this slightly reduces the code size cost of TLS/thread-pointer for
static linking since __init_tp can be inlined into its only caller and
removed. this is analogous to the handling of __init_libc in
__libc_start_main, where the function only has external linkage when
it needs to be called from the dynamic linker.

the implicit-operand form of fucomip is rejected by binutils 2.19 and
perhaps other versions still in use. writing both operands explicitly
fixes the issue. there is no change to the resulting output.

commit a732e80d was the source of this
regression.

use CAS instead of swap since it's lighter for most archs, and keep
EBUSY in the lock value so that the old value obtained by CAS can be
used directly as the return value for pthread_spin_trylock.

the motivation for this change is that the extra declaration (with or
without visibility) using "struct _IO_FILE" instead of "FILE" seems to
trigger a bug in gcc 3.x where it considers the types mismatched.
however, this change also results in slightly better code and it is
valid because (1) these three objects are constant, and (2) applying
the & operator to any of them is invalid C, since they are not even
specified to be objects. thus it does not matter if the application
and libc see different addresses for them, as long as the (initial,
unchanging) value is seen the same by both.

these are used as hidden by asm files (and such use is the whole
reason they exist), but their actual definitions were not hidden.

commit f9cccfc1 left behind the part
in libc.c; remove it too.

part of the goal here is to eliminate use of the ATTR_LIBC_VISIBILITY
macro outside of libc.h, since it was never intended to be 'public'.

these were hacks to work around toolchains that could not properly
optimize PIC accesses based on visibility and would generate GOT
lookups even for hidden data, which broke the old dynamic linker.
since commit f3ddd173 it no longer
matters; the dynamic linker does not assume accessibility of this data
until stage 3.

when the non-stub duplocale code was added as part of the locale
framework in commit 0bc03091, the old
code to memcpy the old locale object to the new one was left behind.
the conditional for the memcpy no longer makes sense, because the
conditions are now always-true when it's reached, and the memcpy is
wrong because it clobbers the new->messages_name pointer setup just
above.

since the messages_name and ctype_utf8 members have already been
copied, all that remains is the cat[] array. these pointers are
volatile, so using memcpy to copy them is formally wrong; use a for
loop instead.

the first switch already returns in the F_SETLKW code path so it need
not be handled in the second switch. moreover the code in the second
switch is wrong for the F_SETLKW command: it's not cancellable.

the leak was found by static analysis (reported by Alexander Monakov),
not tested/observed, but seems to have occured both when failing due
to O_EXCL, and in a race condition with O_CREAT but not O_EXCL where a
semaphore by the same name was created concurrently.

the allocating path which can fail is for dynamic TLS, which can only
occur at runtime, and the check for runtime was already made in the
outer conditional.

commit 637dd2d3 introduced the checks
for RTLD_DEFAULT and RTLD_NEXT here, claiming they fixed a regression,
but the above conditional block clearly already covered these cases,
and removing the checks produces no difference in the generated code.

this is implemented via the build system and does not affect source
files. the idea is to use protected or hidden visibility to prevent
the compiler from pessimizing function calls within a shared (or
position-independent static) libc in the form of overhead setting up
for a call through the PLT. the ld-time symbol binding via the
-Bsymbolic-functions option already optimized out the PLT itself, but
not the code in the caller needed to support a call through the PLT.
on some archs this overhead can be substantial; on others it's
trivial.

analogous to commit 646cb9a4 for sh.

these are perfectly fine with ld-time symbol binding, but otherwise
result in textrels. they cannot be replaced with @PLT jump targets
because the PLT thunks require a GOT register to be setup, so use a
hidden alias instead.

analogous to commit 646cb9a4 for sh.

these are perfectly fine with ld-time symbol binding, but if the calls
go through a PLT thunk, they are invalid because the caller does not
setup a GOT register. use a hidden alias to bypass the issue.

analogous to commit 8ed66ecb for i386.

none of these are actual textrels because of ld-time binding performed
by -Bsymbolic-functions, but I'm changing them with the goal of making
ld-time binding purely an optimization rather than relying on it for
semantic purposes.

