Rich Felker [Sun, 3 Jun 2012 00:04:27 +0000 (20:04 -0400)]
remove implementation-reserved bits when saving signal mask
this fix is necessary because a program could be started with some of
the implementation-reserved signals masked (e.g. due to exec having
been called from a signal handler, or from a non-musl program) and
then could obtain an invalid-to-use-later sigset_t as the old/saved
signal mask.
Rich Felker [Sat, 2 Jun 2012 23:56:18 +0000 (19:56 -0400)]
remove no-longer-needed unblocking of signals in pthread_create
this action is now performed in pthread_self initialization; it must
be performed there in case the first call to pthread_create is from a
signal handler, in which case the old signal mask could be restored on
return from the signal.
Rich Felker [Sat, 2 Jun 2012 21:49:14 +0000 (17:49 -0400)]
add LSB abi junk for ctype functions
this should be the last major fix needed to support running
glibc-linked conforming POSIX programs with musl in place of glibc, as
long as musl provides the features they need and they don't use
pthread cancellation (which is implemented as c++ exceptions in glibc,
and fundamentally incompatible with musl).
Rich Felker [Sat, 2 Jun 2012 20:51:04 +0000 (16:51 -0400)]
declare environ in unistd.h when _GNU_SOURCE feature test macro is used
lots of broken programs expect this, and it's gotten to the point of
being a troubleshooting FAQ topic. best to just fix it.
Rich Felker [Sat, 2 Jun 2012 20:48:22 +0000 (16:48 -0400)]
use fistpll mnemonic instead of fistpq (more widely supported) on x86_64 too
this was fixed previously on i386 but the corresponding code on x86_64
was missed.
Rich Felker [Fri, 1 Jun 2012 03:32:09 +0000 (23:32 -0400)]
add LSB ABI __xstat, etc. junk
Rich Felker [Fri, 1 Jun 2012 03:12:31 +0000 (23:12 -0400)]
enable LARGEFILE64 aliases
these will NOT be used when compiling with -D_LARGEFILE64_SOURCE on
musl; instead, they exist in the hopes of eventually being able to run
some glibc-linked apps with musl sitting in place of glibc.
also remove the (apparently incorrect) fcntl alias.
Rich Felker [Thu, 31 May 2012 01:19:32 +0000 (21:19 -0400)]
fix musl-gcc wrapper to work with -pie
linking the wrong crt1.o resulted in textrels and thus crashing
Rich Felker [Tue, 29 May 2012 02:54:27 +0000 (22:54 -0400)]
there is no such GNU function fpurge, only __fpurge.
no idea where I got the idea fpurge should exist...
Rich Felker [Tue, 29 May 2012 02:53:24 +0000 (22:53 -0400)]
add prototype for BSD/GNU stdio *_unlocked extension functions
also fix up distinction of what is GNU-only and what's GNU+BSD
Rich Felker [Tue, 29 May 2012 02:46:28 +0000 (22:46 -0400)]
remove duplicate lfs64 cruft in stdio.h
nsz [Mon, 28 May 2012 20:55:03 +0000 (22:55 +0200)]
math: fix nextafter definition in tgmath.h
Rich Felker [Sun, 27 May 2012 20:01:44 +0000 (16:01 -0400)]
add ldd and main program loading support to dynamic linker
Rich Felker [Sun, 27 May 2012 18:49:55 +0000 (14:49 -0400)]
cleanup dynamic linker start code cruft
two actual issues: one is that __dynlink no longer wants/needs a GOT
pointer argument, so the code to generate that argument can be
removed. the other issue was that in the i386 code, argc/argv were
being loaded into registers that would be call-clobbered, then copied
to preserved registers, rather than just being loaded into the proper
call-preserved registers to begin with.
this cleanup is in preparation for adding new dynamic linker
functionality (ability to explicitly invoke the dynamic linker to run
a program).
