2023-06-13 11:09:01 +00:00
|
|
|
#!/usr/bin/env sh
|
2023-09-25 01:19:30 +00:00
|
|
|
# SPDX-License-Identifier: GPL-3.0-only
|
2022-11-14 00:51:12 +00:00
|
|
|
# SPDX-FileCopyrightText: 2022 Caleb La Grange <thonkpeasant@protonmail.com>
|
2023-06-13 11:09:01 +00:00
|
|
|
# SPDX-FileCopyrightText: 2022 Ferass El Hafidi <vitali64pmemail@protonmail.com>
|
NEW MAINBOARD: HP EliteBook 820 G2
This is of Broadwell platform, one generation above Haswell.
Of note: this uses HP Sure Start. Although the flash is 16MB,
our CBFS section (and IFD configuration) assumes 12MB flash,
so the final 4MB will be left unflashed on installation,
after blanking the private flash. The coreboot documents have
more information about this.
Some minor design changes in lbmk were made, to accomodate
this port:
Support for extracting refcode binaries added (pulled from
Google recovery images). The refcode file is an ELF that
initialises the MRC and the PCH. It is also responsible for
enabling or disabling the Intel GbE device, where Google
does not enable it, but lbmk modifies it per the instructions
on the coreboot documentation, so as to enable Intel GbE.
Google's recovery image stores the refcode as a stage file,
but coreboot changed the format (for CBFS files) after 4.13
so coreboot 4.13's cbfstool is used to extract refcode. This
realisation made me also change the script logic to use a
cbfstool and ifdtool version matching the coreboot tree, for
all parts of lbmk, whereas lbmk previously used only the
default tree for cbfstool/ifdtool, on insertion and deletion
of vendor files - it was 81dc20e744 that broke extraction of
refcode on google's recovery images, where google used an older
version of cbfstool to insert the files in their coreboot ROMs.
A further backported patch has been added, copying coreboot
revision f22f408956 which is a build fix from Nico Huber.
Iru Cai submitted an ACPI bugfix after the revision lbmk
currently uses, for coreboot/default, and this fix is
needed for rebooting to work on Linux 6.1 or higher. This
patch has been backported to lbmk, while it still uses the
same October 2023 revision of coreboot.
Broadwell MRC is inserted at the same offset as Haswell,
so I didn't need to tweak that.
Signed-off-by: Leah Rowe <leah@libreboot.org>
2024-01-07 13:25:33 +00:00
|
|
|
# SPDX-FileCopyrightText: 2023-2024 Leah Rowe <leah@libreboot.org>
|
2022-11-14 00:51:12 +00:00
|
|
|
|
2023-08-23 17:56:31 +00:00
|
|
|
. "include/err.sh"
|
2023-09-27 20:46:20 +00:00
|
|
|
. "include/option.sh"
|
2023-05-14 04:42:59 +00:00
|
|
|
|
2023-10-15 10:22:43 +00:00
|
|
|
nvmutil="util/nvmutil/nvm"
|
|
|
|
|
NEW MAINBOARD: HP EliteBook 820 G2
This is of Broadwell platform, one generation above Haswell.
Of note: this uses HP Sure Start. Although the flash is 16MB,
our CBFS section (and IFD configuration) assumes 12MB flash,
so the final 4MB will be left unflashed on installation,
after blanking the private flash. The coreboot documents have
more information about this.
Some minor design changes in lbmk were made, to accomodate
this port:
Support for extracting refcode binaries added (pulled from
Google recovery images). The refcode file is an ELF that
initialises the MRC and the PCH. It is also responsible for
enabling or disabling the Intel GbE device, where Google
does not enable it, but lbmk modifies it per the instructions
on the coreboot documentation, so as to enable Intel GbE.
Google's recovery image stores the refcode as a stage file,
but coreboot changed the format (for CBFS files) after 4.13
so coreboot 4.13's cbfstool is used to extract refcode. This
realisation made me also change the script logic to use a
cbfstool and ifdtool version matching the coreboot tree, for
all parts of lbmk, whereas lbmk previously used only the
default tree for cbfstool/ifdtool, on insertion and deletion
of vendor files - it was 81dc20e744 that broke extraction of
refcode on google's recovery images, where google used an older
version of cbfstool to insert the files in their coreboot ROMs.
A further backported patch has been added, copying coreboot
revision f22f408956 which is a build fix from Nico Huber.
Iru Cai submitted an ACPI bugfix after the revision lbmk
currently uses, for coreboot/default, and this fix is
needed for rebooting to work on Linux 6.1 or higher. This
patch has been backported to lbmk, while it still uses the
same October 2023 revision of coreboot.
Broadwell MRC is inserted at the same offset as Haswell,
so I didn't need to tweak that.
