[REQUEST] Asus ROG Strix GL702ZC BIOS Unlock

[Odin Vex]

@Roland83, are you saying you've tried a 3700X, and it failed? If the init routines are similar enough, it should work without any update. This would let us know that the init routines changed, which'd mean we'd definitely need ASUS to step in.

[Roland83]

@Roland83, are you saying you've tried a 3700X, and it failed? If the init routines are similar enough, it should work without any update. This would let us know that the init routines changed, which'd mean we'd definitely need ASUS to step in.

Yes, i have tried my 5 3600 from my desktop rig in the gl702zc. A few months ago with my unmodded 306 bios. Laptop turns on, but nothing more. A few weeks ago, one member from the notebook review forum which i sent my modded bios did change the microcode for the ryzen 3xxx series in my bios. But its the same behavior like the unmodded bios. Agesa 1.0.0.7 which is included in the 306 bios doesn´t support matisse. You can use the programm "smu checker". Just drag your bios over the *.exe and it will tell you in a new window which cpu series are supported. In the gl702zc bios , only the summit ridge cpu´s are supported.
The only way is that asus must release a new bios because they have the tools to make a new bios. No one of the bios modders have experience to change the modules for the agesa code to make new cpu´s work in this laptop. That´s the point.
So everyone must penetrate the asus suport to make new bios. It will be a hard fight, but it could be a little chance that we can win.

[Odin Vex]

@Roland83
Microcode would never bring support, that is CPU patch code alone, not BIOS. I wish people would stop pushing those rumors out on boards. You can push microcode through Windows (done automatically, usually), Linux (grub, for example), (it virtually overlays CPU microcode, best done during boot phase, lasts only one boot because microcode is read-only in field for most people, not flashable without custom software kept under lock and key and undead solders from Germany during WW2 deep inside some temple of...eh *darn censorship*). Anyway...the 3600, being of the same TDP and socket type, would highly suggest the init routines for Zen and Zen 2 are indeed different enough that either every Zen setup would need a BIOS update or at least this board. But you're not accounting for core count, which might be an issue. The BIOS on this board runs custom P-states and power regulation, specific to 8C16T (8 cores, 16 threads). Someone should try a 3700X to eliminate (or confirm). As for the AGESA code, you can drop in new modules and just SPI-flash the final result, but good luck with UEFI and such, I don't know how it'd respond. SMU Checker only looks for microcode, useless to determine which CPUs are supported, only useful to determine which CPUs *were* supported on BIOS build. For instance, if the 1700 and it's family/stepping were changed and rebranded but was the same family (Zen), it'd work, but the BIOS wouldn't recognize it. Init has *nothing* to do with microcode (which is only used to fix bugs and to ensure a 'known working configuration of microcode for the BIOS during boot'). I do agree, we could try to bug ASUS into doing something for consumers instead of their own pockets, for once. I despise how quickly they out-mode things artificially. I had to go around providing patched DLLs for their ASUS Aura software because the morons kept assuming certain API calls always returned certain values and they'd crash. I hosted them on my Github before finally leaving Lamedows behind for Linux full-time. What we'd need is either ASUS to provide a BIOS *or* someone to find a BIOS that runs on a similar enough board that supports Zen 2. Even if the BIOS doesn't work when tried, the modules could be extracted and pushed in after adjusting symbols to compensate as much as possible. No guarantees of functionalities, but who knows. (Or someone to try the 3700X on stock in hopes the init would work for it too.)

Edit: I should mention that we're not running server motherboards, which some builders artificially limit, such as Dell Poweredge servers. They do that to prevent upgrading outside of their licensing fees and bull. Those *do* whitelist by CPUID info, not by microcode. Just wanted to point that out for those possibly wondering.

Edit: I'd also like to see ASUS fix the BIOS TPM issue. On Linux, it interprets the ACPI region correctly and by TCG CRB 2.0 TPM spec, but the BIOS is incorrect. [ACPI region does not cover the entire command/response buffer. [mem 0xdcdae000-0xdcdaefff flags 0x200] vs dcdae000 4000].

[Roland83]

I put my 5 3600 in it and with this cpu the laptop is turning on, but nothing more. Only power led is on and screen is off.
Should we try to make bios crossflash? Maybe from an newer rog or tuf gaming laptop? There is zen and zen+ support in it an new agesa code. When they use 16mb rom chips on those laptops like the gl702zc the bios image should work. I could test this only to see if it will running or not with another bios. What do you think about it?

[Odin Vex]

A BIOS cross-flash would be useless, so much ACPI/DSDT info&addresses and so much more would just fail all over the place. You'd need to extract the AGESA init module(s) and possibly more, then merge into GL702ZC BIOS and begin working on making it work, changing things possibly. The only kind of cross-development I'd suggest even attempting would be an open BIOS, and even that'd be tediously unproductive for quite a long time.

[Roland83]

Yeah, i think you´re right about a cross flash, it would not work. I think that´s nearly impossible to get zen+ to work in this machine. I have absolutely no experience in modding/editing those things. Very sad that there are no informations which modules had to be edited.

[Odin Vex]

It isn't that difficult, there have been BIOSes out there that have been updated by manufacturers to add later Zens, they could be broken down, it's just that the few who could (myself) are far too busy or unable to test/replace due to budget if something goes wrong.

[Roland83]

Okay, you wrote it could be possible. Maybe we can work together to solve this? What software do i need and can you give me some instructions, only if you have time for this? I want to learn this a bit how to modify an uefi bios.

[draklrd]

I can also test as well i have gotten very good at re-flashing my bios back with a spi programmer and at this point if i can't get a newer cpu on this laptop its getting very close to retirement anyway.

[Roland83]

I think that i got all the tools i need. I played a bit with uefi tool to get a closer look into the modules. Now we need to find the necessary modules that had to be edited. Uefu tool, hxd and maybe ubu should be all the tools to work with. If someone have any informations which modules are important, reply in this thread.