(UEFI) Dell XPS 15z L511z modded BIOS - and HOWTO

[kasar]

thanks timewalker

I will asemble the last conections of the programer and will test the conections with a multimeter.

I hope things goes fine ^^

---

well, as for the zenner diode, I didnt had any other options arround home.

all the electronics stores are far from home, I also had to get the printer port from another old project I wasnt using at all.

I checked the voltage with a multimeter and it output 3.40v

will it bring issues?

as for the cable, well, it would be nice to be able the printer port from my old laptop to the xps 15 internal programer port I added, and with the current size, they have to get really close :o

[TimeWalker]

Nah, should be fine. Besides while flashing the gain can go higher than 3.3, so having an extra is a good option.
I have an electronic store 2 minutes away just around the corner but man the components are pricey there .. I would rather drive to another part of the city and probably save more money (including gas) in the end ..

The length of the cable worries me though.. you can read properly but you may have issued while writing.. at least that what I had when I last reprogrammed my DIR-300..

Also please note that you can't probably directly dump, edit and flash the contents of the nand .. at least this is not the case on UEFI desktop boards. Service Tag, Express service code and a bunch of other DMI entries will be wiped, because these are just stored as variables which are pulled from somewhere to be initialized in the firmware. This may not necessarily be the case with our machines, but it's always good to take caution ...

[kasar]

ok, I completed the programer cable






well, like you said, the cable lenght plus the internal wiring I did yesterday on the laptop may be an issue, but well, lets cross fingers and see what we can do

mmm, what do you mean with I cant dump,edit, and flash?

I though by doing this we backup the whole content of the chip, including protected areas.

so, in case I dump, unlock the descriptor, and flash, should it work or it will lead problems?

anyway I think next step is test the cable by doing several dumps and comparing them, if they are identical then dumping method + cable + laptop internal wiring = good, if not, well, something went wrong ^^

I supose nothing can goes wrong by just dumping the chip content, right?

[TimeWalker]

Make sure you eliminate all the power sources of the mobo and that should do it.

Sent from my LG-P500 using Tapatalk 2

[kasar]

holly sh1t.

all the [censored] afternon to get the thing to work :o

well, I managed to get it working, or at least partially working.

I made several attempts, tried using a laptop printer port, it failed.


to make it work I finally ended removing all the resistors, conecting to a PC parallel port, and conecting it with the xps machine powered on (yeah, I know it isnt good, but for some reason it just worked like this way.)


mmm, I attached the multimeter to the vcc and gnd pins while the machine was working and it showed 5 volts.

so I supose I will also have to remove the zenner diode to give the thing 5v like the laptop does, then I will try with the xps off, battery and ac power off too.

oh, when I said it partially worked, I said that because I finally made it recognize the chip, however made 3 different backups, and all of them were different D:

this means data corruption probably, so I will not flash anything yet until I get like 5 consecutive dumps with the same content.

here is the backups I made using the programer cable.

http://www.mediafire.com/?5i3mgju3es3gt6a

I will try to remove the diode and try again with the machine off ^^

wish me luck

[kasar]

ok, the removing the diode didnt helped.

well, it allowed to let me notice about something.
when I plugged the programer into the laptop, I notice the internal usb bluettoh led I installed inside was on, very weak anyway, this means the power from the programer were splited arround the whole laptop, and this isnt good.

this may need another mod to the laptop, cut the vcc from the bios chip, and wire it by a a switch.

since I added a multi swich with 4 different switches and I just used 1 for the cmos reset, I think I can use the second one to connect the vcc to the whole motherboard or not by using that switch.

so when I connect the flasher, I use the switch and avoid the power to split over the whole motherboard.

anyway I'm getting a little tired, maybe another day, not feeling like taking apart again the whole laptop and soldering.

maybe another day.

I also noticed, when the laptop were turned on, it was the only way of make the software read from the chip, it detected the info, but for some reason it apears to disconect after some time. I mean the chip become undetectable, and any current dump start getting lot of FF FF FF from the part it shutsdown.

this means something is unstable with this setup.

omg, this stuff can be pretty annoying sometimes >.<

[TimeWalker]

Thats why chips require removal, because the input power from the flasher gets redistributed to the circuit of the mobo. Maybe try lifting the Vcc leg? Also when laptop is on BIOS and all the reegions are shadowed, since you can't switch the chip's state (not to mentioned wp) the dumps will never be proper. Also reduce the lenght of the headers cabling if possible just to verify you are not getting bad dumps due to that. Good luck!

Sent from my LG-P500 using Tapatalk 2

[jkbuha]

Good luck kasar - respect for trying.

We've come a long way in 6 weeks since we first unlocked the primary stages of the bios - I'm sure we'll find our way around this as well.

[TimeWalker]

Actually @kasar

1. Your FD dump l502x_descriptor_dump.rom (that you've shared earlier) completely matches the FD part of the dumped firmware you got from the flasher (starts at 00000000h end at 00000FFFh) in all 3 dumps you took.
2. After FD region there is a 36FFFEh byte (220Kb) long sector of data, this is GbE region presumably .. but I have nothing to compare it to.. I had the intel server board region but can't find it any longer.
3. ME region starts at 00038000h goes until 17FFFFh presumably (1343487 bytes = 1.28Mb, while the 7.1.52.1176 ME firmware is 1335296 bytes .. so we have like 8 Kb of free space ... which won't fit v8 ME firmware = 1560576 bytes) this region is corrupted with random data (FF) because of the way you dumped. GbE will probably be corrupted the same way, hence these two regions differ in all 3 dumps because R/W security is enabled and those regions can't be dumped properly.
4. BIOS region starts right after ME at 00180000h and ends exactly at the end of the dump at 003FFFFFh (2621440 bytes) - I don't have the dump of your on-chip BIOS region to compare against.

[kasar]

mmm, so aparently i'm going to the right way.

that chip seems to have all regions we need, so as soon as I have the thing working to have a propper backup it will be good

well, I will try adding a second switch, it will control if the vcc leg is controlled to the mobo or not, so I will have control if the chip is distrubuding energy to the motherboard.

also this switch will control if the laptop can boot or not,may be usefull


will also try to remove unnecesary size of the cable or internal part.

anyway, since it dumped some good parts, doesnt that means the cable size is ok?

note, I notice the date and most adjustments go away when I activate the battery swtich, however the password still there, I supose it is stored somewhere inside the chip? this also should allow us to reset the computrace setting by accesing the whole chip, right?