Samsung laptops bricked by using UEFI

Discussion in 'Samsung' started by yknyong1, Jan 30, 2013.

Thread Status:
Not open for further replies.
  1. John Ratsey

    John Ratsey Moderately inquisitive Super Moderator

    Reputations:
    7,162
    Messages:
    28,880
    Likes Received:
    1,937
    Trophy Points:
    581
    But would a problem with the battery disconnect switch stop the computer booting when plugged into mains power?

    John
     
  2. black83

    black83 Notebook Consultant

    Reputations:
    256
    Messages:
    169
    Likes Received:
    2
    Trophy Points:
    31
    Hm, my 770Z5E has sub options to use "UEFI and CSM", "UEFI" and "CSM" when I disable secure boot.
    On UEFI it doesn't find any other bootable devices, on CSM it won't install/boot Win8.
    In combo mode all works fine. Maybe that's how Samsung fixed this problem ?

    Still fast boot needs to be disabled to have it find USB devices.
    After your done change "fast boot" back to enabled, or the Samsung updater software /SW update) won't recognize your notebook and hide drivers from you. Took me some hours to figure out why I had no drivers offered once I installed Win8 but had them on the install before. The updater makes everything work perfectly, just don't let windows update HD4000 drivers or you end up in a crash when the lid is closed/opened (sleep).

    It's fresh installs from MSDNAA isos. With secure boot it won't boot that way anymore, you have to use the Samsung supplied (repair) image if you want secure boot. For fresh installs you have to use uefi and csm mode. Makes no difference for me, booting is done within 5 seconds using Samsungs SSD 840 pro, sleep mode is gone after 2 secs.

    Well my notebook is unchained now from Microsofts tyrany and hope it's unbrickable :)
     
  3. ace_g

    ace_g Newbie

    Reputations:
    0
    Messages:
    9
    Likes Received:
    0
    Trophy Points:
    5
    If I remember correctly, my machine wouldn't boot at all with UEFI until I reformatted the HD using GPT. And, the reverse was also true, it wouldn't recognize the HD as bootable when using UEFI with MBR. Which is probably why, when I sent it to Samsung for a repair, they swapped the motherboard (which defaults to standard BIOS) and then thought they had to swap my (GPT formatted) HD. (When, in fact, they only needed to set UEFI to ON in the BIOS.)

    Right now, it seems that the UEFI situation is not quite ready for prime-time. Particularly if it's bricking computers. And I really don't see the problem (Samsung's) with running GPT on standard BIOS - my old XPS M1210 and Inspiron e1705 machines are both GPT-formatted and they're triple-boot... but of course there's that PITA ghost MBR partition table in the Windows 7 OSes - so Windows only "sees" the first four physical partitions. This means that all of the Windows-accessible logical drives need to be located "up front".

    This is a very frustrating situation for a multi-platform developer - or a consultant who works in different OSes as a matter of course.

    Hmmm, good question. I only provided that as a possibility since the OP was, himself, wondering if that could be the problem. This could be determined empirically by holding down the battery reset switch on a working S7 and while doing so, attempt to plug in and power up the machine. Another option might be for the OP to try powering up his machine using the adapter with the battery unplugged - I haven't tried that one either, but it's pretty simple to set up.

    Since the battery-reset switch does disable power up from battery until the AC adapter is plugged in, it's not a stretch to postulate that if the switch remains depressed it could very well be wired to also inhibit operation while under mains power - either on purpose or as an unintended consequence of its particular circuit design.

    Finally, it is a reset switch, it may also assert a reset signal to the MB that actually resets the processor along with "disconnecting" power. That could, by definition, certainly prevent booting and/or power up.

    In fact it may well be that the open circuit (or closed circuit) created by depressing the reset switch also causes some kind of interlock condition in the power management circuitry that prevents the power up signal (from the power switch) from reaching the power-up logic on the MB. This brings to mind a number of experimental scenarios that could be set up as a means to "reverse engineer" the power/battery/reset circuit's functional specification. But, having spent far too much time "inside" my own S7, I'll gladly leave those projects for someone who needs the answers.
     
