Rich Felker [Sun, 24 Jul 2011 04:54:36 +0000 (00:54 -0400)]
const correctness on function pointer
Rich Felker [Sun, 24 Jul 2011 04:26:12 +0000 (00:26 -0400)]
simplify dynamic linker startup
instead of creating temp dso objects on the stack and moving them to
the heap if dlopen/dlsym are used, use static objects to begin with,
and just donate them to malloc if we no longer need them.
Rich Felker [Sun, 24 Jul 2011 03:45:33 +0000 (23:45 -0400)]
some preliminaries for vdso clock support
these changes also make it so clock_gettime(CLOCK_REALTIME, &ts) works
even on pre-2.6 kernels, emulated via the gettimeofday syscall. there
is no cost for the fallback check, as it falls under the error case
that already must be checked for storing the error code in errno, but
which would normally be hidden inside __syscall_ret.
Rich Felker [Fri, 22 Jul 2011 04:25:56 +0000 (00:25 -0400)]
check for fd exhaustion in forkpty
we cannot report failure after forking, so the idea is to ensure prior
to fork that fd 0,1,2 exist. this will prevent dup2 from possibly
hitting a resource limit and failing in the child process. fcntl
rather than dup2 is used prior to forking to avoid race conditions.
Rich Felker [Fri, 22 Jul 2011 04:23:36 +0000 (00:23 -0400)]
incorrect check for open failure in openpty function
-1, not 0, indicates failure
Rich Felker [Fri, 22 Jul 2011 02:44:05 +0000 (22:44 -0400)]
socket headers macro adjustment - workaround for buggy programs
some program was undefining AF_NETLINK and thereby breaking AF_ROUTE...
Rich Felker [Fri, 22 Jul 2011 01:15:14 +0000 (21:15 -0400)]
fix errno value when fdopendir is given an invalid file descriptor
this resolves an issue reported by Vasiliy Kulikov
Rich Felker [Sun, 17 Jul 2011 03:17:17 +0000 (23:17 -0400)]
ensure in fork that child gets its own new robust mutex list
Rich Felker [Sun, 17 Jul 2011 01:24:02 +0000 (21:24 -0400)]
fix logic error in fread
fread was calling f->read without checking that the file was in
reading mode. this could:
1. crash, if f->read was a null pointer
2. cause unwanted blocking on a terminal already at eof
3. allow reading on a write-only file
Rich Felker [Fri, 15 Jul 2011 02:11:00 +0000 (22:11 -0400)]
fix various bugs in new integer parser framework
1. my interpretation of subject sequence definition was wrong. adjust
parser to conform to the standard.
2. some code for handling tail overflow case was missing (forgot to
finish writing it).
3. typo (= instead of ==) caused ERANGE to wrongly behave like EINVAL
Rich Felker [Thu, 14 Jul 2011 05:12:05 +0000 (01:12 -0400)]
fix wcsto[iu]max with high characters
stopping without letting the parser see a stop character prevented
getting a result. so treat all high chars as the null character and
pass them into the parser.
also eliminated ugly tmp var using compound literals.
Rich Felker [Thu, 14 Jul 2011 04:51:45 +0000 (00:51 -0400)]
new restartable integer parsing framework.
this fixes a number of bugs in integer parsing due to lazy haphazard
wrapping, as well as some misinterpretations of the standard. the new
parser is able to work character-at-a-time or on whole strings, making
it easy to support the wide functions without unbounded space for
conversion. it will also be possible to update scanf to use the new
parser.
Rich Felker [Wed, 13 Jul 2011 00:30:04 +0000 (20:30 -0400)]
gb18030 support in iconv (only from, not to)
also support (and restrict to subsets) older chinese sets, and
explicitly refuse to convert to cjk (since there's no code for it yet)
Rich Felker [Tue, 12 Jul 2011 06:52:06 +0000 (02:52 -0400)]
"implement" getnetbyaddr and getnetbyname
these are useless legacy functions but some old software contains
cruft that expects them to exist...
Rich Felker [Tue, 12 Jul 2011 06:43:24 +0000 (02:43 -0400)]
legacy japanese charset support in iconv (only from, not to)
Rich Felker [Tue, 12 Jul 2011 04:31:39 +0000 (00:31 -0400)]
simplify iconv and support more legacy codepages
Rich Felker [Sat, 9 Jul 2011 22:06:59 +0000 (18:06 -0400)]
add missing signalfd flags
Rich Felker [Mon, 4 Jul 2011 15:55:52 +0000 (11:55 -0400)]
printf: "if a precision is specified, the '0' flag shall be ignored."
