Re: (aic78xx) ReBoot Problem

Robert LaCasse schrieb:
On Mon, 05 Nov 2007 21:11:17 -0800, Manfred Preussig <CerhffvtZ@xxxxxx>ROT13
wrote:

|>On Wed, 31 Oct 2007 20:28:38 -0800, Robert LaCasse <maj-ic12@xxxxxxxxxxxx>
|>wrote:
|>
|>Manfred Preussig <CerhffvtZ@xxxxxx> Emailed:
|>
|>Hello,
|>first of all: it's not a good idea to let people only answer directly |>since the thread will get incomplete this way.

I agree with you.....but, I like both email and NG posts...since I
may miss a post from "bad servers" or something.....
o.k. But what you said (or I understand from it) is to write directly to you ... in spite of writing into the NG.

My Agent Reader supports both email and NG posts as once with a
"Suggest" reply method like Xnews but G2, MS, may be different report...

|>Let's take a look on this:
|>Robert LaCasse schrieb:
|>> (aic78xx) Boot Problem
|>>
|>> I have a 3 scsi cards on this comp, 2 are 2940/UW and all use the
|>> AHA aic78xx.mpd driver.
|>> |>> Problem is after some 3 reboots or so, the computer BSoD blue screen
|>> crashes with a [vxd vtdi error code] so I reinstalled the TCP/IP manager
|>> which the vtdi.vxd file refers to on win95/bmide (busmaster) w/98+ emulation
|>> system, but nothing improved.
|>> |>> The bootlog txt says it crashes when attempting to reload the
|>> aic78xx.mpd driver for the other card, and always crashes at that point.
|>> |>> The AHA Adapter 0 (aic78xx) has 3 internal peripheral disks on it
|>> with the system set to Designated Drive Letter/Removable/no int13...
|>> |>> Crash ..........Bootlog.txt
|>> |>> [000B745F] DEVICEINIT = VDEF |>> [000B745F] DEVICEINITSUCCESS = VDEF |>> [000B7461] Initing hsflop.pdr
|>> [000B7463] Init Success hsflop.pdr
|>> [000B7464] Initing drvwq117.vxd
|>> [000B7474] Init Success drvwq117.vxd
|>> [000B7474] Initing aic78xx.mpd
|>That's the first adaptec?
|>> [000B74E6] Init Success aic78xx.mpd
|>> [000B74E6] Initing aic78xx.mpd
|>That's the second one?
|>> |>> ...........END=Blue Screen Crash [vxd vtdi error code] |>That's an error of the vxd-incorporation driver itself. Possibly it does |>something in parallel so it thinks of the wrong source of crash.

Most definitely since I reinstalled all the Standard and Upgraded
versions of the TCP/IP manager, and still the same [vxd vtdi error code]....

Another proof of some Win/Scsi incompatibility, is on this crash
screen....when I renew the registry to see what happened, there is a bunch
of COPIES completely unrelated/unused backup programs, as well as IE5.5
Internet Explorer files stowed in my C:\temp or C:\Windows\temp from some
hidden crash files or maybe extracted from the Swap file..........
Take a look on the lines down here and in the other post. That's a good advise that the system corrupts the drive. It wirtes wrong data. The FAT is one of the parts re-written at shut down since the registry which allways have to be stored needs clusters to save it in. So the control scheme has to be updated too.

|>> |>> |>> |>> |>> SCSI MECHANIC BootLog of 3 AHA cards:
|>What's the type of the third one? I haven't find the type number yet?
|>> |>> SCSI Cmd: Adapter 0 (aic78xx)
|>> Target: 0:0
|>> SCSI Cdb: 12 01 80 00 ff 00 [Inquiry]
|>> Transfer Ct: 0xff
|>> Actual Ct: 0x12
|>> DataBuffer: 00 80 00 0e 30 30 32 32 33 36 39 35 30 30 30 30
|>> 50 4a 00 00 b0 45 a5 07 00 00 00 00 00 00 00 00
|>> 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
|>> 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
|>> Cmd Status: 1 (SCSI Request Completed Ok!) Error!
|>> Host Status: 12 (Data Under/Over Run)
|>> Target Status:0 (Ok)
|>What the other answerers don't look on and don't count on: The INQUIRY |>is the first command the SCSI system sets on a new-found device. It's |>the boot time ans the first device: Where does the additional data |>beyond the '4a'-value in the second line come from? Of course the buffer |>is zeroed-off before the allocation routine hands it out to the driver. |>The 'b0 45 a5 07' data are additional and came from something Possibly |>from the device since the buffer is reported as 'ff' in size so |>additional data to send and to recieve is no problem on ... in spite of |>the fact, that the device only says that this are 0x12, not 0x18.
|>

