Comments (12)
tlaurion
commented on August 24, 2024
The following is linked to the linux payload being too big to be added to coreboot. How did you get around it?
Intel ACPI Component Architecture
ASL+ Optimizing Compiler version 20160318-64
Copyright (c) 2000 - 2016 Intel Corporation
coreboot toolchain v1.42 August 11th, 2016
dsdt.aml 720: Method(XDSS, 2)
Remark 2146 - ^ Method Argument is never used (Arg1)
dsdt.aml 1250: Method(POSC,3,Serialized)
Remark 2146 - ^ Method Argument is never used (Arg0)
dsdt.aml 1696: Method (_CRS, 0)
Remark 2120 - ^ Control Method should be made Serialized (due to creation of named objects within)
dsdt.aml 2044: Method(BINF, 2, NotSerialized)
Remark 2120 - ^ Control Method should be made Serialized (due to creation of named objects within)
ASL Input: dsdt.aml - 2726 lines, 48015 bytes, 1201 keywords
AML Output: dsdt.aml - 13847 bytes, 476 named objects, 725 executable opcodes
Compilation complete. 0 Errors, 0 Warnings, 4 Remarks, 394 Optimizations, 2 Constants Folded
IASL build/dsdt.aml disassembled correctly.
HOSTCC cbfstool/cbfstool.o
HOSTCC cbfstool/compress.o
HOSTCC cbfstool/cbfs_image.o
HOSTCC cbfstool/cbfs-mkstage.o
HOSTCC cbfstool/cbfs-mkpayload.o
HOSTCC cbfstool/fit.o
HOSTCC cbfstool/partitioned_file.o
HOSTCC cbfstool/cbfs.o
HOSTCC cbfstool/fsp_relocate.o
HOSTCC cbfstool/lz4_wrapper.o
HOSTCC cbfstool/mem_pool.o
HOSTCC cbfstool/region.o
HOSTCC cbfstool/lzma.o
HOSTCC cbfstool/LzFind.o
HOSTCC cbfstool/LzmaDec.o
HOSTCC cbfstool/LzmaEnc.o
HOSTCC cbfstool/2sha_utility.o
HOSTCC cbfstool/2sha1.o
HOSTCC cbfstool/2sha256.o
HOSTCC cbfstool/2sha512.o
HOSTCC cbfstool/lz4.o
HOSTCC cbfstool/lz4hc.o
HOSTCC cbfstool/lz4frame.o
HOSTCC cbfstool/xxhash.o
HOSTCC cbfstool/linux_trampoline.o
HOSTCC cbfstool/cbfs-payload-linux.o
HOSTCC cbfstool/cbfstool (link)
Performing operation on 'COREBOOT' region...
Created CBFS (capacity = 4194008 bytes)
Performing operation on 'COREBOOT' region...
Performing operation on 'COREBOOT' region...
CBFS fallback/romstage
Performing operation on 'COREBOOT' region...
E: Input file size (102404) greater than page size (65536).
CBFS mrc.cache
Performing operation on 'COREBOOT' region...
CBFS cpu_microcode_blob.bin
Performing operation on 'COREBOOT' region...
CBFS fallback/ramstage
Performing operation on 'COREBOOT' region...
CBFS cmos.default
Performing operation on 'COREBOOT' region...
CBFS cmos_layout.bin
Performing operation on 'COREBOOT' region...
CBFS fallback/dsdt.aml
Performing operation on 'COREBOOT' region...
CBFS fallback/payload
Performing operation on 'COREBOOT' region...
E: The stage file is not in ELF format!
E: Could not add [./bzImage, 4083201 bytes (3987 KB)@0x0]; too big?
E: Failed to add './bzImage' into ROM image.
E: Failed while operating on 'COREBOOT' region!
E: The image will be left unmodified.
