Mac arm fixes#1794
Merged
Merged
Conversation
On Apple Silicon, snapshot_load fixes up the loaded code's data in place in MAP_JIT (W^X) memory; the bare stores fault with SIGBUS (KERN_PROTECTION_FAILURE). Wrap the three load-side write sites in JITDataReadWriteMaybeExecute() / JITDataReadExecute() (same pattern as the other JIT-literal write sites): - the code-literals memcpy, - the fixup_objects walk (walk_temporary_root_objects<fixup_objects_t>), - the fixup_internals walk. (The save path fixes up a RW copy of the buffer, so it never hit this.) This unblocks snapshot load past the W^X SIGBUS. The remaining failure -- the position-independent function-pointer relocation decode (decodeEntryPoint / fixedAddress) -- is separate, pre-existing sc598 relocation work. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…ath quoting Make snapshot save/load/executable work on macOS arm64 (SLAD-SNAPSHOT and SLAD-EXECUTABLE now pass): 1. Relocation decode: the entry-point decode validated the resolved address by comparing its first byte against a saved firstByte and ABORTED on mismatch (decodeEntryPointForCompiledCode) / warned (fixedAddress). But a function's first instruction is frequently a relocatable instruction (e.g. ADRP) whose encoded immediate bytes legitimately differ between the save-time and load-time JIT (different load addresses => different page offsets). The offset/symbol-based decode is correct (verified: snapshot loads and returns the right value); drop the firstByte equality check, which was a false positive on every relocated first insn. 2. Executable link: save-lisp-and-die :executable builds a clang++ command run via system() (a shell); the -Wl,-force_load,<libdir>/libiclasp.a path was unquoted, so a build dir containing a space (e.g. a Dropbox path) split the arg and the link failed. Quote the output, sectcreate, and force_load paths. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…e -L
The executable link command (run via system()) carried only the RELATIVE
-Lboehmprecise/lib from BUILD_LINKFLAGS, so -lclasp resolved only when
save-lisp-and-die :executable ran with CWD=build/; from any other directory the
link failed ("library not found for -lclasp"). Add an absolute, quoted
-L"<_LibDir>" (as the Linux branch already does). The runtime rpath is already
absolute, so the produced executable both links and runs from any CWD.
Verified: created and ran a standalone executable from /tmp (returns the right
value); SLAD-EXECUTABLE now passes from the repo root, not just build/.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Two problems prevented `ninja -C build` from working on an Apple Silicon Mac with a Homebrew toolchain: * boost (used by clbind/config.h) ships no pkg-config file and is therefore not declared as a koga `library`; clasp relies on it being on the default include path. units.lisp only added the Intel Homebrew prefix `/usr/local/include`, so `/opt/homebrew/include` was never searched and the build failed immediately with "'boost/config.hpp' file not found". Add `/opt/homebrew/include` for darwin when it exists (probe-guarded, so Intel and Linux are unaffected). * The rpath was emitted as an unquoted `-Wl,-rpath,<abs path>`. When the build directory contains a space, ninja passes the flag through /bin/sh which then splits it at the space, and the link fails with "no such file or directory: '<tail of path>'". Quote the path. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
On Apple Silicon, the LLVM ORC JITLink memory slab is mapped MAP_JIT and is write-protected (execute mode) per-thread by default; a thread must call pthread_jit_write_protect_np(false) before writing to it. clasp writes Lisp object pointers into each native module's literals vector, which lives in that JIT memory. Those writes were not bracketed by a switch to write mode, so on Apple Silicon they fault with SIGBUS (EXC_BAD_ACCESS code=2, KERN_PROTECTION_FAILURE) -- which manifested as a crash in loadltv::attr_clasp_module_native while loading freshly compiled native FASLs during the "Compiling Clasp native image" bootstrap step. (On x86-64 and Linux the rwx page is genuinely writable, so the bug was latent there.) The helpers JITDataReadWriteMaybeExecute()/JITDataReadExecute() already exist for exactly this but had no callers. Bracket every write into JIT-resident literals memory with them: * core::core__literals_vset (compiler.cc) * llvmo::code_literal_set (code.cc) * loadltv::attr_clasp_module_native (loadltv.cc) * loadltv::attr_clasp_function_native_estranged (loadltv.cc) * snapshot-load literal relocation memcpy (snapshotSaveLoad.cc) Reads of the literals vector are left untouched: MAP_JIT memory is readable in execute mode, only writes fault. The write-mode window is kept as small as possible (the bare store) so no JIT code is executed while the thread is in write mode. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
On Apple Silicon, JIT'd code/data lives in MAP_JIT memory that is write-protected (execute mode) by default; a thread must switch it to write mode (pthread_jit_write_protect_np) around any store or the store faults with SIGBUS (KERN_PROTECTION_FAILURE). setf_jit_lookup_t (llvmo:jit-lookup-t setf) stores a Lisp function pointer into a JIT-emitted global. make-callback / clasp-ffi:%defcallback reach it via (setf (llvm-sys:jit-lookup-t dylib varname) function), and on arm64-darwin it SIGBUSes at compile time -- this is what makes the defcallback-native regression test (CFFI-DEFCALLBACK) crash during compile-file. Wrap the store in JITDataReadWriteMaybeExecute()/JITDataReadExecute(), matching the existing W^X guards on the other JIT-literal write sites (core__literals_vset, loadltv op_setf_literals / attr_clasp_module_native). Verified on macOS arm64 (native boehmprecise image): CFFI-DEFCALLBACK now passes (previously a Bus error at compile time); no regressions. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The only thing we were using it for is a correctness check. If that check is not valid due to relocation, there' no point in putting the byte in.
This was referenced Jun 12, 2026
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.
This incorporates most of @dg1sbg's remaining PRs (#1768, #1771, #1786, #1788) relating to Mac: