Previously, if the builder exported $VERSION in their environment (as
past Gitian-building docs told them to), but their HEAD does not
actually point to v$VERSION, their build outputs will differ from those
of other builders.
This is because the contrib/guix/guix-* scripts only ever act on the
current git worktree, and does not try to check out $VERSION if $VERSION
is set in the environment.
Setting $VERSION only makes the scripts pretend like the current
worktree is $VERSION.
This problem was seen in jonatack's attestation for all.SHA256SUMS,
where only his bitcoin-22.0rc3-osx-signed.dmg differed from everyone
else's.
Here is my deduced sequence of events:
1. Aug 27th: He guix-builds 22.0rc3 and uploads his attestations up to
guix.sigs
2. Aug 30th, sometime after POSIX time 1630310848: he pulls the latest
changes from master in the same worktree where he guix-built 22.0rc3
and ends up at 7be143a960
3. Aug 30th, sometime before POSIX time 1630315907: With his worktree
still on 7be143a960, he guix-codesigns. Normally, this would result
in outputs going in guix-build-7be143a960e2, but he had
VERSION=22.0rc3 in his environment, so the guix-* scripts pretended
like he was building 22.0rc3, and used 22.0rc3's guix-build directory
to locate un-codesigned outputs and dump codesigned ones.
However, our SOURCE_DATE_EPOCH defaults to the POSIX time of HEAD
(7be143a960), which made all timestamps in the resulting codesigned
DMG 1630310848, 7be143a960e2's POSIX timestamp. This differs from the
POSIX timestamp of 22.0rc3, which is 1630348517. Note that the
windows codesigning procedure does not consider SOURCE_DATE_EPOCH.
We resolve this by only allowing VERSION overrides via the FORCE_VERSION
environment variable.
The SHA256SUMS file can be used in a sha256sum -c command to verify
downloaded binaries. However users are likely to download just a single
file and not place this file in the correct directory relative to the
SHA256SUMS file for the simple verification command to work. By not
including the directory name in the SHA256SUMS file, it will be easier
for users to verify downloaded binaries.
Co-authored-by: Carl Dong <contact@carldong.me>
90b3e482e9 release: Release with separate SHA256SUMS and sig files (Carl Dong)
Pull request description:
This allows us to:
- remove the rfc4880 EOL hacks, and
- release with a SHA256SUMS.asc file that's a combination of all signer signatures
ACKs for top commit:
achow101:
ACK 90b3e482e9
laanwj:
Concept and code review ACK 90b3e482e9
Tree-SHA512: 5d5086063d303aa0cbd590e5fdf2ae8f555e25f4e43bf67545e33384449b990e94834c711622530ad0eb3dcc83f52746884a5081dadb0acff8dd799cfadafac7
9b313dfef1 guix: Ensure EPOCH_SOURCE_DATE does not include GPG information (Andrew Chow)
43225f0a2a guix: Remove extra \r from all.SHA256SUMS line ending (Andrew Chow)
d080c27066 guix, doc: Add a note that codesigners need to rebuild after tagging (Andrew Chow)
4a466388a0 guix: Allow changing the base manifest in guix-verify (Andrew Chow)
33455c7696 guix: Make all.SHA256SUMS rather than codesigned.SHA256SUMS (Andrew Chow)
Pull request description:
`guix-verify` expects `all.SHA256SUMS` but `guix-attest` produces `codesigned.SHA256SUMS`. Since `all.SHA256SUMS` makes more sense (as the file contains all the sha256sums, not just the codesigned ones), `guix-attest` has been changed to output a file of that name.
As a quality of life improvement, `guix-verify` can take `SIGNER` and use the signer's manifest as the base to compare against. This makes it easier to compare a single person's attestations with everyone else's and can make it more obvious when one builder is clearly mismatching with everyone else.
Lastly `release-process.md` is updated with a note about a gotcha that can cause a mismatch in the codesigned attestation.
ACKs for top commit:
fanquake:
ACK 9b313dfef1
Tree-SHA512: 0d60627def38288dbd3059ad1e72cad224f9205da11b1a561c082ef28250a074df5cc5f2797c91a7be027bc486a3fda3319c2e496a8724e5b539337236c6f990
If the user has set log.showSignature=true in their git config, then the
git log will always output GPG signature information. Since git log is
used to set EPOCH_SOURCE_DATE, this will mistakenly have GPG signature
information in it which causes issues for the build. To avoid this
issue, we override the config and force log.showSignature=false.
guix-attest mistakenly added an extra \r to the line endings in
all.SHA256SUMS, causing guix-verify to erroneously fail.
