Jump to content

thop1544

Registered User
  • Posts

    39
  • Joined

  • Last visited

  • Days Won

    1

Posts posted by thop1544

  1. @Suntor

    This has requested many times, but the consensus seems to be that, on account of no new features of note having apparently been introduced in A12 or A13, the answer is no.

    @svl7

    Sorry to ask twice, but I didn't find an answer in any of the pages of this thread before or after last September (when I last mentioned it): is there a way to expose and control the AC adapter’s authentication handshake? Would this be possible using either the current A11 modified BIOS, or maybe with some (guided) modifications to it? I would do this myself if I knew what to look for, and would of course be able to test the changes to whatever BIOS was necessary.

    I'm on my second 240W AC adapter from Dell, and it recently quit working just like the last one did. I'm a few months out of warranty at this point though, which means having to pay about $65 USD to buy a third 240W, which is just as liable stop functioning properly in the same manner.

    I’ve used both Flextronics and Delta adapters, with two different motherboards, but still inevitably have the same issue. The BIOS no longer recognize it as a 240W adapter, and as such refuses to either charge the battery or run the CPU and GPU at full clocks.

    The problem seems to lie in a failure of a parasitically powered ID chip in the adapter itself (which is delivered by the middle pin visible when looking down the plug).

    I don’t know if the authentication is too much of a hardware-level thing unfortunately, but there is some information about how Dell’s ID chip works with these adapters here:

    330 Watt power supply for Alienware M17x » imsolidstate

    330W power supply for M17x update » imsolidstate

    http://forum.techinferno.com/alienware-m17x-aw-17/1948-240w-psu-id-chip-330w-psu-2.html

    http://web.archive.org/web/20130102101025/http://www.laptops-battery.co.uk/blog/dell-ac-power-adapter-type-cannot-be-determined-solution/

    • Thumbs Up 1
  2. Thanks a lot for this, which is finally a good fix for the newer Enduro/UEFI/etc. enabled 7970m cards. FWIW though, at least on the xxx.022.xxx vBIOS the checksum bits start somewhat later in the file (0xFFFF) than where the OP's screenshots show and are formatted a little differently.

    Does anyone know what algo ATIWinFlash uses for generating the CRC though? I would rather not have to boot a Windows box each time I need to check that, just for the sake of a couple of bytes.

  3. As far as I know, the not charging problem results from irregularitieswith a parasitically powered microchip in the AC adapter. I've hadtrouble with it too, but the best solution I've 'found' to work isplugging in the adapter, waiting for the blue ring to appear, thenunplugging and plugging back in the cord to the wall, fast enough tokeep the blue light on. Then connect it to the laptop, and it seemsto work.

    Dellwent to pretty exorbitant lengths to ensure that only an exact (Dell) adapter can be used for each laptop.

  4. if the card came with a .017 vBIOS then you should just use the patcher against that older version; it doesn't work with anything newer. Though, if you're getting thermal shutdown at 0.950V without heightened clocks then you've likely got bigger issues. I'd suggest checking your paste, fans and heatsinks first, especially for dust and warping.

  5. You'll have to grab the proprietary drivers, which is probably most easily done from their site. As things currently stand, "Nvidia's" FOSS drivers will never get close to what the AMD and Intel stacks offer, especially with acceleration: re-clocking is broken under nouveau.

  6. Well running the latest Arch with the testing repos enabled, audio works properly (under ALSA) by simply plugging into the secondary headphones jack. I've yet to test this with pulseaudio.

    What's more troubling is how the 7970m has no s3tc/dxtn support in the radeonsi FOSS driver, and indeed no hardware acceleration at all unless using Glamor. I guess that's supposed to be merged with kernel 3.10 but be dependent on the next Mesa, which isn't going to be ready for about 5 months. I've opened some threads across the Arch and openSUSE forums detailing these issues, and I can point anyone there who is interested. There are some upstream bug reports open as well, by others.

  7. But how? I've had audio jack detection issues across multiple kernels / distros since the time of Ubuntu 12.04 LTS (which is just the first one I tried with this M17x R4).

    I'm currently dual-booting W8 with Arch running kernel 3.9.1 (from the testing repos) and KDE, to answer the OP, but the issues persist under Linux for me.

  8. Crap. It doesn't look like the new patcher is coming at all. Anyone fancy swapping a dell 021 bios 7970m for a dell 017 bios 7970m?

