seems to be a related problem with APM and
>having USB support enabled.
>
>Its entirely possible that both problems stem from poor USB compliance with
>my monitor - it's still a pretty new technology
>for most hardware vendors.
>
>J
>
>Bill Drake wrote in message ...
MrScary
>SCSIADAPTER=Adaptec AHA-2940U2/AHA-2940U2W PCI SCSI Controller bios ver
>2.00.0;
Hi,
I was just curious about something. Me and a few friends are debating
wether or not to upgrade to this card. I like it and was wondering if this
is the Bios revision that came with it or did you have to flash it to this
level?
Attila Szabo
Microsoft-MVP (DTS)
Bell Atlantic-Services Technician
a.k.a.' MrScary '
Steve Prentice
beta 6 and I've had no problems what so ever, it's a great board. Only thing
is it can temp monitor, but there's no CPU monitor lead in the box, and
asus don't make them yet, but you can make one yourself for about $2
with a thermistor brought from Radio Shack or wherever.
Steve
MrScary wrote in message ...
Barry Allard
MrScary wrote in message ...
> Barry Allard wrote in message ...
>>SCSIADAPTER=Adaptec AHA-2940U2/AHA-2940U2W PCI SCSI Controller bios ver
>>2.00.0;
>
>
>Hi,
>
> I was just curious about something. Me and a few friends are debating
>wether or not to upgrade to this card. I like it and was wondering if this
>is the Bios revision that came with it or did you have to flash it to this
>level?
It is the orig bios that came with it in 2940 Ultra2 Kit. (ver 2.00.0)
Adaptec told me that 2.00.1 should be available. SCAM PnP is not supported.
(even though it claims to have it on the b#@$-s%*$ box. SCAM isn't really
needed since how hard is it to pick a number between 0 and 15? The
controller has to negotiate with the device anyhoo. That's PnP enough!
Have you found the mysterious vaporware ACPI upgrade (1) yet?
Barry Allard
all...@ix.netcom.com
(1)= http://support.microsoft.com/support/kb/articles/q190/7/15.asp
MrScary
How did you know I was hunting for vxds?
No luck as of yet. Had a theory that the latest Sindwinder software had some
of these included but it didn't pan out. Any more info I find I will be
posted ASAP in this group.
Thanks for the feedback on the card!!
Attila Szabo
Microsoft-MVP (DTS)
Bell Atlantic-Services Technician
a.k.a.' MrScary '
Barry Allard wrote in message ...
Barry Allard
-Barry Allard
MrScary wrote in message <#HuKFwI0...@uppssnewspub04.moswest.msn.net>...
MrScary
...but you seem interested in all this ACPI techno stuff....have been to
these sites yet?
http://www.microsoft.com/hwdev/desinit/acpifaq2.HTM
http://www.microsoft.com/hwdev/desinit/hiberfat32.htm
http://www.microsoft.com/hwdev/ONNOW.HTM#HCT
Advanced Configuration and Power Interface
Hardware Compatibilty Test for
Microsoft Windows 98 and Microsoft Windows NT 5
This package contains the Advanced Configuration and Power Interface (ACPI)
Hardware Compatibility Test (HCT) suite for Microsoft Windows98 and
Microsoft Windows NT version 5.0. It is intended to verify system hardware
and fimware(BIOS) compliance with the ACPI(v1.0), PC97 and PC98
specification compliance as well as operational compatibility with Microsoft
Windows98 and Windows NT5. The HCT also includes viewing facilities for the
data structures and namespace provided by the firmware. The namespace viewer
has limited capabilites for testing control methods, however it is not
always safe to the sytstem, so this facility should be used with caution.
Additional Information
For information about this test and the initiatives it covers:
http://www.microsoft.com/hwdev
http://www.microsoft.com/hwdev/onnow.htm
For information regarding ACPI:
//http://www.teleport.com/~acpi
For information on this and other HCTs:
http://www.microsoft.com/hwtest
The latest version of this test is available at:
http://www.microsoft.com/hwtest/testkits
Questions or problems with the HCT should be sent to WHQ...@microsoft.com
>>And so there's no mistake the test tool is here.....if you dare NOT for
the
faint of heart.
http://www.microsoft.com/hwdev/acpihct.htm
What this test does
This test package covers system features that are specific to ACPI only. It
will validate ACPI data structures, hardware control registers, and the
system description provided through ACPI Machine Language(AML) Namespace.
The test will verify the presence of required features, versions and functio
nal attributes to insure that Microsoft Windows operating systems can use
the system facilities provided.
What this test does not do
This test does not cover operational features that are not specific to ACPI.
Device configurations, docking station operation, device removal and device
arrival are examples of features that are tested with other Hardware
Compatiblity Tests available from Microsoft. This test is only one of a
large suite of tests used in the process of verifying system compatibility
with Windows operating systems. This test is not intended to be a full
standalone tests. Please contact Windows Hardware Quality Lab (WHQL) for
information about additional HCTs.