Rich Felker [Sat, 26 May 2012 22:04:17 +0000 (18:04 -0400)]
fix overrun (n essentially ignored) in wcsncmp
bug report and solution by Richard Pennington
Rich Felker [Sat, 26 May 2012 22:02:45 +0000 (18:02 -0400)]
fix failure of mbsinit(0) (not UB; required to return nonzero)
issue reported by Richard Pennington; slightly simpler fix applied
Rich Felker [Sat, 26 May 2012 22:01:34 +0000 (18:01 -0400)]
fix failure of strrchr(str, 0)
bug report and solution by Richard Pennington
Rich Felker [Sat, 26 May 2012 02:44:34 +0000 (22:44 -0400)]
avoid using pthread cleanup push/pop in stdio when not needed
unfortunately in dynamic-linked programs, these macros cause
pthread_self to be initialized, which costs a couple syscalls, and
(much worse) would necessarily fail, crash, and burn on ancient (2.4
and earlier) kernels where setting up a thread pointer does not work.
i'd like to do this in a more generic way that avoids all use of
cleanup push/pop before pthread_self has been successfully called and
avoids ugly if/else constructs like the one in this commit, but for
now, this will suffice.
Rich Felker [Sat, 26 May 2012 02:34:09 +0000 (22:34 -0400)]
ensure pthread-internal signals are unblocked before threads are used
if the process started with these signals blocked, cancellation could
fail or setxid could deadlock. there is no way to globally unblock
them after threads have been created. by unblocking them in the
pthread_self initialization for the main thread, we ensure that
they're unblocked before any other threads are created and also
outside of any signal handler context (sigaction initialized
pthread_self), which is important so that return from a signal handler
won't re-block them.
Rich Felker [Fri, 25 May 2012 14:45:05 +0000 (10:45 -0400)]
fix regex on arm
TRE has a broken assumption that wchar_t is signed, which is a sane
expectation, but not required by the standard, and false on ARM's ABI.
i leave tre_char_t as wchar_t for now, since a pointer to it is
directly passed to functions that need pointer to wchar_t. it does not
seem to break anything. and since the maximum unicode scalar value is
0x10ffff, just use that explicitly rather than using the max value of
any particular C type.
Rich Felker [Fri, 25 May 2012 04:59:31 +0000 (00:59 -0400)]
remove cruft from pthread structure (old cancellation stuff)
Rich Felker [Fri, 25 May 2012 04:35:09 +0000 (00:35 -0400)]
remove leftover cp_sp cruft from cancellation code, fix small bug
the bug was that cancellation requests which arrived while a
cancellation point was interrupted by a signal handler would not be
acted upon when the signal handler returns. this was because cp_sp was
never set; it's no longer needed or used.
instead, just always re-raise the signal when cancellation was not
acted upon. this wastes a tiny amount of time in the rare case where
it even matters, but it ensures correctness and simplifies the code.
Rich Felker [Thu, 24 May 2012 16:27:51 +0000 (12:27 -0400)]
fix arm syscall.h to reflect which syscalls actually exist (on EABI)
Rich Felker [Thu, 24 May 2012 16:21:34 +0000 (12:21 -0400)]
avoid deprecated (by linux) alarm syscall; use setitimer instead
Rich Felker [Thu, 24 May 2012 14:55:58 +0000 (10:55 -0400)]
linux deprecated SYS_utime on some archs, so use SYS_utimes instead
the old code could be kept for cases where SYS_utime is available, but
it's not really worth the ifdef ugliness. and better to avoid
deprecated stuff just in case the kernel devs ever get crazy enough to
start removing it from archs where it was part of the ABI and breaking
static bins...
Rich Felker [Wed, 23 May 2012 19:58:53 +0000 (15:58 -0400)]
fix bad opcode in arm syscall_cp_asm
Rich Felker [Wed, 23 May 2012 19:45:41 +0000 (15:45 -0400)]
fix issue with longjmp out of signal handlers and cancellation
stale state information indicating that a thread was possibly blocked
at a cancellation point could get left behind if longjmp was used to
exit a signal handler that interrupted a cancellation point.
to fix the issue, we throw away the state information entirely and
simply compare the saved instruction pointer to a range of code
addresses in the __syscall_cp_asm function. all the ugly PIC work
(which becomes minimal anyway with this approach) is defered to
cancellation time instead of happening at every syscall, which should
improve performance too.
this commit also fixes cancellation on arm, which was mildly broken
(race condition, not checking cancellation flag once inside the
cancellation point zone). apparently i forgot to implement that. the
new arm code is untested, but appears correct; i'll test and fix it
later if there are problems.