Co-Authored-By: Carl Dong <contact@carldong.me>
9f01feda0a guix/build: Remove vestigial SKIPATTEST.TAG (Carl Dong)
Pull request description:
No longer needed or referenced by anything. A relic from prior to the great hierarchy overhaul of #22182
ACKs for top commit:
achow101:
ACK 9f01feda0a
fanquake:
ACK 9f01feda0a
Tree-SHA512: a94cf63f0c5cb8dbacf1025b6c0e81b219c2a3c93b3cbcefc239ccde29e602ecd4b717b1d93dbe53cb791a5017236fb09823c034aec42b0c31894fc9e0ab8b21
e6a94d4446 guix: Bump to version-1.3.0 from upstream (Carl Dong)
90fd13b954 guix: Pin kernel header version (Carl Dong)
Pull request description:
```
- Use 4.19 for riscv64 (earliest LTS release w/ riscv64 support)
- Use 4.9 for all others (second-oldest LTS release, released in
combination with glibc glibc 2.24 in Debian stretch)
```
```
The chosen commit is the HEAD of Guix's version-1.3.0 branch as of July
15th, 2021.
Also fix visual indenting.
```
-----
This + the documentation PR should make our Guix system ready for release!
ACKs for top commit:
MarcoFalke:
review ACK e6a94d4446 to change to vanilla guix. Did not review the kernel change.
laanwj:
ACK e6a94d4446
fanquake:
ACK e6a94d4446
Tree-SHA512: a175e4ddb3ee786a39f5e800ce336932ad2f6797a3a28400a6f723875d0f19833fd36cedc41b3580e4604110517211bd9f557be36adf7265fd8e591c434ae032
fac4814106 doc/release-process: Add torrent creation details (Carl Dong)
5d24cc3d82 guix/INSTALL: Guix installs init scripts in libdir (Carl Dong)
5da2ee49d5 guix/INSTALL: Add coreutils/inotify-dir-recreate troubleshooting (Carl Dong)
318c60700b guix: Adapt release-process.md to new Guix process (Carl Dong)
fcab35b229 guix-attest: Produce and sign normalized documents (Carl Dong)
c2541fd0ca guix: Overhaul README (Carl Dong)
46ce6ce378 tree-wide: Rename gitian-keys to builder-keys (Carl Dong)
fc4f8449f3 guix: Update various check_tools lists (Carl Dong)
263220a85c guix: Check for a sane services database (Carl Dong)
Pull request description:
Based on: #21462
Keeping the README in one file so that it's easy to search through. Will add more jumping links later so navigation is easier.
Current TODOs:
- [x] Shell installer option: prompt user to re-login for `/etc/profile.d` entry to be picked up
- [x] Binary tarball option: prompt user to create `/etc/profile.d` entry and re-login
- [x] Fanquake docker option: complete section
- [x] Arch Linux AUR option: prompt to start `guix-daemon-latest` unit after finishing "optional setup" section
- [x] Building from source option: Insert dependency tree diagram that I made
- [x] Building from source option: redo sectioning, kind of a mess right now
- [x] Optional setup: make clear which parts are only needed if building from source
- [x] Workaround 1 for GnuTLS: perhaps mention how to remove Guix build farm's key
- [x] Overall (after everything): Make the links work.
Note to self: wherever possible, tell user how to check that something is true rather than branching by installation option.
ACKs for top commit:
fanquake:
ACK fac4814106 - going to go ahead and merge this now. It's a lot of documentation, and could probably be nit-picked / improved further, however, that can continue over the next few weeks. I'm sure more (backportable) improvements / clarifications will be made while we progress through RCs towards a new release.
Tree-SHA512: dc46c0ecdfc67c7c7743ca26e4a603eb3f54adbf81be2f4c1f4c20577ebb84b5250b9c9ec89c0e9860337ab1c7cff94d7963c603287267deecfe1cd987fa070a
That way we can easily combine the document and detached signature to
produce cleartext signature files for upload during the release process.
See subsequent commits which modify doc/release-process.md for more
details.
1edddf5de4 Avoid GCC 7.1 ABI change warning in guix build (Pieter Wuille)
Pull request description:
The arm-linux-gnueabihf guix build output is littered with warnings like:
```
/gnu/store/7a96hdqdb2qi8a39f09n84xjy2hr23rs-gcc-cross-arm-linux-gnueabihf-8.4.0/include/c++/bits/stl_vector.h:1085:4: note:
parameter passing for argument of type '__gnu_cxx::__normal_iterator<CRecipient*, std::vector<CRecipient> >' changed in GCC 7.1
```
These are irrelevant for us. Disable them using `-Wno-psabi`.
