We use eb2c04185f (2024-11-21 1.3+ bugfixes)
Where meteor lake is 5e0d9b04a0 is from 1.3 (3 weeks ago)
Signed-off-by: Thierry Laurion <insurgo@riseup.net>
Made sure all linux configs are in good oldconfig state by:
docker run -e DISPLAY=$DISPLAY --network host --rm -ti -v $(pwd):$(pwd) -w $(pwd) tlaurion/heads-dev-env:v0.2.3
find ./boards/ -type f | awk -F "/" {'print $3'} | sort | uniq | while read board; do make BOARD=$board linux.modify_and_save_oldconfig_in_place; done
Signed-off-by: Thierry Laurion <insurgo@riseup.net>
cryptsetup2 2.6.1 is a new release that supports reencryption of Q4.2 release LUKS2 volumes created at installation.
This is a critical feature for the Qubes OS 4.2 release for added data at rest protection
Cryptsetup 2.6.x internal changes:
- Argon2 used externally and internally: requires a lot of RAM and CPU to derivate passphrase to key validated in key slots.
- This is used to rate limit efficiently bruteforcing of LUKS key slots, requiring each offline brute force attempt to consume ~15-30 seconds per attempt
- OF course, strong passphrases are still recommended, but bruteforcing LUKSv2 containers with Argon2 would require immense time, ram and CPU even to bruteforce low entropy passphrase/PINs.
- passphrase change doesn't permit LUKS key slot specification anymore: key slot rotates (new one consusumed per op: then old one wiped internally. EG: LUKS key slot 1 created, then 0 deleted)
- reencryption doesn't permit old call arguments. No more direct-io; inadmissively slow through AIO (async) calls, need workarounds for good enough perfs (arguments + newer kernel with cloudfare fixes in tree)
cryptsetup 2.6.1 requires:
- lvm2 2.03.23, which is also included in this PR.
- requires libaio, which is also included in this PR (could be hacked out but deep dependency at first sight: left in)
- requires util-linux 2.39
- patches for reproducible builds are included for above 3 packages.
luks-functions was updated to support the new cryptsetup2 version calls/changes
- reencryption happen in direct-io, offline mode and without locking, requiring linux 5.10.9+ to bypass linux queues
- from tests, this is best for performance and reliability in single-user mode
- LUKS container ops now validate Disk Recovery Key (DRK) passphrase prior and DRK key slot prior of going forward if needed, failing early.
- Heads don't expect DRK to be in static key slot anymore, and finds the DRK key slot dynamically.
- If reencrytipn/passphrase change: make sure all LUKS containers on same block device can be unlocked with same DRK
- Reencryption: requires to know which key slot to reencrypt.
- Find LUKS key slot that unlocks with DRK passphrase unlock prior of reencrypt call
- Passphrase change: no slot can be passed, but key slot of DRK rotates.
kexec-seal-key
- TPM LUKS Disk Unlock Key key slots have changed to be set in max slots per LUKS version (LUKSv1:7 /LUKSv2: 31)
- If key slot != default LUKS version's keyslot outside of DRK key slot: prompt the user before wiping that key slot, otherwise wipe automatically
- This takes for granted that the DRK key slot alone is needed on the system and Heads controls the LUKS key slots.
- If user has something else going on, ie: Using USB Security dongle + TPM DUK, then the user will need to say no when wiping keys.
- It was suggested to leave LUKS key slots outside of DRK alone, but then: what to do when all key slots would be used?
- Alternative implementation could be to only prompt users to wipe keyslots other then DRK when key slots are all used (LUKSv1: 0-7, LUKSv2: 0-31)
- But then cleanup would need to happen prior of operations (LUKS passphrase change, TPM DUK setup) and could be problematic.
- LUKS containers now checked to be same LUKS version prior of permitting to set TPM DUK and will refuse to go forward of different versions.