Signed-off-by: Leah Rowe <leah@libreboot.org>
2024-01-07 13:25:33 +00:00
|
|
|
eval "$(setvars "" archive rom modifygbe nukemode release new_mac tree)"
|
blobs/inject: fix checksum validation if no-ucode
on e6400_4mb, the release build scripts remove nvidia's vga
rom which is used on dgpu models. however, microcode is also
removed in separately copied rom images
the inject script was inserting vgaroms directly into these
no-microcode roms, but the microcode blob is bigger than the
vga rom, and cbfstool inserts into the first available free
spot within cbfs, so it was inserting into the spot where
cpu microcode went. this caused the rom checksum to not match
what was generated during build/release/roms being executed
the only real fix is to guarantee offsets within cbfs for all
files, by recording what offsets were used and then calculating
that during insertion
so this patch is a workaround, but fixes the issue. the workaround
is: don't insert blobs directly on no-microcode roms, instead
insert only on microcode-based roms, then re-copy those roms
and remove microcode in aptly named copies
it's a bit more convoluted, but works perfectly fine.
Signed-off-by: Leah Rowe <leah@libreboot.org>
2023-09-09 19:05:11 +00:00
|
|
|
|
2023-05-14 04:42:59 +00:00
|
|
|
main()
|
|
|
|
{
|
2023-08-23 17:56:31 +00:00
|
|
|
[ $# -lt 1 ] && err "No options specified."
|
2023-12-28 16:34:45 +00:00
|
|
|
[ "${1}" = "listboards" ] && eval "items config/coreboot || :; exit 0"
|
2022-11-14 00:51:12 +00:00
|
|
|
|
2023-05-14 04:42:59 +00:00
|
|
|
archive="${1}"
|
|
|
|
|
2023-10-14 05:30:31 +00:00
|
|
|
while getopts n:r:b:m: option; do
|
2023-07-29 06:24:18 +00:00
|
|
|
case "${option}" in
|
2023-10-14 02:30:52 +00:00
|
|
|
n) nukemode="${OPTARG}" ;;
|
2023-09-26 00:34:10 +00:00
|
|
|
r) rom=${OPTARG} ;;
|
|
|
|
b) board=${OPTARG} ;;
|
|
|
|
m) modifygbe=true
|
|
|
|
new_mac=${OPTARG} ;;
|
2023-12-16 07:56:26 +00:00
|
|
|
*) : ;;
|
2023-07-29 06:24:18 +00:00
|
|
|
esac
|
2023-05-14 04:42:59 +00:00
|
|
|
done
|
2022-11-14 00:51:12 +00:00
|
|
|
|
2023-05-14 04:42:59 +00:00
|
|
|
check_board
|
|
|
|
build_dependencies
|
2023-10-06 23:57:55 +00:00
|
|
|
inject_vendorfiles
|
2023-10-14 02:30:52 +00:00
|
|
|
[ "${nukemode}" = "nuke" ] && return 0
|
2023-08-16 23:43:34 +00:00
|
|
|
printf "Friendly reminder (this is *not* an error message):\n"
|
2023-12-24 09:04:36 +00:00
|
|
|
printf "Please ensure that the files were inserted correctly.\n"
|
2022-11-14 00:51:12 +00:00
|
|
|
}
|
|
|
|
|
2023-05-14 04:42:59 +00:00
|
|
|
check_board()
|
|
|
|
{
|
2023-11-01 10:34:05 +00:00
|
|
|
failcheck="n"
|
|
|
|
check_release "${archive}" || failcheck="y"
|
|
|
|
if [ "${failcheck}" = "y" ]; then
|
2023-12-28 16:34:45 +00:00
|
|
|
[ -f "$rom" ] || err "check_board \"$rom\": invalid path"
|
|
|
|
[ -z "${rom+x}" ] && err "check_board: no rom specified"
|
|
|
|
[ -n "${board+x}" ] || board=$(detect_board "${rom}")
|
2023-05-14 04:42:59 +00:00
|
|
|
else
|
2023-10-15 10:22:43 +00:00
|
|
|
release="y"
|
much, much stricter, more verbose error handling
lbmk is much more likely to crash now, in error conditions,
which is a boon for further auditing.
also: in "fetch", remove the downloaded program
if fail() was called.
this would also be done for gnulib, when downloading
grub, but done in such a way that gnulib goes first.
where calls to err write "ERROR" in the string, they
no longer say "ERROR" because the "err" function itself
now does that automatically.
also: listmodes/listoptions (in "lbmk") now reports an
error if no scripts and/or directories are found.
also: where a warning is given, but not an error, i've
gone through in some places and redirected the output
to stderr, not stdout
as part of error checks: running anything as root, except
for the "./build dependencies *" commands, is no longer
permitted and lbmk will throw an error
mrc downloads: debugfs output no longer redirected to /dev/null,
and stderr no longer redirected to stdout. everything is verbose.
certain non-error states are also more verbose. for example,
patch_rom in blobs/inject will now state when injection succeeds
certain actual errors(bugs) were fixed:
for example, build/release/roms now correctly prepares the blobs
hash files for a given target, containing only the files and
checksums in the list. Previously, a printf message was included.