    Only if that swap involves me getting a .017 7970m!

    The patch could come eventually, but there have been mentions of somewhat confounding voltage irregularities in the new vBIOS packages. At least times were good before Dell/AMD started breaking everything :upset:

  9. Were you talking to me or answering someone else?

    You. I'm trying to get more info before I resign to having Dell refund my M17x R4. Still, unless they're willing to swap for 680m parts (which would ruin FOSS Linux driver chances) then I think svl7 might have had the final word about voltage irregularities/UEFI vBIOS issues plaguing (any) newer 7970m chips.

  10. What was the p/n on the 7970m chip it came off? I'm working through a fifth 7970m replacement from Dell now, but it seems they're out of all but the 113-C42904A1-119 parts what with these bad .022 vBIOSs. I messed with hex editing to look for the differences in some of the older vBIOS mods, but it's a bit beyond my league.

  11. There's unfortunately still not a lot of news on if this will be possible, which is probably at least in part related to some newer 7970m chips having voltage irregularities. I've now gone through four cards ever since Dell replaced my happy .017 vBIOS equipped card with a newer one. I need a vBIOS overclock when using Linux, but as none of the cards I've had since will run these older vBIOS files I guess I'll just keep [ab]using my warranty until something changes. I'm sure upset with Dell that their second highest model from their flagship brand is less capable as my production machine than this $500, 2.5 year old HP though :sour:

  12. Yeah, Nvidia is the way to go for linux users.. AMD's driver support is a little lacking.

    Unless you want to overclock or are using the open source drivers, in which case the situation is completely reversed. Catalyst proprietary has overclocking support unlike Nvidia's binaries, and Nouveau has about no actual acceleration, while 3.8.0's Radeon is showing better performance than Catalyst in some situations.

    Anyway, does someone have a changelog for the new A11 M17x R4 BIOS?

    And sorry, but what exactly does FPT stand for? Something to do with Fortran? Intel Flash Programming Tool? I think the latter might be available under DOS, but will that really be able to safely bypass the firmware validity check that we had to use A05 to get around before?

  13. You can't use the key combo in Linux as far as I know, boot Windows if you need to switch over. The unlocked BIOS shows you the same option the key combo does though, which is switching between PEG and SG for driving video output. PCI will crash it, and at least for me IGFX doesn't get 3d initialized properly.

    Why not just keep it on PEG? Bumblebee is a crazy amount of work if it isn't absolutely necessary (like with the truly hobbled Optimus laptops), and will never be more than a workaround.

  14. I ordered my m17x on the 16th January and I have flashed my 7970m no problem.

    May have been a bad ROM?

    That was my first thought but the vBIOS I flashed was the exact same one as I successfully used with a different 7970m.

    Alienware m17x R4 Linux Cuda Issues:

    conspired together to make Alienware inaccessible to Linux...

    Actually (some) Alienware laptops are among the last to fully support Linux, with the ability to switch exclusively over to the dGPU and avoid this forever broken switchable graphics implementation. Using the BIOS/FN + F7 method should work properly, though I have had issues getting the iGPU working normally when switched to with the unlocked BIOS.

    Lol dang triple post about software issues? You should probably talk to Alienware Tech Support; improper CUDA functioning on Windows involves a platform they're fully expected to support. Besides, Bumblebee is slow and your solution would only work for a Debian derivative, not to mention there isn't any VS support under Linux.

  15. The new M17x R4 laptops are definitely shipping with this new ???.022.???.??? vBIOS on 7970m parts. I just bricked a fresh one from Dell, after trying to flash the modded 940/1350 vBIOS (version .017) which worked perfectly well on an older card of mine. I have a stock vBIOS dump I made (with ATIFlash) from the new card first, if anyone needs it.

    It will probably be possible to mess with these vBIOS files in the future, right? This (brand new) part from Dell did seem a bit unstable even before flashing, so maybe that's the reason I can't boot it with one of the older vBIOS files flashed? It seems strange there would be a problem if I didn't have a bad part to begin with. I'll probably just return this and use my Asus full time again otherwise, it's been my brother's dev box lately but he won't mind... :pirate:

    The R4 is stuck with its iGPU for now, at least. Maybe Dell will let me buy/trade in/etc. for a GTX 680m or something. This will be my 3rd 7970m replacement, and only the 2nd 7970m I've tried flashing; working chips of these seem hard for me to get.

×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use. We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.