not heavily tested, but at least they don't seem to break anything on
soft float targets with or without coprocessors. they check the auxv
AT_HWCAP flags to determine which coprocessor, if any, is available.

this allows config.mak to override the default -lgcc for building with
other compilers such as pcc/clang/etc.

since the correct declaration was not visible, and since the
representation of the types wchar_t and wint_t always match, a
compiler would have to go out of its way to make this bug manifest,
but better to fix it anyway.

the original code was wrongly based on how it would be done in thumb
mode, but that's not needed because musl's asm only targets arm.

it's expected that this will be needed/useful only in asm, so I've
given it its own symbol that can be addressed in pc-relative ways from
asm rather than adding a field in the __libc structure which would
require hard-coding the offset wherever it's used.

this seems counter-intuitive since sem_trywait is supposed to just try
once, not wait for the semaphore. however, the retry loop is not a
wait. instead, it's to handle the case where the value changes due to
a simultaneous post or wait from another thread while the semaphore
value remains positive. in such a case, it's absolutely wrong for
sem_trywait to fail with EAGAIN because the semaphore is not busy.

this is needed in case -lgcc is passed explicitly on the link command
line, for example if the wrapper is being used to build musl itself.

based on patches by orc and Isaac Dunham, with some fixes. sys/io.h
exists and contains prototypes for these functions regardless of
whether the target arch has them; this is a bit unorthodox but I don't
think it will break anything. the function definitions do not exist
unless the appropriate SYS_* syscall number macro is defined, which
should make sure configure scripts looking for these functions don't
find them on other systems.

presently, sys/io.h does not have the inb/outb/etc. port io
macros/functions. I'd be surprised if ioperm/iopl are useful without
them, so they probably need to be added at some point in appropriate
bits/io.h files...

based on patches by orc and Isaac Dunham.

based on patch by orc and Isaac Dunham, with some fixes.

based on patch by orc and Isaac Dunham, with some details fixed.

128 is the size in bytes, not longs.

also fix the alignment of jmp_buf to meet the abi. linux always
emulates fpu on mips if it's not present, so enabling this code
unconditionally is "safe" but may be slow. in the long term it may be
preferable to find a way to disable it on soft float builds.

untested but should be correct..

not sure this is the best fix but it should work

use the main program's PT_INTERP header if possible, since this is
sure to be a correct (and hopefully absolute) pathname.

sc was overwriting the result

the fields in the mcontext_t are long long (for no good reason) even
on 32-bit mips, so the offset of the instruction pointer (as a word)
varies depending on endianness.

the kernel wrongly expects the cmsg length field to be size_t instead
of socklen_t. in order to work around the issue, we have to impose a
length limit and copy to a local buffer. the length limit should be
more than sufficient for any real-world use; these headers are only
used for passing file descriptors and permissions between processes
over unix sockets.

this fix is easier than trying to reorder the header stuff

signal handling was very broken because of this

like arm, mips requires 64-bit arguments to be "aligned" on an even
register boundary.

these could have caused memory corruption due to invalid accesses to
the next field. all should be fixed now; I found the errors with fgrep
-r '__lock(&', which is bogus since the argument should be an array.

it had not been updated for the futex-based locks

otherwise offs in ucontext_t will be wrong, and break code that
inspects or modifies the signal makes (including cancellation code).

after the thread unmaps its own stack/thread structure, the kernel,
performing child tid clear and futex wake, could clobber a new mapping
made at the same location as the just-removed thread's tid field.
disable kernel clearing of child tid to prevent this.

the mips abi reserves stack space equal to the size of the in-register
args for the callee to save the args, if desired. this would cause the
beginning of the thread structure to be clobbered...

the old code worked in qemu app-level emulation, but not on real
kernels where the clone syscall does not copy the register values to
the new thread. save arguments on the new thread stack instead.