8db1b59233
PR#6542: choeger's patch for type_cases (only generalize when useful)
2015-12-08 19:47:08 +09:00
78a7a55609
PR#6542: solve new performance problems du to path compression and copy
...
of types for the delayed icheck for unused cases
2015-12-08 19:45:42 +09:00
9d4b3a4513
Fix PR#7067: performance regression when compiling large nested structures in native code.
2015-12-07 15:07:50 +01:00
0358b397ee
Merge pull request #242 from c-cube/use_asciidoc
...
Use asciidoc for text files
2015-12-06 19:59:49 +01:00
8179c9e8f1
Merge pull request #288 from c-cube/fix-7026
...
PR#7026: optimize assignments with non-parametrized polymorphic variants
2015-12-06 19:58:58 +01:00
f3d3c221b1
move README.win32 to asciidoc
2015-12-06 19:31:15 +01:00
066022aa8a
move install file to asciidoc
2015-12-06 19:31:15 +01:00
62da50c7ba
use asciidoc for text files
2015-12-06 19:31:15 +01:00
6c5665393e
fix 0007026
...
- remove write barrier for constant poly variants
- add tests
2015-12-06 19:24:31 +01:00
b349cb64e9
fix minor arm/arch.ml failure -- waiting for 'match end'
2015-12-06 18:22:21 +01:00
c6b5a28d77
Merge pull request #320 from gasche/revert-user-defined-indexing-operators
...
Revert user defined indexing operators
2015-12-06 17:56:40 +01:00
9f7f3cefb7
PR#6854: Manual example does not initialize compare_ext field
2015-12-06 17:54:48 +01:00
0939a59850
PR#6289: Unix.utimes uses the current time only if both arguments are exactly 0.0
...
Also: use utimes() in preference to utime() so as to get sub-second resolution.
2015-12-06 17:35:34 +01:00
9d78aa722f
PR#6780: Poor error message for wrong -farch and -ffpu options (ocamlopt, ARM)
2015-12-06 16:57:17 +01:00
64146625f6
PR#7071: m-nt.h: don't #define inline for MSVC if we're compiling C++ code
2015-12-06 16:37:32 +01:00
8e6606d729
PR#6108, PR#6802: fail cleanly if dynlink.cma or ocamltoplevel.cma are loaded inside the toplevel loop
2015-12-06 12:59:23 +01:00
45343f6a39
PR#6989: in Str library, make sure that all \(...\) groups are binding and can be consulted with Str.matched_group.
...
There used to be a limitation to 32 binding groups, which is now lifted by dynamically allocating the groups data if it is too big.
Also:
- Modernize strstubs.c to use the Caml namespace
- Fixed a potential bug with backreferences \\n referring to a nonexistent group.
2015-12-06 12:43:02 +01:00
71c4f8b68a
Merge pull request #308 from rneswold/patches
...
Add (experimental) support for NetBSD/arm.
2015-12-06 10:21:07 +01:00
2893da8be1
Mention addition of NetBSD/arm support in "Feature wishes" section of
...
"Changes" file.
2015-12-05 10:18:01 -06:00
d59bc44b08
NetBSD doesn't have gethostbyname_r or gethostbyaddr_r, yet gets configured as
...
having them. The config script doesn't fail when there's no prototype to the
functions. It should maybe be compiled as "warnings as errors".
(obtained from PKGSRC patches.)
2015-12-05 10:16:49 -06:00
2914b95cbb
Add support for NetBSD/arm.
2015-12-05 10:16:48 -06:00
b2449a3120
Document hexadecimal floating-point literals.
2015-12-05 12:10:36 +01:00
4d478a0cfd
PR#6296: Some documentation on the floating-point representations recognized by Pervasives.float_of_string
2015-12-05 11:50:25 +01:00
9895ec921c
Travis CI: disable opam and oasis build which seem broken for now
2015-12-04 22:05:51 +01:00
6b7f81caf5
Switch to -custom for bytecode tests.
2015-12-04 18:29:44 +01:00
f443b1b617
Merge branch 'trunk' of ssh://github.com/ocaml/ocaml into trunk
2015-12-04 16:51:52 +01:00
d94488d7b5
Fix Changelog.
2015-12-04 16:47:23 +01:00
840f7ca506
Fix testsuite: use binary channels.
2015-12-04 16:46:52 +01:00
0388ef46d9
Fix testsuite: do not require a globally installed ocamlrun.