2 cards here the same bios at ctrl-a, and cards are selected from
the menu......I'm using mostly defaults (to start) then config, and no
(boot-up) device seek from Asus Board CMOS....

There is no AHAdapter card bios reserved till windows allocates one
for itself, and there are no conflicts in the (Win-System Properties/Device
Manager) settings.....
That's not really interesting since what the BIOS possibly does before windows don't know. What I mean it's the first contact from the driver to the drive and the hba. At this time the buffer to take the data from the device is allocated from windows memory management. This zeroes out the whole buffer before it present it to the driver since there can be kernel data in from use before.
I'm lost here I guess (0x12, not 0x18) ..
It's quite simple: The data lines consists of 16 bytes each. So data count of 0x12 means one line and two. But as you can see the last byte different from zero is the eighth one in second line so the data written there are 0x18 bytes in count. These could be old data from former uses of the data buffer but windows zeroes out the buffer before getting to the driver. It could be old data from former calls this driver did (since there is no zero-out) but the inquiry allways is the first action so there is nothing the driver can have done before. So where the data came from?

And of course the same is correct on the SCSI Mechanic too.

|>The second thing is: '12 01' is not the standard inquiry but an enhanced |>one (EVDP-Bit set). At boot time I think the standard inquiry data are |>needed since there for instance are the names of vendor and product are |>reported. So why it begans with the enhanced inquiry?

This is a (SCSI Mechanic 3.0) Bus Scan report...so it is more
inquisitive about it's query than Windows report.log
And what's in the report? To check for boot time faults we need boot time info I think (if avialable).

|>> |>> |>> SCSI Cmd: Adapter 0 (aic78xx)
|>> Target: 1:0
|>> SCSI Cdb: 12 01 80 00 ff 00 [Inquiry]
|>> Transfer Ct: 0xff
|>> Actual Ct: 0x12
|>> DataBuffer: 00 80 00 0e 4a 4c 31 35 37 31 35 36 30 50 50 42
|>> 4a 52 00 00 40 42 a5 07 00 00 00 00 00 00 00 00
|>> 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
|>> 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
|>> Cmd Status: 1 (SCSI Request Completed Ok!) Error!
|>> Host Status: 12 (Data Under/Over Run)
|>> Target Status:0 (Ok)
|>That's the same. The four last bytes not 0 are overrun-data ... and |>different from these seen above so no one can think of old data or so.
|>> |>> |>> SCSI Cmd: Adapter 0 (aic78xx)
|>> Target: 8:0
|>> SCSI Cdb: 12 01 80 00 ff 00 [Inquiry]
|>> Transfer Ct: 0xff
|>> Actual Ct: 0x26
|>> DataBuffer: 00 80 00 22 50 43 42 3d 32 30 31 31 33 30 30 30
|>> 30 32 20 20 3b 20 48 44 41 3d 31 38 34 37 30 37
|>> 33 31 33 37 34 39 34 37 00 00 00 00 00 00 00 00
|>> 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
|>> Cmd Status: 1 (SCSI Request Completed Ok!) Error!
|>> Host Status: 12 (Data Under/Over Run)
|>> Target Status:0 (Ok)
|>> |>And for the third one. The data count in third line is 8 so 0x26 is |>incorrect too.
|>> |>> OTHER CARDS:
|>> |>> SCSI Cmd: Adapter 1 (sparrow)
|>But what's that? I thought you have wrote it is an adaptec adaptor too |>and what is this sparrow at all? I don't know any adaptor from any |>source with such a name but of course I don't know all adaptors so can |>someone tell me what this is?
|>> Target: 4:0
|>> SCSI Cdb: 12 01 80 00 ff 00 [Inquiry]
|>> Transfer Ct: 0xff
|>> Actual Ct: 0xb
|>> DataBuffer: 01 80 00 07 47 43 30 31 35 5a 56 00 00 00 00 00
|>> 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
|>> 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
|>> 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
|>> Cmd Status: 1 (SCSI Request Completed Ok!) Error!
|>> Host Status: 12 (Data Under/Over Run)
|>> Target Status:0 (Ok)
|>Here's the count error a little bit more astonishing since here it seems |>to be correct ... but possibly not the driver.
|>> |>> |>> SCSI Cmd: Adapter 1 (sparrow)
|>> Target: 6:0
|>> SCSI Cdb: 12 01 80 00 ff 00 [Inquiry]
|>> Transfer Ct: 0xff
|>> Actual Ct: 0xb
|>> DataBuffer: 01 80 00 07 47 43 30 30 56 53 44 00 00 00 00 00
|>> 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
|>> 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
|>> 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
|>> Cmd Status: 1 (SCSI Request Completed Ok!) Error!
|>> Host Status: 12 (Data Under/Over Run)
|>> Target Status:0 (Ok)
|>> |>Same way too.
|>> |>> SCSI Cmd: Adapter 2 (ESDI_506)
|>That's really a missmatch. ESDI_506 is one of the predecessors of the |>IDE interface and is defined in early 80's I think. It is out-of-date |>for long time and I think it's today impossible to find any device for |>this interface anywhere outside a museum. So this is of course no |>ESDI_506 adaptor.

You right, this is an IDE Bus installed CD-ROM device that (SCSI
Mechanic 3.0) and most SCSI fix problems see as an adapter....
But ESDI? Windows is crazy sometimes.

|>> Target: 0:0
|>> SCSI Cdb: 00 00 00 00 00 00 [Test Unit Ready]
|>> Transfer Ct: 0x0
|>> Actual Ct: 0x0
|>> Host Status: 0 (Ok)
|>> Target Status:0 (Ok)
|>Of course no reaktion. Test Unit Ready would get no reaction anyway on ESDI.
|>> |>> |>> SCSI Cmd: Adapter 2 (ESDI_506)
|>> Target: 0:0
|>> SCSI Cdb: 00 00 00 00 00 00 [Test Unit Ready]
|>> Transfer Ct: 0x0
|>> Actual Ct: 0x0
|>> Host Status: 0 (Ok)
|>> Target Status:0 (Ok)
|>> |>> |>> SCSI Cmd: Adapter 2 (ESDI_506)
|>> Target: 0:0
|>> SCSI Cdb: 00 00 00 00 00 00 [Test Unit Ready]
|>> Transfer Ct: 0x0
|>> Actual Ct: 0x0
|>> Host Status: 0 (Ok)
|>> Target Status:0 (Ok)
|>> |>> |>> SCSI Cmd: Adapter 2 (ESDI_506)
|>> Target: 0:0
|>> SCSI Cdb: 5a 00 05 00 00 00 00 00 48 00 [Mode Sense Long]
|>> Transfer Ct: 0x48
|>> Actual Ct: 0x48
|>> DataBuffer: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
|>> 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
|>> 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
|>> 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
|>> Cmd Status: 4 (SCSI Request Error!)
|>> Host Status: 0 (Ok)
|>> Target Status:2 (Check Condition)
|>> Sense Data: 70 00 05 00 00 00 00 0a 00 00 00 00 24 00 00 00 |>> 00 00 |>> Illegal Request Error.
|>That's correct. Since ESDI doesn't support such a command it is an |>illegal request.
|>> Additional Sense: Invalid Field in CDB
|>> Sense Qualifier: 0
|>> |>> |>> ======================================
|>> |>> The only way I can reboot into windows after a crash, is to
|>> reinstall a former registry which is the same as the one being used, and it
|>> logs in...but that is a real drag ater a few thousand times.
|>> |>> Thanx in Advance
|>
|>What I think of: You use a windows95-System. What is the computer |>itself? Which kind of sytem is it? The processor?
|>
********************
* System Summary *
********************