Rich Felker [Wed, 23 May 2012 18:13:54 +0000 (14:13 -0400)]
simplify cancellation push/pop slightly
no need to pass unnecessary extra arguments on to the core code in
pthread_create.c. this just wastes cycles and code bloat.
Rich Felker [Wed, 23 May 2012 04:10:39 +0000 (00:10 -0400)]
debloat jmp_buf in _GNU_SOURCE mode
i originally made it the same size as the bloated GNU version, which
contains space for saved signal mask, but this makes some structures
containing jmp_buf become much larger for no benefit. we will never
use the signal mask field with plain setjmp; sigsetjmp serves that
purpose.
Rich Felker [Wed, 23 May 2012 02:43:27 +0000 (22:43 -0400)]
remove everything related to forkall
i made a best attempt, but the intended semantics of this function are
fundamentally contradictory. there is no consistent way to handle
ownership of locks when forking a multi-threaded process. the code
could have worked by accident for programs that only used normal
mutexes and nothing else (since they don't actually store or care
about their owner), but that's about it. broken-by-design interfaces
that aren't even in glibc (only solaris) don't belong in musl.
Rich Felker [Wed, 23 May 2012 02:28:17 +0000 (22:28 -0400)]
some feature test fixes for unistd.h
Rich Felker [Wed, 23 May 2012 02:12:10 +0000 (22:12 -0400)]
fix missing _BSD_SOURCE support in bits/*.h
this is actually rather ugly, and would get even uglier if we ever
want to support further feature test macros. at some point i may
factor the bits headers into separate files for C base, POSIX base,
and nonstandard extensions (the only distinctions that seem to matter
now) and then the logic for which to include can go in the main header
rather than being duplicated for each arch. the downside of this is
that it would result in more files having to be opened during
compilation, so as long as the ugliness does not grow, i'm inclined to
leave it alone for now.
Rich Felker [Wed, 23 May 2012 02:07:42 +0000 (22:07 -0400)]
_GNU_SOURCE implies all BSD features except ones GNU rejects
Rich Felker [Wed, 23 May 2012 02:04:55 +0000 (22:04 -0400)]
various header cleanups, some related to _BSD_SOURCE addition
there is no reason to avoid multiple identical macro definitions; this
is perfectly legal C, and even with the maximal warning options
enabled, gcc does not issue any warning for it.
Rich Felker [Wed, 23 May 2012 01:54:19 +0000 (21:54 -0400)]
bsd_signal is a legacy (removed) XSI function, not needed in _BSD_SOURCE
its only purpose was for use on non-BSD systems that implement sysv
semantics for signal() by default.
Rich Felker [Wed, 23 May 2012 01:52:08 +0000 (21:52 -0400)]
support _BSD_SOURCE feature test macro
patch by Isaac Dunham. matched closely (maybe not exact) to glibc's
idea of what _BSD_SOURCE should make visible.
Rich Felker [Tue, 22 May 2012 03:55:36 +0000 (23:55 -0400)]
fix typo in utimes function that made it mess up file times
Rich Felker [Tue, 22 May 2012 02:51:30 +0000 (22:51 -0400)]
fix out-of-bounds array access in pthread barriers on 64-bit
it's ok to overlap with integer slot 3 on 32-bit because only slots
0-2 are used on process-local barriers.
Rich Felker [Mon, 21 May 2012 02:56:06 +0000 (22:56 -0400)]
move getpass decl to the right place
Rich Felker [Thu, 17 May 2012 00:14:30 +0000 (20:14 -0400)]
fix misplaced semicolon in preprocessor directive (#undef h_errno)
Rich Felker [Mon, 14 May 2012 18:52:55 +0000 (14:52 -0400)]
yet another try to get the check for gcc right...
