Acer MXM BIOS Mods and discussion (was: x920G vs. 9600M GT)

***NOTE: If you have a 9600M GT or similar card, that works with the Microsoft standard VGA driver, but produces a lockup/BSOD/black screen as soon as you install recent Nvidia drivers, read this. This is a known bug in Nvidia drivers, and a workaround is available in the form of a VBIOS patch. This is not a problem with the BIOSes in this thread.

Greetings!

I think I found the reason behind GeForce 9600M GT's reluctance to work with Acer Aspire 5920G/6920G/8920G laptops. Apparently it all boils down to a corrupt MXM structure in the BIOS.

The 6920G and 8920G <s>share the same BIOS</s> EDIT: They actually use different BIOSes, despite containing the same motherboard (though the MXM structure is the same in both), and indeed, their symptoms are also the same: no display on the internal LCD during POST. However, connecting an external display to the HDMI port allows POST messages to show up on that screen.

So, I looked into the MXM structure embedded in the 6920G's BIOS (armed with the recently leaked specification for the MXM structure), and found the following:
The output device structure for the HDMI port of the 6/8920G is 0x00F9F78A2120, which parses to:
Device type: TMDS (DVI or HDMI)
DDC port: DDC-B (MXM pins 230 & 232)
Connector: HDMI Type A socket
Connector location: External
Link type: Single-link DVI/HDMI connected to TMDS link A
HDMI audio source: No audio
Digital signal drive strength: Normal
Digital reserved field 1 (2 bits, should always be 11): 01 (incorrect, violates the MXM spec!)
Output select GPIO: None
DDC select GPIO: None
Device detection GPIO: #0, active low (logic low on GPIO pin 0 indicates a display plugged into the HDMI port)
System hot plug notification: Disabled

The culprit is apparently Device detection GPIO (marked in red), which is presumably always low/0 on this machine, indicating a device plugged in! Because the boot display order is VGA first, then HDMI, finally internal display; connecting a display to HDMI (or fooling the MXM VGA into thinking one is connected, as it is happening now) suppresses POST display on the internal LCD, and directs it to the HDMI port. When Windows finally boots up, it recognizes the internal display and makes it primary (Windows always considers the internal display primary by default, regardless of the MXM structure), so display in Windows is normal.

So, the MXM VGA basically thinks that an HDMI display is plugged in all the time, and proceeds to make that imaginary display primary.

The 5920G (and its little brother, the 4720G) fares even worse: it flat out refuses to POST with a 9600M GT plugged in! Again, a quick look at the MXM structure (hidden in BIOSCOD02.ROM inside the BIOS image) reveals that something is amiss, even more so than with the 6/8920G:
-The checksum of the MXM structure is wrong.
-On this notebook, the boot display preference order is different: internal first, then analog VGA, then HDMI, and finally TV-Out. So, the problem is not the same as with the 6920G. However:
-The internal display's output device struct claims that the internal display can also produce sound through a SPDIF connection, and that the link between the internal panel and the computer is long-distance, requiring extra signal power. A field identified as "reserved, set to all ones" is set to all zeros.
-According to their respective output device structs in the MXM structure, the VGA, HDMI and TV outputs are all internal connectors.
-The VGA output has a "TV format" field set to NTSC-M, despite not being a TV output. The Digital Connection field is set to "Reserved for future uses, do not use" instead of "Not applicable".
-The HDMI link again has the reserved field zeroed out.
-The TV out is identified as an HDTV YPbPr component output, but the TV format is "legacy 525-line NTSC-M" (should be "480i" for an HD connector). And again, Digital connection is "Reserved for future uses". But here comes the real bummer:
-All outputs (including the internal display!) mark GPIO#0 for device detection, interpreting logic 0 as Connected, and 1 as Disconnected. This is wrong in many ways, e.g. one GPIO line can only be used by one display; the internal display has no device detection (it is always plugged in); the board has no GPIO expander, so GPIO#0 doesn't even exist, and is always read as 0; and finally the HDMI output also marks this GPIO as its Output selection GPIO - which is impossible, as a GPIO can't be in more than one mode at the same time.
-The MXM connector claims to be MXM-HE, capable of delivering 3.2W to regular and 12W to HE cards. I the reality, the connector is MXM-II, so it doesn't support HE cards at all; and I'm pretty sure it can handle more than 3.2W when dealing with regular Type II cards. (I've modded this one to 32W.)
-The final two fields are a vendor-specific field for PCI\VEN_0401 (a nonexistent vendor), and a weird 4-byte field that appears to be completely random junk. Disassembling BIOSCOD02.ROM in IDA revealed that this field is not really junk - it is the beginning of the x86 machine code that responds to MXM interrupts. Apparently when the BIOS was built, the end of the already-wrong MXM structure got overwritten with BIOS code!