in the case of memmove's call to memcpy, making it explicit that the
memmove asm is assuming the forward-copying behavior of the memcpy asm
is desirable anyway; in case memcpy is ever changed, the semantic
mismatch would be apparent while editing memmcpy.s.

this fixes truncation of error messages containing long pathnames or
symbol names.

the dlerror state was previously required by POSIX to be global. the
resolution of bug 97 relaxed the requirements to allow thread-safe
implementations of dlerror with thread-local state and message buffer.

this reverts the commit f29fea00
which was based on a bug in C99 and POSIX and did not match IEEE-754
http://www.open-std.org/jtc1/sc22/wg14/www/docs/n1515.pdf

these functions are never called directly; only their addresses are
used, so PLT indirections should never happen unless a broken
application tries to redefine them, but it's still best to make them
hidden.

this change is made in preparation to support linking without
-Bsymbolic-functions.

commit 5fc14878 attempted to fix it,
but neglected the fact that mips has branch delay slots.

the braf instruction's destination register is an offset from the
address of the braf instruction plus 4 (or equivalently, the address
of the next instruction after the delay slot). the code for dlsym was
incorrectly computing the offset to pass using the address of the
delay slot itself. in other places, a label was placed after the delay
slot, but I find this confusing. putting the label on the branch
instruction itself, and manually adding 4, makes it more clear which
branch the offset in the constant pool goes with.

even hidden functions need @PLT symbol references; otherwise an
absolute address is produced instead of a PC-relative one.

this caused the dynamic linker/startup code to abort when r0 happened
to contain a negative value.

the conventional way to implement sigsetjmp is to save the signal mask
then tail-call to setjmp; siglongjmp then restores the signal mask and
calls longjmp. the problem with this approach is that a signal already
pending, or arriving between unmasking of signals and restoration of
the saved stack pointer, will have its signal handler run on the stack
that was active before siglongjmp was called. this can lead to
unbounded stack usage when siglongjmp is used to leave a signal
handler.

in the new design, sigsetjmp saves its own return address inside the
extended part of the sigjmp_buf (outside the __jmp_buf part used by
setjmp) then calls setjmp to save a jmp_buf inside its own execution.
it then tail-calls to __sigsetjmp_tail, which uses the return value of
setjmp to determine whether to save the current signal mask or restore
a previously-saved mask.

as an added bonus, this design makes it so that siglongjmp and longjmp
are identical. this is useful because the __longjmp_chk function we
need to add for ABI-compatibility assumes siglongjmp and longjmp are
the same, but for different reasons -- it was designed assuming either
can access a flag just past the __jmp_buf indicating whether the
signal masked was saved, and act on that flag. however, early versions
of musl did not have space past the __jmp_buf for the non-sigjmp_buf
version of jmp_buf, so our setjmp cannot store such a flag without
risking clobbering memory on (very) old binaries.

previously, the dynamic tlsdesc lookup functions and the i386
special-ABI ___tls_get_addr (3 underscores) function called
__tls_get_addr when the slot they wanted was not already setup;
__tls_get_addr would then in turn also see that it's not setup and
call __tls_get_new.

calling __tls_get_new directly is both more efficient and avoids the
issue of calling a non-hidden (public API/ABI) function from asm.

for the special i386 function, a weak reference to __tls_get_new is
used since this function is not defined when static linking (the code
path that needs it is unreachable in static-linked programs).

applying the attribute to a weak_alias macro was a hack. instead use a
separate declaration to apply the visibility, and consolidate
declarations together to avoid having visibility mess all over the
file.

in a few places, non-hidden symbols were referenced from asm in ways
that assumed ld-time binding. while these is no semantic reason these
symbols need to be hidden, fixing the references without making them
hidden was going to be ugly, and hidden reduces some bloat anyway.

in the asm files, .global/.hidden directives have been moved to the
top to unclutter the actual code.

at the point of call it was declared hidden, but the definition was
not hidden. for some toolchains this inconsistency produced textrels
without ld-time binding.

otherwise the call instruction in the inline syscall asm results in
textrels without ld-time binding.