07-26-2013, 09:45 AM
On to my other project...(this one should be much more doable)
My room mate has a Fujitsu Lifebook T4220 (we call it a wannabe transformer, for being a laptop-come-tablet). Presently he is running a slightly lower end Core 2 Duo, and i purchased a T9550 to go in it and get him a little more power and better L3 cache. Ignoring all the complaints from Phoenix bios Editor, I've been able to pull the UPDATE.ROM which contains the CPU microcode. I'm working with a hex editor to implant the most current code in to support the T9550. Now, this -should- work just fine, as the specific chipset (Intel G35 I think?) is used in several other brands that support the processor, despite the chipset itself being limited to 800mhz bus with a 1066mhz bus processor...given that the rate is about 50/50, I firmly believe that the issue is more on whether or not a given manufacturer cared enough to update the bios to support said processor.
I am again using Intel's .dat package of microcodes for Linux usage (apparently they don't think I'm smart enough to hex edit this stuff in! Well poo on them!). I am trying to identify exactly WHICH microcode string I need, and worse comes to worst, I'll use my mother's Dell Vostro 1520 to do it (needed a new motherboard, and the replacement should be here in the next few days...it -does- support the T9550 even with the bus differences, although I haven't looked to see if they are using the same chipset, as a fried motherboard makes it hard to boot up windows and look with AIDA64). I'm looking at an id of 000006FBh for the Lifebook's present Core2...which my intel pack has 7 different files for. Does anyone know how to break down the file codes, or have suggestions for where to look?
Examples:
cpu000006fb_plat00000004_ver000000bc_date20101003.bin
cpu000006fb_plat00000008_ver000000bb_date20101003.bin
It's this plat and ver codes that are [censored] me up in these projects, as I cannot differentiate what they mean offhand...or should i just try to copy/paste all 7 codes in? Does anyone have suggestions on packing multiple code strings into a .rom file for packing into a bios? The UPDATE.ROM is straightforward enough but only appears to contain a single microcode info patch...can I just copy/paste all 7 codes into a single .rom, and have that work, or will I have to alter headers/footers? I mean, worst comes to worst, I'll just try it and see...just would prefer to have a little more info to work with so I don't brick his system...though the way the Phoenix editor seems to work, if it doesn't see anything, it just won't load any of it...
Still...advice and assistance would be lovely if anyone wishes to help. Either way, i'll update my progress as I go...
So after looking at the .ROM file, yes, they are just posted in sequence (I found, and subsequently updated, the information for 000006FA chipset via header, after pasting the first 000006FB chipset information in and watching it completely overwrite everything that had been in those locations prior to the paste...). I just pasted in all 7 microcode sets for the 000006FB identifier...saved the new .rom, and inserted it into Phoenix bios editor...and viola! So...first step down, as it seems to have liked my microcode building, and compiled a new BIOS...now, provided the T9550 Core2 is still a 000006FB and not 000006FD (and I'm half tempted to paste those in 2, just to be certain! What could go wrong, right?) this should work. I might still add in the 000006FD codes just to make sure, before flashing it over, and seeing if it still works...since if that step goes fine with the CURRENT processor, that's a good step...especially since the bloody thing has to be practically completely dismantled to install the new processor.
At least this is proving easier than the other project. When it's all said and done, I'll post up the bios file a la mod, and y'all what have one of these things can have an update to play with...I suppose the next step would be updating the SLIC for S&G, but not really certain I -need- to, as he's already running Windows 7...
...anyone wanna suggest the best flasher and recovery setup for what I'm scheming? I'm in (to me) completely new territory doing all of this, but [censored] this is fun to learn!
Ok, just to keep entertaining everyone who's laughing at my silliness with updates...