I've made an attempt to fix these errors, but I have no Acer 5/6/8920G to test them, <s>so the following modded BIOSes are untested</s> Update: 6920G+9600M GT confirmed working by lizzarddos, 5920G confirmed by multiple users to help 9500M GS, 9600M GS and Quadro cards, 5920G + 9600M GT confirmed by u6b36ef, 9920G personally tested by me, so there is no reason why the 4720G would not work.

(BIOSes last updated 2011-05-29.)
Aspire 4720(Z)G (Phoenix): request download ticket | ifile.it
Aspire 5520G (Phoenix, with unlocked menus): http://ifile.it/bi2qtc6/IC50133_MXM.WPH
Aspire 5720(Z)G, 7720(Z)G (InsydeH2O, with unlocked menus and integrated EFI shell, probably fixes external display issues, updated 2011-11-24): http://ifile.it/3wlxm6j/icl50_1.45_black_edition_40.fd
Aspire 5920G (Phoenix, with unlocked menus, updated 2011-11-06): http://ifile.it/noxplab/5920G_MXM_V3.WPH
Aspire 6920G (Phoenix, with unlocked menus): <!-- request download ticket | ifile.it previous: --> request download ticket | ifile.it
Aspire 8920G (Phoenix, with unlocked menus): <!-- request download ticket | ifile.it previous: --> request download ticket | ifile.it
NEW! Aspire 9920G (Phoenix, with unlocked menus, updated 2011-11-06, use with driver v275.33 and these VBIOSes): http://ifile.it/ej0c3oi/CV120_MXM_V2.WPH
Extensa 5620G, Travelmate 5720G (Phoenix, with unlocked menus, experimental): request download ticket | ifile.it Link fixed 2012-03-28
<s>iFile.it folder containing all current BIOSes: BIOS Mods | ifile.it</s> This folder contains newer BIOSes than the ones above, and one of them (5920G) has produced a bad flash. Consider these to be beta BIOSes until further notice!
Please test them and report any results here!

Note: <s>The 5520G, 7520G and 9920G are affected, too - however, the BIOSes on these machines don't contain an MXM structure at all, so unless someone figures out how to add an interrupt handler for INT 15h, EAX=0x5F80 to BIOSCOD02.ROM, they cannot be patched. (Note: Even if someone finds the interrupt handler for INT 15h, EAX=0xE820 - memory map, that would be of great help.)</s> 5520G and 9920G are available now. The 5520G BIOS may also work on the 7520G, but this is untested.

The Aspire 5720G (ICL50) BIOS is NOT required for 9600M GT upgrades - it primarily fixes issues with ATI cards refusing to drive external displays. It is also usable on Aspire 5315, 5320, 5715(Z)(G), 5720(Z)(G), 7320, 7720(Z)(G) and eMachines E510 machines, including IGP ones, to unlock extra tweaking menus. It does not, however, enable upgrading IGP ICL50s to MXM cards - that is still physically impossible.

Thanks,
.NetRolller 3D

EDIT by downloads on January 25th 2014:
.NetRolller 3D kindly re-uploaded these BIOSes here:
https://mega.co.nz/#F!iJ5FwAoR!PqYOcbC_r9dciJ5sut6AfA

 

Holy smokes, .NetRoller3D, that might just be the ticket!
I'll put this topic in the top post of the Acer MXM thread so it gets more eyes.

 

Finally a possible fix! Unfortunately, I've burned my 9600m gt due to extreme overclock. But don't fret, I've just received a replacement today! Gonna try it out tonight.

 

@lizzarddos: Eagerly awaiting the results.

@TehSuigi: Thanks! BTW, you got my nickname wrong in the MXM cards topic... Also, "[Note: The 9600M GT is NOT compatible with the 5920G]" is still there after the 5920G's list (though it may worth keeping it there until my fix is tested on the 5920G).

 

Just a small favor though, could you do the tweak on this modded bios v1.16 for acer 6920G?

