821fba4722
default.nix: formatting
2 years ago
ad5cee037a
remove accidentally-checked-in file
...
I accidentally checked in "ectool-patches-of-unclear-provenance.patch",
which was never used in any way by the build expressions.
This file contained my local patches to a very old version of
ectool, some of which came from Debian and some of which I wrote
myself. The `ectool.nix` expression uses a newer version of ectool,
which has upstreamed the relevant changes from Debian. So those
patches no longer need to be carried. The other patches delete
functionality which I don't need, but other people might, so those
patches won't be included in ownerboot.
2 years ago
8438f297b4
Relicense: GPL-v2-or-v3 instead of GPL-v2-or-later
...
At this point I am the only person who has contributed to ownerboot,
so I'm able to make this correction unilaterally. I wanted to do
this while I still had the chance.
2 years ago
9924354136
README: add todo item for moving to module-merge-based kconfig
2 years ago
276723681d
kevin/linux.config: enable ath9k spectral scan and debugfs
2 years ago
0f15fd86a8
kgpe/linux.config: remove sb700 watchdog-related settings
...
The KGPE-D16 southbridge contains a watchdog; the kernel module for
this watchdog is sp5100_tco.
Unfortunately this watchdog keeps its counter in the main system CPU
RAM. This means that the IOMMU will block the watchdog's attempts
to read and write its own counter (through the SMBus, which is PCI
device 0000:00:14.0), resulting in massive amounts of dmesg spam
like this:
```
pci 0000:00:14.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x0004 address=0xfdf9103300 flags=0x0030]
piix4_smbus 0000:00:14.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x0004 address=0xfdf9103300 flags=0x0030]
```
Setting the IOMMU to allow these accesses is a headache, and
pointless anyways: the KGPE-D16 board disallows the use of the
watchdog by the CPU; it is reserved for the exclusive use of the
BMC. Attempting to enable the watchdog does nothing; the enable bit
will not turn on. Recent kernels check for this situation and
report when it is detected:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=f7781b067522aa269213e8025c80cbed1868d349
Fortunately there is *another* watchdog on the motherboard, inside
the `W83667HG-A` SuperIO chip, and it is supported by
`w83627hf_wdt.ko`. Subsequent commits to ownerboot will use *that*
watchdog instead.
Dump of my collection of possibly-related links:
- https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=266325
- https://forums.freebsd.org/threads/ivhd-errors-after-upgrade-host-to-13-1.85234/
- https://lists.linuxfoundation.org/pipermail/iommu/2015-October/014523.html
- https://lkml.org/lkml/2014/4/22/562
- https://bugzilla.kernel.org/show_bug.cgi?id=170741
- https://marc.info/?l=linuxbios&m=152035277020470&w=2
2 years ago
6cf5223d41
ecfw: move flashing scripts to separate derivation
2 years ago
e5824b756d
ecfw: fix comment
2 years ago
4cf0b70ae3
ecfw: fix version attr
2 years ago
cab2ef5655
README.md: update roadmap
2 years ago
1ce5815fc1
README.md: move roadmap to top of page
2 years ago
036e75e542
README.md: link to more details about ecfw write protection
2 years ago
bdebb940c9
platform/kevin/ecfw: init
2 years ago
1b9d35c10d
flashrom: append -chromebook to pname and $out/bin-name when forChromebook==true
2 years ago
9ff6fba57c
ectool: remove redundant EXTRA_CFLAGS
2 years ago
9e592c9751
ectool: make boardName a non-package argument
2 years ago
30ac75903b
src/util/ectool/: init
2 years ago
c40ed36e6b
src/util/flashrom: move from src/flashrom
2 years ago
709bb2104d
src/plat/*: allow to override hostPlatform
2 years ago
0bdf6e9355
src/plat/*: pass `lib` as a parameter instead of reimporting it
2 years ago
ed0d745ca4
kgpe: add missing `mkdir` and path component
2 years ago
d90863f36b
coreboot: use python3
2 years ago
2bf7048655
README.md: fix typo
2 years ago
400ca67b14
README.md: fix link
2 years ago
98ff0a6132
README.md: current plans
2 years ago
a0ba72a30a
README.md: mention nvramtool
2 years ago
364ab1d34d
README.md: shout out to PrawnOS
2 years ago
f0eef409de
README.md: mention google's flashrom-wp chromebook fork
2 years ago
576e50fc85
flashrom: add option for google chromebook fork
2 years ago
6243fa9245
src/coreboot: expose src attr in passthru.src
2 years ago
018e75000a
optionally build nvramtool, put it and the .layout files in the initramfs
2 years ago
7a93f3a128
update coreboot-toolchain to deal with changes in nixpkgs 21.11
2 years ago
427d6beaae
update nixpkgs to 21.11
2 years ago
bceb4820f4
factor out lib as an optional argument
2 years ago
5fa23513f9
factor out pkgsFun as an optional argument
2 years ago
7fa5116288
doc/owner-controlled.md: mention debian and gnuk
2 years ago
ece2e5c052
add links
2 years ago
8eaa402ba1
doc/platform/kevin/notes.md: mention CONFIG_INITRAMFS_SOURCE issue
2 years ago
b8729694c2
doc/platform/am1i/notes.md: mention iommu situation
2 years ago
0bf512385d
coreboot: add linux-command-line parameter for CONFIG_LINUX_COMMAND_LINE
2 years ago
184d2421b2
kgpe: include microcode updates for 63xx cpus
2 years ago
9658f9f551
kgpe/coreboot.config: remove extraneous CONFIG_LOCALVERSION
2 years ago
1732d9c14e
src/coreboot: add console-loglevel parameter
2 years ago
040bcdbd79
docs: add platform-specific notes for each platform
2 years ago
44caf3407c
src/kernel: use 5.10.148 on all platforms
...
This commit bumps the kgpe and am1i platforms to 5.10.148 for the
pre-kexec() kernel, so all platforms now use the same version.
2 years ago
8c8c0017ca
em100: fix meta.maintainers
2 years ago
a180b832e7
kgpe: flashrom.layout no longer needed
...
Since a6cd35, ownerboot includes a patch to flashrom which allows
nested (but non-overlapping) fmap regions, so the flashrom.layout
file is no longer necessary.
2 years ago
1a0e0922ac
ok, then care about science
2 years ago
2c6a837b48
src/kevin/atf: roll back ~9months to fix suspend-to-ram
...
ATF v1.6 on gru-kevin causes the laptop to reset itself instead of
waking up from suspend-to-ram. The cause of this problem is
something in the ~835 commits prior to the v1.6 release.
For now, let's simply use an older commit from upstream;
suspend-to-ram is pretty important for laptops.
TODO: git bisect and revert only the commits that cause this problem.
2 years ago
4a2a89d88a
src/kernel: bump non-gru-kevin kernels for Linux wifi RCE
...
This bumps the kernel version on non-gru-kevin to 5.10.148, which
has fixes for the notorious Linux kernel wifi RCE exploits:
CVE-2022-41674, CVE-2022-42719, and CVE-2022-42720.
2 years ago
Adam Joseph