Rich Felker [Mon, 4 Jul 2011 05:57:00 +0000 (01:57 -0400)]
zero precision with zero value should not inhibit prefix/width printing
Rich Felker [Mon, 4 Jul 2011 05:01:58 +0000 (01:01 -0400)]
printf("%#x",0) should print 0 not 0x0
Rich Felker [Sun, 3 Jul 2011 23:26:12 +0000 (19:26 -0400)]
iconv was not returning -1 on most failure
this broke most uses of iconv in real-world programs, especially
glib's iconv wrappers.
Rich Felker [Sun, 3 Jul 2011 20:41:20 +0000 (16:41 -0400)]
0.7.12 release notes
Rich Felker [Sat, 2 Jul 2011 02:40:00 +0000 (22:40 -0400)]
fix dlopen UB due to longjmp/volatile rules violation
Rich Felker [Fri, 1 Jul 2011 03:12:07 +0000 (23:12 -0400)]
res_search symbol, aliased to res_query for now (better than nothing)
Rich Felker [Fri, 1 Jul 2011 03:02:27 +0000 (23:02 -0400)]
simple rpath support (no token expansion yet) for dynamic linker
Rich Felker [Fri, 1 Jul 2011 00:23:24 +0000 (20:23 -0400)]
fill in junk in stropts.h
STREAMS are utterly useless as far as I can tell, but some software
was apparently broken by the presence of stropts.h but lack of macros
it's supposed to define...
Rich Felker [Thu, 30 Jun 2011 20:37:51 +0000 (16:37 -0400)]
fix error in previous ld80 fpclassify commit
Rich Felker [Thu, 30 Jun 2011 20:31:43 +0000 (16:31 -0400)]
catch invalid ld80 bit patterns and treat them as nan
this should not be necessary - the invalid bit patterns cannot be
created except through type punning. however, some broken gnu software
is passing them to printf and triggering dangerous stack-smashing, so
let's catch them anyway...
Rich Felker [Thu, 30 Jun 2011 17:27:08 +0000 (13:27 -0400)]
fix logic in __fwriting
Rich Felker [Thu, 30 Jun 2011 16:44:48 +0000 (12:44 -0400)]
add and consolidate nasty stdio_ext junk
hopefully this resolves the rest of the issues with hideously
nonportable hacks in programs that use gnulib.
Rich Felker [Thu, 30 Jun 2011 15:42:33 +0000 (11:42 -0400)]
implement the nonstandard GNU function fpurge
this is a really ugly and backwards function, but its presence will
prevent lots of broken gnulib software from trying to define its own
version of fpurge and thereby failing to build or worse.
Rich Felker [Thu, 30 Jun 2011 12:11:06 +0000 (08:11 -0400)]
fix buffer overrun in getgrent code when there are no group members
Rich Felker [Wed, 29 Jun 2011 23:26:30 +0000 (19:26 -0400)]
posix_memalign should fail if size is not a multiple of sizeof(void *)
Rich Felker [Wed, 29 Jun 2011 21:13:01 +0000 (17:13 -0400)]
avoid errors in ucontext.h when no feature test macros are defined
Rich Felker [Wed, 29 Jun 2011 19:29:52 +0000 (15:29 -0400)]
locking support for random() prng
these interfaces are required to be thread-safe even though they are
not state-free. the random number sequence is shared across all
threads.
Rich Felker [Wed, 29 Jun 2011 04:42:42 +0000 (00:42 -0400)]
work around linux bug in mprotect
per POSIX: The mprotect() function shall change the access protections
to be that specified by prot for those whole pages containing any part
of the address space of the process starting at address addr and
continuing for len bytes.
on the other hand, linux mprotect fails with EINVAL if the base
address and/or length is not page-aligned, so we have to align them
before making the syscall.