Aspire 6920G BIOS_1.16 MOD.rar

P.S. The modded bios is from the Bios mod Request Thread in case you're wondering. http://forum.notebookreview.com/acer/480992-acer-laptop-phoenix-bios-bios-mod-request-129.html

 

@lizzarddos: Here we go: request download ticket | ifile.it

 

Important note to those who downloaded the BIOSes before 13:00 UTC, 2011/05/06: Please re-download your BIOS, my previous uploads were incorrect, and did not contain the rebuilt MXM structures. This has been corrected in the re-uploaded BIOSes.
(Also, the 6920G BIOS now unlocks additional Setup menus, courtesy of lizzarddos.)

 

Ok I've swap in my replacement 9600M GT and flashed the modded bios and

:::drumrolls:::

It freakin' works!!! Thanks so much NET roller this is the best thing that happened ever since I upgraded. Mystery of the black screen solved FINALLY! Can't thank you enough for this amazing discovery. WOW

To anyone hesitating to try, this is guranteed to work on an Acer 6920G with 9600M GT upgrade!

 

Thanks for the confirmation.

And remember, none of this would have been possible by the people behind the borec.cz website, who have been kind enough to upload the MXM 2.1 specification, normally only available to MXM-SIG members.

In case any of you borec.cz staff are reading this topic, whoever you are - THANKS!

 

@.NetRolller_3D: Sorry about the mix-up; will correct with the next update (not at my PC right now)
@lizzarddos: Screenshots please! This is Acer MXM history we're making here!!

Now I'm tempted to do the same; maybe to celebrate the end of my 6920G's warranty?

 

Move into the light young one

 

Stuff like this always made me wonder why should we trust the hardware guy on getting things right...
I just need half of your 1337 RE skills.

 

I've said it before, and I'll say it again: when the manufacturer fails, the community will succeed.

 

Here are the screenshots.





Whose planning to upgrade now to the fully working 9600M GT on Acer 6920G?

Do note that there are 2 revisions of the 9600MGT, A1 and C1. I bought the A1 revision second time round thinking that there was no difference. How wrong was I. I could only manage a stable OC of 650/850/1625. The C1 revision could go much higher at 780/1000/1800. Though the C1 does cost more than the A1 revision. 7.1K vs 8K 3D Mark 06 . Being cheapskate not good huh

 

could you be a little specific on how to update the bios ? do i update the 9600m bios or the 6920g bios ? i mean do i update the video bios or the motherboard's bios or both ? and what program should i use ; nvflash for the 9600m and the crisis disk for 6920g ? i know the question seems kind of stupid; i only update "normal" graphic card bioses and motherboard bioses

 

Can you try this with the BIOS of an MSI gx620?

I am trying to get an ATI 4850 to work on internal display.

Ironically the 4850 card was made for the gx720, and of course works on its internal display. Maybe you could do a comparison?

 

@marlanu: Just the 6920G motherboard BIOS; the MXM card has nothing wrong with it, it's the way MXM is implemented in the 6920G's BIOS that needs fixing. Use the WinPhlash utility packaged with Acer's own BIOS updates from their website.

 

@ .NetRolller 3D,

That "MXM Graphics Module Specification" document was a good read.

I did recently open up a BIOS (video or system - I can't remember), file to see if I could alter the fan switch on temps, for another laptop. The file was a .BIN file and was a page of machine code, so I shut it down again.

Anyway, I read through your first posting on this thread quite a few times. It's a challenging read.

It's a scary idea flashing a modded SYSTEM BIOS file, but it's been done now for a 6920 - congratulations.
- Is the 6920/8920 fix sure to work on the 8920?
- I am nervous is the 5920 solution going to work. You were clear in making the point the 5920 structure was worse. I may try it though, it seems a good idea. It looks like your fix will cure the 9500M GS problem with the 5920. (Based on refusing to POST behaviour of both the 9500M GS, and the 9600M GT.) Confusingly though as I said on MXM Cards and Models, some have 9500M GS working in the 5920. (These were 9500M GS fitted by Acer under warranty when their 8600M GT's failed). That is with the same SBIOS we are all using - V1.381.
- What does (based on a SLIC2.1-modded BIOS): for the 5920 modded BIOS mean.

Also, what did you mean please, when you said you,
"-The TV out is identified as an HDTV YPbPr component output, but the TV format is "legacy 525-line NTSC-M" (should be "480i" for an HD connector). "
The 5920 is HD to 720 vertical resolution, at 1200x800. (Sorry if I have my units and understanding wrong.)

