Extend macOS portability layer for Apple Silicon#9070
Closed
yeabat8tcap wants to merge 3947 commits intofiredancer-io:macos-integrationfrom
Closed
Extend macOS portability layer for Apple Silicon#9070yeabat8tcap wants to merge 3947 commits intofiredancer-io:macos-integrationfrom
yeabat8tcap wants to merge 3947 commits intofiredancer-io:macos-integrationfrom
Conversation
Fixes a bug that results in unrelated fork graph nodes getting dropped on accdb_cancel.
…o#8763) * bundle: fix STEM_BURST by buffering transactions in a deque unit tests zero copy design more hardening * tile * PR comments * use stem abstraction * PR comments * remove stale metrics * remove stack copies * add tests * plumb bundle topology depth
Reduce git diff churn when reordering metrics by deriving metric indices using an enum
…cer-io#8782) * runtime: fix missing serialization and address in builtin-to-core-bpf migration new_target_program_account() constructs an UpgradeableLoaderState::Program state but never serializes it into acc->data, leaving the account with all-zero data after migration. The sister function new_target_program_data_account() correctly calls fd_bpf_upgradeable_loader_state_encode(). In Agave, AccountSharedData::new_data() handles both allocation and serialization atomically. Additionally, the call site in migrate_builtin_to_core_bpf1() never sets new_target_program->addr to the builtin program address, so tmp_account_store() writes the account to an uninitialized address. Agave stores to target.program_address explicitly. Add the encode call in new_target_program_account() and set the address at the call site. * add missing addr assignment * flamenco, bpf migration: implement verified_build_hash check * flamenco, bpf migration: fix logic for when buffer has authority but config doesn't specify one * fixup * flamenco, ledger: add enshrine_slashing_program ledger * flamenco, bpf migration: add discriminant check --------- Co-authored-by: nlgripto <155116201+nlgripto@users.noreply.github.com>
get_token_to_eol prevents oob read, but would still except numeric prefixes -- and is just wrong usage
the write in fd_runtime_pre_execute_check makes more sense, also given early returns
…cer-io#9038) * use flock for submodules * test vectors temp remove
* stakes: eliminate O(n_delegations) loop in fd_stakes_activate_epoch
* firedancer: clean up vote_keyed_lsched and chacha20 * firedancer: clean up switch_to_chacha8_turbine from codebase --------- Co-authored-by: Emanuele Cesena <ecesena@jumptrading.com>
This commit implements a comprehensive portability layer to enable native Firedancer builds on macOS ARM64. Key changes include: - Core Libraries: Successfully compiled libfd_util, libfd_ballet, libfd_tango, libfd_flamenco, and libfd_disco on macOS. - Shared Memory: Replaced Linux hugepages with mlock-backed anonymous memory via shm_open(SHM_ANON). - Threading: Implemented Mach thread priority (QoS) and affinity shims for fd_tile. - Agave Integration: Patched x86_64 rdtsc intrinsics for ARM64 and resolved RocksDB linkage issues. - System Shims: Implemented getrandom-to-getentropy, clock_nanosleep, and networking shims (mmsghdr, recvmmsg). - SANDBOX & SECCOMP: Stubbed Linux-specific process isolation and security features for macOS compatibility. - Type Compatibility: Resolved global ulong vs uint64_t pointer mismatches and type-generic overflow built-ins. - Build System: Introduced apple_silicon.mk and updated Local.mk rules for Darwin targets. Work in progress: Finalizing fdctl linkage and verifying Gossiping/RPC.
Contributor
|
Thank you for the contribution. This is a bit too big too review or merge (600K lines of changes), but I'd be happy to accept a series of PRs with smaller changes. For example, the first PR could just add |
Contributor
|
For this kind of contribution specifically, we cannot accept LLM-generated patches, because of the asymmetric human review cost. |
Author
|
Ahh I see, okay. Is it better to create a separate branch and cherry pick from there? I can be able to do that. Let me know, thanks. @ripatel-fd |
Contributor
Yes that would be great, thanks! |
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
20 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.
This PR merges the Apple Silicon (M-series) portability layer from the firedancer-apple fork into the macos-integration branch. It includes: