-
Notifications
You must be signed in to change notification settings - Fork 78
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Move ocaml/
into /
(a/k/a pivot_root)
#3166
Conversation
… locations in the same way as errors) (ocaml-flambda#2787)
…#2902) * Progress fixing typecore * saving work * Finish fixing typecore * Fix issue with typemod picked up in rebase * Fix to buggy merge conflict resolution that caused us to hit 'assert false' in update_type
* Fix conflicts in w53 test * backport upstream #13170 * Add compilers reference file
…da#2905) * refactor string-like access primitives to make it easier to add primitives for unboxed access without duplicating as much code Signed-off-by: David Vulakh <[email protected]> * hoist out [indexing_primitives] don't close over [mode] when constructing the [indexing_primitives] map, and hoist the computation out of [lookup_primitive] (previously we were recomputing the cartesian product each time we entered [lookup_primitive]) Signed-off-by: David Vulakh <[email protected]> * make fmt Signed-off-by: David Vulakh <[email protected]> * label [index] and [bound] in [check_bound] Signed-off-by: David Vulakh <[email protected]> * move annotation in [actual_max_length_for_string] Signed-off-by: David Vulakh <[email protected]> * move annotation in [string_like_load] Signed-off-by: David Vulakh <[email protected]> * replace cartesian product helper with let operator Signed-off-by: David Vulakh <[email protected]> * review changes - change type of [bound_kind] in [bound] - move some type constraints to arguments - return old comment about bounds checks - move [is_substring] to [misc] Signed-off-by: David Vulakh <[email protected]> --------- Signed-off-by: David Vulakh <[email protected]>
* Initial state where runtime builds * Review (cherry picked from commit e41be34) * Review of conflicts * Remove XXX in memprof files (cherry picked from commit 58e7b16) * ptrdiff_t (cherry picked from commit e4b4388) * domain.c:778 to stay deleted (cherry picked from commit f05df07) * Fix memprof bit in domain.c:domain_create (cherry picked from commit 34925f4) * Revert to upstream compaction behaviour in gc_ctrl.c (cherry picked from commit 7740dc4) * Fix for gettimeofday (cherry picked from commit d2f2330) * Typo fix (cherry picked from commit 2db34fd) * Remove XXX in caml_enter_blocking_section * Remove message in caml_enter_blocking_section for async exn raise * Remove XXX in caml_ba_alloc * adopt_orphaned_work * caml_alloc_custom_mem * Try to fix DEBUG code in platform.c * Resurrect CHECK_STACK_ALIGNMENT macro * Dune build fixes for the runtime - may need a bit more tidying * git rm s390x.S win32.c power.S riscv.S
…ren't ignored (ocaml-flambda#2918) (cherry picked from commit 3cae2fb)
* Modify cms format * Convert cms_uid_to_loc to be an array * Add crs * Add comment to need_to_clear_env * Remove debugging type annotation * Convert directly to array * Remove loadpath from cms * Guard initial_env behind -bin-annot-occurrences * Revert "Convert directly to array" This reverts commit fc1857f. * Directly use array * Revert "Directly use array" This reverts commit 4ccb2ae. * Revert "Convert cms_uid_to_loc to be an array" This reverts commit 808b35b.