Rich Felker [Wed, 29 Jun 2011 04:29:08 +0000 (00:29 -0400)]
textrel support, cheap and ugly
Rich Felker [Wed, 29 Jun 2011 02:06:58 +0000 (22:06 -0400)]
release notes for 0.7.11
Rich Felker [Tue, 28 Jun 2011 23:40:14 +0000 (19:40 -0400)]
reclaim the memory wasted by dynamic linking for use by malloc
Rich Felker [Tue, 28 Jun 2011 18:41:41 +0000 (14:41 -0400)]
use type directives for fenv asm functions
Rich Felker [Tue, 28 Jun 2011 18:20:41 +0000 (14:20 -0400)]
use load address from elf header if possible
this is mostly useless for shared libs (though it could help for
prelink-like purposes); the intended use case is for adding support
for calling the dynamic linker directly to run a program, as in:
./libc.so ./a.out foo
this usage is not yet supported.
Rich Felker [Tue, 28 Jun 2011 18:13:51 +0000 (14:13 -0400)]
make dynamic linker relocate the main program image last, after all libs
prior to this change, copy relocations for initialized pointer
variables would not reflect the relocated contents of the pointer.
Rich Felker [Tue, 28 Jun 2011 12:27:38 +0000 (08:27 -0400)]
fix a few bugs from last dynamic linking build system commit
some cruft was left and DESTDIR was not being used correctly.
Rich Felker [Tue, 28 Jun 2011 02:34:47 +0000 (22:34 -0400)]
minor updates to INSTALL documentation
Rich Felker [Tue, 28 Jun 2011 01:38:11 +0000 (21:38 -0400)]
cleanup shared library build system to be more $HOME-local-install friendly
the path for the dynamic linker is now configurable, and failure to
install the symlink for it will not stop the build.
Rich Felker [Mon, 27 Jun 2011 05:02:28 +0000 (01:02 -0400)]
further fixup dlfcn.h
Rich Felker [Mon, 27 Jun 2011 05:01:19 +0000 (01:01 -0400)]
match LSB/glibc constants for dynamic loader
Rich Felker [Mon, 27 Jun 2011 02:39:34 +0000 (22:39 -0400)]
fix stale pointer issue in dynamic linker with dlopen
Rich Felker [Mon, 27 Jun 2011 02:09:32 +0000 (22:09 -0400)]
don't leave the lock held on dlopen failure..
Rich Felker [Mon, 27 Jun 2011 01:50:01 +0000 (21:50 -0400)]
add RTLD_DEFAULT support
Rich Felker [Mon, 27 Jun 2011 01:36:44 +0000 (21:36 -0400)]
in dlopen: don't use null pointer
deps can be null if a library has no dependencies (such as libc itself)
Rich Felker [Mon, 27 Jun 2011 01:21:04 +0000 (21:21 -0400)]
fix resolving symbols in objects loaded in RTLD_LOCAL mode
basically we temporarily make the library and all its dependencies
part of the global namespace but only for the duration of performing
relocations, then return them to their former state.
Rich Felker [Sun, 26 Jun 2011 23:23:28 +0000 (19:23 -0400)]
experimental dlopen/dlsym and dynlink changes needed to support them
Rich Felker [Sun, 26 Jun 2011 21:41:34 +0000 (17:41 -0400)]
type directives for x86_64 math asm
Rich Felker [Sun, 26 Jun 2011 21:39:17 +0000 (17:39 -0400)]
error handling in dynamic linking
some of the code is not yet used, and is in preparation for dlopen
which needs to be able to handle failure loading libraries without
terminating the program.
Rich Felker [Sun, 26 Jun 2011 20:34:05 +0000 (16:34 -0400)]
fix useless use of potentially-uninitialized mode variable in sem_open
Rich Felker [Sun, 26 Jun 2011 20:12:43 +0000 (16:12 -0400)]
eliminate OOB array hacks in malloc
Rich Felker [Sun, 26 Jun 2011 19:52:37 +0000 (15:52 -0400)]
use .type directives for math asm (needed for dynamic linking to work)
Rich Felker [Sun, 26 Jun 2011 02:36:21 +0000 (22:36 -0400)]
fix some symbol resolution issues in dynamic linker
1. search was wrongly beginning with lib itself rather than dso head
2. inconsistent resolution of function pointers for functions in plt
Rich Felker [Sat, 25 Jun 2011 22:57:17 +0000 (18:57 -0400)]
fix dynamic linker issue in musl-gcc wrapper
Rich Felker [Sat, 25 Jun 2011 22:54:33 +0000 (18:54 -0400)]
wordexp cannot use we_offs unless WRDE_DOOFFS flag is set
previously, a potentially-indeterminate value from we_offs was being
used, resulting in wrong we_wordc and subsequent crashes in the
caller.