Rich Felker [Mon, 14 May 2012 12:14:27 +0000 (08:14 -0400)]
fix error in last configure change (lack of escaping)
Rich Felker [Mon, 14 May 2012 04:10:28 +0000 (00:10 -0400)]
useless lastlog path just to make some stuff happy
Rich Felker [Mon, 14 May 2012 04:01:48 +0000 (00:01 -0400)]
missing limit LOGIN_NAME_MAX
Rich Felker [Sun, 13 May 2012 23:59:28 +0000 (19:59 -0400)]
correct the check for gcc (previous version failed for cross compilers)
Rich Felker [Sun, 13 May 2012 21:20:01 +0000 (17:20 -0400)]
remove some no-op end of string tests from regex parser
these are cruft from the original code which used an explicit string
length rather than null termination. i blindly converted all the
checks to null terminator checks, without noticing that in several
cases, the subsequent switch statement would automatically handle the
null byte correctly.
Rich Felker [Sun, 13 May 2012 21:16:10 +0000 (17:16 -0400)]
another BRE fix: in ^*, * is literal
i don't understand why this has to be conditional on being in BRE
mode, but enabling this code unconditionally breaks a huge number of
ERE test cases.
Rich Felker [Sun, 13 May 2012 04:44:35 +0000 (00:44 -0400)]
let sysconf accurately report # of cpus available
i've been trying out openmp and it seems like it won't be much use
without this...
Rich Felker [Sun, 13 May 2012 03:45:07 +0000 (23:45 -0400)]
use __h_errno_location for h_errno
we do not bother making h_errno thread-local since the only interfaces
that use it are inherently non-thread-safe. but still use the
potentially-thread-local ABI to access it just to avoid lock-in.
Rich Felker [Sun, 13 May 2012 03:38:04 +0000 (23:38 -0400)]
susv4 removed gethostbyname, etc. legacy cruft.
Rich Felker [Sun, 13 May 2012 03:34:39 +0000 (23:34 -0400)]
namespace cleanup - NI_* is NOT reserved by netdb.h
Rich Felker [Sun, 13 May 2012 03:31:52 +0000 (23:31 -0400)]
some gnu junk in netdb.h
Rich Felker [Sun, 13 May 2012 02:18:34 +0000 (22:18 -0400)]
fix missing va_list for vsyslog
nsz [Sat, 12 May 2012 23:50:53 +0000 (01:50 +0200)]
search: add comments to tsearch_avl.c
nsz [Sat, 12 May 2012 23:34:20 +0000 (01:34 +0200)]
search: add tdestroy (gnu extension)
Rich Felker [Fri, 11 May 2012 15:05:41 +0000 (11:05 -0400)]
add missing IN6_ARE_ADDR_EQUAL
written to avoid multiple conditional jumps and avoid ugly repetitive
lines in the header file.
Rich Felker [Fri, 11 May 2012 03:32:28 +0000 (23:32 -0400)]
add one more bogus legacy header
this one is for program(s|ers) who haven't heard of uint16_t and
uint32_t (which are obviously the correct types for use in such
situations, as they're the argument/return types for ntohs/htons and
ntohl/htonl).
Rich Felker [Fri, 11 May 2012 02:41:54 +0000 (22:41 -0400)]
move vsyslog out of SYSLOG_NAMES conditional
Rich Felker [Fri, 11 May 2012 02:25:14 +0000 (22:25 -0400)]
Merge remote-tracking branch 'nsz/master'
Rich Felker [Fri, 11 May 2012 02:16:15 +0000 (22:16 -0400)]
remove __lock dependency from exit
there's no sense in using a powerful lock in exit, because it will
never be unlocked. a thread that arrives at exit while exit is already
in progress just needs to hang forever. use the pause syscall for this
because it's cheap and easy and universally available.
Rich Felker [Fri, 11 May 2012 01:51:36 +0000 (21:51 -0400)]
fix missing static (namespace clash)
Rich Felker [Thu, 10 May 2012 16:10:44 +0000 (12:10 -0400)]
fix missing parens in bit op macros (param.h)
Rich Felker [Thu, 10 May 2012 15:59:07 +0000 (11:59 -0400)]
and another bug in setbit, etc. macros..
