[Request] Dell Vostro 260 - Add Ivy Bridge support

BIOS/UEFI Modding BIOS Modding Guides and Problems
101

What updater from windows do you use, when trying mod BIOS? Please link me to it. I think you’re correct, the stock exe is not packed via AMI tools like some of the others, and the stock exe contains an HDR file without any normal flash/update tools.
I checked the 3010 BIOS ab it’s is still Aptio IV based

Thanks for report back with several test methods. Seems some other GOP and or vBIOS is needed then, because it should be booting perfectly with SB using onboard or PCIE, although onboard is not much of a concern it still should be working.
Sounds like Ivy is booting properly as always since the early mods, but no display. Does the OptiPlex with Ivy work with onboard GFX and PCIE, and what about Sandy, all OK too?
If yes, I will use it as a display source, and transfer all settings from that ones vBIOS to a copy in the 260 BIOS too (including whatever GOPs are there for the onboards).
I checked GOP/vBIOS info with UBU and I see it’s using nearly the same GOP pair I put in this latest test BIOS, and exact same vBIOS but with different settings, so using it’s settings and vBIOS may help (if other modules are not what’s holding it back too)

UBU-CompareNew260-3010.png



Yes, it’s very complex, and as you saw mentioned by plutomaniac early on, none of this may ever be possible, because normally this kind of change over (Major ME Version update) must first be done by the manufacturer because of all the other changes to other BIOS modules.
After they did that, then we could update the files we discuss to whatever version we wanted and it would still work.

102

The updater I used is O3010A22.exe

I’m not sure if Ivy is actually booting; I don’t see any USB initialization (keyboard lights or mouse), nor does it boot to OS (by HD indicator lights or network stack coming up). I have a PCI tester card that shows various init codes throughout the boot process, but the board has no PCI slots…

On the Optiplex BIOS, Sandy and Ivy boot with onboard. I didn’t try PCIE (it didn’t even cross my mind! - thanks for mentioning it), but I assume it would be fine. I will verify a bit later today.

I’m game for exhausting all possibilities; for every bin, there will be solder, flux & hot air :slight_smile:

103

OK, that’s the one I have too, it’s just HDR and some other stuff that I can’t tell how it’s flashing or preparing the flash for reboot
BIOS has already been proven to be booting into OS early on by other users, networking into PC as it’s running and all was fine.

Wish you had a clip, I hate to think of popping off and putting back on BIOS every time just for simple tests. If you leave the 260 BIOS in and just use Sandy for a while, aside from Ivy CPU swap test, you can leave BIOS on the board and just flash using AFU

I’ll wait to do anything until I hear back from you about the 3010 PCIE test results.

104


This is decidedly not the case for me, just tried it again. Sandy boots with PCIE (not onboard), Ivy doesn’t - at all. No USB init, no LAN. The proc in use is now an i5-3470 (which is the same I used for the 3010 BIOS). I’m not sure why it’s worked fine for others and not sure how to diagnose why it doesn’t for me, but I can assure you it doesn’t POST with Ivy.

Another strange thing I noticed: this BIOS (the one you most recently uploaded) won’t boot with AHCI! This is an MBR Windows install, not UEFI. If AHCI is on, I get "Hard disk failure" and the system won’t boot from the drive. If I set it to IDE, it will boot.

If I can find one, I may install a socket for the flash chip, but for now I don’t mind using hot air.

I will now swap back for the 3010 8MB BIOS and test onboard+PCIE on both Sandy and Ivy.

EDIT: Both Sandy and Ivy work with both onboard and PCIE on the 3010 8MB BIOS. If we’re unable to come to a resolution with 4MB, resolving issue #3 in post #96 would at least get a nicely functional system; the rest (microcode, modules) can probably be updated manually to the latest as we have the entire 8MB bin, removing the need for updating the BIOS via Dell’s tool. Not the best solution, but the best I can think of!

I looked up gloobox’s posts and found a thread describing the same issues I ran into with the 3010 BIOS and I see you’ve done modifications to it. I will now try those in order.

EDIT2: I have tried M1, no POST (turns on briefly, then off). This is with a Sandy CPU. Are these 3010 BIOS mods or 990?

EDIT3: I believe I answered my own question, they’re 3010 mods. With M2, this seems to have worked. I encountered several lines of "ERROR: Parsing HII string database failed". However, as the first line was green, I figured this was manufacturing mode. The system no longer complained about "Power Button Cable failure" and "Front I/O Cable failure" on every reboot. That said, as soon as I hit ALT+F to exit manufacturing mode, the system would halt again. This time there would be no green line of text, instead a group of 5 and another group of 3 "ERROR: Parsing HII string database failed". I figure the last three are the "press F1 to continue, F2 for setup, F5 for diag" lines. I’m not sure whether there are consequences to leaving it in manufacturing mode, nor what they would be.

MBoth.bin behaves the same as M2.bin

EDIT4: The same behaviour holds with the original 3010 bin, except the relevant strings are loaded and displayed. The system will happily continue booting as long as manufacturing mode is left on; once exited, every reboot requires a press of F1 to continue.