ACKs for top commit:
laanwj:
ACK 1edddf5de4
hebasto:
ACK 1edddf5de4, after thorough reading related materials, I agree this change can be merged. As I mentioned above, I have been compiling my arm-32bit binaries with `-Wno-psabi` flag for two years, and no related flaws were observed.
Tree-SHA512: 485c7500547ac5da567ad23847341c18ff832607f5a1002676404cc647e437cf3445b6894ecff5b52929ca52bea946c06bd90eace1997c895e56204e787065e4
- Use 4.19 for riscv64 (earliest LTS release w/ riscv64 support)
- Use 4.9 for all others (second-oldest LTS release, released in
combination with glibc glibc 2.24 in Debian stretch)
On bare systems, it is possible to be lacking a services database. Check
for basic entries before attempting a build.
See the error message in the diff for more context.
We use these flags in our test-security-check make target, but they are
only available because debian patches them in.
We can patch them in for our Guix builds so that we can check the sanity
of our security/symbol checking suite before running them.
Now that our Guix builds are performed on glibc 2.24 and 2.27 (RISCV),
we no-longer need to pass the --enable-glibc-back-compat option.
Replace it with --disable-threadlocal, to prevent the usage of symbols
from glibc 2.18.
None of the binaries produced required symbols later than 2.17, and 2.27
(RISCV).
Our 'bitcoin-linux-g++' definition better integrates with our depends
system than the stock linux-g++-64 definition.
This fixes a bug whereby Guix builds on x86_64 for x86_64 did not
produce a QMinimalIntegrationPlugin and led to bitcoin-qt not being
built.
Support for riscv64 in glibc landed in 2.27 so it's unavoidable that we
use 2.27.
Running a Bitcoin build with toolchains based on 2.24 for platforms
other than riscv64 seem to produce binaries which do not have 2.17
symbols. So use 2.24 since it's more recent and maintained by Debian
Stretch.
aa80b5759d scripts: check macOS SDK version is set (fanquake)
c972345bac scripts: check minimum required Windows version is set (fanquake)
29615aef52 scripts: check minimum required macOS vesion is set (fanquake)
8732f7b6c9 scripts: LIEF 0.11.5 (fanquake)
Pull request description:
macOS:
We use a compile flag ([-mmacosx-version-min=10.14](https://github.com/bitcoin/bitcoin/blob/master/depends/hosts/darwin.mk#L96)) to set the minimum required version of macOS needed to run our binaries. This adds a sanity check that the version is being set as expected.
Clangs Darwin driver should infer the SDK version used during compilation, and forward that through to the linker. Add a check that this has been done, and the expected SDK version is set. Should help prevent issues like #21771 in future.
Windows:
We use linker flags ([-Wl,--major/minor-subsystem-version](https://github.com/bitcoin/bitcoin/blob/master/configure.ac#L683)) to set the minimum required version of Windows needed to run our binaries. This adds a sanity check that the version is being set as expected.
Gitian builds:
```bash
# macOS:
8b6fcd61d75001c37b2af3fceb5ae09f5d2fe85e97d361f684214bd91c27954a bitcoin-f015e1c2cac9-osx-unsigned.dmg
3c1e412bc7f5a7a5d0f78e2cd84b7096831414e1304c1307211aa3e135d89bbf bitcoin-f015e1c2cac9-osx-unsigned.tar.gz
50b7b2804e8481f63c69c78e3e8a71c0d811bf2db8895dd6d3edae9c46a738ae bitcoin-f015e1c2cac9-osx64.tar.gz
fe6b5c0a550096b76b6727efee30e85b60163a41c83f21868c849fdd9876b675 src/bitcoin-f015e1c2cac9.tar.gz
8a20f21b20673dfc8c23e22b20ae0839bcaf65bf0e02f62381cdf5e7922936f0 bitcoin-core-osx-22-res.yml
# Windows:
b01fcdc2a5673387050d6c6c4f96f1d350976a121155fde3f76c2af309111f9d bitcoin-f015e1c2cac9-win-unsigned.tar.gz
b95bdcbef638804030671d2332d58011f8c4ed4c1db87d6ffd211515c32c9d02 bitcoin-f015e1c2cac9-win64-debug.zip
350bf180252d24a3d40f05e22398fec7bb00e06d812204eb5a421100a8e10638 bitcoin-f015e1c2cac9-win64-setup-unsigned.exe
2730ddabe246d99913c9a779e97edcadb2d55309933d46f1dffd0d23ecf9aae5 bitcoin-f015e1c2cac9-win64.zip
fe6b5c0a550096b76b6727efee30e85b60163a41c83f21868c849fdd9876b675 src/bitcoin-f015e1c2cac9.tar.gz