< Hardware >
Processor: GenuineIntel Family 6 Model 6 338 MHz Math Support: On Chip
BIOS: Award Software 05/19/00
So windows even don't have the correct MB drivers. That's a little bit a problem too.
Bus Type: PCI
Ports: 1 Parallel, 2 Serial
Memory: 278 MB (47% Utilized)
Floppy Disks: 1.44 MB, 1.2 MB
Hard Disks: 7.87 GB
Multimedia: Sound, CD-ROM
Video: 800x600 in 64K Colors, NVIDIA RIVA TNT Ver. 4.0
What is the driver for this? It's a windows 98 or windows ME driver too I think. The card is newer (a lot newer I think) than the os. GDI.EXE (for graphics) can do some damage to the system too.

< Operating System >
Windows: 4.0 (Build 1111a)
DOS: 7.10
Net Clients: None


All windows drivers were re-upgraded after I mistakenly installed
from WinB/C to the downgrade-update files....
That I don't understand!

Has been upgraded with *Bus Master* to read the Processor/pci
"System Devices"........... configs

|>And then: What's about the dates: The mpd-driver for the AIC, what is |>the date of it? Is it newer then windows itself?
|>
;
; aic78xx.inf -- This file contains descriptions of all the Adaptec SCSI
; adapters supported by the 7800 Family Manager Set
;
; Copyright (c) 1993-1998 Adaptec, Inc.
;
;**********************************************************
;
; Version data
;
[Version]
signature="$Windows 95$"
It's for Win95 designed. That's o.k. I think.
Class=SCSIAdapter
Provider=%ADP%
CatalogFile=AIC78xx.cat

;**********************************************************
;
; Class install
;
[ClassInstall]
Addreg=SCSIReg

;**********************************************************
;
; SCSIReg data
;
[SCSIReg]
HKR,,,,%SCSIClassName%
HKR,,EnumPropPages,,"iosclass.dll,SCSIEnumPropPages"
HKR,,Icon,,-10


;**********************************************************
;
; Manufacturer list
;
[Manufacturer]
%ADP%=ADAPTEC

;**********************************************************
;
; List of device supported by aic78xx.mpd
;

Yes Newer Than Windows.....Aic78xx.mpd is 9-16-99


LOG0143: *****************************************************************
LOG0144: Starting ASPICHK installer on Mon Nov 05 21:39:54 2007
ASPICHK0438: OS = Windows 95 (4.00.1111)
UPGRADE1004: c:\windows\system\wnaspi32.dll dated 09/10/1999 20:06:00,
45056 bytes.
UPGRADE1062: c:\windows\system\wnaspi32.dll version 4.6.0.1021
Do you have the Win32-extension on it? Win95 itself can not use 32-bit dll's since it is 16 bit (not the vxd's of course).
UPGRADE1004: c:\windows\system\winaspi.dll dated 09/10/1999 20:06:00, 5600
bytes.
UPGRADE1062: c:\windows\system\winaspi.dll version 4.6.0.1021
UPGRADE1004: c:\windows\system\iosubsys\apix.vxd dated 09/10/1999 20:06:00,
22603 bytes.
UPGRADE1062: c:\windows\system\iosubsys\apix.vxd version 4.6.0.1021
UPGRADE1004: c:\windows\system\aspienum.vxd dated 09/10/1999 20:06:00, 7743
bytes.
UPGRADE1062: c:\windows\system\aspienum.vxd version 4.6.0.1021
ASPICHK0582: ASPI is properly installed and is fully operational. A total
of 4 host adapters have been detected.
That's correct of course with the ATAPI device.
LOG0218: Closing installer log on Mon Nov 05 21:39:55 2007
LOG0219:
*****************************************************************


|>greetings
|>
|>Manfred


If there is any chance to do so pull off all of the SCSI hba in spite of the one for the disks. If possible try changing the hba for the disks (you have three if these -take an advance of this). Remember you have SCSI: even the slowest one can drive the hdd -possible you need an adaptor from 50 to 68 but that's it. The drives can work with 50 too.

Of course there are the standard suggestions like: check termination; only two are allowed but less are bad too and so on. But I think you have checked these first so we haven't to pay more attention to this than this small paragraph.

greetings

Manfred

.