Rich Felker [Thu, 10 May 2012 15:55:16 +0000 (11:55 -0400)]
fix typo in sys/param.h that broke setbit, etc. macros
this is all junk, but some programs use it.
Rich Felker [Wed, 9 May 2012 15:47:06 +0000 (11:47 -0400)]
omit declaration of basename wrongly interpreted as prototype in C++
the non-prototype declaration of basename in string.h is an ugly
compromise to avoid breaking 2 types of broken software:
1. programs which assume basename is declared in string.h and thus
would suffer from dangerous pointer-truncation if an implicit
declaration were used.
2. programs which include string.h with _GNU_SOURCE defined but then
declare their own prototype for basename using the incorrect GNU
signature for the function (which would clash with a correct
prototype).
however, since C++ does not have non-prototype declarations and
interprets them as prototypes for a function with no arguments, we
must omit it when compiling C++ code. thankfully, all known broken
apps that suffer from the above issues are written in C, not C++.
nsz [Mon, 7 May 2012 22:22:56 +0000 (00:22 +0200)]
math: fix remquo.c when x==-y and a subnormal remainder bug as well
backported fix from freebsd:
http://svnweb.FreeBSD.org/base?view=revision&revision=233973
Rich Felker [Mon, 7 May 2012 22:05:50 +0000 (18:05 -0400)]
some assemblers don't like fistpq; use the alt. mnemonic fistpll
Rich Felker [Mon, 7 May 2012 21:55:13 +0000 (17:55 -0400)]
fix error checking for \ at end of regex (this was broken previously)
Rich Felker [Mon, 7 May 2012 21:50:32 +0000 (17:50 -0400)]
fix copy and paste error in regex code causing mishandling of \) in BRE
Rich Felker [Mon, 7 May 2012 21:43:38 +0000 (17:43 -0400)]
fix regex breakage in last commit (failure to handle empty regex, etc.)
Rich Felker [Mon, 7 May 2012 18:50:49 +0000 (14:50 -0400)]
fix ugly bugs in TRE regex parser
1. * in BRE is not special at the beginning of the regex or a
subexpression. this broke ncurses' build scripts.
2. \\( in BRE is a literal \ followed by a literal (, not a literal \
followed by a subexpression opener.
3. the ^ in \\(^ in BRE is a literal ^ only at the beginning of the
entire BRE. POSIX allows treating it as an anchor at the beginning of
a subexpression, but TRE's code for checking if it was at the
beginning of a subexpression was wrong, and fixing it for the sake of
supporting a non-portable usage was too much trouble when just
removing this non-portable behavior was much easier.
this patch also moved lots of the ugly logic for empty atom checking
out of the default/literal case and into new cases for the relevant
characters. this should make parsing faster and make the code smaller.
if nothing else it's a lot more readable/logical.
at some point i'd like to revisit and overhaul lots of this code...
Rich Felker [Sun, 6 May 2012 21:19:37 +0000 (17:19 -0400)]
update release info for 0.9.0
Rich Felker [Sun, 6 May 2012 20:35:32 +0000 (16:35 -0400)]
some extra legacy header stuff
nsz [Sun, 6 May 2012 19:24:28 +0000 (21:24 +0200)]
add FORCE_EVAL macro to evaluate float expr for their side effect
updated nextafter* to use FORCE_EVAL, it can be used in many other
places in the math code to improve readability.
Rich Felker [Sun, 6 May 2012 18:48:20 +0000 (14:48 -0400)]
fix unused variable warnings in new nextafter/nexttoward code
apparently initializing a variable is not "using" it but assigning to
it is "using" it. i don't really like this fix, but it's better than
trying to make a bigger cleanup just before a release, and it should
work fine (tested against nsz's math tests).
Rich Felker [Sun, 6 May 2012 17:40:19 +0000 (13:40 -0400)]
take byte order from gcc if gcc has defined it
this only works with gcc 4.6 and later, but it allows us to support
non-default endianness on archs like arm, mips, ppc, etc. that can do
both without having separate header sets for both variants, and it
saves one #include even on fixed-endianness archs like x86.