105

Please test with BIOS from post #5, confirmed working to network into in post #10, see if it works like that for you too. That is weird about AHCI thing! Is that BIOS in #99 with that issue?
I will start fresh now, so nothing can get messed up on my end, since I have the old folder on this with lots of edits and random files edited and extracted etc.

Yes, if we can’t get it working, I will update microcode, modules etc and work on figuring out the issues, specifically #3

What is M1, M2, Mboth earlier BIOS I posted I assume, those looks like names I would use, but none I noticed that I did use in BIOS made in my folder on this.
Ohh, I see your mentioned link to other thread now, and looking at that, he never reported back on results, so I can’t confirm what’s going on there (let me check my folder on this and see if I can figure that out)
It’s good to have a report back on this from someone finally, thanks Now I can look into further fixing the issue based on your report above.
* Edit, checking my edits here, Funny to see Mboth boots/works, but M1 does not boot? MBoth contains exact edit as M1 plus additional edit from M2, logically MBoth should fails as it contains the M1 edit… so it’s

M1-M2-Mboth.png


Will look this over more in assembly now * Checked, I can make other changes instead of the ones I made, and different routines could be edited as well, all this is in one same module (DellErrorLogConfig)

I will start fresh right now, with 260 BIOS again, putting in ME8, vBIOS and GOP from the 3010, microcodes. If that fails, then I think we have to move forward with 3010 BIOS instead.


* Edit @turbo9000 - Here is new test, this has updated CPU microcodes (always), SNB/Ivy vBIOS from 3010 inserted into CMSCORE, old sandy vBIOS removed (Desktop & mobile), EFI GOP from 3010 for IVY ONLY (no room for both)
So, when using Sandy, this one may only work with PCIE card. Additionally included is a second mod, where I left the Sandy mobile orom in place, all rest same as above (Could not fit Sandy GOP in here either)

NewTry.png


http://s000.tinyupload.com/index.php?fil…897728969197525

106

Re. AHCI: Yes, it was the BIOS from #99 with that issue.

Results:
M1.bin retry (in case first attempt was a bad resolder): Ivy endless powercycle (on for 2s, off for 3s) - same result. Did not test Sandy.
BIOS from post #5: please reupload, no longer there
M8IGOPIVB.bin: Sandy PCIE boots, AHCI works, Ivy no boot onboard or PCIE (powers on and sits there, no USB init, no LAN, no HD)
go2.bin: Sandy onboard & PCIE boots, AHCI works, Ivy no boot onboard or PCIE (powers on and sits there, no USB init, no LAN, no HD)

I’d like to test that BIOS from #5… It’s strange that Ivy works on that one for others, I’d really like to try it for myself and see if it works, as it seems to me we’re getting stuck before even reaching the vBIOS stage with these. If it does work, it’d make a good baseline.

107

That’s not a good sign! Here is BIOS from #5 again, also updated link on that post.
http://s000.tinyupload.com/index.php?fil…896483348377267

I think the issue may be “Made from dump” vs “Made from stock BIOS” and how it’s flashed in. Everyone else here is flashing, and this makes a huge difference when flashing in stock edited vs programming in stock edited.
These are made from stock BIOS, so should really only be flashed in, on top of (Full) BIOS that was not programmed in from a stock BIOS.
Stock BIOS is missing NVRAM volumes and tons of entries inside NVRAM, when you program this in vs flashing you loose everything in NVRAM that was there originally (This is tons of info and settings).
So if these also fail for you, then you need to program back your original BIOS that was dumped first from this system, it needs to have a full non-stock NVRAM volume (Or two).
If you are unsure what this means vs a stock one when looking at NVRAM contents, upload what you think is your very first dump from this system and I will let you know if it’s really a dump from non-stock programmed BIOS or actual full dump from the board that’s only had BIOS flashed in before (it’s a huge difference)
Flashing in now vs dump wont help, unless you first have a complete/full (non stock/empty) NVRAM filled BIOS on the board before flashing.

108

I see what you mean! I’ll have to read through the thread to figure out how to flash from DOS. My original dump (immediately after initial removal without any flash attempts) is in post #91.

I will test the #5 BIOS and see what happens; if it doesn’t work, I’ll reflash my stock, flash #5 from DOS, and test - if Ivy works, I’ll repeat tests with all the other 4MB BIOSes by flashing in DOS.

109

In the package above in text I described how to flash in the modifications.txt, but yes you may also need to read through the thread because I cannot remember which method ended up working for users earlier.
Thanks, I had not used your dump for any of this, I didn’t think about the stock vs dumped actual (NVRAM stuff) until I mentioned above after realizing you had issues with BIOS files that others didn’t have issues with.
And yes, I checked your backup, it has dual NVRAM volumes, fully populated as expected. All BIOS I have sent for you here have only one empty stock NVRAM volume, so this is likely the main and huge issue/difference between your testing and others previously who flashed instead of programmed.