  4. rimunroe

    rimunroe Newbie

    Reputations:
    0
    Messages:
    4
    Likes Received:
    0
    Trophy Points:
    5
    I encountered this bug a few weeks ago on my NP300E5C-A01UB and am attempting to unbrick the laptop by reflashing. In the process of doing this, my friend and I pulled a copy of the borked firmware, and I have uploaded it to GitHub: https://github.com/rimunroe/SamsungUEFI

    Also, if anyone here can share a copy of the vendor supplied firmware I would *greatly* appreciate it since then we could reflash.
     
  5. John Ratsey

    John Ratsey Moderately inquisitive Super Moderator

    Reputations:
    7,162
    Messages:
    28,880
    Likes Received:
    1,937
    Trophy Points:
    581
    You can try following the guidance in this thread to get an earlier firmware which will put you back on the path to the latest.

    John
     
  6. rimunroe

    rimunroe Newbie

    Reputations:
    0
    Messages:
    4
    Likes Received:
    0
    Trophy Points:
    5
    Thanks, that pointed us in the right direction. We may have a working image now. We'll try flashing it tomorrow.
     
  7. EVA Alepou

    EVA Alepou Newbie

    Reputations:
    0
    Messages:
    1
    Likes Received:
    0
    Trophy Points:
    5
    Hello. I had a similar experience with an NP300-E5C by using Acronis True Image WD Edition (recovery software) to perform an HDD backup. The BIOS locked fully after the second use and I was unable to access it or boot from CD/DVD, USB Stick or even use the internal recovery software. Windows was still booting but the Safe mode was not usable. I sent it for service and they replaced the motherboard. This proves that any UEFI Software can brick those laptops, even recovery software.
     
  8. black83

    black83 Notebook Consultant

    Reputations:
    256
    Messages:
    169
    Likes Received:
    2
    Trophy Points:
    31
    I can say now that installing Linux didn't brick anyting on my 770Z5E.
    Had a fresh Windows 8 install with disabled secure boot and sub option set to "CSM and UEFI".

    Had to remove GPT partiton table stray with "fixparts" tool first because Linux couldn't find my MBR partitions
     
  9. Dannemand

    Dannemand Decidedly Moderate Super Moderator

    Reputations:
    11,267
    Messages:
    4,355
    Likes Received:
    2,005
    Trophy Points:
    231
    @black83: Just to make sure I understand you correctly: You're saying you converted your disk to MBR (or replaced it with an MBR partitioned SSD); and with the CSM and UEFI setting in BIOS you are able to boot Windows (and Linux) from an MBR disk. Is that correct? Do you have an EFI partition on the disk?

    I ask because according to this MS technet article Windows in UEFI mode isn't able to boot from MBR disks:

    Admittedly, it also says:

    But I've understood that while MBR was supported on external disks (using CSM), Windows in UEFI mode had to boot from a GPT disk.

    I have an old Sandy Bridge model myself and don't have any of those BIOS options. But I am trying to understand them in order to help answer questions around here.

    Thanks you for helping me understand :eek:
     
  10. Ubertom

    Ubertom Newbie

    Reputations:
    0
    Messages:
    2
    Likes Received:
    0
    Trophy Points:
    5
    So it looks like I've bricked my Samsung NP900X3B-A02US by attempting to boot into Ubuntu from USB.

    I've tried holding the battery disconnect button(the tiny hole at the back?) along with the power button for 1 minute(actually tried for two just to be safe). This did not work, I've done it again and I'm going to leave it for the next 48 hours. This will reset the CMOS? How is this possible if the CMOS is stored using it's own battery? Won't I have to wait until that runs out as well?

    Also where is this information coming from? There's almost nothing I can find relating to this on the internet apart from a few articles and this thread? How can I contact Samsung to fix this? Having a really hard time here, any help would be much appreciated.
     
Loading...
Thread Status:
Not open for further replies.

Share This Page