Thanks..

 

@niffcreature: I've checked the MXM structure in this BIOS, and it appears to be completely valid (apart from a minor nit in the VGA output's descriptor, which may cause some very-strict VBIOSes to reject the MXM structure - but that would result in no POST at all, not just the internal display not working; and on top of that, the display preference in the MXM structure is internal->VGA->HDMI, so an error in VGA shall not break the internal display).
What is still possible is that it doesn't match the actual hardware of the GX620; I've tried to get a schematic to verify this, but it's apparently only available on paysites, and I'm not going to pay for it - if you will, you can Google for "MSI GX620 schematic" and purchase it, but AFAIK even discussing schematics is forbidden on this forum.

Note that the BIOS I tested was version 1.14 - if you are running a different version, I can't vouch for that.

Also, it is possible that the 4850's VBIOS hardcodes certain parameters for the GX720, instead of getting them from the MXM structure - try getting a different VBIOS, preferably an ATI reference one.

@ub6b36ef:
-The 8920G is just a 6920G in a different (bigger) case, AFAI<!-- comment -->K. (Like the 7720G is really just a bigger 5720G.)
-IIRC there are 2 revisions of the 9500M GS, with different VBIOSes. One of them may either use hardcoded MXM parameters (entirely independent of the MXM structure), or contain fallbacks in case the MXM structure is corrupt. The modded SBIOS should get both revisions to work.
-The SLIC-modded comment refers to the fact that the BIOS I used for modding already contained an SLIC mod, so it should probably still be present in the modded BIOS.
-The TV comment refers to the TV-out connector, not the internal display. It is incorrectly marked as defaulting to legacy NTSC - however, the closest thing to legacy NTSC that exists in the HD world is 480i. This is unrelated to the fact that the internal display is good up to 720p.


As a bonus, here is an official 6920/8920 BIOS image (not modded) that contains a 64-bit WinPhlash (usable for flashing the modded BIOS):
ftp://ftp.work.acer-euro.com/notebook/aspire_8920G/vista/Bios/v1.14.zip

EDIT: Acronym auto-expansion sucks!

 

so this will also work on my
Acer Aspire 6920G
Core 2 Duo T8300 -with a GeForce 9500M GS full hd

did you try this bios @Tehsuigi

 

@ .NetRolller 3D

Looks good.
Thankyou for replying.

Yes the mainboard in the 7720 is also the same in the 5715 too.
I wondered though if I flash an 8920 with your modded SBIOS, will it smile back saying, "I'm a 6920 now".

Figuring it through, does it mean that the system information is independant of the SBIOS? It probably makes sense to you. When you look in CPU-Z's mainboard tab though, it shows the model name. In the case of the 6920, it's "6920" (ref lizzarddos screenshots). Also, when you look in the SBIOS (at boot (f2)) under the information tab, there's - 'Product name' - which states the laptop model. Then if the SBIOS is identical, doesn't it follow there must be a mainboard revision difference, or something different. Is it worth it to, or did you allready look at the 6920 and the 8920 SBIOS' code side by side.

(Additional info: I emailed to CPU-Z about a CPU. They replied that CPU-Z reads the CPU info from the CPU, not the SBIOS.) Maybe then CPU-Z reads the board info in the same way. Maybe that hints to a difference in board versions, eg in the 7720 and 5715, so CPU-Z and SBIOS can read different models.

U6b36ef.

 

<s>Not sure about PhoenixBIOS, but InsydeH2O (the one my 5720G has) uses a not-to-flash region in the BIOS flash chip (NVRAM) to store model identifiers & serial numbers. Phoenix likely does the same.</s> I take that back. Downloaded v1.16 for the 8920G, and it IS indeed different (though mostly single-byte changes, and of course, changes to the DMI strings). So, no, the 6920G BIOS probably doesn't fit the 8920G correctly, despite the motherboard being the same.

 

@ .NetRolller 3D

That makes perfect sense now.

- Phew. No worries this time though. (I'm not likely to flash eight-nine-twenny for a while; seems like no-one else has tried it.) It looks like you have added an individual modified sBIOS for the 8920, to your opening page; cool.