aa60d7a753e8cb2d4323cfbbf4d964ad3645e74c918cccd66862888f8646d80f bitcoin-core-win-22-res.yml
```
ACKs for top commit:
hebasto:
ACK aa80b5759d, tested by breaking tests:
Tree-SHA512: 10150219910e8131715fbfe20edaa15778387616ef3bfe1a5152c7acd3958fe8f88c74961c3d3641074eb72824680c22764bb1dc01a19e92e946c2d4962a8d2c
e2c40a4ed5 guix-attest: Error out if SHA256SUMS is unexpected (Carl Dong)
4cc35daed5 Rewrite guix-{attest,verify} for new hier (Carl Dong)
28a9c9b839 Make SHA256SUMS fragment right after build (Carl Dong)
Pull request description:
Based on: #22075
Code reviewers: I recommend reading the new `guix-{attest,verify}` files instead of trying to read the diff
The following changes resolve many usability improvements which were pointed out to me:
1. Some maintainers like to extract their "uncodesigned tarball" inside the `output/` directory, resulting in the older `guix-attest` mistakenly attesting to the extracted contents
2. Maintainers whose GPG keys reside on an external smartcard often need to physically interact with the smartcard as a way to approve the signing operation, having one signature per platform means a lot of fidgeting
3. Maintainers wishing to sign on a separate machine now has the option of transferring only a subtree of `output/`, namely `output/*/SHA256SUMS.part`, in order to perform a signature (you may need to specify an `$OUTDIR_BASE` env var)
4. An `all.SHA256SUMS` file should be usable as the base `SHA256SUMS` in bitcoin core torrents and on the release server.
For those who sign on an separate machine than the one you do builds on, the following steps will work:
1. `env GUIX_SIGS_REPO=/home/achow101/guix.sigs SIGNER=achow101 NO_SIGN=1 ./contrib/guix/guix-attest`
2. Copy `/home/achow101/guix.sigs/<tag>/achow101` (which does not yet have signatures) to signing machine
3. Sign the `SHA256SUMS` files:
```bash
for i in "<path-to-achow101>/*.SHA256SUMS"; do
gpg --detach-sign --local-user "<your-key-here>" --armor --output "$i"{.asc,}
done
```
5. Upload `<path-to-achow101>` (now with signatures) to `guix.sigs`
-----
After this change, output directories will now include a `SHA256SUMS.part` fragment, created immediately after a successful build:
```
output
└── x86_64-w64-mingw32
├── bitcoin-4e069f7589da-win64-debug.zip
├── bitcoin-4e069f7589da-win64-setup-unsigned.exe
├── bitcoin-4e069f7589da-win64.zip
├── bitcoin-4e069f7589da-win-unsigned.tar.gz
└── SHA256SUMS.part
```
These `SHA256SUMS.part` fragments look something like:
```
3ebd7262b1a0a5bb757fef1f70e7e14033c70f98c059bc4dbfee5d1992b25825 dist-archive/bitcoin-4e069f7589da.tar.gz
def2e7d3de5ab3e3f955344e75151df4f33713f9101f5295bd13c9375bdf633b x86_64-w64-mingw32/bitcoin-4e069f7589da-win64-debug.zip
643049fe3ee4a4e83a1739607e67b11b7c9b1a66208a6f35a9ff634ba795500e x86_64-w64-mingw32/bitcoin-4e069f7589da-win64-setup-unsigned.exe
a247a1ccec0ccc2e138c648284bd01f6a761f2d8d6d07d91b5b4a6670ec3f288 x86_64-w64-mingw32/bitcoin-4e069f7589da-win-unsigned.tar.gz
fab76a836dcc592e39c04fd2396696633fb6eb56e39ecbf6c909bd173ed4280c x86_64-w64-mingw32/bitcoin-4e069f7589da-win64.zip
```
Meaning that they are valid `SHA256SUMS` files when `sha256sum --check`'d at the `guix-build-*/output` directory level
When `guix-attest` is invoked, these `SHA256SUMS.part` files are combined and sorted (by `-k2`, `LC_ALL=C`) to create:
1. `noncodesigned.SHA256SUMS` for a manifest of all non-codesigned outputs, and
3. `all.SHA256SUMS` for a manifest of all outputs including non-codesigned outputs
Then both files are signed, resulting in the following `guix.sigs` hierarchy:
```
4e069f7589da/
└── dongcarl
├── all.SHA256SUMS
├── all.SHA256SUMS.asc
├── noncodesigned.SHA256SUMS
└── noncodesigned.SHA256SUMS.asc
```
ACKs for top commit:
achow101:
ACK e2c40a4ed5
hebasto:
ACK e2c40a4ed5, tested on Linux Mint 20.1 (x86_64) with and w/o `NO_SIGN=1`. Changes in `contrib/guix/libexec/codesign.sh` and `contrib/guix/guix-verify` are reviewed only.
