add BLST_PORTABLE to CGO_CFLAGS to fix SIGILL issue with blst#49
Merged
roberto-bayardo merged 1 commit intodevnet-v3from Nov 7, 2022
Merged
add BLST_PORTABLE to CGO_CFLAGS to fix SIGILL issue with blst#49roberto-bayardo merged 1 commit intodevnet-v3from
roberto-bayardo merged 1 commit intodevnet-v3from
Conversation
Owner
|
one qq; is docker meant to directly consume the .env fille? I couldn't find any documentation on how workflows use it |
Collaborator
Author
|
https://docs.docker.com/compose/environment-variables/ This is where I saw it mentioned. |
Owner
|
i could be wrong, but i think the working directory of ci/docker is at the project root. The .env won't be picked up unless you move it to the root. |
Collaborator
Author
|
Ok let me see if the test will fail if I comment it out. |
roberto-bayardo
force-pushed
the
sigill-fix
branch
2 times, most recently
from
50b80a2 to
804a156
Compare
Collaborator
Author
|
Yeah test is passing regardless, I guess we're back on a more modern CPU for our builds! I moved it into the ci.yml file via an "env:" section which I suppose is better anyway (more explicit). |
Owner
|
I dont think this will work either because docker doesnt use the env of the caller/parent process for runs. I'd update the go docker script to provide the env file. |
Collaborator
Author
|
Ugg good point. I didn't want to put it in the dockerfile since I was hoping to scope it only to the CI, but perhaps that's the simplest solution. In fact it's already in the dockerfile only commented out! |
Owner
|
yeah updating the Dockerfile sounds fine by me. |
roberto-bayardo
force-pushed
the
sigill-fix
branch
from
804a156 to
b6ba5c3
Compare
roberto-bayardo
force-pushed
the
sigill-fix
branch
from
b6ba5c3 to
2f34ce3
Compare
Collaborator
Author
|
@Inphi ptal I believe this should work while isolating it to the CI |
roberto-bayardo
added a commit
that referenced
this pull request
Nov 7, 2022
roberto-bayardo
added a commit
that referenced
this pull request
Nov 17, 2022
…ongoing development (#60) * create devnet v3 branch including submodules + new submodule for erigon client * Add Lodestar * Supply chain config * Cache all docker work before yarn build * Build packages individually to allow them to cache * We must also build the CLI * Fix clean * Update to head of Lodestar * Fun through Capella * Latest Lodestar * latest prsym * Remove nonfunctional setting * Bump Lodestar * update geth to latest merge w/ master (#46) * Make all beacon chain config values explicit and move to common directory (#48) * Add common beacon chain config file * Move PRESET_BASE to top of config file * bump service start timeout from 1 to 2 min * Revert timeout change from 2 minutes to 60 sec Co-authored-by: Roberto Bayardo <bayardo@alum.mit.edu> * add BLST_PORTABLE to CGO_CFLAGS to fix SIGILL issue with blst (#49) * Update lodestar submodule. Copy trusted setup * Update devnet v3 with latest subrepos and new test from master (#51) * Add historical sync test (#44) Asserts that the beacon node follower can sync blobs during initial-sync. This patch also refactors the test harness to accommodate dynamic nodes. * update subrepos to latest Co-authored-by: Inphi <mlaw2501@gmail.com> * Update test harness to start either prysm or lodestar * Fix the Prysm communication * Fix the Lodestar wiring * Fix up validator <-> beacon node communication * Start switching tests to real beacon API * fee-market tests pass against Prysm without gRPC * Fix initial-sync test * Switch prysm to my branch which has GetBlockJSON * Configure Prysm to upgrade later (as though it had Capella). Also run Lodestar pre-4844 test in CI * Fix make lodestar-up * Remove Prysm-specific chain config. We used a shared config now. * Point Prysm back at the devnet-v3 branch * Update geth to latest from devnet-v3 * Put erigon back * Remove preset base, let it take the default * Attempt to parallelize CI * idunno lol * Try two jobs * Comment failing Lodestar CI test * Fix clean * Use basename pwd instead * Go back to using --project-name, but correctly supply it to docker compose down also * upgrade subrepos to latest (#56) * update geth & prysm subrepos to latest, pointing at eip-4844 instead of devnet-v3 (#57) * update subrepos to latest, move back to using eip-4844 branch for development * add 30 min timeout to some tests that sometimes hang * fix CI, Makefile, and update Prysm & geth subrepos Note: the new prysm & geth subrepos use a different kzg trusted setup matching, which now matches the latest consensus spec Co-authored-by: dancoffman <coffman@coinbase.com> Co-authored-by: Jimmy Chen <jchen.tc@gmail.com> Co-authored-by: Inphi <mlaw2501@gmail.com> Co-authored-by: Daniel Coffman <dgcoffman@gmail.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Fixes issue seen in #48