Now, with this new code: blobutil/inject rightly verifies hashes.
doing all of this in one giant patch is cleaner
than 100 patches changing each file. even this is yet part
of a much larger audit going on in the Libreboot project.
Signed-off-by: Leah Rowe <leah@libreboot.org>
2023-08-24 19:19:41 +00:00
|
|
|
board=$(detect_board "${archive}")
|
2022-11-14 00:51:12 +00:00
|
|
|
fi
|
|
|
|
|
2023-05-14 04:42:59 +00:00
|
|
|
boarddir="${cbcfgsdir}/${board}"
|
NEW MAINBOARD: HP EliteBook 820 G2
This is of Broadwell platform, one generation above Haswell.
Of note: this uses HP Sure Start. Although the flash is 16MB,
our CBFS section (and IFD configuration) assumes 12MB flash,
so the final 4MB will be left unflashed on installation,
after blanking the private flash. The coreboot documents have
more information about this.
Some minor design changes in lbmk were made, to accomodate
this port:
Support for extracting refcode binaries added (pulled from
Google recovery images). The refcode file is an ELF that
initialises the MRC and the PCH. It is also responsible for
enabling or disabling the Intel GbE device, where Google
does not enable it, but lbmk modifies it per the instructions
on the coreboot documentation, so as to enable Intel GbE.
Google's recovery image stores the refcode as a stage file,
but coreboot changed the format (for CBFS files) after 4.13
so coreboot 4.13's cbfstool is used to extract refcode. This
realisation made me also change the script logic to use a
cbfstool and ifdtool version matching the coreboot tree, for
all parts of lbmk, whereas lbmk previously used only the
default tree for cbfstool/ifdtool, on insertion and deletion
of vendor files - it was 81dc20e744 that broke extraction of
refcode on google's recovery images, where google used an older
version of cbfstool to insert the files in their coreboot ROMs.
A further backported patch has been added, copying coreboot
revision f22f408956 which is a build fix from Nico Huber.
Iru Cai submitted an ACPI bugfix after the revision lbmk
currently uses, for coreboot/default, and this fix is
needed for rebooting to work on Linux 6.1 or higher. This
patch has been backported to lbmk, while it still uses the
same October 2023 revision of coreboot.
Broadwell MRC is inserted at the same offset as Haswell,
so I didn't need to tweak that.
Signed-off-by: Leah Rowe <leah@libreboot.org>
2024-01-07 13:25:33 +00:00
|
|
|
[ -d "$boarddir" ] || err "check_board: board $board missing"
|
|
|
|
[ -f "$boarddir/target.cfg" ] || \
|
|
|
|
err "check_board $board: target.cfg missing"
|
|
|
|
. "$boarddir/target.cfg" 2>/dev/null
|
|
|
|
[ -z "$tree" ] && err "check_board $board: tree undefined"; return 0
|
2022-11-14 00:51:12 +00:00
|
|
|
}
|
|
|
|
|
2023-05-14 04:42:59 +00:00
|
|
|
check_release()
|
|
|
|
{
|
2023-08-21 18:41:49 +00:00
|
|
|
[ -f "${archive}" ] || return 1
|
|
|
|
[ "${archive##*.}" = "xz" ] || return 1
|
|
|
|
printf "%s\n" "Release archive ${archive} detected"
|
2022-11-14 00:51:12 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
# This function tries to determine the board from the filename of the rom.
|
|
|
|
# It will only succeed if the filename is not changed from the build/download
|
2023-05-14 04:42:59 +00:00
|
|
|
detect_board()
|
|
|
|
{
|
2023-08-27 08:25:50 +00:00
|
|
|
path="${1}"
|
2023-12-16 07:56:26 +00:00
|
|
|
filename=$(basename "${path}")
|
2022-11-14 00:51:12 +00:00
|
|
|
case ${filename} in
|
2023-07-29 06:24:18 +00:00
|
|
|
grub_*)
|
2023-08-21 18:41:49 +00:00
|
|
|
board=$(echo "${filename}" | cut -d '_' -f2-3) ;;
|
2023-07-29 06:24:18 +00:00
|
|
|
seabios_withgrub_*)
|
2023-08-21 18:41:49 +00:00
|
|
|
board=$(echo "${filename}" | cut -d '_' -f3-4) ;;
|
2023-07-29 06:24:18 +00:00
|
|
|
*.tar.xz)
|
2023-04-03 00:06:46 +00:00
|
|
|
_stripped_prefix=${filename#*_}
|
2023-08-21 18:41:49 +00:00
|
|
|
board="${_stripped_prefix%.tar.xz}" ;;
|
2023-07-29 06:24:18 +00:00
|
|
|
*)
|
2023-10-20 03:10:50 +00:00
|
|
|
err "detect_board $filename: could not detect board type"
|
2022-11-14 00:51:12 +00:00
|
|
|
esac
|
2023-09-09 22:27:44 +00:00
|
|
|
printf "%s\n" "${board}"
|
2022-11-14 00:51:12 +00:00
|
|
|
}
|
|
|
|
|
2023-05-14 04:42:59 +00:00
|
|
|
build_dependencies()
|
|
|
|
{
|
NEW MAINBOARD: HP EliteBook 820 G2
This is of Broadwell platform, one generation above Haswell.
