BIOS Updates forum

I'm experiencing some weird stuff in here, and I got some beliefs that it might be because of BIOS ...
So, first I would like to know a bit about BIOS structure.


I've reflashed the BIOS, loaded defaults, then I made a backup copy of the written BIOS,I went in windows and hashed the original image and the backup image, and they differ. So I used the WinHEX to search for differences and he found 905 8bit hex values different between the orig and backup !!!!! That's a good 100 lines ASM code ... or BIOS settings? Also if I compare BIOS image again after a while, new changes are made in the ROM, so it periodicaly changes. Do they also store the clock in the flash or it's just the code updating itself ... ?
If they do store settings and clock in the flash, I'm really disappointed by the engineers. Why would you do that, just so you render the BIOS flash dead sooner than if it were written less often, and make user buy a new mobo because of that?


So does the BIOS save the configurations on the flash(old ROM) along with the BIOS? If that's the case, why do you loose your settings if you remove the battery? Flash is not volatile.


EDIT: Ok, just found out that RTC is the "guy" that takes care of time, not BIOS, sooo .. what are those constant changes in my BIOS ROM?

please send both images in a zip-file to elhvb@outlook.com so I can compare where the differences are. most likely this is plug-n-play information.

Yep, ACPI information, put there by the operating system. More info on the bios in general:
http://en.wikipedia.org/wiki/BIOS#IBM_P ... BIOS_chips

More info on EEPROM (the technique used in bios chips):
http://en.wikipedia.org/wiki/EEPROM

The manufacturers usually specify the maximum number of rewrites being 1 million or more

Even with 10 reboots per day, that would equal about 2750 years...

Is this an Asus M3A board btw?
http://www.asus.com/Motherboards/AMD_AM2Plus/M3A/

Please check for bad capacitors:
http://www.badcaps.net

Well, BACKUP_1.ROM was made immediately after flashing the BIOS, no OS booted on, so the changes in ACPI were made by EZ Flasher, not OS (probably to accommodate the specific version/revision of mobo).
But can you tell what are the 30 bytes changed between BACKUP_1.ROM and BACKUP_2.ROM ?

Also, I couldn't read anywhere, but I figured out that BIOS settings are not saved on ROM, it must be saved on some RAM powered by the battery, probably on same chip with RTC. (there is some vague info here)

edwin wrote:Is this an Asus M3A board btw?
http://www.asus.com/Motherboards/AMD_AM2Plus/M3A/

Please check for bad capacitors:
http://www.badcaps.net

Yes, that's the mobo .. and no, there are no bad caps, at least no physical indication that any of them are failing.

That's most likely the last registered time for your disks or something similar. I am wondering what the odd behaviour is about because I think it's more a driver/program misbehaving.

Ok, you asked .. here is the full story

Some time ago (1 year) I started to experience a very weird behavior from the system, that is, whenever I boot on Windows I can't click on Start menu button and some other graphic elements on the screen, as if there is a transparent graphic layer above almost all graphic elements, but I can use the keyboard with no problem (opening start menu, etc).
To solve the issue, all I have to do (in Windows) is to log off and then log back in, after that everything comes back to normal and pointer device can access all graphic elements.

To narrow down any driver/OS/virus (software) problem, I've unplugged the HDD and booted on Live CDs and Windows Setup CDs to test if the problem persists, and the same thing happens. It doesn't matter what it is, WinXP, Win7, Linux, RecoveryCD, all have the same issue with the pointer device. To make it work, I have to do some other tricky things (opening menus, etc) since Log off is no longer an option on a Live CD.

Next I went for the hardware, and I've tried a different video card (the current one is on PCI-E and the tested one was on PCI) and nothing changed. I've measured and tested different PSU, I've used different RAM modules, even the mobo was recently changed (because the old one was slowly dieing-> random POST fail until one day it completely refuse to POST; I suspect the BIOS flash) and nothing changed.

The only component that I didn't test yet is the CPU (couldn't find a spare AM2 CPU). So the only suspects remaining are:
1.CPU with faulty functions related to GPU, that does not affect the rest of functionalities, and some how get fixed when certain stuff are reloaded in RAM (log off, log on)
2.a very clever BIOS virus that slowly over a year damaged the BIOS flash(during this time I had numerous POST fail with a beep code indicating missing/broken video card, no matter what video card I used), kept a copy in MBR of the disk and now injected itself in the new mobo .. oh, the new mobo does NOT have POST fails.

Isn't this science fiction or what?

Oh .. and one more think, if I do a hard reboot with power cut off, the issue disappears, but if I don't cut off the power between reboots, issue persists

You tried a different pointer device and a different power supply ofcourse?

That's like the first thing I did.

Gremlins? No seriously, this does sound quite weird. Normally, hotflashing your bios chip in another board with a known-good clean bios, should take care of any infections in there. If you then use any LiveCD without your original disk attached to the system, you still have the issue, you may have yet again a defective board, this may even be a design flaw (provided you had the M3A before as well).

The only other possibility may be a near-flat CMOS battery or yet again, bad capacitors but not bulging from the top, rather pushing down the seal at the bottom. What marking do the larger capacitors have on your board? Are they from a reputable brand or of a known bad-brand?

That has to be one hell of a coincidence. Before, I had Asus M2N-SLI Deluxe for 6 years and now I have M3A, bought few weeks a go. Issue appeared 1 year a go, no behavior changes shown after mobo was changed.
Both mobos don't have any signs of bulging or leaking capacitors. Button cell batteries are fine, 3.01V measured outside circuit.
Everything has been replaced(different mobo, different PSU, different video card tested with different connector, 4 RAM modules swapped one by one). The only hardware untested is the CPU. I might take a visit to a service company this week to test a different AM2 CPU, though CPU usually don't cause such persistent errors, and rather random ones.

Here is another funny thing, if I use an old serial mouse(yeah, I still have one of those around ) issue is gone .. o_0, if I use ANY USB or PS/2 mouse, the weird clicking issue kicks in.

I forgot to point this out, that when I encounter this issue, the mouse buttons are not completely unresponsive, as if driver/hardware device would not work, instead it's unresponsive only with some specific graphics elements. What I mean is that for instance in Windows I can select icons on desktop and pop up the context menu, but I can't interact through the mouse with the task bar or with the windows, even if I hover the start button with the mouse it does not trigger the highlighting animation. It's as if there is a invisible graphic layer over task bar and windows. I do the log off-log on thing with the keyboard, problem solved.

In Live CDs pretty much the same thing happens(only some graphic elements are click-able). To make it work, I usually open a context menu (with the keyboard inside a window) and I click with the mouse on one of the menus. After that it seems to work (when I hover buttons inside the window animations are triggered), but if the graphic layer is changed (different window) it starts again to block my mouse.

In Windows, if I click on a foremost window (title bar is solid color) with the mouse before "log off-log on", it becomes pale color, as if another window would become foremost, but there are no other windows.

If a different CPU won't fix it, I'll be clueless ...

Have you ever tried running something like this?

http://support.kaspersky.com/faq/?qid=208285998

Cool recovery disk, but I tried it and same story as with any other multi-layer graphic interface.
I tried a couple of IT service companies, but apparently no one is willing to put their used CPU in my motherboard unless I buy it .. can't blame them.

For this one I really suggest you go to a good repair shop and swap out stuff one by one until you find what is broken. I would say you have a broken windows install but then the live CDs should still work normal UNLESS those are based on the same driverset/install you are running your normal windows from.

Just working out the more exotic stuff:
Are you unhooking your harddrive when using a liveCD?
Does this also happen in Linux liveCDs?
Have you tried another monitor/screen?
Have you tried another keyboard?