2015-12-04 16:46:36 +01:00
1ff6db10bf
Merge branch 'trunk' of https://github.com/bvaugon/ocaml into bvaugon-trunk
2015-12-04 16:31:59 +01:00
cce9cb4546
fix `ocaml -stdin` (was broken by 0296d022)
2015-12-04 16:20:55 +01:00
660dd6784f
fix min_int test
2015-12-04 15:48:05 +01:00
1fa12312a1
Merge pull request #329 from bschommer/array-functions
...
Add array functions: for_all, exists, mem, memq.
2015-12-04 15:47:33 +01:00
83dcf62ebf
Removed for_all2, exists2 and find again.
2015-12-04 15:41:48 +01:00
89d116c514
Merge branch 'trunk' of ssh://github.com/ocaml/ocaml into trunk
2015-12-04 13:57:39 +01:00
7ce026e815
Disable a test that does not work on 32-bit.
2015-12-04 13:55:35 +01:00
c278a3d49e
Merge pull request #325 from AltGr/gc_stats
...
Change v=0x400 in OCAMLRUNPARAM to print more Gc stats on exit
2015-12-04 13:38:21 +01:00
24e42d4d22
Add v=0x400 to OCAMLRUNPARAM to print more Gc stats on exit
...
In a format following that of Gc.print_stat. I chose to print only the "quick_stat"
values rather than call gc_ctrl.c::heap_stats because it's lighter, and the extra
information is typically not very useful at program exit.
Also adds documentation for the 0x400 flag (in man and Gc module)
This replaces the previous undocumented 0x400 that only displayed the
total (minwords + majwords - prowords) and with a different format,
since keeping both wouldn't provide more information.
2015-12-04 21:09:27 +09:00
a631e0858c
More tests for array functions.
...
The testsuite now covers more corner cases and includes also test
for float arrays and matrices.
2015-12-04 12:27:13 +01:00
e9bb7e117b
Update Changes.
2015-12-04 11:30:46 +01:00
7a082785be
Fixed typos and changed exists2 and for_all2.
...
Now exist2 and for_all2 only raise the invalid_arg excpetion when
the end of one array is reached.
Also some new tests in the testsuite.
2015-12-04 11:15:40 +01:00
dcc1421e1e
Added exists2, for_all2, mem, memq and find.
...
The functions are implemented as their pendants in the List module.
2015-12-04 10:10:36 +01:00
54d01d0d9c
Merge branch 'trunk' of ssh://github.com/ocaml/ocaml into trunk
2015-12-04 09:01:40 +01:00
66e864a53f
Use raise_notrace instead of raise to implement Hashtbl.replace.
...
This significantly improves the benchmark below:
let ht = Hashtbl.create 100 in
for x = 1 to 2500000 do
Hashtbl.replace ht x ();
Hashtbl.remove ht x
done
(about 20% faster, still not as good as #328 , which is about 35% faster
than trunk on this one)
In addition to the speedup, using raise_notrace should avoid destroying
the current stack trace when Hashtbl.replace is used in an exception
handler.
2015-12-04 08:57:16 +01:00
7f23ad6692
Early exit versions of exists and for_all.
...
The functions now return earlier when either an element is found that
does satisfy the condition in the case of exists or an element that does
not satisfy the condition in the case of for_all.
2015-12-04 01:46:25 +01:00
5e3964da86
Fix "weak hash of serialised closures" from mantis #0005942 .
2015-12-04 00:51:44 +01:00
35c25fb353
Added tests for Array.exists and Array.for_all
...
For each function there are two sets of tests, one on an empty and one on
an full set of elements. They use some simple predicates to test whether
for_all and exists work as expected.
2015-12-04 00:41:29 +01:00
d3aa77961c
Add exist and for_all to array.
...
Implemented an exist and a for_all function for arrays with the same
interfaces as the ones for lists.
2015-12-04 00:41:29 +01:00
e8625747fa
PR#6805 Duplicated expression in case of hole in a non-failing switch.
2015-12-03 17:52:50 +01:00
34a14a1193
update Change log
...
GPR#170 can break existing programs
especially ppx rewriter as the Parsetree is updated
2015-12-03 17:20:49 +01:00
Jacques Garrigue
alainfrisch
Gabriel Scherer
Simon Cruanes
Xavier Leroy
Rich Neswold
Damien Doligez
Bernhard Schommer
Louis Gesbert
Benoit Vaugon
maranget
Hugo Heuzard