Program back in your dump from 91, then retest any/all BIOS via flashing not programming. However, ME flashing will possibly be an issue, especially changing versions like we did . I may need to remake some BIOS for you using your dump instead.
Or you can use this guide to edit your BIOS from 91 before you program it back in, to unlock FD fully, then you can use FPT from ME System Tools V7 package to FPT flash ME8 (ONLY).
Then carry on with rest of BIOS testing by flashing using AFU. Here is command to flash ME region only using FPT >> FPTw.exe -me -f source.bin (For source, you can use any mod BIOS here with ME8, or extract ME region.bin from any mod BIOS here and use that as source)

Don’t use FPT/method to flash BIOS or BIOS region, unless it’s created from your dump, otherwise you’ll have same stock NVRAM issue as with programmer (FPT programs in exactly what you give it, AFU does not and leaves your NVRAM in place).
If you do use FPT to flash in ME8, only use one of the mod BIOS here as source for the ME region, since it contains all the proper settings transferred in already.

110

OK, sounds good. The steps I took:

1. Modified 260s_stock.bin with the guide mentioned above (FFFF’d the 3 regions to unlock FD)
2. Flashed to chip, soldered, booted, loaded defaults
3. Extracted ME.bin from M8IGOPIVB.bin
4. Loaded to USB along with ME7 tools
5. “fpt -me -f ME.bin” - was successful, verified MEv8 with MEINFO after poweroff, sit for 30s, poweron
6. Used AFUDOS ES 3.06.01 with "afudos M8IGOPIVB.bin /p /b /k"

This resulted in a Sandy boot via onboard and PCIE, but dead on Ivy in both cases.

I will try the same steps but with the BIOS in #5 next. I assume it’s OK to flash over the existing one with AFUDOS?

EDIT: Tried this; Sandy boots PCIE, Ivy does not (same as all other attempts).
Tried MS-A10_MEvBIvy2.bin as well out of curiosity, same behaviour - Sandy boots PCIE, Ivy nothing. Keep in mind I did not try onboard for either attempts, though I don’t think it’d make a difference.

111

Onboard may be required, in the sky/kaby/coffee mods they had to create PCIE mod to fix PCIE function, and only in latest BIOS I’ve created for you past few days has correct GOP modules for Ivy, earlier ones do not have GOP changed only vBIOS (I think)

What is this BIOS M8IGOPIVB << From early on, or one I sent you recently? Either way, once you flash this, you loose all that NVRAM stuff I was discussing when you used all those flags with AFU.
I am not 100% sure, you’ll have to test by flashing, dumping, compare with 260s_stock.bin in UEFITool at first volume (NVRAM, expand), but I think only one you want to use is /p - /b is not needed, and /k may overwrite NVRAM (Unsure), but I’m also unsure if /p does that as well.
Before doing anything to try and correct this or redo things, dump current BIOS via FPT and look at the NVRAM volume, is it populated with lots of stuff like 260s_stock.bin, or empty with only 8-12 items like you see in a stock download update BIOS?

Good thing is now you have unlocked FD, so you can re-write 3010 BIOS any time too when you are ready, just be sure any time you flash in a full BIOS (not region) that it also has unlocked FD
If you always make sure FD is unlocked in the file you are about to flash, you’ll never have to pop the chip off again.

Maybe @Flagers can also help to speculate why it worked for him but not for you? He could also clarify which method he was using to flash, that might make a difference too if you’re using different method now, like other AFU version etc

112

I would like to use a rx 550 on my dell 620 but for that i would need a uefi bios.
i7 2600 + 0Gdg8y
Sorry for my english (google translator)

113

@Lost_N_BIOS
When flashed with the 3010(change the 8MB bios chip),it could support ivb and uefi boot mode.But the machine just got 2 sata interfaces available.

For convenience,I think 4MB bios modded is better,and it could flash with software but NOT desolder.

114

@gloobox - hide the one you don’t need, problem solved

115

@Lost_N_BIOS
Thanks!
I’d got the mobo,and I had programmer.And I 'll tried and give you some pic.And I will going to do more functional test.

116

@gloobox - Yes, once you get it going, send me pic of the BIOS and circle the setting you need hid, I can hide for you if you aren’t sure how

117

@tiagocx1337
You could mod your vga bios,and disable the uefi boot in your vga bios.I think modding the bios of your mobo is harder than vga bios.

118

@Lost_N_BIOS
My bad!
I launched it,and it showed “3010” in the left upper corner on the screen interface of the bios .

119

@Lost_N_BIOS
I found the mobo got was OptiPlex 390 from the tag in mobo,but it’s bios chip was changed with the 8MB bios chip.
On the other hand,I could not update any bios version in the official website neither in DOS nor windows.

I met several mobos of dell had the problem,such as 3010.3020.7010,they had a similar place that the bios chip was desoldered and flashed by bios file dumped on other mobo(same model).
I think something in the original bios must have gone?or dell must exam some ID and SN for match?

120

@gloobox - Dump BIOS for each system, can be FPT or programmer, then look in last volume for FID (Flash ID), make this match the BIOS you want to flash in, this often what allows or blocks flash by applications.
Sometimes though, this info needs edited in more than one place in BIOS.