TODO:
- async (AIO) calls are not used. direct-io is used instead. libaio could be hacked out
- this could be subject to future work
Notes:
- time to deprecated legacy boards the do not enough space for the new space requirements
- x230-legacy, x230-legacy-flash, x230-hotp-legacy
- t430-legacy, t430-legacy-flash, t430-hotp-legacy already deprecated
Unrelated:
- typos fixes found along the way
Signed-off-by: Thierry Laurion <insurgo@riseup.net>
Cloudfare patches to speed up LUKS encryption were upstreamed into linux kernel and backported to 5.10.9: https://github.com/cloudflare/linux/issues/1#issuecomment-763085915
Therefore, we bump to latest of 5.10.x (bump from 5.10.5 which doesn't contain the fixes)
Trace:
sed -i 's/5.10.5/5.10.214/g' boards/*/*.config
find ./boards/*/*.config | awk -F "/" {'print $3'}| while read board; do echo "make BOARD=$board linux"; make BOARD=$board linux; echo make BOARD=$board linux.save_in_oldconfig_format_in_place || make BOARD=$board linux.modify_and_save_oldconfig_in_place; done
git status | grep modified | awk -F ":" {'print $2'}| xargs git add
git commit --signoff
- Move patches from 5.10.5 -> 5.10.214
- Add linux kernel hash and version under modules/linux
- Change board configs accordingly
Signed-off-by: Thierry Laurion <insurgo@riseup.net>
Instead of requesting additional ramstage heap, cherry-pick patches for
newer JPEG decoder that does not require heap allocation.
Signed-off-by: Jonathon Hall <jonathon.hall@puri.sm>
Signed-off-by: Thierry Laurion <insurgo@riseup.net>
librems vs mainline have:
- mei module not compiled in
- Streamline modules/linux Makefile helpers so that one asks on console for new Kconfig options to be y/n, another one to do olddefconfig (accept new defaults)
- Exercice proves again that oldconfig exposes new things added into kernel by default as opposed to defconfig format.
- Add TCP Syn cookies to all linux configs (all boards add CDC tethering AFAIK. Add this protection by default)
- Remove unneeded network card drivers from librems common and unify
- Remove unneeded microsoft surface drivers from librems common and unify
- Remove WMI embedded Binary MOF driver CONFIG_WMI_BMOF
- Unify removed Kconfig options from x230 ported to 6.1.8 to librems common
- Verify qemu (AMD) changes working (note, there is clock source watchdog that would need to be investigated seperately for QEMU TCG mode, that is, not KVM)
- Review crypto backend requirements/unify once more
- Removed bunch of unused stuff under QEMU Q35 (AMD)
Q/A:
- CONFIG_RANDSTRUCT_NONE vs CONFIG_RANDSTRUCT_FULL? CONFIG_RANDSTRUCT_NONE now.
- CONFIG_LDISC_AUTOLOAD=y?
- CONFIG_PTP_1588_CLOCK_OPTIONAL=y?
- CONFIG_X86_THERMAL_VECTOR=y?
- ACPI-WMI (Windows Management Instrumentation) mapper device (PNP0C14) enabled by ACPI_VIDEO and depended by DRM drivers.
- ACPI_VIDEO seems needed.
- CONFIG_INPUT_VIVALDIFMAP=y anabled by CONFIG_KEYBOARD_ATKBD (AT/PS2 Keyboard)
Signed-off-by: Thierry Laurion <insurgo@riseup.net>
Reuse the toolchain from coreboot-24.02.01 for coreboot-purism (based
on 24.02.01) instead of compiling a separate toolchain.
Signed-off-by: Jonathon Hall <jonathon.hall@puri.sm>
As discussed in: linuxboot#1704
there is no need to include iotools module for nitropads.
Since there is no board using it, and we see no reason to use
it in the future (the EC udpate will not require it, as update
will be server by coreboot in the future), drop the module as well.
Signed-off-by: Maciej Pijanowski <maciej.pijanowski@3mdeb.com>
Update Purism coreboot to 24.02.01-Purism-1.