Makefile.inc:779: recipe for target 'build/coreboot.pre' failed
make[1]: *** [build/coreboot.pre] Error 1
make[1]: Leaving directory '/media/user/7725fbc2-4b2d-4ade-9f1d-337ad8203dc3/Qubes/heads/build/coreboot-git'
Makefile:87: recipe for target '/media/user/7725fbc2-4b2d-4ade-9f1d-337ad8203dc3/Qubes/heads/build/coreboot-git/build/coreboot.rom' failed
from heads.
osresearch
commented on August 24, 2024
I'm doing a clean checkout build now to figure out why the initrd/bin directory isn't created. The Makefile is supposed to create the directory if it doesn't exist:
#
# Install a file into the initrd, if it changed from
# the destination file.
#
define install =
@if [ ! -d "$(dir $$@)" ]; \
then mkdir "$(dir $$@)"; \
fi
cmp --quiet "$$@" "$$<" || \
cp -a "$$<" "$$@"
endef
from heads.
osresearch
commented on August 24, 2024
How large is your Linux kernel and initrd? The resulting bzImage on ubuntu 15.10 is 3.7MB, so there is only a few hundred KB of spare room in the 4MB image.
In other news, I've managed to reduce the space the ME uses in the second SPI flash chip, which will free up 6MB for the coreboot payload. One thought is to use the 4 MB one as a write-protected recovery image and store the real kernel in the lower one.
from heads.
osresearch
commented on August 24, 2024
On the directory error, $(dir $$@) was not expanding correctly. I've just pushed a fix for it. 24ef1e1
from heads.
tlaurion
commented on August 24, 2024
Have a missing libqrencode.so.3 when building linux (currently rebuilding, will repost with results about size when done, but precedent post has sizes in bold)
[...]
find initrd/bin -type f -a ! -name '*.sh' -print0 \
| xargs -0 strip
./populate-lib \
./initrd/lib/x86_64-linux-gnu/ \
initrd/bin/* \
initrd/sbin/* \
initrd/bin/qrenc: 'libqrencode.so.3' libqrencode.so.3 => not found
[...]
in heads root:
ls initrd.cpio -al
-rw-rw-r-- 1 user user 6701280 Sep 26 17:32 initrd.cpio
from heads.
osresearch
commented on August 24, 2024
That is quite a bit larger than mine.
diamond:~/build/clean/heads: ls -Fla initrd.cpio
-rw-rw-r-- 1 hudson hudson 5658120 Sep 26 17:01 initrd.cpio
The qrencode looks like a ldd mistake on my part -- it is linked against ../qrencode-3.4.4/.libs/libqrencode.so, but the initrd building pulls in mine from /usr/lib.
To help track down what files are so large in your initrd, here is what I end up with in a clean checkout:
-rw-rw-r-- 1 hudson hudson 110 Sep 26 15:37 ./initrd/.ash_history
-rwxrwxr-x 1 hudson hudson 10376 Sep 26 17:01 ./initrd/bin/base32*
-rwxr-xr-x 1 hudson hudson 306992 Sep 26 17:01 ./initrd/bin/busybox*
-rwxrwxr-x 1 hudson hudson 23240 Sep 26 17:01 ./initrd/bin/cbmem*
-rwxrwxr-x 1 hudson hudson 10400 Sep 26 17:01 ./initrd/bin/clearown*