Running Tests
Start by running ACPIINST.EXE which will perform an initial scan of the data
structures before presenting the installation options dialog. At this point
it is fairly certain that the OS will boot in ACPI mode, although it may not
function very well if there are errors in the description of the devices in
the system. In extreme cases the OS may crash and not be able to recover.
Should this occur, a fresh installation will be required.
Windows 98 compatibilty testing requires the BIOS memory map test to insure
the BIOS is correctly describing the physical memory in the system and how
it is used. There are several requirements on how the BIOS describes the
memory present in the system to prevent conflicts and data corruption from
DOS programs which use EMS or XMS memory. The setup options screen provides
two methods for running this test: A test floppy and run from startup files.
The test floppy is the preferred method, since it doesn't mess with the
startup files. Running the test from boot will require manual removal and is
therefore not recomended for most users. If the test floppy is selected, the
user will be prompted for a blank, formatted, floppy to be insterted into
drive A on which to place test programs. The user will be prompted to remove
the diskette and re-boot the system after the test is run. The diskette will
be requested later. This test is not required for Windows NT testing.
On the subsequent boot, after login on networked systems, the main
functionality test will begin. For x86 based systems the program will, with
systems running Windows 98, ask for the memory map disk. The test will
process the system description and check several features.
If message waiting indicator and/or system status indicator devices are
listed, the user will be asked to determine if the proper indications are
given. In this case, the user must be familliar with the expected
indications to properly answer the questions. There is additional
information in the release notes section below.
For systems that support passive cooling (clock throttling), a test of this
process will be performed. This test does not yet support multiple
processors. There is additional information in the release notes section.
When the testing is completed, a brief summary will be displayed. The viewer
windows are then available, they are described below.
About the ACPI Viewers
The ACPI HCT view windows can help you track down problems in your ACPI
implementation.
ACPI Namespace Window
ACPI Table Window
ACPI HCT Log File
Using the ACPI Namespace Window
The ACPI Namespace window displays a tree-view representation of the ACPI
Namespace. The left hand pane contains the tree and the right hand pane
contains information about the currently selected object in the tree. In
sonme display resolutions, the splitter bar between the two views is placed
against the left edge of the window, and the view appears as a single, blank
window. To see the tree view, place the mouse cursor on the left edge of the
screen and drag it to the right. Note: the typical splitter cursor does not
appear. You may expand the tree and see the objects. To get information on a
particular object, select it. There is a "feature" available that will allow
you to "run" control methods and get the results. This will only work for
methods that do not require parameters. There is no plans for a mechanism to
provide parameters to the methods. DANGER: Control Methods are used to
change properties of the system. If you run a method that turns off the boot
hard drive, windows will crash. Use this feature cautiously!
To evaluate a namespace object, double-click it in the tree view. The result
will be displayed in the right-hand pane of the window.
Using the ACPI Table Window
The ACPI Table Window displays the contents of the ACPI Tables. A view is
built and saved on the drive and then notepad is opened with that file. The
file will not be deleted as a convienience in case it might be useful
elsewhere.
Using the ACPI HCT Log File Window
This window displays any errors or warnings that are generated by the HCT.
Notepad will be opened with the log file.
Exiting the ACPI HCT
To close the HCT, you can do one of the following: - Click the “X” button in
the upper-right corner - Choose “Exit” from the File menu - Press Alt+F4
Release Notes<\H2>
* Activating ACPI on Windows 98 is not a pretty sight, however it only looks
bad. Hardware detection is known to crash when done on some systems. We are
investigating this currently, and it does not affect the system or the test.
The subsequent boot will cause Windows to re- discover the system using ACPI
instead of the previous mechanism (there are several to choose from) that it
knew. This is likely to prompt for the Windows 98 CD, and the CD ROM drive
will not be available. For this reason, it is best to have the Windows 98
distribution on the local drive and to have installed from that image. If
the CD is prompted for, it is quite likely that the file it is looking for
is already in the correct location, and you just need to point the installer
to them. The three most likely directories are \WINDOWS, \WINDOWS\SYSTEM and
\WINDOWS\SYSTEM32\DRIVERS. If this sort of thing frightens you, it is best
to leave this job to a trained technician.
* The clock throttle test is noted to have some inaccuracy on some systems.
This will sometimes result in an incorrect error that one or two of the
settings had no effect, and a subsequent error that the following setting
showes less throttling. Systems that demonstrate this error are not
nessesarily disqualified from recieving a Windows 98 / Windows NT 5 logo.
These errors must be evaluated by WHQL to determine if they are valid or
not.