- Also you have added notes about other laptops, including the 9920, and truly the 9920 has some serious clash.
If you have not allready seen it, take a look at this page/thread. http://forum.notebookreview.com/acer/475196-9920g-gpu-options-9.html The crux of the matter so far is, it's only known to respond to one individual VBIOS, on one card.

Previously I wondered if the clash of the 9920 was related by the SBIOS to the unique inverter it uses. (The inverter is about 13cmx5cm.) Probably though it seems more likely the SBIOS now since you looked over it, and saw flaws.

Thanks again,
U6b36ef.

- I had a look with a 'hex editor' at a SBIOS again, and still can only see hex. If you can explain if it's possible to view the SBIOS in English (higher level language), that would be most appreciated. (Additionally, maybe I can find what you need for the 9920/7520/5520 mod.)

 

Yeah, it should work on your machine.
I'm not going to try it because I really have no need to upgrade; this thing's getting turned into a media center as soon as I get my next system.

 

@u6b36ef: You first need to extract the BIOS using PhoenixTool or Phoenix BIOS Editor. Then, use a disassembler such as IDA on the extracted BIOSCODxx.ROM files.

 

@ .NetRolller 3D
Aha! Thanks. I had tried Free-Hex-Editor-neo; the disassembler was not in the free version. I hadn't tried an extraction to get the file. I just opened it from an SBIOS file on my desktop, with a .bin reader, and free-Hex-etc. Adios MXM person.

U6b36ef

 

well my 8920G has died again.
this is the 2nd 9500M card to fail, leaving the charger on over night seems to kill the VGA (anyone know why?)

I guess I could try the 'corrected' BIOS but it wont be for about a month, scouring ebay for a good deal on a 9600GT

-.NetRoller, does the 8920G BIOS have VT enabled?

 

Probably overheating-related, or maybe your charger has inadequate input or output filtering, resulting in voltage spikes or ripple in the output. Check the charger for Fuhjyyus and other crapacitors.

And yes, the BIOS is VT-enabled (or has an option to enable VT).

EDIT: @TehSuigi: I see you updated MXM models and cards again. You *almost* got my nickname right this time. (That's a space, not an underscore.)

 

Your name is what I say it is! Now sit down and take your kudos like man.

 

@.NetRolller 3D - Great job, man. Quite a lot of people would be happy

 

Is anyone willing to test this on a 5920G?

 

@ Lizzarddos,

Please can you post info about what temperatures do you have with your current 9600M GT on stock clocks? Also, how high did temps go when you overclocked your previous C1 version 9600M GT.

 

What do you mean to test? To check if the BIOS works at all, or just the 9600M GT thing?

 

Stock clocks
Idle:46-49c
Peak:63c

Oc'd (650/850/1625)
Idle:49-50c
Peak:66c

Old card, C1 version, Oc'd (780/1000/1800)
Idle:53c
Peak:70-72c

 

Whether it accepts previously picky cards, like the 9600M GT. I really hope it does NOT break previously working cards - but given that the end of the MXM structure was overwritten with machine code in the original BIOS (requiring me to shorten the structure), it is possible that even the byte that became the new checksum byte was originally code, so regression testing may also be needed.

 

It may break previously working Acer proprietary cards - but it sounds like it definitely will still work with the standard 9600m GT cards as well as the HP fx 770m cards flashed to standard BIOS...

Apparently getting kudos for figuring that out is not a requirement as it isn't important enough to put on the MXM page

 

hi everyone, i search for help and end up here

just recently bought nvidia fx700 quadro board, upgrade for my acer aspire 5720g but there is no image at all on internal or external displays BUT i can hear windows boot sound
i need that quadro, is there any way i can see something

thankyou for any help

btw, i bought it on ebay and have no idea if it works at all, but seller is 100% possitive and accepts my request for return

 

Well, if you remove the SLIC 2.1 mod (I don't want to mess-up my Vista registration), I'll give it a go. I don't have 9600M GT, so it would be just it-runs/it-doesn't test

 

@ Lizzarddos,

Given you rep one for doing the systemBIOS flash first. Can't stop 'laughing out loud' (cslol), when I think what you did.

Thanks for the temp info. Staggered to be honest, how low the readings are that you have. The 8600M GT does 50'C to 77'C. Statistically with the tech (lithography) size of the 9600m reduced to 65nm (over 80nm of the 8600m), lower temps could be forseen. Contradictory though the 9600m has more transistors on chip, and clocking 275MHz faster. (Power consumption is kind of reported the same.)