Of note: this uses HP Sure Start. Although the flash is 16MB,
our CBFS section (and IFD configuration) assumes 12MB flash,
so the final 4MB will be left unflashed on installation,
after blanking the private flash. The coreboot documents have
more information about this.
Some minor design changes in lbmk were made, to accomodate
this port:
Support for extracting refcode binaries added (pulled from
Google recovery images). The refcode file is an ELF that
initialises the MRC and the PCH. It is also responsible for
enabling or disabling the Intel GbE device, where Google
does not enable it, but lbmk modifies it per the instructions
on the coreboot documentation, so as to enable Intel GbE.
Google's recovery image stores the refcode as a stage file,
but coreboot changed the format (for CBFS files) after 4.13
so coreboot 4.13's cbfstool is used to extract refcode. This
realisation made me also change the script logic to use a
cbfstool and ifdtool version matching the coreboot tree, for
all parts of lbmk, whereas lbmk previously used only the
default tree for cbfstool/ifdtool, on insertion and deletion
of vendor files - it was 81dc20e744 that broke extraction of
refcode on google's recovery images, where google used an older
version of cbfstool to insert the files in their coreboot ROMs.
A further backported patch has been added, copying coreboot
revision f22f408956 which is a build fix from Nico Huber.
Iru Cai submitted an ACPI bugfix after the revision lbmk
currently uses, for coreboot/default, and this fix is
needed for rebooting to work on Linux 6.1 or higher. This
patch has been backported to lbmk, while it still uses the
same October 2023 revision of coreboot.
Broadwell MRC is inserted at the same offset as Haswell,
so I didn't need to tweak that.
Signed-off-by: Leah Rowe <leah@libreboot.org>
2024-01-07 13:25:33 +00:00
|
|
|
cbdir="src/coreboot/$tree"
|
|
|
|
cbfstool="cbutils/$tree/cbfstool"
|
|
|
|
ifdtool="cbutils/$tree/ifdtool"
|
|
|
|
[ -d "${cbdir}" ] || x_ ./update trees -f coreboot $tree
|
2023-10-13 02:20:33 +00:00
|
|
|
if [ ! -f "${cbfstool}" ] || [ ! -f "${ifdtool}" ]; then
|
NEW MAINBOARD: HP EliteBook 820 G2
This is of Broadwell platform, one generation above Haswell.
Of note: this uses HP Sure Start. Although the flash is 16MB,
our CBFS section (and IFD configuration) assumes 12MB flash,
so the final 4MB will be left unflashed on installation,
after blanking the private flash. The coreboot documents have
more information about this.
Some minor design changes in lbmk were made, to accomodate
this port:
Support for extracting refcode binaries added (pulled from
Google recovery images). The refcode file is an ELF that
initialises the MRC and the PCH. It is also responsible for
enabling or disabling the Intel GbE device, where Google
does not enable it, but lbmk modifies it per the instructions
on the coreboot documentation, so as to enable Intel GbE.
Google's recovery image stores the refcode as a stage file,
but coreboot changed the format (for CBFS files) after 4.13
so coreboot 4.13's cbfstool is used to extract refcode. This
realisation made me also change the script logic to use a
cbfstool and ifdtool version matching the coreboot tree, for
all parts of lbmk, whereas lbmk previously used only the
default tree for cbfstool/ifdtool, on insertion and deletion
of vendor files - it was 81dc20e744 that broke extraction of
refcode on google's recovery images, where google used an older
version of cbfstool to insert the files in their coreboot ROMs.
A further backported patch has been added, copying coreboot
revision f22f408956 which is a build fix from Nico Huber.
Iru Cai submitted an ACPI bugfix after the revision lbmk
currently uses, for coreboot/default, and this fix is
needed for rebooting to work on Linux 6.1 or higher. This
patch has been backported to lbmk, while it still uses the
same October 2023 revision of coreboot.
Broadwell MRC is inserted at the same offset as Haswell,
so I didn't need to tweak that.