* The System Status and Message Waiting indicator test has caused a great
deal of confusion, finger pointing and general panic. The indicators in
question are for user notification only. The tester must understand what the
system vendor defines the indications to be. Microsoft requires that a
system have a status indicator that will inform the user of the power state
of the system. Microsoft requires three unique indications: On, Off, and
Sleeping. Microsoft defines only that the Off state has the indicator off.
The other two states are up to the system designer. Microsoft does not
define HOW this indicator be implemented. ACPI defines a mechanismy by which
the system status indicator MAY be controlled. This is an optional feature
of ACPI. If present, the OS will use this method. The indicator is an output
only device and does not have a mechanism to check it's status, so the
tester must be asked to check and report. For this test to have any meaning,
the tester must know what the system designer defined the indications to be.
The message waiting indicator is an optional feature for both Microsoft and
ACPI. Neither Microsoft nor ACPI define what the indications will be for any
number of messages. ACPI defines 0 messages must be different than >0
messages, and that any number is valid. The test number chosen is 2, being a
value greater than 0. As with the system status indicator, this is purely a
user notification device that cannot report back it's status, so the tester
must be querried.
Barry Allard
> Barry Allard wrote in message ...
>>SCSIADAPTER=Adaptec AHA-2940U2/AHA-2940U2W PCI SCSI Controller bios ver
>>2.00.0;
>
>
I've downloaded the AHA-2940U2W BIOS Update.
ftp://ftp.adaptec.com/pub/BBS/hardware/2940u2up.exe
Here goes nothing...
Barry
MrScary
There's a few things you should know.
Use an old Dos 6.22 system disk if you can...if you use a Win9x system boot
diskette remember to envoke the safe menu and select "Safe Mode Command
Prompt only" before AFU is started.
Remove Drvspace.bin and /or Dblspace.bin from any OS version system boot
diskette that you use
Use the utility included in AFU to backup your existing bios FIRST. After
that select "Erase current bios" and THEN apply the new bios flash.
Barry Allard
Barry Allard wrote in message ...
>MrScary wrote in message ...
>> Barry Allard wrote in message ...
>>>SCSIADAPTER=Adaptec AHA-2940U2/AHA-2940U2W PCI SCSI Controller bios ver
>>>2.00.0;
>>
>>
>
>I've downloaded the AHA-2940U2W BIOS Update.
>ftp://ftp.adaptec.com/pub/BBS/hardware/2940u2up.exe
>Here goes nothing...
>
>Barry
It worked, bios now rev. 2.00.1 !
Barry
Barry Allard
MrScary
Barry Allard wrote in message ...
>It worked, bios now rev. 2.00.1 !
>
>Barry
>
>
>
Barry Allard
I've been using win95 (A, B, and/or C) as boot disks for flash, etc.
Haven't had a problem yet. That's on 486, pentium, pent pro, pent II, k6,
k5 brand names, clones, and customs.
I could press shift-F5 on a win95 or win98 machine to get a dos clean boot,
but for a host adapter, i'd rather boot from floppy or the win98 cd-rom.
speaking of the win98 cd-rom, you can boot to win98 dos, without a hard
drive, from it if your cd drive support the bootable cd-rom format (el
torrito).
MrScary wrote in message ...
>Wait!!!
>
> There's a few things you should know.
>
>Use an old Dos 6.22 system disk if you can...if you use a Win9x system boot
>diskette remember to envoke the safe menu and select "Safe Mode Command
>Prompt only" before AFU is started.
>
>Remove Drvspace.bin and /or Dblspace.bin from any OS version system boot
>diskette that you use
>
>Use the utility included in AFU to backup your existing bios FIRST. After
>that select "Erase current bios" and THEN apply the new bios flash.
>
>
>>
>>>
>>
>
thanks,
Barry Allard
all...@ix.netcom.com
MrScary
disaster it never occured to me that you've been down this road enough to
walk it alone sorry
Barry Allard wrote in message ...
>I used to use dos 6.22 disks minus the *.bin file(s) Recently, at work,
>I've been using win95 (A, B, and/or C) as boot disks for flash, etc.
>Haven't had a problem yet. That's on 486, pentium, pent pro, pent II, k6,
>k5 brand names, clones, and customs.
>
>I could press shift-F5 on a win95 or win98 machine to get a dos clean boot,
>but for a host adapter, i'd rather boot from floppy or the win98 cd-rom.
Actually I meant using a floppy and still envoking the safe menu.......but
its all water under the bridge now.
>speaking of the win98 cd-rom, you can boot to win98 dos, without a hard
>drive, from it if your cd drive support the bootable cd-rom format (el
>torrito).
OEM is the only bootable cd but is not recommended for use by the general
public in any case.
I've got a funny feeling there is even a higher revision available but
it would require getting the ok on your retail SCSI card by Adaptec's
Support Dept. first. They give you access to the password protected files on
their server once the check for compatability and elegibility for support is
satisfied. Its still free save for the time invested.
FWIW