@ .NetRolller 3D

Oh yeah - gave you +one rep; veritably deserved.
I was hanging onto the edge just recently thinking of the flash for the 5920. (You scared me with this new stuff you posted today though.) Maybe I should learn blind flashing for a repair in case something goes wrong. I have a no-POST 9500M GS which, as you discussed should be got to work by your SBIOS mod.

A flash on an 8920 seems straightforward since Lizzarddos made the 6920. Now you have the distinction between the 8920 SBIOS and the 6920 SBIOS, it should be ok. (That's presuming the 8920 SBIOS mod mirrors the 6920 SBIOS mod, yeah?)

Allthough I used the disassembler you suggested, I still can't see past columns of hex. It looks better than it did before I disassembled, which was one long block of hex. I can't make it look the way it appears in the MXM pdf. document, all strings and brackets etc. No worries though, I think I might give up, even though I really wanted to get to changing fan switch on temp. (That was just about last page in the MXM document, btw).

U6b36ef.

 

@ juodvarnis: You probably have a broken card.
The 5720G has an InsydeH2O "BIOS" (actually an EFI), and indeed there are errors in the MXM structure, similar to the ones in the 6920G. Oddly enough, I've never encountered any MXM incompatibilities with this machine before.

But anyway, here is a BIOS for 5720G and other ICL50/ICK70-based Acers/eMachines with fixed MXM structure (+ SLIC2.1 + unlocked Advanced menus): request download ticket | ifile.it

Won't brick/regress anything (tested on my 5720G), but I have no FX770M to test. Use InsydeFlash to flash (but make a backup of your old BIOS!); if anything goes wrong, use the Fn+Esc method to recover.

@triturbo: Just checked, all SLICs in these BIOSes are from Acer, so they should not break activation. (SLIC2.1 BIOS can activate SLIC2.0 OS, just not vice-versa.)

@u6b36ef: Create a Phoenix crisis disk with an unmodded BIOS first. That should allow you to recover (given that I have not modified the bootblock).
And yes, the 8920 mod is exactly the same as the 6920 (these 2 notebooks contain the exact same MXM structure, bit-for-bit).

About the disassembler: If you just want to see the MXM structure, merely search for "MXM_" in a hex editor in BIOSCOD02.ROM. Currently, there is no way to parse it programmatically - you need to read it with the help of the documentation. I'm planning to write a tool to dump the MXM structure found in a BIOS module in human-readable form later.
However, fan temps switching should actually be managed by the ACPI AML, unless it is being controlled directly by the EC (as done on the 5720G).

 

thnx, netroller, i already found your another post with link to this bios, updated, but the card dont work anyway, so i will return it to seller and will take cheaper good ol' gf9600m gt
btw i saw few post here from people with same problem (but still no solution), i doubt that the card is broken

sad, as i work with 3ds max mainly and never had any quadro or firegl card, always modding drivers to force pro capabilities (now hd2600>firegl5600 on w7u 64bit)

starting save money for aspire 3830TG

 

There is no way you could have saw this BIOS in another post of mine, as I built it right before posting it here today - you've probably downloaded a predecessor of this BIOS, which contained all fixes in this one (Advanced menu, VT enable, SLIC, maybe a few more I've forgotten), except for the MXM fix.
(The easiest way to tell if you are running the correct BIOS is that the MXM-patched BIOS should display "MXM Edition" in green under the Intel logo, when quick/silent boot is enabled. My previous mods display "Advanced Menu Edition", or "Advanced Menu Edition V2" under the logo, while stock Acer BIOS has no text there at all.)

However, it's easily possible that the card is broken - the usual problem with 770M cards in a 5720G is "MXM structure not found or invalid, wait 30 seconds", and not the inability to get a display.

 

true, that was v2. mxm edition no help. thanx anyway...

 

well my 9600GT C1 is on its way.
its the DDR2 version which was way cheaper than DDR3.

