The unpac monorepo manager self-hosting as a monorepo using unpac
0
fork

Configure Feed

Select the types of activity you want to include in your feed.

Fixing typos (#12955)

authored by

Fourchaux and committed by
GitHub
7f3ecc86 a36a5337

+25 -24
+2 -2
Changes
··· 1767 1767 1768 1768 - #11846: Mark rbx as destroyed at C call for Win64 (mingw-w64 and Cygwin64). 1769 1769 Reserve the shadow store for the ABI in the c_stack_link struct instead of 1770 - explictly when calling C functions. This simultaneously reduces the number of 1770 + explicitly when calling C functions. This simultaneously reduces the number of 1771 1771 stack pointer manipulations and also fixes a bug when calling noalloc 1772 1772 functions where the shadow store was not being reserved. 1773 1773 (David Allsopp, report by Vesa Karvonen, review by Xavier Leroy and ··· 2783 2783 - #8516: Change representation of class signatures 2784 2784 (Leo White, review by Thomas Refis) 2785 2785 2786 - - #9444: -dtypedtree, print more explictly extra nodes in pattern ast. 2786 + - #9444: -dtypedtree, print more explicitly extra nodes in pattern ast. 2787 2787 (Frédéric Bour, review by Gabriel Scherer) 2788 2788 2789 2789 - #10337: Normalize type_expr nodes on access
+1 -1
Makefile.common
··· 284 284 # Each program foo is characterised by the foo_LIBRARIES and foo_SOURCES 285 285 # variables. The following macros provide the infrastructure to build foo 286 286 # from the object files whose names are derived from these two 287 - # varialbes. In particular, the following macros define several 287 + # variables. In particular, the following macros define several 288 288 # variables whose names are prefixed with foo_ to compute the 289 289 # different lists of files used to build foo. 290 290
+2 -2
asmcomp/stackframegen.ml
··· 42 42 by treating them as non-tail calls, even if they are implemented as 43 43 tail calls. 44 44 45 - This method can be overriden in [Stackframe] to implement target-specific 45 + This method can be overridden in [Stackframe] to implement target-specific 46 46 behaviors. *) 47 47 48 48 method is_call = function ··· 61 61 This is the case if it contains calls, but also if it allocates 62 62 variables on the stack. 63 63 64 - This method can be overriden in [Stackframe] to implement target-specific 64 + This method can be overridden in [Stackframe] to implement target-specific 65 65 behaviors. *) 66 66 67 67 method frame_required f contains_calls =
+1 -1
manual/src/cmds/intf-c.etex
··· 2412 2412 It is possible to annotate with "[\@untagged]" any immediate type, i.e. types 2413 2413 that are represented like "int". This includes "bool", "char", any variant 2414 2414 type with only constant constructors. Note: this does not include 2415 - "Unix.file_descr" which is not represented as an interger on all platforms. 2415 + "Unix.file_descr" which is not represented as an integer on all platforms. 2416 2416 2417 2417 \subsection{ss:c-direct-call}{Direct C call} 2418 2418
+1 -1
parsing/builtin_attributes.mli
··· 172 172 (** [attr_equals_builtin attr s] is true if the name of the attribute is [s] or 173 173 ["ocaml." ^ s]. This is useful for manually inspecting attribute names, but 174 174 note that doing so will not result in marking the attribute used for the 175 - purpose of warning 53, so it is usually preferrable to use [has_attribute] 175 + purpose of warning 53, so it is usually preferable to use [has_attribute] 176 176 or [select_attributes]. *) 177 177 val attr_equals_builtin : Parsetree.attribute -> string -> bool 178 178
+1 -1
release-info/introduction.md
··· 114 114 * stable feature sets 115 115 * only emergency bug fixes 116 116 * documentation PRs postponed to after the release 117 - * inteded for wide testing (and detecting deployment or production issues among 117 + * intended for wide testing (and detecting deployment or production issues among 118 118 large private code base) 119 119 120 120 Starting from the first alpha release, there is a small team effort to try to
+1 -1
runtime/caml/domain.h
··· 118 118 [caml_try_run_on_all_domains*] runners, it will 119 119 run on all participant domains in parallel. 120 120 121 - The "STW critical section" is the runtime interval betweeen the 121 + The "STW critical section" is the runtime interval between the 122 122 start of the execution of the STW callback and the last barrier in 123 123 the callback. During this interval, mutator code from registered 124 124 participants cannot be running in parallel.
+1 -1
runtime/lf_skiplist.c
··· 354 354 /* attentive readers will have noticed that we assume memory is aligned to 355 355 * atleast even addresses. This is certainly the case on glibc amd64 and 356 356 * Visual C++ on Windows though I can find no guarantees for other 357 - platorms. */ 357 + platforms. */ 358 358 struct lf_skipcell *new_cell = caml_stat_alloc( 359 359 SIZEOF_LF_SKIPCELL + (top_level + 1) * sizeof(struct lf_skipcell *)); 360 360 new_cell->top_level = top_level;
+1 -1
runtime/memprof.c
··· 591 591 lambda]. We could use more a involved algorithm, but this should 592 592 be good enough since, in the average use case, [lambda] <= 0.01 and 593 593 therefore the generation of the binomial variable is amortized by 594 - the initialialization of the corresponding block. 594 + the initialization of the corresponding block. 595 595 596 596 If needed, we could use algorithm BTRS from the paper: 597 597 Hormann, Wolfgang. "The generation of binomial random variates."
+1 -1
runtime/tsan.c