Rich Felker [Sun, 6 May 2012 13:42:23 +0000 (09:42 -0400)]
Merge remote-tracking branch 'nsz/master'
Rich Felker [Sun, 6 May 2012 13:03:19 +0000 (09:03 -0400)]
add isastream (obsolete STREAMS junk)
apparently some packages see stropts.h and want to be able to use
this. the implementation checks that the file descriptor is valid by
using fcntl/F_GETFD so it can report an error if not (as specified).
nsz [Sun, 6 May 2012 11:08:59 +0000 (13:08 +0200)]
math: nextafter and nexttoward cleanup
make nexttoward, nexttowardf independent of long double representation.
fix nextafterl: it did not raise underflow flag when the result was 0.
Rich Felker [Sun, 6 May 2012 02:22:46 +0000 (22:22 -0400)]
fix definitions of FP_ILOGB constants
two issues: (1) the type was wrong (unsigned instead of signed int),
and (2) the value of FP_ILOGBNAN should be INT_MIN rather than INT_MAX
to match the ABI. this is also much more useful since INT_MAX
corresponds to a valid input (infinity). the standard would allow us
to set FP_ILOGB0 to -INT_MAX instead of INT_MIN, which would give us
distinct values for ilogb(0) and ilogb(NAN), but the benefit seems way
too small to justify ignoring the ABI.
note that the macro is just a "portable" (to any twos complement
system where signed and unsigned int have the same width) way to write
INT_MIN without needing limits.h. it's valid to use this method since
these macros are not required to work in #if directives.
Rich Felker [Sat, 5 May 2012 21:18:31 +0000 (17:18 -0400)]
add -frounding-math to build
Rich Felker [Sat, 5 May 2012 06:39:51 +0000 (02:39 -0400)]
update readme and release notes for 0.9.0 release (not yet final)
Rich Felker [Sat, 5 May 2012 04:16:21 +0000 (00:16 -0400)]
relicense musl under MIT license
Rich Felker [Sat, 5 May 2012 04:02:04 +0000 (00:02 -0400)]
update license of njk contributed code (x86_64 asm)
these changes are based on the following communication via email:
"I hereby grant that all of the code I have contributed to musl on or
before April 23, 2012 may be licensed under the terms of the following
MIT license:
Copyright (c) 2011-2012 Nicholas J. Kain
Permission is hereby granted, free of charge, to any person obtaining
a copy of this software and associated documentation files (the
"Software"), to deal in the Software without restriction, including
without limitation the rights to use, copy, modify, merge, publish,
distribute, sublicense, and/or sell copies of the Software, and to
permit persons to whom the Software is furnished to do so, subject to
the following conditions:
The above copyright notice and this permission notice shall be
included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE."
Rich Felker [Sat, 5 May 2012 03:53:50 +0000 (23:53 -0400)]
update INSTALL file to reflect configure script usage
Rich Felker [Sat, 5 May 2012 03:24:51 +0000 (23:24 -0400)]
initial commit of configure script
this script is not based on autoconf; however it attempts to follow
the same interface contracts for ease of integration with build
systems. it is also not necessary to use musl. manually written
config.mak files are still supported, as is building without any
config.mak at all as long as you are happy with the default options
and you supply at least ARCH on the command line to make.
Rich Felker [Sat, 5 May 2012 02:51:59 +0000 (22:51 -0400)]
make pthread stacks non-executable
this change is necessary or pthread_create will always fail on
security-hardened kernels. i considered first trying to make the stack
executable and simply retrying without execute permissions when the
first try fails, but (1) this would incur a serious performance
penalty on hardened systems, and (2) having the stack be executable is
just a bad idea from a security standpoint.
if there is real-world "GNU C" code that uses nested functions with
threads, and it can't be fixed, we'll have to consider other ways of
solving the problem, but for now this seems like the best fix.