Rich Felker [Sat, 25 Jun 2011 22:18:57 +0000 (18:18 -0400)]
XSI search.h API implementation by Szabolcs Nagy
Rich Felker [Sat, 25 Jun 2011 21:49:16 +0000 (17:49 -0400)]
handle library paths better (ignore empty path components, etc.)
Rich Felker [Sat, 25 Jun 2011 19:38:00 +0000 (15:38 -0400)]
move all limits that don't vary out of bits/limits.h, into main limits.h
Rich Felker [Sat, 25 Jun 2011 16:26:08 +0000 (12:26 -0400)]
fix possible (questionable) strict aliasing violations and ptr arith
Rich Felker [Sat, 25 Jun 2011 13:23:36 +0000 (09:23 -0400)]
const correctness in mq_notify
why did gcc allow this invalid assignment to compile in the first place?
Rich Felker [Sat, 25 Jun 2011 05:56:34 +0000 (01:56 -0400)]
proper path searching for dynamic linker
first, use $LD_LIBRARY_PATH unless suid. if that fails, read path from
/etc/ld-musl-$ARCH.path and fallback to a builtin default.
Rich Felker [Sat, 25 Jun 2011 05:38:25 +0000 (01:38 -0400)]
optimize opendir using O_CLOEXEC
Rich Felker [Sat, 25 Jun 2011 04:47:28 +0000 (00:47 -0400)]
discard dso descriptors after performing relocations
eventually (once dlopen exists) this behavior will be conditional on
dlopen/dlsym not being reachable.
Rich Felker [Sat, 25 Jun 2011 04:18:19 +0000 (00:18 -0400)]
keep track of which dsos have been relocated
Rich Felker [Fri, 24 Jun 2011 22:32:30 +0000 (18:32 -0400)]
use soname in ld-musl.so to prevent filename appearing in DT_NEEDEDs
Rich Felker [Fri, 24 Jun 2011 18:02:47 +0000 (14:02 -0400)]
use symlink rather than bogus linker script for libc.so
the linker script caused a bogus DT_NEEDED entry
Rich Felker [Fri, 24 Jun 2011 02:13:47 +0000 (22:13 -0400)]
adapt build/install/gcc-wrapper systems for dynamic linking support
Rich Felker [Fri, 24 Jun 2011 02:04:06 +0000 (22:04 -0400)]
prepare support for LD_LIBRARY_PATH (checking suid/sgid safety)
the use of this test will be much stricter than glibc and other
typical implementations; the environment will not be honored
whatsoever unless the program is confirmed non-suid/sgid by the aux
vector the kernel passed in. no fallback to slow syscall-based
checking is used if the kernel fails to provide the information; we
simply assume the worst (suid) in this case and refuse to honor
environment.
Rich Felker [Thu, 23 Jun 2011 22:02:23 +0000 (18:02 -0400)]
initial commit of prng implementation by Szabolcs Nagy
Rich Felker [Mon, 20 Jun 2011 02:42:10 +0000 (22:42 -0400)]
make ldso asm more uniform with rest of codebase (no unnecessary suffixes)
Rich Felker [Sun, 19 Jun 2011 12:33:27 +0000 (08:33 -0400)]
x86_64: ensure that dtor pointer passed to app is 0 so it won't be used
leaving it uninitialized caused unpredictable crashes or worse due to
calling an indeterminate function pointer.
Rich Felker [Sun, 19 Jun 2011 02:52:01 +0000 (22:52 -0400)]
dynamic linker 64bit fix: hash table entries are always 32bit
Rich Felker [Sat, 18 Jun 2011 23:48:42 +0000 (19:48 -0400)]
experimental dynamic linker!
some notes:
- library search path is hard coded
- x86_64 code is untested and may not work
- dlopen/dlsym is not yet implemented
- relocations in read-only memory won't work
Rich Felker [Sat, 18 Jun 2011 11:41:14 +0000 (07:41 -0400)]
fix memory leak on failure in realpath
Rich Felker [Sat, 18 Jun 2011 11:39:05 +0000 (07:39 -0400)]
copy-paste error in prctl.h
Rich Felker [Thu, 16 Jun 2011 21:11:35 +0000 (17:11 -0400)]
fix some struct padding to match LSB/glibc ABI where it may be helpful
Rich Felker [Thu, 16 Jun 2011 20:53:11 +0000 (16:53 -0400)]
duplicate re_nsub in LSB/glibc ABI compatible location
Rich Felker [Wed, 15 Jun 2011 03:15:08 +0000 (23:15 -0400)]
restore use of .type in asm, but use modern @function (vs %function)
this seems to be necessary to make the linker accept the functions in
a shared library (perhaps to generate PLT entries?)
strictly speaking libc-internal asm should not need it. i might clean
that up later.