Signed-off-by: Leah Rowe <leah@libreboot.org>
2024-01-07 13:25:33 +00:00
|
|
|
x_ ./update trees -b coreboot utils $tree
|
2023-10-13 02:20:33 +00:00
|
|
|
fi
|
2023-12-24 09:04:36 +00:00
|
|
|
[ -z "$new_mac" ] || [ -f "$nvmutil" ] || x_ make -C util/nvmutil
|
2023-12-28 16:34:45 +00:00
|
|
|
[ "$nukemode" = "nuke" ] || x_ ./vendor download $board; return 0
|
2023-05-14 04:42:59 +00:00
|
|
|
}
|
2023-05-06 20:21:42 +00:00
|
|
|
|
2023-10-06 23:57:55 +00:00
|
|
|
inject_vendorfiles()
|
2023-05-14 04:42:59 +00:00
|
|
|
{
|
2023-12-28 16:34:45 +00:00
|
|
|
[ "${release}" != "y" ] && eval "patch_rom \"$rom\"; return 0"
|
2023-10-03 11:59:35 +00:00
|
|
|
patch_release_roms
|
2023-04-03 00:06:46 +00:00
|
|
|
}
|
|
|
|
|
2023-05-14 04:42:59 +00:00
|
|
|
patch_release_roms()
|
|
|
|
{
|
2023-09-05 00:49:35 +00:00
|
|
|
_tmpdir="tmp/romdir"
|
2023-12-22 09:46:43 +00:00
|
|
|
remkdir "${_tmpdir}"
|
2023-10-20 03:10:50 +00:00
|
|
|
tar -xf "${archive}" -C "${_tmpdir}" || \
|
2023-12-24 09:04:36 +00:00
|
|
|
err "patch_release_roms: !tar -xf \"$archive\" -C \"$_tmpdir\""
|
2023-05-14 04:42:59 +00:00
|
|
|
|
2023-08-27 08:25:50 +00:00
|
|
|
for x in "${_tmpdir}"/bin/*/*.rom ; do
|
2023-09-09 22:31:20 +00:00
|
|
|
printf "patching rom: %s\n" "$x"
|
2023-10-20 03:10:50 +00:00
|
|
|
patch_rom "${x}"
|
2023-04-03 00:06:46 +00:00
|
|
|
done
|
|
|
|
|
2023-05-14 04:42:59 +00:00
|
|
|
(
|
2023-12-23 07:11:12 +00:00
|
|
|
cd "${_tmpdir}/bin/"* || \
|
|
|
|
err "patch_release_roms: !cd ${_tmpdir}/bin/*"
|
2023-09-09 15:39:26 +00:00
|
|
|
|
|
|
|
# NOTE: For compatibility with older rom releases, defer to sha1
|
2023-12-22 13:05:32 +00:00
|
|
|
[ "${nukemode}" = "nuke" ] || sha512sum --status -c vendorhashes || \
|
|
|
|
sha1sum --status -c vendorhashes || sha512sum --status -c \
|
|
|
|
blobhashes || sha1sum --status -c blobhashes || \
|
much, much stricter, more verbose error handling
lbmk is much more likely to crash now, in error conditions,
which is a boon for further auditing.
also: in "fetch", remove the downloaded program
if fail() was called.
this would also be done for gnulib, when downloading
grub, but done in such a way that gnulib goes first.
where calls to err write "ERROR" in the string, they
no longer say "ERROR" because the "err" function itself
now does that automatically.
also: listmodes/listoptions (in "lbmk") now reports an
error if no scripts and/or directories are found.
also: where a warning is given, but not an error, i've
gone through in some places and redirected the output
to stderr, not stdout
as part of error checks: running anything as root, except
for the "./build dependencies *" commands, is no longer
permitted and lbmk will throw an error
mrc downloads: debugfs output no longer redirected to /dev/null,
and stderr no longer redirected to stdout. everything is verbose.
certain non-error states are also more verbose. for example,
patch_rom in blobs/inject will now state when injection succeeds
certain actual errors(bugs) were fixed:
for example, build/release/roms now correctly prepares the blobs
hash files for a given target, containing only the files and
checksums in the list. Previously, a printf message was included.
Now, with this new code: blobutil/inject rightly verifies hashes.
doing all of this in one giant patch is cleaner
than 100 patches changing each file. even this is yet part
of a much larger audit going on in the Libreboot project.
Signed-off-by: Leah Rowe <leah@libreboot.org>
2023-08-24 19:19:41 +00:00
|
|
|
err "patch_release_roms: ROMs did not match expected hashes"
|
2023-12-11 05:21:27 +00:00
|
|
|
) || err "can't verify vendor hashes"
|
2023-04-03 00:06:46 +00:00
|
|
|
|
2023-12-28 16:34:45 +00:00
|
|
|
[ "${modifygbe}" = "true" ] && \
|
2023-08-27 08:25:50 +00:00
|
|
|
for x in "${_tmpdir}"/bin/*/*.rom ; do
|
|
|
|
modify_gbe "${x}"
|
2023-04-03 00:06:46 +00:00
|
|
|
done
|
|
|
|
|
2023-10-01 05:33:43 +00:00
|
|
|
[ -d bin/release ] || x_ mkdir -p bin/release
|
|
|
|
x_ mv "${_tmpdir}"/bin/* bin/release/
|
2023-12-28 16:34:45 +00:00
|
|
|
x_ rm -Rf "${_tmpdir}"
|
much, much stricter, more verbose error handling
lbmk is much more likely to crash now, in error conditions,
which is a boon for further auditing.
also: in "fetch", remove the downloaded program
if fail() was called.
this would also be done for gnulib, when downloading
grub, but done in such a way that gnulib goes first.