Remove CFLAGS overrides when building coreboot. These overrides break
24.02.01, which added (and needs) --param=min-pagesize=1024. This has
happened repeatedly in the past since Heads has to duplicate coreboot's
CFLAGS if it overrides them.
Specifically, the build fails with this error:
src/commonlib/include/commonlib/endian.h:27:26: error: array subscript 1 is outside array bounds of 'void[0]' [-Werror=array-bounds=]
27 | *(uint8_t *)dest = val;
| ~~~~~~~~~~~~~~~~~^~~~~
In function 'setup_default_ebda':
cc1: note: source object is likely at address zero
That's because coreboot is attempting to write to EBDA at physical
address 0x40e, just above 1024. That is a valid address for x86, but
it's too close to 0 by default for GCC, --param-min-pagesize=1024
allows writes to physical addresses above 1024.
coreboot shouldn't need any of the usual Heads CFLAGS overrides for
reproducibility; it is already reproducible.
Fix indentation in modules/coreboot. Make accepted it before because
the indented lines followed a variable assignment, so they couldn't
be part of a recipe. That assignment is now gone, so they're now
interprted as part of a recipe for the `.configured` target just above,
they should not be indented.
Add nss to flake.nix, needed as of 24.02.01.
Update Librem coreboot configs for 24.02.01-Purism-1. Notably, the
board Kconfig changed for Mini v2 in coreboot, so this is needed for
correct builds.
Signed-off-by: Jonathon Hall <jonathon.hall@puri.sm>
- Revert gnupg toolstack version bump to prior of #1661 merge (2.4.2 -> 2.4.0). Version bump not needed for reproducibility.
- Investigation and upstream discussions will take their time resolving invalid time issue introduced by between 2.4.0 and latest gnupg, fix regression first under master)
- oem-factory-reset
- Adding DO_WITH_DEBUG to oem-factory-reset for all its gpg calls. If failing in debug mode, /tmp/debug.txt contains calls and errors
- Wipe keyrings only (*.gpg, *.kbx) not conf files under gpg homedir (keep initrd/.gnupg/*.conf)
- flake.nix
- switch build derivative from qemu and qemu_kvm to qemu_full to have qemu-img tool which was missing to run qemu boards (v0.1.8 docker)
- add gnupg so that qemu boards can call inject_gpg to inject public key in absence of flashrom+pflash support for internal flashing
- flake.lock: Updated nix pinned package list under flake.lock with 'nix flake update' so qemu_full builds
- README.md: have consistent docker testing + release (push) notes
- .circleci/config.yml: depend on docker v0.1.8 (qemu_full built with canokey-qemu lib support, diffoscopeMinimal and gnupg for proper qemu testing)
TODO:
- some fd2 instead of fd1?!
- oem-factory-resest has whiptail_or_die which sets whiptail box to HEIGHT 0. This doesn't show a scrolling window on gpg errors which is problematic with fbwhiptail, not whiptail
Signed-off-by: Thierry Laurion <insurgo@riseup.net>
Specifying LIBS_BASE causes flashrom's Makefile to link in an RPATH,
using the Heads workspace path, which is not what we want. It does
other things too, but we already pass the parts we need to the make
invocation for flashrom.
Signed-off-by: Jonathon Hall <jonathon.hall@puri.sm>
Signed-off-by: Thierry Laurion <insurgo@riseup.net>
Thanks to @JonathonHall-Purism, that pointed to me that sysroot was
neglected in tpm2-tools configure step.
I wonder why this is not respected if not forced with --with-sysroot and
TSS2_ESYS_3_0_LIBS="-ltss2-esys -L$(INSTALL)/lib"?
Signed-off-by: Thierry Laurion <insurgo@riseup.net>
Signed-off-by: Manuel Mendez <github@i.m.mmlb.dev>
Signed-off-by: Thierry Laurion <insurgo@riseup.net>