I couldnt justify paying over $200 USD.

thats half the price of an AMD based Netbook.

will update in a week or so

 

@mookill You should have gotten the DDR3 version as it is way faster, 1000-1200 3dmarks difference vs the ddr2 version (Just remembering from the top of my head, someone correct me if I'm wrong). If it was me, I'd rather go for the older revision A1 and grab the DDR3. Just my humble opinion

 

About the 5920 flash uncertainty.
-Is it that the part of the MXM structure that was missing from from the original SBIOS - is the problem? (Meaning you don't know what original modifications were given to the original MXM structure.) We have seen that the 9920 SBIOS/VBIOS construct is unique, and also the clashing 9500M GS BIOS's. Do you suspect the 5920 standard cards, may have some unique SBIOS MXM structure? I had been kind of figuring the MXM was the mostly meant to be the same for each card, excluding 9920 for example.
-Or is it the command of the checksum byte.

Anyway I have read about Phoenix flashing and discovered some have difficulty in getting boot from USB stick. Is it simply a case of dropping the Phoenix crisis files onto a USB stick, since I don't have a floppy. Once I understand preparation of the USB stick, I'll be OK from there. (I realise VBIOS flash sticks need to be bootable. Yet I can't get the picture of what to do 'exactly' for the Phoenix flash disc preparation, if I need to repair.)

I have a G84 GPU (8600m gt) that I don't mind using for testing. It actually needs oven baking since it has not picture at the moment. If I bake it and then it works, I don't mind if it dies under your BIOS test. I don't want to use the working card I have in use at the moment, (since you think something might break.) Also I have the 9500M GS that doesn't like the current 5920 SBIOS, which I could try with.

 

Neither. The problem is that the code for responding to MXM interrupts, including the one for retrieving the MXM structure, overlaps the MXM structure itself in this BIOS (presumably due to confusion around what the "length" field of the MXM structure means - in v2.x and up, it is the size of the MXM structure in bytes, excluding the header, wheile in MXM v1.x, it was including the header). To overcome this, I made the MXM structure shorter (0x29 bytes data + 8 bytes header), but to achieve this, I had to set the new final byte of the shortened MXM structure (the checksum byte) so that the bytewise sum of the entire structure is evenly divisible by 256 (i.e. the checksum of the whole structure is 0). However, it is possible that the checksum byte itself was also intended to be code, so overwriting it with an actual checksum may have made the MXM interrupt handler nonfunctional.

You do NOT need to do a VBIOS flash, but an SBIOS one. Basically, for a Phoenix BIOS, there is a special "Phoenix crisis disk utility" to create a floppy disk that can be used for recovery. (You need an USB floppy driver for this to work - AFAIK no flash disk or CD/DVD support exists, only floppy.) You can find the necessary tools @ Phoenix Bios Recovery - The Elder Geek on Windows

Also, could you please dump the first 1MB of your system memory using a Linux live CD (with the help of /dev/mem) and post the result here? This would allow me to figure out the exact beginning of the MXM interrupt handling code, and adjust the MXM structure to prevent overwriting any machine code.

I'm absolutely sure it won't damage the MXM card - but it may make the laptop unbootable, if I got the end offset of the MXM structure wrong. This is when you may need a crisis recovery disk. (It won't physically damage the motherboard either.)

 

@ .NetRolller 3D

Yeah I know it's SBIOS flash. I'm stuck at how to make the USB stick ready to go. I saw the 'Elder Geek' link first when I went looking. Since I had no registration with that site (I can't download, so) I continued searching. I copied instructions for SBIOS flashing from four separate sites. They all concurred. Plus I downloaded the software.

[I have a USB drive and the files. Others say they struggled to get USB going and had to resort to floppy drive.]
My question was:
Is the USB to be formatted to a bootable drive 'like' when you make a VBIOS flash drive, or;
Do I just dump the Phoenix Crisis files (+BIOS file) onto a USB drive, then it's good to go?


Quote .NetRolller,
" Also, could you please dump the first 1MB of your system memory using a Linux live CD."
I've never seen this before. Please be patient while I sort it.

 

AFAIK there should be an executable in the file you downloaded which can generate a crisis disk image from a provided WPH file, that you can then write to a floppy disk with rawrite (or perhaps it writes directly to a floppy disk). You can NOT use anything other than a floppy AFAIK when dealing with PhoenixBIOS - the boot block requires a specially formatted floppy disk for recovery. (USB drives and CDs are AFAIK for InsydeH2O only, perhaps also for AMI or Award.)

About the 1MB dump: You boot up from a live CD, then do "dd if=/dev/mem of=~/memory.dmp bs=1024 count=1024" as root. The dump file will be in ~/memory.dmp.

 

@ NetRoller 3D, I got message, dd opening '/dev/mem': Permission denied. Am going to keep trying/looking.