Sun

Sun Ultra 40, stuck on 211 keyboard error

Hi All,

I tried this one via SunHelp / Rescue but so far with luck, so let's see if someone here has an idea:


I recently acquired cheap a Sun Ultra 40 that was on an early firmware.

With some help I managed to find bios version 1.50 and successfully updated my Ultra 40 M2.

Initially it worked fine but not too long after the update it seems to have mysteriously developed an unsolvable 'error 211 Keyboard error' on startup, basically stopping the machine from booting past the initial bios screen. The error pops up along with the hint to press F1 to resume or F2 to enter the bios but the keystrokes don't produce any result.

I tried 2 different keyboards (in addition to the microsoft keyboard that I was using initially and worked fine before), cleaning out the usb ports, resetting the cmos via the motherboardb jumper and replacing the motherboard battery, all without success. As part of these experiments the machine was also disconnected from power for an hour or so.

To be clear, the error also pops up when nothing at all is connected to the computer, so with all USB ports empty. No change when using either of the back or the front USB's and different keyboard combo's.

If anybody has any hints or further tricks to try to shock the workstation to ignore this error that would be excellent!
Failing a solution the machine is essentially useless.

Thanks in advance!
JP
:Tezro: :Fuel: :O2: :320: :1600SW:
Not sure what the manufacturer of the firmware in your Ultra 40 M2 is, but for BIOS setups from Award there is an option - named "Halt On" usually located under "Standard CMOS features" - where you can configure on which errors the machine should halt on startup. This can also be configured to not halt on any errors and might workaround your problem.
:Indy: :O2: :Octane: :Octane2: :O200: = :O200: - :O200: = :O200: (O200 cluster w/2 GIGAchannel cabinets)
[ ( hp ) ] 712/80 c3000 (dead) :hpserv: (J5600) c3700 c3750 c8000 rp2470 :rx2600: (rx2620) rx4640
| d | i | g | i | t | a | l | AXPpci33 AlphaStation 200 AlphaStation 255 PWS 500au AlphaServer DS20E AlphaServer DS25
C O B A L T Qube 2 Qube 3 RaQ RaQ 2 RaQ 4r RaQ XTR
Hi Johnnym - main problem is that I can't access the bios setup in any way...no keyboard presses are registered. Cheers.
:Tezro: :Fuel: :O2: :320: :1600SW:
Hi.

BIOS battery down or low capacity?
TABLE 5-3 Battery Specifications
Specification Value
Voltage 3 VDC
Type CR 2032


And 1.50 is the most recent version of the BIOS? For U40 M1 it should be at least 1.60.

Sources:
https://docs.oracle.com/cd/E19127-01/ul ... index.html
https://docs.oracle.com/cd/E19127-01/ul ... Id-1006784
During POST, the BIOS outputs the error code to I/O port 80h. The workstation
motherboard is equipped with a two segment LED display that displays the current
value of port 80h.
If the BIOS detects a terminal error condition, it issues a terminal-error beep code,
attempts to display the error code on upper left corner of the screen and on the port
80h LED display, and halts POST. If the system hangs before the BIOS can process
the error, the value displayed at port 80h is the last test performed.
211 is 0xD3. - Mmm, nothing according to the service manual.
or maybe w/o 2 will be 11 "General dispatcher for alternate register initialization. Set initial POST values for other hardware devices defined in the register tables."

It's helpful to put all unrequired hardware out of the system (HDD, SCSI-Ctrl) and restore the BIOS defaults by switch/jumper on the motherboard.

Which hex code is shown by POST on the two segment LED (motherboard)?
Sun Solaris 2.4 @ SPARCstation 2 and SNI PCD-4H . Migration path: NetBSD
It is getting stuck on hex code 60, which:

60

Determine and test the amount of extended memory available. Determine if memory exists by writing to a few strategic locations and see if the data can be read back. If so, perform an address-line test and a RAM test on the memory. Save the total extended memory size in the CMOS at cmosExtende

I tried reseating all memory sticks, no change. I have replace the motherboard battery already. 1.5 is nearly the newest bios, newest one I could find these days. Anyway, the machine booted for a few times after applying it....

Cheers
:Tezro: :Fuel: :O2: :320: :1600SW:
Hombre71 wrote: Hi Johnnym - main problem is that I can't access the bios setup in any way...no keyboard presses are registered. Cheers.

Sorry, I overlooked:
Hombre71 wrote: but the keystrokes don't produce any result.


Hombre71 wrote: [...]
I tried reseating all memory sticks, no change. I have replace the motherboard battery already. 1.5 is nearly the newest bios, newest one I could find these days. Anyway, the machine booted for a few times after applying it....

Have you already tried to go down with memory modules to the absolute minimum? I know that Athlon64s can do with a single memory module (single channel) but I'm unsure if Opterons can also do this. If you have two CPUs installed you could also try to remove the second one (although cumbersome and delicate for Socket F CPUs) and test all memory modules one at a time (if supported). If you have two CPUs and don't want to remove the second one you could test two modules at a time (if supported).

Or if you have another machine that supports DDR2 memory, you could install the modules there and test them with memtest86+ for a few complete runs to see if they have obvious defects.

Although possibly unrelated: in the past I had problems with Athlon64 X2 systems using DDR2-800/PC2-6400 modules - Windows XP crashed a few minutes after startup, Linux (Ubuntu) worked most of the time. Although these should be supported by the CPU, it only worked reliably if I installed additional DDR2-667/PC2-5300 modules to slow down the memory bus. So if you have DDR2-400/PC2-3200 modules around you could give it a try, maybe they work in your machine. In the end I found out that one of the DDR2-800/PC2-6400 modules was defective, but interestingly the additional modules seemed to have masked this.
:Indy: :O2: :Octane: :Octane2: :O200: = :O200: - :O200: = :O200: (O200 cluster w/2 GIGAchannel cabinets)
[ ( hp ) ] 712/80 c3000 (dead) :hpserv: (J5600) c3700 c3750 c8000 rp2470 :rx2600: (rx2620) rx4640
| d | i | g | i | t | a | l | AXPpci33 AlphaStation 200 AlphaStation 255 PWS 500au AlphaServer DS20E AlphaServer DS25
C O B A L T Qube 2 Qube 3 RaQ RaQ 2 RaQ 4r RaQ XTR
Good tip by johnnym...give it a try.

Source: https://docs.oracle.com/cd/E19127-01/ul ... Id-1000253
Caution - DIMMs are installed in adjacent pairs . However, if you replace a single DIMM, the new DIMM must be identical to the DIMM that you removed.
(...)
Sun Ultra 40 memory is installed as pairs of DDR400 PC3200 Registered ECC DIMMs. Sun Ultra 40 M2 memory is installed as pairs of DDR2-667 DIMMs.
(...)
The DIMMs in a pair must be identical, from the same manufacturer with the same type and number of memory devices, the same amount of memory per device, and the same memory speed. All installed DIMMs must have the same memory speed.
  • The minimum configuration is two 512-Mbyte DIMMs installed in slots 0 and 1 .

Caution - The Ultra 40 and Ultra 40 M2 DIMMs are numbered the same but located differently, and this difference is critical. On the Ultra 40, DIMM slots 0/1 and 4/5 are closestto their respective CPUs. On the Ultra 40 M2, DIMM slots 0/1 and 4/5 are farthest from their respective CPUs .
Sun Solaris 2.4 @ SPARCstation 2 and SNI PCD-4H . Migration path: NetBSD