Tree-SHA512: 618aacefb0eb6595735a9ab6a98ea6598fce65f9ccf33fa1e7ef93bf140c0f6cfc16e34870c6aa3e4777dd3f004b92a82a994141879870141742df948ec59c1f
683d197970 Use latest signapple commit (Andrew Chow)
Pull request description:
Update gitian and guix to use the same latest signapple commit.
Also changed guix to use the actual repo. The changes from the fork were incorporated upstream.
ACKs for top commit:
fanquake:
ACK 683d197970 - sanity checked that the updated package is built:
Tree-SHA512: a4981f8bbe33e6c5654632bc9b9f6f2f1e675741a19ac7296205e370f1e64a747101ecb632e0cc82a0134e4c2e9ce47b3f7b4d8c8f75f0f06dd069c078303759
108a6be92a guix: Check for disk space availability before building (Carl Dong)
d7dec89091 guix: Remove dest if OUTDIR mv fails (Carl Dong)
Pull request description:
There seems to be some corner cases that can be hit when guix scripts unexpectedly fail in the middle of operation, see: https://gnusha.org/bitcoin-builds/2021-05-24.log
- Perform an early disk space check for `guix-build`
- Overwrite existing output directory after a successful build (the existing one might be malformed), and cleanup output directory if the `mv` somehow fails
ACKs for top commit:
laanwj:
Tested ACK 108a6be92a
achow101:
ACK 108a6be92a
Tree-SHA512: cf6438317da40bf55714cd2d8cce859b3d435cc66cabefe8d4a53552d7880966acfe84ffe8fadf1c80e368ae6b037992258a6d409df85ffc6ce8bf780e98e2e5
a58868d201 build: Makes rcc output always deterministic (Hennadii Stepanov)
Pull request description:
The Qt Resource Compiler ([rcc](https://doc.qt.io/qt-5/rcc.html)) has a command-line option `--format-version` which has the [default value](https://code.qt.io/cgit/qt/qtbase.git/tree/src/tools/rcc/main.cpp?h=5.12.10#n172) 2.
The only difference from `--format-version 1` is adding a [last modified timestamp](https://code.qt.io/cgit/qt/qtbase.git/tree/src/tools/rcc/rcc.cpp?h=5.12.10#n207) to the output file ([credits](https://github.com/bitcoin/bitcoin/pull/21654#issuecomment-819198228) to **fanquake**). That, in turn, forces us to use `QT_RCC_SOURCE_DATE_OVERRIDE=1` to get deterministic builds (#13732).
This change makes rcc output always deterministic by using `--format-version 1` option that makes usage of the
`QT_RCC_SOURCE_DATE_OVERRIDE` needless.
---
Also it improves interaction with ccache:
On master (f6c44e999b):
```
$ make && make clean && ccache --zero-stats && make && ccache --show-stats
...
cache directory /home/hebasto/.ccache
primary config /home/hebasto/.ccache/ccache.conf
secondary config (readonly) /etc/ccache.conf
stats updated Sun Apr 11 15:45:43 2021
stats zeroed Sun Apr 11 15:45:05 2021
cache hit (direct) 638
cache hit (preprocessed) 0
cache miss 1
cache hit rate 99.84 %
called for link 10
cleanups performed 0
files in cache 20023
cache size 13.2 GB
max cache size 15.0 GB
```
The missed file is always `qt/libbitcoinqt_a-qrc_bitcoin_locale.o`.
With this PR:
```
$ make && make clean && ccache --zero-stats && make && ccache --show-stats
...
cache directory /home/hebasto/.ccache
primary config /home/hebasto/.ccache/ccache.conf
secondary config (readonly) /etc/ccache.conf
stats updated Sun Apr 11 15:28:46 2021
stats zeroed Sun Apr 11 15:28:21 2021
cache hit (direct) 639
cache hit (preprocessed) 0
cache miss 0
cache hit rate 100.00 %
called for link 10
cleanups performed 0
files in cache 20012
cache size 13.2 GB
max cache size 15.0 GB
```
ACKs for top commit:
fanquake:
ACK a58868d201
Tree-SHA512: 52f4a3267f41883d13025c0de79b6da22e92d60c729e01b986935c6812bbfe7fadc40b742bd715bfdf09df94af6838d4fbbe8208c6123f366108e38c8e1121c5
Carl Dong