-rwxrwxr-x 1 hudson hudson 10432 Sep 26 17:01 ./initrd/bin/counter_create*
-rwxrwxr-x 1 hudson hudson 10416 Sep 26 17:01 ./initrd/bin/counter_increment*
-rwxrwxr-x 1 hudson hudson 10392 Sep 26 17:01 ./initrd/bin/counter_read*
-rwxrwxr-x 1 hudson hudson 10432 Sep 26 17:01 ./initrd/bin/counter_release*
-rwxr-xr-x 1 hudson hudson 78752 Sep 26 17:01 ./initrd/bin/dmsetup*
-rwxrwxr-x 1 hudson hudson 6280 Sep 26 17:01 ./initrd/bin/forceclear*
-rwxrwxr-x 1 hudson hudson 22832 Sep 26 17:01 ./initrd/bin/getcapability*
-rwxr-xr-x 1 hudson hudson 372400 Sep 26 17:01 ./initrd/bin/gpgv*
-rwxrwxr-x 1 hudson hudson 167144 Sep 26 17:01 ./initrd/bin/kexec*
-rwxrwxr-x 1 hudson hudson 10464 Sep 26 17:01 ./initrd/bin/nv*
-rwxrwxr-x 1 hudson hudson 14544 Sep 26 17:01 ./initrd/bin/nv_definespace*
-rwxrwxr-x 1 hudson hudson 14584 Sep 26 17:01 ./initrd/bin/nv_readvalue*
-rwxrwxr-x 1 hudson hudson 14560 Sep 26 17:01 ./initrd/bin/nv_writevalue*
-rwxrwxr-x 1 hudson hudson 6296 Sep 26 17:01 ./initrd/bin/pcrreset*
-rwxrwxr-x 1 hudson hudson 6280 Sep 26 17:01 ./initrd/bin/physicaldisable*
-rwxrwxr-x 1 hudson hudson 6280 Sep 26 17:01 ./initrd/bin/physicalenable*
-rwxrwxr-x 1 hudson hudson 14496 Sep 26 17:01 ./initrd/bin/physicalpresence*
-rwxrwxr-x 1 hudson hudson 6272 Sep 26 17:01 ./initrd/bin/physicalsetdeactivated*
-rwxrwxr-x 1 hudson hudson 6264 Sep 26 17:01 ./initrd/bin/qrenc*
-rwxrwxr-x 1 hudson hudson 18688 Sep 26 17:01 ./initrd/bin/sealfile2*
-rwxrwxr-x 1 hudson hudson 1238 Sep 26 15:24 ./initrd/bin/sealtotp.sh*
-rwxrwxr-x 1 hudson hudson 14544 Sep 26 17:01 ./initrd/bin/takeown*
-rwxrwxr-x 1 hudson hudson 10408 Sep 26 17:01 ./initrd/bin/totp*
-rwxrwxr-x 1 hudson hudson 10432 Sep 26 17:01 ./initrd/bin/unsealfile*
-rwxrwxr-x 1 hudson hudson 426 Sep 26 15:24 ./initrd/bin/unsealtotp.sh*
-rwxrwxr-x 1 hudson hudson 1095 Sep 26 15:37 ./initrd/init*
-rwxrwxr-x 1 hudson hudson 154376 Sep 26 17:01 ./initrd/lib/x86_64-linux-gnu/ld-linux-x86-64.so.2*
-rwxrwxr-x 1 hudson hudson 66800 Sep 26 17:01 ./initrd/lib/x86_64-linux-gnu/libbz2.so.1.0*
-rwxrwxr-x 1 hudson hudson 1869392 Sep 26 17:01 ./initrd/lib/x86_64-linux-gnu/libc.so.6*
-rwxrwxr-x 1 hudson hudson 294496 Sep 26 17:01 ./initrd/lib/x86_64-linux-gnu/libdevmapper.so.1.02.1*
-rwxrwxr-x 1 hudson hudson 14592 Sep 26 17:01 ./initrd/lib/x86_64-linux-gnu/libdl.so.2*
-rwxrwxr-x 1 hudson hudson 137400 Sep 26 17:01 ./initrd/lib/x86_64-linux-gnu/liblzma.so.5*
-rwxrwxr-x 1 hudson hudson 378336 Sep 26 15:25 ./initrd/lib/x86_64-linux-gnu/libmbedcrypto.so.0*
-rwxrwxr-x 1 hudson hudson 448440 Sep 26 17:01 ./initrd/lib/x86_64-linux-gnu/libpcre.so.3*
-rwxrwxr-x 1 hudson hudson 142080 Sep 26 17:01 ./initrd/lib/x86_64-linux-gnu/libpthread.so.0*
-rwxrwxr-x 1 hudson hudson 47240 Sep 26 17:01 ./initrd/lib/x86_64-linux-gnu/libqrencode.so.3*
