* configure.in (RUBY_SO_NAME): [EXPERIMENTAL] use the given name
literally if --with-soname is specified.
[ruby-core:79972] [Misc #13296]
[ci skip]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@57825 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
We already check for __builtin_mul_overflow in configure but never
actually referred it before. Why not call it if available, because
that should render supposedly-optimial assembly outputs.
Optionally if __builtin_mul_overflow_p is available, which is the case
for recent GCC, use that to detect fixnum overflow. This is much
faster than the previous. On my machine generated assembly of
numeric.c:int_pow reduces from 480 to 448 bytes, according to nm(1).
Also on my machine, following script boosts from 7.819 to 6.929 sec.
time ./miniruby -e 'i=0; while i < 30_000_000 do i += 1; 7 ** 23; end'
Signed-off-by: Urabe, Shyouhei <shyouhei@ruby-lang.org>
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@57784 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
NOTE:
(1) Fixnum's LSB is always 1.
It means you can always run `x - 1` without overflow.
(2) Of course `z = x + (y-1)` may overflow.
Now z's LSB is always 1, and the MSB of true result is also 1.
You can get true result in long as `(1<<63)|(z>>1)`,
and it equals to `(z<<63)|(z>>1)` == `ror(z)`.
GCC and Clang have __builtin_add_ovewflow:
* https://gcc.gnu.org/onlinedocs/gcc/Integer-Overflow-Builtins.html
* https://clang.llvm.org/docs/LanguageExtensions.html#checked-arithmetic-builtins
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@57506 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* configure.in (--with-gmp, --with-jemalloc): use AC_SEARCH_LIBS
to check if no library is required, instead of AC_CHECK_LIB.
[ruby-core:79368] [Bug #13175]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@57490 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* configure.in (crypt.h): crypt_r() was added in FreeBSD 12.0 but
is declared in unistd.h. [ruby-core:78664] [Bug #13038]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@57090 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
The `compatibility_version` should have an API version and the
`current_version` should have a program version of Ruby, but they have
been reversed and the binary compatibility has never worked.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56905 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* file.c: include sys/sysmacros.h for ArchLinux which deprecated
use of major() and minor() in sys/types.h.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56549 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* configure.in (ac_cv_func_round): round(3) in x86_64-w64-mingw32
is not accurate in an edge case. [ruby-core:77794] [Bug #12878]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56534 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* configure.in (RUBY_APPEND_OPTION, RUBY_PREPEND_OPTION): expand
the option to be appended/prepended when matching, as well as
RUBY_APPEND_OPTIONS and RUBY_PREPEND_OPTIONS.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56471 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* configure.in (DLDFLAGS): append --compress-debug-sections=zlib
if available, which reduces the size of LIBRUBY_SO by half or
more.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56445 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
actually check its availability rather to check GCC's version.
* configure.in (WARN_UNUSED_RESULT): moved to here.
* configure.in (RUBY_FUNC_ATTRIBUTE): change function declaration
to return int rather than void, because it makes no sense for a
warn_unused_result attributed function to return void.
Funny thing however is that it also makes no sense for noreturn
attributed function to return int. So there is a fundamental
conflict between them. While I tested this, I confirmed both
GCC 6 and Clang 3.8 prefers int over void to correctly detect
necessary attributes under this setup. Maybe subject to change
in future.
* internal.h (UNINITIALIZED_VAR): renamed to MAYBE_UNUSED, then
moved to configure.in for the same reason we move
WARN_UNUSED_RESULT.
* configure.in (MAYBE_UNUSED): moved to here.
* internal.h (__has_attribute): deleted, because it has no use now.
* string.c (rb_str_enumerate_lines): refactor macro rename.
* string.c (rb_str_enumerate_bytes): ditto.
* string.c (rb_str_enumerate_chars): ditto.
* string.c (rb_str_enumerate_codepoints): ditto.
* thread.c (do_select): ditto.
* vm_backtrace.c (rb_debug_inspector_open): ditto.
* vsnprintf.c (BSD_vfprintf): ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56169 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* configure.in, include/ruby/defines.h (RUBY_USE_SETJMPEX):
include setjmpex.h only when using setjmpex() for RUBY_SETJMP.
the header of mingw32 overrides setjmp() by setjmpex().
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55471 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* missing/crypt.h (struct crypt_data): remove unnecessary member
"initialized".
* missing/crypt.c (des_setkey_r): nothing to be initialized in
crypt_data.
* configure.in (struct crypt_data): check for "initialized" in
struct crypt_data, which may be only in glibc, and isn't on AIX
at least.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55272 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* configure.in: revert r55237. replace crypt, not crypt_r, and
check if crypt is broken more.
* missing/crypt.c: move crypt_r.c
* string.c (rb_str_crypt): use crypt_r if provided by the system.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55245 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* configure.in (XCFLAGS): merge flags only for ruby itself from
ruby_cflags.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55148 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-fexcess-precision=standard and -fp-model precise are set to this now.
* configure.in (cflags): use ruby_cflags.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55142 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* configure.in (RUBY_CHECK_BUILTIN_SETJMP): declare t as NORETURN
to suppress warnings by -Wsuggest-attribute=noreturn.
[ruby-core:75510] [Bug #12383]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55023 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* configure.in (FUNC_STDCALL, FUNC_CDECL, FUNC_FASTCALL): set
macro names explicitly to the old names, which are accidentally
changed at r54985, for backward compatibilities.
fiddle also depends on these names to fallback to ANSI names.
[ruby-core:75494] [Bug #12377]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55015 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* util.c (ruby_qsort): use qsort_s if available, for Microsoft
Visual Studio 2005 (msvcr80.dll) and mingw.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@54974 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
is specified and it implied -fexcess-precision=standard.
Now with -std=gnu99, it should be explicitly specified.
https://gcc.gnu.org/onlinedocs/gcc/Optimize-Options.html
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@54967 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
and define CONSTFUNC and PUREFUNC if available.
Note that I don't add those options as default because
it still shows many false-positive (it seems not to consider
longjmp).
* vm_eval.c (stack_check): get rb_thread_t* as an argument
to avoid duplicate call of GET_THREAD().
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@54952 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
nobu
shyouhei
naruse
knu
shugo
usa
mrkn