cbmk/util/dell-flash-unlock
Leah Rowe 102ce12cea rebase cbmk 9429287 per lbmk c4d90087..f5b04fa5
cbmk 9429287 is the present canoeboot revision, on this day,
two commits after canoeboot 20231107

the cbmk revision was based on lbmk c4d90087, but lbmk
has developed a lot since, right up to f5b04fa5. lbmk
c4d90087 was four commits after libreboot 20231106

this patch brings cbmk up to date, versus lbmk f5b04fa5,
which is 135 commits after libreboot 20231106 (not 4)

therefore, the next canoeboot release shall import lbmk
changes made *after* lbmk revision f5b04fa5. good day!

In English (the above is for my reference, next time
I make a new canoeboot release):

This imports all of the numerous improvements from
Libreboot, sans the non-FSDG-compliant changes. You
can find a full list of such changes in the audit4 page:

https://libreboot.org/news/audit4.html

A full canoeboot-ised changelog will be available in
the next canoeboot release, with these and subsequent
changes. Most notable here is the update to the new
GRUB 2.12 release (instead of 2.12-rc1), and the
improvements Riku made to pico-serprog. And the build
system improvements from lbmk, such as improved, more
generic cmake and autoconf handling.

Canoeboot-specific changes: I also tweaked the deblob
logic, to make it less error-prone. The new design
changes imported into cbmk (based on latest lbmk) somewhat
broke the deblob logic; it was constantly reminding the
user that blobs.list was missing for coreboot,
at config/coreboot/blobs.list - coreboot is a multi-tree
project in both cbmk and lbmk, and the deblob logic was
tuned for single/multi, but was treating coreboot as both.
for simplicity, i removed the check for whether blobs.list
is present. this means that the operator must ensure that
these files are present, in any given revision, where they
are required on a given set of projects (and the files are
all present, in this update to cbmk)

Also of note: the grub.cfg improvements are included in this
cbmk update. The improved grub.cfg can find grub/syslinux
configs by default, not just grub anymore, also finds extlinux,
and will also find them on EFI System Partition - in addition,
UEFI-based install media is also more robust; although cbmk
doesn't provide UEFI configurations on x86, our GRUB palyoad
does still need to work with distro install media, and many
of them now use UEFI-based GRUB configurations in their
installation media, which just happen to work with our GRUB

Signed-off-by: Leah Rowe <leah@libreboot.org>
2024-01-02 11:55:45 +00:00
..
COPYING Canoeboot 20231026 release 2023-10-27 08:21:04 +01:00
Makefile Canoeboot 20231026 release 2023-10-27 08:21:04 +01:00
README.md rebase cbmk 9429287 per lbmk c4d90087..f5b04fa5 2024-01-02 11:55:45 +00:00
accessors.c Canoeboot 20231026 release 2023-10-27 08:21:04 +01:00
accessors.h Canoeboot 20231026 release 2023-10-27 08:21:04 +01:00
dell_flash_unlock.c Canoeboot 20231026 release 2023-10-27 08:21:04 +01:00

README.md

Dell Laptop Internal Flashing

This utility allows you to use flashrom's internal programmer to program the entire BIOS flash chip from software while still running the original Dell BIOS, which normally restricts software writes to the flash chip. It seems like this works on any Dell laptop that has an EC similar to the SMSC MEC5035 on the E6400, which mainly seem to be the Latitude and Precision lines starting from around 2008 (E6400 era).

TL;DR

On Linux, ensure you are booting with the iomem=relaxed kernel parameter. On OpenBSD, ensure you are booting with securelevel set to -1. Run make to compile the utility, and then run sudo ./dell_flash_unlock and follow the directions it outputs.

Confirmed supported devices

  • Latitude E6400, E6500
  • Latitude E6410, E4310
  • Latitude E6420
  • Latitude E6430, E6530
  • Precision M6800

It is likely that any other Latitude/Precision laptops from the same era as devices specifically mentioned in the above list will work as Dell seems to use the same ECs in one generation.

Tested

These systems have been tested, but were reported as not working with dell-flash-unlock. This could be due to user error, a bug in this utility, or the feature not being implemented in Dell's firmware. If you have such a system, please test the utility and report whether or not it actually works for you.

  • Latitude E6220
  • Latitude E6330

Detailed device specific behavior

  • On GM45 era laptops, the expected behavior is that you will run the utility for the first time, which will tell the EC to set the descriptor override on the next boot. Then you will need to shut down the system, after which the system will automatically boot up. You should then re-run the utility to disable SMM, after which you can run flashrom. Finally, you should run the utility a third time to reenable SMM so that shutdown works properly afterwards.
  • On 1st Generation Intel Core systems such as the E6410 and newer, run the utility and shutdown in the same way as the E6400. However, it seems like the EC no longer automatically boots the system. In this case you should manually power it on. It also seems that the firmware does not set the BIOS Lock bit when the descriptor override is set, making the 2nd run after the reboot technically unnecessary. There is no harm in rerunning it though, as the utility can detect when the flash is unlocked and perform the correct steps as necessary.

How it works

There are several ways the firmware can protect itself from being overwritten. One way is the Intel Flash Descriptor (IFD) permissions. On Intel systems, the flash image is divided into several regions such as the IFD itself, Gigabit Ethernet (GBE) non-volative memory, Management Engine (ME) firmware, Platform Data (PD), and the BIOS. The IFD contains a section which specifies the read/write permissions for each SPI controller (such as the host system) and each region of the flash, which are enforced by the chipset.

On the Latitude E6400, the host has read-only access to the IFD, no access to the ME region, and read-write access to the PD, GBE, and BIOS regions. In order for flashrom to write to the entire flash internally, the host needs full permissions to all of these regions. Since the IFD is read only, we cannot change these permissions unless we directly access the chip using an external programmer, which defeats the purpose of internal flashing.

However, Intel chipsets have a pin strap that allows the flash descriptor permissions to be overridden depending on the value of the pin at power on, granting RW permissions to all regions. On the ICH9M chipset on the E6400, this pin is HDA_DOCK_EN/GPIO33, which will enable the override if it is sampled low. This pin happens to be connected to a GPIO controlled by the Embedded Controller (EC), a small microcontroller on the board which handles things like the keyboard, touchpad, LEDs, and other system level tasks. Software can send a certain command to the EC, which tells it to pull GPIO33 low on the next boot.

Although we now have full access according to the IFD permissions, we still cannot flash the whole chip, due to another protection the firmware uses. Before software can update the BIOS, it must change the BIOS Write Enable (BIOSWE) bit in the chipset from 0 to 1. However, if the BIOS Lock Enable (BLE) bit is also set to 1, then changing the BIOSWE bit triggers a System Management Interrupt (SMI). This causes the processor to enter System Management Mode (SMM), a highly privileged x86 execution state which operates transparently to the operating system. The code that SMM runs is provided by the BIOS, which checks the BIOSWE bit and sets it back to 0 before returning control to the OS. This feature is intended to only allow SMM code to update the system firmware. As the switch to SMM suspends the execution of the OS, it appears to the OS that the BIOSWE bit was never set to 1. Unfortunately, the BLE bit cannot be set back to 0 once it is set to 1, so this functionality cannot be disabled after it is first enabled by the BIOS.

Older versions of the E6400 BIOS did not set the BLE bit, allowing flashrom to flash the entire flash chip internally after only setting the descriptor override. However, more recent versions do set it, so we may have hit a dead end unless we force downgrade to an older version (though there is a more convenient method, as we are about to see).

What if there was a way to sidestep the BIOS Lock entirely? As it turns out, there is, and it's called the Global SMI Enable (GBL_SMI_EN) bit. If it's set to 1, then the chipset will generate SMIs, such as when we change BIOSWE with BLE set. If it's 0, then no SMI will be generated, even with the BLE bit set. On the E6400, GBL_SMI_EN is set to 1, and it can be changed back to 0, unlike the BLE bit. But there still might be one bit in the way, the SMI_LOCK bit, which prevents modifications to GBL_SMI_EN when SMI_LOCK is 1. Like the BLE bit, it cannot be changed back to 0 once it set to 1. But we are in luck, as the vendor E6400 BIOS leaves SMI_LOCK unset at 0, allowing us to clear GBL_SMI_EN and disable SMIs, bypassing the BIOS Lock protections.

There are other possible protection mechanisms that the firmware can utilize, such as Protected Range Register settings, which apply access permissions to address ranges of the flash, similar to the IFD. However, the E6400 vendor firmware does not utilize these, so they will not be discussed.

References