-rwxrwxr-x 1 hudson hudson 101232 Sep 26 17:01 ./initrd/lib/x86_64-linux-gnu/libresolv.so.2*
-rwxrwxr-x 1 hudson hudson 31680 Sep 26 17:01 ./initrd/lib/x86_64-linux-gnu/librt.so.1*
-rwxrwxr-x 1 hudson hudson 138400 Sep 26 17:01 ./initrd/lib/x86_64-linux-gnu/libselinux.so.1*
-rwxrwxr-x 1 hudson hudson 368672 Sep 26 15:25 ./initrd/lib/x86_64-linux-gnu/libtpm.so*
-rwxrwxr-x 1 hudson hudson 122624 Sep 26 17:01 ./initrd/lib/x86_64-linux-gnu/libudev.so.1*
-rwxrwxr-x 1 hudson hudson 104824 Sep 26 17:01 ./initrd/lib/x86_64-linux-gnu/libz.so.1*
-rwxrwxr-x 1 hudson hudson 1373 Sep 26 15:37 ./initrd/mount-boot*
-rwxrwxr-x 1 hudson hudson 519 Sep 26 15:37 ./initrd/start-xen*
-rw-rw-r-- 1 hudson hudson 17559 Sep 26 15:37 ./initrd/trustedkeys.gpg
from heads.
tlaurion
commented on August 24, 2024
That's the closest I was able to get from your output with du.
du -ah ./initrd | sort -h
4.0K ./initrd/bin/sealtotp.sh
4.0K ./initrd/bin/unsealtotp.sh
4.0K ./initrd/init
4.0K ./initrd/lib64
4.0K ./initrd/mount-boot
4.0K ./initrd/sbin
4.0K ./initrd/start-xen
8.0K ./initrd/bin/forceclear
8.0K ./initrd/bin/pcrreset
8.0K ./initrd/bin/physicaldisable
8.0K ./initrd/bin/physicalenable
8.0K ./initrd/bin/physicalsetdeactivated
8.0K ./initrd/bin/qrenc
12K ./initrd/bin/base32
12K ./initrd/bin/clearown
12K ./initrd/bin/counter_create
12K ./initrd/bin/counter_increment
12K ./initrd/bin/counter_read
12K ./initrd/bin/counter_release
12K ./initrd/bin/nv
12K ./initrd/bin/totp
12K ./initrd/bin/nv
12K ./initrd/bin/totp
12K ./initrd/bin/unsealfile
16K ./initrd/bin/nv_definespace
16K ./initrd/bin/nv_readvalue
16K ./initrd/bin/nv_writevalue
16K ./initrd/bin/physicalpresence
16K ./initrd/bin/takeown
16K ./initrd/lib/x86_64-linux-gnu/libdl.so.2
20K ./initrd/bin/sealfile2
20K ./initrd/trustedkeys.gpg
24K ./initrd/bin/cbmem
24K ./initrd/bin/getcapability
32K ./initrd/lib/x86_64-linux-gnu/librt.so.1
68K ./initrd/lib/x86_64-linux-gnu/libbz2.so.1.0
100K ./initrd/lib/x86_64-linux-gnu/libresolv.so.2
104K ./initrd/lib/x86_64-linux-gnu/libz.so.1
124K ./initrd/lib/x86_64-linux-gnu/libudev.so.1
128K ./initrd/lib/x86_64-linux-gnu/libselinux.so.1
136K ./initrd/lib/x86_64-linux-gnu/libpthread.so.0
148K ./initrd/bin/dmsetup
160K ./initrd/bin/kexec
160K ./initrd/lib/x86_64-linux-gnu/ld-linux-x86-64.so.2
300K ./initrd/bin/busybox
356K ./initrd/lib/x86_64-linux-gnu/libdevmapper.so.1.02.1
364K ./initrd/bin/gpgv
364K ./initrd/lib/x86_64-linux-gnu/libtpm.so
372K ./initrd/lib/x86_64-linux-gnu/libmbedcrypto.so.0
448K ./initrd/lib/x86_64-linux-gnu/libpcre.so.3
1.1M ./initrd/lib/x86_64-linux-gnu/libm.so.6
1.3M ./initrd/bin
1.8M ./initrd/lib/x86_64-linux-gnu/libc.so.6
5.2M ./initrd/lib
5.2M ./initrd/lib/x86_64-linux-gnu
6.5M ./initrd
from heads.