Rich Felker [Wed, 15 Jun 2011 03:14:22 +0000 (23:14 -0400)]
__syscall5 inline is having trouble with PIC; just use the function for now
Rich Felker [Tue, 14 Jun 2011 05:35:51 +0000 (01:35 -0400)]
fix race condition in pthread_kill
if thread id was reused by the kernel between the time pthread_kill
read it from the userspace pthread_t object and the time of the tgkill
syscall, a signal could be sent to the wrong thread. the tgkill
syscall was supposed to prevent this race (versus the old tkill
syscall) but it can't; it can only help in the case where the tid is
reused in a different process, but not when the tid is reused in the
same process.
the only solution i can see is an extra lock to prevent threads from
exiting while another thread is trying to pthread_kill them. it should
be very very cheap in the non-contended case.
Rich Felker [Tue, 14 Jun 2011 05:25:17 +0000 (01:25 -0400)]
run dtors before taking the exit-lock in pthread exit
previously a long-running dtor could cause pthread_detach to block.
Rich Felker [Tue, 14 Jun 2011 05:23:42 +0000 (01:23 -0400)]
minor locking optimizations
Rich Felker [Tue, 14 Jun 2011 00:52:01 +0000 (20:52 -0400)]
avoid 64bit warnings when using pointers as entropy for temp names
Rich Felker [Tue, 14 Jun 2011 00:37:52 +0000 (20:37 -0400)]
fix sigset macro for 64-bit systems (<< was overflowing due to wrong type)
Rich Felker [Tue, 14 Jun 2011 00:28:14 +0000 (20:28 -0400)]
remove all .size and .type directives for functions from the asm
these are useless and have caused problems for users trying to build
with non-gnu tools like tcc's assembler.
Rich Felker [Tue, 14 Jun 2011 00:21:21 +0000 (20:21 -0400)]
remove old useless timezone.s file (unused)
Rich Felker [Mon, 13 Jun 2011 18:06:04 +0000 (14:06 -0400)]
fix wrong type for wcsrchr argument 2
Rich Felker [Mon, 13 Jun 2011 17:37:43 +0000 (13:37 -0400)]
x86_64 fenv support (untested; at least known to build successfully)
Rich Felker [Mon, 13 Jun 2011 17:37:16 +0000 (13:37 -0400)]
fix fesetround - it was writing to status word instead of control word
Rich Felker [Sun, 12 Jun 2011 19:58:15 +0000 (15:58 -0400)]
floating point environment, untested
at present the i386 code does not support sse floating point, which is
not part of the standard i386 abi. while it may be desirable to
support it later, doing so will reduce performance and require some
tricks to probe if sse support is present.
this first commit is i386-only, but it should be trivial to port the
asm to x86_64.
Rich Felker [Sun, 12 Jun 2011 14:53:42 +0000 (10:53 -0400)]
malloc: cast size down to int in bin_index functions
even if size_t was 32-bit already, the fact that the value was
unsigned and that gcc is too stupid to figure out it would be positive
as a signed quantity (due to the immediately-prior arithmetic and
conditionals) results in gcc compiling the integer-to-float conversion
as zero extension to 64 bits followed by an "fildll" (64 bit)
instruction rather than a simple "fildl" (32 bit) instruction on x86.
reportedly fildll is very slow on certain p4-class machines; even if
not, the new code is slightly smaller.
Rich Felker [Sun, 12 Jun 2011 14:25:29 +0000 (10:25 -0400)]
another return value fix for mktemp...
Rich Felker [Fri, 10 Jun 2011 01:47:24 +0000 (21:47 -0400)]
add netpacket/packet.h
looks like busybox is going to want it, and apparently some other
low-level network software does too...
Rich Felker [Wed, 8 Jun 2011 20:47:08 +0000 (16:47 -0400)]
it's called getgrgid_r, not getgruid_r...