Rich Felker [Sat, 5 May 2012 01:54:57 +0000 (21:54 -0400)]
improve the build rules for installing /lib/ld-musl-$ARCH.so
these new rules should avoid spurious error messages when the
directory (usually /lib) and the dynamic linker symlink already exist,
and minimize the spam when they can't be created.
Rich Felker [Sat, 5 May 2012 00:56:01 +0000 (20:56 -0400)]
Merge remote-tracking branch 'nsz/master'
Rich Felker [Sat, 5 May 2012 00:18:18 +0000 (20:18 -0400)]
fix error reporting for dlsym with global symbols
nsz [Fri, 4 May 2012 23:11:56 +0000 (01:11 +0200)]
math: change the formula used for acos.s
old: 2*atan2(sqrt(1-x),sqrt(1+x))
new: atan2(fabs(sqrt((1-x)*(1+x))),x)
improvements:
* all edge cases are fixed (sign of zero in downward rounding)
* a bit faster (here a single call is about 131ns vs 162ns)
* a bit more precise (at most 1ulp error on 1M uniform random
samples in [0,1), the old formula gave some 2ulp errors as well)
Rich Felker [Fri, 4 May 2012 05:26:43 +0000 (01:26 -0400)]
fix uninitialized var in vfwprintf printing 0-prec string
this could lead to spurious failures of wide printf functions
Rich Felker [Fri, 4 May 2012 04:31:25 +0000 (00:31 -0400)]
add *64 junk for sys/*.h headers
Rich Felker [Fri, 4 May 2012 04:13:23 +0000 (00:13 -0400)]
add support for ugly *64 functions with _LARGEFILE64_SOURCE
musl does not support legacy 32-bit-off_t whatsoever. off_t is always
64 bit, and correct programs that use off_t and the standard functions
will just work out of the box. (on glibc, they would require
-D_FILE_OFFSET_BITS=64 to work.) however, some programs instead define
_LARGEFILE64_SOURCE and use alternate versions of all the standard
types and functions with "64" appended to their names.
we do not want code to actually get linked against these functions
(it's ugly and inconsistent), so macros are used instead of prototypes
with weak aliases in the library itself. eventually the weak aliases
may be added at the library level for the sake of using code that was
originally built against glibc, but the macros will still be the
desired solution in the headers.
Rich Felker [Fri, 4 May 2012 03:19:29 +0000 (23:19 -0400)]
avoid setting nondefault scheduler too
Rich Felker [Fri, 4 May 2012 03:18:26 +0000 (23:18 -0400)]
implement stub versions of sched_*
these actually work, but for now they prohibit actually setting
priority levels and report min/max priority as 0.
Rich Felker [Fri, 4 May 2012 02:27:36 +0000 (22:27 -0400)]
uglify headers for the sake of junk that compiles with gcc -std=c89/-ansi
Rich Felker [Fri, 4 May 2012 02:12:46 +0000 (22:12 -0400)]
add additional compatibility union member for ipv6 addresses
in6_* is in the reserved namespace, so this is valid
Rich Felker [Fri, 4 May 2012 00:42:45 +0000 (20:42 -0400)]
overhaul SSP support to use a real canary
pthread structure has been adjusted to match the glibc/GCC abi for
where the canary is stored on i386 and x86_64. it will need variants
for other archs to provide the added security of the canary's entropy,
but even without that it still works as well as the old "minimal" ssp
support. eventually such changes will be made anyway, since they are
also needed for GCC/C11 thread-local storage support (not yet
implemented).
care is taken not to attempt initializing the thread pointer unless
the program actually uses SSP (by reference to __stack_chk_fail).
Rich Felker [Fri, 4 May 2012 00:35:11 +0000 (20:35 -0400)]
make all .o files depend on *_impl.h and libc.h
hopefully the annoyance of this will be minimal. these files all
define internal interfaces which can change at any time; if different
modules are using different versions of the interfaces, the library
will badly break. ideally we would scan and add the dependency only
for C files that actually reference the affected interfaces, but for
now, err on the side of caution and force a rebuild of everything if
any of them have changed.
this commit is in preparation for the upcoming ssp overhaul commit,
which will change internals of the pthread struct.