tlaurion
commented on August 24, 2024
I seem to have the file /initrd/lib/x86_64-linux-gnu/libm.so.6 that you don't have.
from heads.
osresearch
commented on August 24, 2024
I used find ./initrd -type f -print0 | xargs -0 ls -Fla to generate my list.
Can you tell which binary is depending on libm? We really don't have space for another 1MB library... ldd initrd/bin/* should tell you which one brought it in.
from heads.
osresearch
commented on August 24, 2024
I built on fedora-23 (qubes 3.2) and found that dmsetup is linked against libm.so.6 and that is making the initrd too large. Commenting it out in the makefile shrank initrd+bzImage from 4.5MB to 4.1MB; cleaning the lib directory by running rm initrd/lib/x86_64-linux-gnu/* removed the extra libraries (libelf, libdw, libcap, etc). And now it fits into 3.3MB compressed.
--- a/Makefile
+++ b/Makefile
@@ -151,7 +151,7 @@ $(build)/$(coreboot_dir)/util/cbmem/cbmem: $(build)/$(coreboot_dir)/.canary
# Mounting dm-verity file systems requires dm-verity to be installed
# We use gpgv to verify the signature on the root h ash.
# Both of these should be brought in as modules instead of from /sbin
-initrd_bins += initrd/bin/dmsetup
+#initrd_bins += initrd/bin/dmsetup
initrd/bin/dmsetup: /sbin/dmsetup
cp "$<" "$@"
initrd_bins += initrd/bin/gpgv
from heads.
tlaurion
commented on August 24, 2024
Bingo. x230.rom exists :)
What about x230.me.bin ?
from heads.
osresearch
commented on August 24, 2024
Since the dm-verity setup (issue #6) isn't ready, you should be able to build without dmsetup and have an image to test with. Good luck!
from heads.
Related Issues (20)
- Revise CircleCI caching strategies
- Interestingly enough, it's flashrom now that is not reproducible between local builds and CircleCI with nix docker image HOT 1
- nitropad-nv41 v0.2.0-2147-g1e583e0 rom kernel panic HOT 8
- Nix buildstack next step : have and use qemu-canokey non-existing pre-compiled qemu/kvm in docker image
- Enable Lenovo Thinkpad T530-unmaintained-build for testing and reporting functionality of build HOT 4
- OEM Factory Reset GPG user authentication: key_to_card fails with « invalid time » HOT 5
- bash: line 1: /home/user/heads/crossgcc/x86/bin/x86_64-linux-musl-gcc: No such file or directory HOT 2
- Bring colors to the shell console "'a la" systemd : "[...] doing action" -> "[OK] action" (green) || "[FAILED] action" (red) HOT 5
- Bring back Lenovo Thinkpad T530 to supported/tested - images are working fine HOT 5
- t420 (display port does not work after external flashing) HOT 11
- Review CONFIG_ONBOARD_VGA_IS_PRIMARY enablement for dGPU enabled boards HOT 2
- Merge heads-wiki repo into a heads-wiki subdir of this project to render at osresearch.net and archive for reference? (long-term plan) HOT 1
- Force reboot between tpm reset+sign /boot and DUK for non-HOTP deployments ?
- Error message misleading HOT 1
- Language selection
- Programmer selection HOT 6
- Update nv41/ns50 intel_iommu=igfx_off (librem 11 = GOP igfx)
- Switch all Linux config schedulers out of performance schedulers
- Remove unneeded iotools
- nv41/ns50 boards: rebrand nitrokey boards as novacustom HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from heads.