* Promote tests that remove redundant kind from error message history * Rename Primitive to Builtin * Remove Const.Sort and Const.Layout aliases * Remove pointless const definition * Remove stale crs about eg layout_float64 * Remove legacy layout usage in printing * Remove legacy layout usage in extension universe check * Rename has_imported_history to is_imported * Remove Legacy module * Remove unused builtins * Move const_of_user_written_annotation into Const module * Update comment * Remove redundant layout definition * Remove unnecessary layout aliases * Document jkind printing * Rename ModeParser to Mode_parser * Fix cr formatting * Remove stale cr * Remove incorrect CRs * Replace any with abstract layout in error message * Assert layout is value in value_kind_of_value_jkind * Don't print : value * Add CRs to remove annotation * Improve debug printing * Share parsing logic between modes and jkinds * Simplify parsing * Add tests * Create test for not calling caml_modify * Reformat files * Add comments to caml_modify test * Rename external_record to internal_record * Hide external_variant's immediacy * Add inlining test for caml_modify * Add copyright header * Clean up typemodifier * Fix prining typo * Add CR to Builtin * Stop calling crossing of nullability mode-crossing * Improve warning 211 message * Restore PRs about incorrect error messages * Cleanup typemodifier more * Fix bug in required maturity level for modes * Fix floatarray kind and promote tests
* Revert "Move float32 back to beta (ocaml-flambda#2843)" This reverts commit 4846e6c. * fix test * upstream_compatible test * upstream_compatible simd test
…l-flambda#2906) * unboxed access primitives for string-likes index by: int32#,int64#,nativeint# access width: int16,int32,float32,int64,int128,int32#,float32#,int64# for bigstring, string, bytes Signed-off-by: David Vulakh <[email protected]> * make fmt Signed-off-by: David Vulakh <[email protected]> * remove [CAMLprim] from macro names Signed-off-by: David Vulakh <[email protected]> * tweak formatting in [lambda.ml{,i}] Signed-off-by: David Vulakh <[email protected]> * improve generated tests Signed-off-by: David Vulakh <[email protected]> * better data in generated tests Signed-off-by: David Vulakh <[email protected]> * add more tests of edge-case indices Signed-off-by: David Vulakh <[email protected]> * reflect float32 -> stable in tests fixes ci Signed-off-by: David Vulakh <[email protected]> --------- Signed-off-by: David Vulakh <[email protected]>
…a#2935) Revert "Remove "scheduling" leftovers (ocaml-flambda#2857)" This reverts commit ef78547.
* Resolve conflicts in ocamltest * Restore the "do_something" mechanism. * Further fixes for ocamltest --------- Co-authored-by: Xavier Clerc <[email protected]>
* Fixes to domain.c bad merges (cherry picked from commit ace16c4) * Remove caml_reset_young_limit line
* Remove jane syntax for layouts * Reshuffle jkind annotations * Fix location mapping/iterating omission * Fix chamelon * remove dead code from `Jane_syntax_parsing` * Update ocaml/parsing/ast_helper.mli Co-authored-by: Chris Casinghino <[email protected]> * Feedback from review * more feedback for review --------- Co-authored-by: Chris Casinghino <[email protected]>
* Markdelay support for multi-domain programs - Domains now begin marking simultaneously at some minor GC during the major cycle - Before marking, the write barrier is disabled and fresh allocations are UNMARKED - Change to ephemeron logic: per-cycle ephemeron data structure reset is now done at marking start rather than at the start of the cycle. (In particular, this makes orphaned work simpler to deal with) * Re-enable parallel tests Includes several systhreads tests that rely on the interaction between systhreads, domains and backup threads. * Add a new test (tests/parallel/churn.ml) stressing cross-domain promotion
…a#3146) * Squelch quotes with -error-style short * Something mangles comments and can't deal with quote marks
Lift continuations (with cli option) By default, no continuation is lifted, can be configured with a command line flag. When the continuatin lifting budget is non-zero, we need to use the advanced meet, because there are known issue that occurs when using the regular meet together with continuation lifting. About the fixed test: Adding a Sys.opaque_identity makes the lifetime of the result of `quick_stat` does not depend on whether flambda2 lifts some continuation or not (in turn, this difference of lifetime induces a difference in the tests results, since it counts the number of allocations in the major heap).
Add [Debuginfo.compare_outer_first]
Add a function to append [Clflags.directory] to filename and use it
* Don't include zero_alloc info in string_of_scope by default It was a mistake to include it, it's only needed for debug printing. * Make this argument not optional so we can see where it's used. * Update call sites * Format
Parser Change ChecklistThis PR modifies the parser. Please check that the following tests are updated:
This test should have examples of every new bit of syntax you are adding. Feel free to just check the box if your PR does not actually change the syntax (because it is refactoring the parser, say). |
…mbda#3164) * Force runtime4 symbol separator * Fix asm stubs * dynlink_nat.c * Update comment * Fix question.reference * Frame pointers tests
233fafd
to
732fb10
Compare
12a2b76
to
3595623
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This will be merged manually
Pushing this as an experiment. This may not be something that can meaningfully be put through the PR process but it should give us an idea of what rebasing PRs onto this will look like.