where calls to err write "ERROR" in the string, they
no longer say "ERROR" because the "err" function itself
now does that automatically.
also: listmodes/listoptions (in "lbmk") now reports an
error if no scripts and/or directories are found.
also: where a warning is given, but not an error, i've
gone through in some places and redirected the output
to stderr, not stdout
as part of error checks: running anything as root, except
for the "./build dependencies *" commands, is no longer
permitted and lbmk will throw an error
mrc downloads: debugfs output no longer redirected to /dev/null,
and stderr no longer redirected to stdout. everything is verbose.
certain non-error states are also more verbose. for example,
patch_rom in blobs/inject will now state when injection succeeds
certain actual errors(bugs) were fixed:
for example, build/release/roms now correctly prepares the blobs
hash files for a given target, containing only the files and
checksums in the list. Previously, a printf message was included.
Now, with this new code: blobutil/inject rightly verifies hashes.
doing all of this in one giant patch is cleaner
than 100 patches changing each file. even this is yet part
of a much larger audit going on in the Libreboot project.
Signed-off-by: Leah Rowe <leah@libreboot.org>
2023-08-24 19:19:41 +00:00
|
|
|
|
|
|
|
printf "Success! Your ROMs are in bin/release\n"
|
2023-04-03 00:06:46 +00:00
|
|
|
}
|
|
|
|
|
2023-05-14 04:42:59 +00:00
|
|
|
patch_rom()
|
|
|
|
{
|
|
|
|
rom="${1}"
|
|
|
|
|
2023-12-22 13:05:32 +00:00
|
|
|
check_defconfig "$boarddir" && err "patch_rom $boarddir: no configs"
|
2023-09-29 21:34:34 +00:00
|
|
|
|
2023-05-14 04:42:59 +00:00
|
|
|
set -- "${boarddir}/config/"*
|
2023-08-31 16:16:54 +00:00
|
|
|
. "${1}" 2>/dev/null
|
2023-05-14 04:42:59 +00:00
|
|
|
|
2023-08-21 18:41:49 +00:00
|
|
|
[ "$CONFIG_HAVE_MRC" = "y" ] && \
|
2023-10-14 05:30:31 +00:00
|
|
|
inject "mrc.bin" "${CONFIG_MRC_FILE}" "mrc" "0xfffa0000"
|
NEW MAINBOARD: HP EliteBook 820 G2
This is of Broadwell platform, one generation above Haswell.
Of note: this uses HP Sure Start. Although the flash is 16MB,
our CBFS section (and IFD configuration) assumes 12MB flash,
so the final 4MB will be left unflashed on installation,
after blanking the private flash. The coreboot documents have
more information about this.
Some minor design changes in lbmk were made, to accomodate
this port:
Support for extracting refcode binaries added (pulled from
Google recovery images). The refcode file is an ELF that
initialises the MRC and the PCH. It is also responsible for
enabling or disabling the Intel GbE device, where Google
does not enable it, but lbmk modifies it per the instructions
on the coreboot documentation, so as to enable Intel GbE.
Google's recovery image stores the refcode as a stage file,
but coreboot changed the format (for CBFS files) after 4.13
so coreboot 4.13's cbfstool is used to extract refcode. This
realisation made me also change the script logic to use a
cbfstool and ifdtool version matching the coreboot tree, for
all parts of lbmk, whereas lbmk previously used only the
default tree for cbfstool/ifdtool, on insertion and deletion
of vendor files - it was 81dc20e744 that broke extraction of
refcode on google's recovery images, where google used an older
version of cbfstool to insert the files in their coreboot ROMs.
A further backported patch has been added, copying coreboot
revision f22f408956 which is a build fix from Nico Huber.
Iru Cai submitted an ACPI bugfix after the revision lbmk
currently uses, for coreboot/default, and this fix is
needed for rebooting to work on Linux 6.1 or higher. This
patch has been backported to lbmk, while it still uses the
same October 2023 revision of coreboot.
Broadwell MRC is inserted at the same offset as Haswell,
so I didn't need to tweak that.
Signed-off-by: Leah Rowe <leah@libreboot.org>
2024-01-07 13:25:33 +00:00
|
|
|
[ -n "$CONFIG_HAVE_REFCODE_BLOB" ] && \
|
|
|
|
inject "fallback/refcode" "$CONFIG_REFCODE_BLOB_FILE" "stage"
|
2023-08-21 18:41:49 +00:00
|
|
|
[ "${CONFIG_HAVE_ME_BIN}" = "y" ] && \
|
2023-10-14 05:30:31 +00:00
|
|
|
inject "IFD" "${CONFIG_ME_BIN_PATH}" "me"
|
2023-08-21 18:41:49 +00:00
|
|
|
[ "${CONFIG_KBC1126_FIRMWARE}" = "y" ] && \
|
2023-10-14 05:30:31 +00:00
|
|
|
inject "ecfw1.bin" "$CONFIG_KBC1126_FW1" "raw" \
|
|
|
|
"${CONFIG_KBC1126_FW1_OFFSET}" && \
|
|
|
|
inject "ecfw2.bin" "$CONFIG_KBC1126_FW2" "raw" \
|
|
|
|
"${CONFIG_KBC1126_FW2_OFFSET}"
|
2023-12-22 13:05:32 +00:00
|
|
|
[ -n "$CONFIG_VGA_BIOS_FILE" ] && [ -n "$CONFIG_VGA_BIOS_ID" ] && \
|
2023-10-14 05:30:31 +00:00
|
|
|
inject "pci${CONFIG_VGA_BIOS_ID}.rom" \
|
|
|
|
"${CONFIG_VGA_BIOS_FILE}" "optionrom"
|
2023-08-21 18:41:49 +00:00
|
|
|
[ "${CONFIG_INCLUDE_SMSC_SCH5545_EC_FW}" = "y" ] && \
|
2023-12-16 07:56:26 +00:00
|
|
|
[ -n "${CONFIG_SMSC_SCH5545_EC_FW_FILE}" ] && \
|
2023-10-14 05:30:31 +00:00
|
|
|
inject "sch5545_ecfw.bin" "$CONFIG_SMSC_SCH5545_EC_FW_FILE" raw
|
2023-10-15 10:22:43 +00:00
|
|
|
[ "${modifygbe}" = "true" ] && ! [ "${release}" = "y" ] && \
|
2023-10-14 05:30:31 +00:00
|
|
|
inject "IFD" "${CONFIG_GBE_BIN_PATH}" "GbE"
|
much, much stricter, more verbose error handling
lbmk is much more likely to crash now, in error conditions,
which is a boon for further auditing.
also: in "fetch", remove the downloaded program
if fail() was called.
this would also be done for gnulib, when downloading
grub, but done in such a way that gnulib goes first.
where calls to err write "ERROR" in the string, they
no longer say "ERROR" because the "err" function itself
now does that automatically.
also: listmodes/listoptions (in "lbmk") now reports an
error if no scripts and/or directories are found.
also: where a warning is given, but not an error, i've
gone through in some places and redirected the output
to stderr, not stdout
as part of error checks: running anything as root, except
for the "./build dependencies *" commands, is no longer
permitted and lbmk will throw an error
mrc downloads: debugfs output no longer redirected to /dev/null,
and stderr no longer redirected to stdout. everything is verbose.
certain non-error states are also more verbose. for example,
patch_rom in blobs/inject will now state when injection succeeds
certain actual errors(bugs) were fixed:
for example, build/release/roms now correctly prepares the blobs
hash files for a given target, containing only the files and
checksums in the list. Previously, a printf message was included.
Now, with this new code: blobutil/inject rightly verifies hashes.
doing all of this in one giant patch is cleaner
than 100 patches changing each file. even this is yet part
of a much larger audit going on in the Libreboot project.
Signed-off-by: Leah Rowe <leah@libreboot.org>
2023-08-24 19:19:41 +00:00
|
|
|
|
|
|
|
printf "ROM image successfully patched: %s\n" "${rom}"
|
2023-05-14 04:42:59 +00:00
|
|
|
}
|
|
|
|
|
2023-10-14 05:30:31 +00:00
|
|
|
inject()
|
2023-05-14 04:42:59 +00:00
|
|
|
{
|
2023-10-14 05:30:31 +00:00
|
|
|
[ $# -lt 3 ] && \
|
|
|
|
err "inject $@, $rom: usage: inject name path type (offset)"
|
|
|
|
|
|
|
|
eval "$(setvars "" cbfsname _dest _t _offset)"
|
|
|
|
cbfsname="${1}"
|
|
|
|
_dest="${2##*../}"
|
|
|
|
_t="${3}"
|
|
|
|
[ $# -gt 3 ] && _offset="-b ${4}" && [ -z "${4}" ] && \
|
|
|
|
err "inject $@, $rom: offset passed, but empty (not defined)"
|
|
|
|
|
|
|
|
[ -z "${_dest}" ] && err "inject $@, ${rom}: empty destination path"
|
|
|
|
[ ! -f "${_dest}" ] && [ "${nukemode}" != "nuke" ] && \
|
|
|
|
err "inject_${dl_type}: file missing, ${_dest}"
|
|
|
|
|
|
|
|
[ "$nukemode" = "nuke" ] || \
|
2023-12-28 16:34:45 +00:00
|
|
|
printf "Inserting %s/%s in file: %s\n" "$cbfsname" "$_t" "$rom"
|
2023-10-14 05:30:31 +00:00
|
|
|
|
|
|
|
if [ "${_t}" = "GbE" ]; then
|
|
|
|
x_ mkdir -p tmp
|
2023-10-20 03:10:50 +00:00
|
|
|
cp "${_dest}" "tmp/gbe.bin" || \
|
|
|
|
err "inject: !cp \"${_dest}\" \"tmp/gbe.bin\""
|
2023-10-14 05:30:31 +00:00
|
|
|
_dest="tmp/gbe.bin"
|
2023-10-20 03:10:50 +00:00
|
|
|
"${nvmutil}" "${_dest}" setmac "${new_mac}" || \
|
|
|
|
err "inject ${_dest}: can't change mac address"
|
2023-10-14 02:30:52 +00:00
|
|
|
fi
|
2023-10-14 05:30:31 +00:00
|
|
|
if [ "${cbfsname}" = "IFD" ]; then
|
|
|
|
if [ "${nukemode}" != "nuke" ]; then
|
2023-12-24 09:04:36 +00:00
|
|
|
"$ifdtool" -i ${_t}:${_dest} "$rom" -O "$rom" || \
|
2023-10-20 03:10:50 +00:00
|
|
|
err "inject: can't insert $_t ($dest) into $rom"
|
2023-10-14 05:30:31 +00:00
|
|
|
else
|
2023-12-24 09:04:36 +00:00
|
|
|
"$ifdtool" --nuke $_t "$rom" -O "$rom" || \
|
|
|
|
err "inject $rom: can't nuke $_t in IFD"
|
2023-10-14 05:30:31 +00:00
|
|
|
fi
|
2023-10-14 02:30:52 +00:00
|
|
|
else
|
2023-10-14 05:30:31 +00:00
|
|
|
if [ "${nukemode}" != "nuke" ]; then
|
NEW MAINBOARD: HP EliteBook 820 G2
This is of Broadwell platform, one generation above Haswell.