Apparently adding in the 12k to add in the 000006FD codes made the bios image too large...so, this shall take some tweaking. I think first I'll worry about if it works with the 000006FB codes as done so far, and then remove the 000006FA code to make room for the 000006FD codes one at a time...or just remove the first half of them, OR just put the 000006FD codes and dump the FA and FB codes should the D's be needed, as I just cannot see him downgrading with how much work this can come to be...nevermind that if he goes backwards, his basic bios as downloadable from Fujitsu would work just fine...live and learn! I still think it's silly that 12k was more than the bios image has left to work with lol
My room mate has a Fujitsu Lifebook T4220 (we call it a wannabe transformer, for being a laptop-come-tablet). Presently he is running a slightly lower end Core 2 Duo, and i purchased a T9550 to go in it and get him a little more power and better L3 cache. Ignoring all the complaints from Phoenix bios Editor, I've been able to pull the UPDATE.ROM which contains the CPU microcode. I'm working with a hex editor to implant the most current code in to support the T9550. Now, this -should- work just fine, as the specific chipset (Intel G35 I think?) is used in several other brands that support the processor, despite the chipset itself being limited to 800mhz bus with a 1066mhz bus processor...given that the rate is about 50/50, I firmly believe that the issue is more on whether or not a given manufacturer cared enough to update the bios to support said processor.
I am again using Intel's .dat package of microcodes for Linux usage (apparently they don't think I'm smart enough to hex edit this stuff in! Well poo on them!). I am trying to identify exactly WHICH microcode string I need, and worse comes to worst, I'll use my mother's Dell Vostro 1520 to do it (needed a new motherboard, and the replacement should be here in the next few days...it -does- support the T9550 even with the bus differences, although I haven't looked to see if they are using the same chipset, as a fried motherboard makes it hard to boot up windows and look with AIDA64). I'm looking at an id of 000006FBh for the Lifebook's present Core2...which my intel pack has 7 different files for. Does anyone know how to break down the file codes, or have suggestions for where to look?
Examples:
cpu000006fb_plat00000004_ver000000bc_date20101003.bin
cpu000006fb_plat00000008_ver000000bb_date20101003.bin
It's this plat and ver codes that are [censored] me up in these projects, as I cannot differentiate what they mean offhand...or should i just try to copy/paste all 7 codes in? Does anyone have suggestions on packing multiple code strings into a .rom file for packing into a bios? The UPDATE.ROM is straightforward enough but only appears to contain a single microcode info patch...can I just copy/paste all 7 codes into a single .rom, and have that work, or will I have to alter headers/footers? I mean, worst comes to worst, I'll just try it and see...just would prefer to have a little more info to work with so I don't brick his system...though the way the Phoenix editor seems to work, if it doesn't see anything, it just won't load any of it...
Still...advice and assistance would be lovely if anyone wishes to help. Either way, i'll update my progress as I go...
So after looking at the .ROM file, yes, they are just posted in sequence (I found, and subsequently updated, the information for 000006FA chipset via header, after pasting the first 000006FB chipset information in and watching it completely overwrite everything that had been in those locations prior to the paste...). I just pasted in all 7 microcode sets for the 000006FB identifier...saved the new .rom, and inserted it into Phoenix bios editor...and viola! So...first step down, as it seems to have liked my microcode building, and compiled a new BIOS...now, provided the T9550 Core2 is still a 000006FB and not 000006FD (and I'm half tempted to paste those in 2, just to be certain! What could go wrong, right?) this should work. I might still add in the 000006FD codes just to make sure, before flashing it over, and seeing if it still works...since if that step goes fine with the CURRENT processor, that's a good step...especially since the bloody thing has to be practically completely dismantled to install the new processor.
At least this is proving easier than the other project. When it's all said and done, I'll post up the bios file a la mod, and y'all what have one of these things can have an update to play with...I suppose the next step would be updating the SLIC for S&G, but not really certain I -need- to, as he's already running Windows 7...
...anyone wanna suggest the best flasher and recovery setup for what I'm scheming? I'm in (to me) completely new territory doing all of this, but [censored] this is fun to learn!
Ok, just to keep entertaining everyone who's laughing at my silliness with updates...
Apparently adding in the 12k to add in the 000006FD codes made the bios image too large...so, this shall take some tweaking. I think first I'll worry about if it works with the 000006FB codes as done so far, and then remove the 000006FA code to make room for the 000006FD codes one at a time...or just remove the first half of them, OR just put the 000006FD codes and dump the FA and FB codes should the D's be needed, as I just cannot see him downgrading with how much work this can come to be...nevermind that if he goes backwards, his basic bios as downloadable from Fujitsu would work just fine...live and learn! I still think it's silly that 12k was more than the bios image has left to work with lol