··· 159 159 the FFI rules on the condition that the GC does not run between the 160 160 allocation and the end of initialization) and a conflicting access is made 161 161 from OCaml after publication to other threads. There should be no data 162 - race thanks to data dependency (see [MMOC] coment in memory.c), but TSan 162 + race thanks to data dependency (see [MMOC] comment in memory.c), but TSan 163 163 does not take data dependencies into account. 164 164 - A field is accessed from C with `Field`, or more generally using a 165 165 `volatile value *` or a relaxed atomic access, and that field is modified
+1 -1
stdlib/CONTRIBUTING.md
··· 16 16 17 17 Obviously, proposals made to evolve the standard library will be 18 18 evaluated with very high standards, similar to those applied to the 19 - evolution of the surface langage, and much higher than those for 19 + evolution of the surface language, and much higher than those for 20 20 internal compiler changes (optimizations, etc). 21 21 22 22 A key property of the standard library is its stability. Backward
+1 -1
stdlib/dynarray.ml
··· 483 483 reserved-but-not-yet-filled space, it will get a clean "missing 484 484 element" error. This is worse than with the fill-then-notify 485 485 approach where the new elements would only become visible 486 - (to iterators, for removal, etc.) alltogether at the end of 486 + (to iterators, for removal, etc.) altogether at the end of 487 487 loop. 488 488 489 489 To summarise, "reserve before fill" is better on add-add races,
+1 -1
stdlib/marshal.mli
··· 191 191 Care must be taken when marshaling a mutable value that may be modified by 192 192 a different domain. Mutating a value that is being marshaled (i.e., turned 193 193 into a sequence of bytes) is a programming error and might result in 194 - suprising values (when unmarshaling) due to tearing, since marshaling 194 + surprising values (when unmarshaling) due to tearing, since marshaling 195 195 involves byte-per-byte copy. 196 196 *)
+1 -1
stdlib/random.ml
··· 143 143 * we use rejection sampling on the greatest interval [ [0, k*n-1] ] 144 144 * that fits in [ [0, mask] ]. That is, we reject the 145 145 * sample if it falls outside of this interval, and draw again. 146 - * This is what the test below does, while carefuly avoiding 146 + * This is what the test below does, while carefully avoiding 147 147 * overflows and sparing a division [mask / n]. *) 148 148 if r - v > mask - n + 1 then int_aux s n mask else v 149 149
+1 -1
testsuite/tests/basic/stringmatch.ml
··· 31 31 assert (tst02 "\000\000\000\003" = 3) ; 32 32 () 33 33 34 - (* Keword reckognition *) 34 + (* Keyword recognition *) 35 35 36 36 let s00 = "get_const" 37 37 let t00 = "set_congt"
+1 -1
testsuite/tests/frame-pointers/fp_backtrace.c
··· 126 126 if (!is_from_executable(symbol, execname)) 127 127 goto skip; 128 128 129 - /* Exctract the full function name */ 129 + /* Extract the full function name */ 130 130 regmatch_t funcname = func_name_from_symbol(symbol); 131 131 if (funcname.rm_so == -1) 132 132 goto skip;
+1 -1
testsuite/tests/typing-modules/inclusion_errors.ml
··· 1329 1329 type w = private float 1330 1330 type q = private (int * w) 1331 1331 type u = private (int * q) 1332 - module M : sig (* Confussing error message :( *) 1332 + module M : sig (* Confusing error message :( *) 1333 1333 type t = private (int * (int * int)) 1334 1334 end = struct 1335 1335 type t = private u
+1 -1
tools/objinfo.ml
··· 435 435 436 436 let arg_list = [ 437 437 "-quiet", Arg.Set quiet, 438 - " Only print explicitely required information"; 438 + " Only print explicitly required information"; 439 439 "-no-approx", Arg.Set no_approx, 440 440 " Do not print module approximation information"; 441 441 "-no-code", Arg.Set no_code,
+1 -1
typing/typecore.ml
··· 4737 4737 [type_argument] on the cases, and discard the cases' 4738 4738 inferred type in favor of the constrained type. (Function 4739 4739 cases aren't inferred, so [type_argument] would just call 4740 - [type_expect] straightaway, so we do the same here.) 4740 + [type_expect] straight away, so we do the same here.) 4741 4741 - [type_without_constraint]: If there is just a coercion and 4742 4742 no constraint, call [type_exp] on the cases and surface the 4743 4743 cases' inferred type to [type_constraint_expect]. *)
+1 -1
typing/value_rec_check.ml
··· 154 154 (* Note on module presence: 155 155 For absent modules (i.e. module aliases), the module being bound 156 156 does not have a physical representation, but its size can still be 157 - derived from the alias itself, so we can re-use the same code as 157 + derived from the alias itself, so we can reuse the same code as 158 158 for modules that are present. *) 159 159 let size = classify_module_expression env mexp in 160 160 let env = Ident.add mid size env in
+1 -1
utils/linkdeps.mli
··· 51 51 52 52 val check : t -> error option 53 53 (** [check t] should be called once all the compilation units to be linked 54 - have been added. It retuns some error if: 54 + have been added. It returns some error if: 55 55 - There are some missing implementations 56 56 and [complete] is [true] 57 57 - Some implementation appear
+2 -1
utils/local_store.mli
··· 14 14 (**************************************************************************) 15 15 16 16 (** This module provides some facilities for creating references (and hash 17 - tables) which can easily be snapshoted and restored to an arbitrary version. 17 + tables) which can easily be snapshotted and restored to an arbitrary 18 + version. 18 19 19 20 It is used throughout the frontend (read: typechecker), to register all 20 21 (well, hopefully) the global state. Thus making it easy for tools like