Of note: this uses HP Sure Start. Although the flash is 16MB,
our CBFS section (and IFD configuration) assumes 12MB flash,
so the final 4MB will be left unflashed on installation,
after blanking the private flash. The coreboot documents have
more information about this.
Some minor design changes in lbmk were made, to accomodate
this port:
Support for extracting refcode binaries added (pulled from
Google recovery images). The refcode file is an ELF that
initialises the MRC and the PCH. It is also responsible for
enabling or disabling the Intel GbE device, where Google
does not enable it, but lbmk modifies it per the instructions
on the coreboot documentation, so as to enable Intel GbE.
Google's recovery image stores the refcode as a stage file,
but coreboot changed the format (for CBFS files) after 4.13
so coreboot 4.13's cbfstool is used to extract refcode. This
realisation made me also change the script logic to use a
cbfstool and ifdtool version matching the coreboot tree, for
all parts of lbmk, whereas lbmk previously used only the
default tree for cbfstool/ifdtool, on insertion and deletion
of vendor files - it was 81dc20e744 that broke extraction of
refcode on google's recovery images, where google used an older
version of cbfstool to insert the files in their coreboot ROMs.
A further backported patch has been added, copying coreboot
revision f22f408956 which is a build fix from Nico Huber.
Iru Cai submitted an ACPI bugfix after the revision lbmk
currently uses, for coreboot/default, and this fix is
needed for rebooting to work on Linux 6.1 or higher. This
patch has been backported to lbmk, while it still uses the
same October 2023 revision of coreboot.
Broadwell MRC is inserted at the same offset as Haswell,
so I didn't need to tweak that.
Signed-off-by: Leah Rowe <leah@libreboot.org>
2024-01-07 13:25:33 +00:00
|
|
|
if [ "$_t" = "stage" ]; then # broadwell refcode
|
|
|
|
"$cbfstool" "$rom" add-stage -f "$_dest" \
|
|
|
|
-n "$cbfsname" -t stage -c lzma
|
|
|
|
else
|
|
|
|
"$cbfstool" "$rom" add -f "$_dest" \
|
|
|
|
-n "$cbfsname" -t $_t $_offset || \
|
|
|
|
err "$rom: can't insert $_t file $_dest"
|
|
|
|
fi
|
2023-10-14 05:30:31 +00:00
|
|
|
else
|
2023-12-24 09:04:36 +00:00
|
|
|
"$cbfstool" "$rom" remove -n "$cbfsname" || \
|
|
|
|
err "inject $rom: can't remove $cbfsname"
|
2023-10-14 05:30:31 +00:00
|
|
|
fi
|
2023-10-14 02:30:52 +00:00
|
|
|
fi
|
2023-05-14 04:42:59 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
usage()
|
|
|
|
{
|
|
|
|
cat <<- EOF
|
2023-10-19 23:17:30 +00:00
|
|
|
USAGE: ./vendor inject -r [rom path] -b [boardname] -m [macaddress]
|
|
|
|
Example: ./vendor inject -r x230_12mb.rom -b x230_12mb
|
2023-05-14 04:42:59 +00:00
|
|
|
|
|
|
|
Adding a macadress to the gbe is optional.
|
|
|
|
If the [-m] parameter is left blank, the gbe will not be touched.
|
|
|
|
|
2023-10-19 23:17:30 +00:00
|
|
|
Type './vendor inject listboards' to get a list of valid boards
|
2023-05-14 04:42:59 +00:00
|
|
|
EOF
|
|
|
|
}
|
|
|
|
|
|
|
|
main $@
|