i386: punt all config options that just control -march

You can set these yourself in your CFLAGS, and our list is incomplete
and pointless (we agreed some time ago to stop doing this).

We keep around CONFIG_[3456]86 because it is used to select between
different optimized subdirs of handcoded assembly.

Also take this opportunity to up the default to i686.  If you're using
an older cpu, then you can handle changing this yourself.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>

libubacktrace: fix backtrace for statically linked application

libgcc_s.so's unwinder could not access unwind tables of statically
linked binaries, so we really want to use _Unwind_* stuff from
libgcc_eh.a.
It required to build backtrace.c differentiating between shared and
static case.

Signed-off-by: Filippo Arcidiacono <filippo.arcidiacono@st.com>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>

libdl: fix dlopen implementation from statically linked application

Calling dlopen from statically linked application is actually broken,
because _dl_find_hash enters into an infinite loop when trying to
resolve symbols. In this case it doesn't need to extend the global
scope, it is readyto be used as it is, because _dl_loaded_modules already points
to the dlopened library.

The patch also fixesi a typo in __LDSO_LD_LIBRARY_PATH__ macro, that was
preventing to get the actual value of the LD_LIBRARY_PATH.

Signed-off-by: Filippo Arcidiacono <filippo.arcidiacono@st.com>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>

libc: deal with aux vect inside __uClibc_main only if !SHARED

It's not safe to use the aux vect inside __uClibc_main if we are running
with shared libraries, because it could have been already modified.
For example, if some constructor plays with environment variables by
using unsetenv, the modifications done into the stack to unset an
environment variable, have impacts on the aux vect due to the extra NULL
entries added.
Due to this, __uClibc_main is not able to detect where the aux vect
starts, so all the entries that are used by __uClibc_main (AT_UID,
AT_EUID, AT_GID, AT_EGID, AT_PAGESZ and possibly other arch specific)
are impacted.

Same side effect on the aux vect is caused by the ld.so when running a
SUID program with some of the unsecure environment variables set, that
will be unset by the ld.so itself.

In order to fix this issue, it needs to handle aux vect entries into
__uClibc_main only if SHARED is not defined.
In SHARED case, libc refers to __dl_secure and _dl_pagesize as initialised
by the ld.so where the aux vext is still untouched.

Signed-off-by: Filippo Arcidiacono <filippo.arcidiacono@st.com>
Reviewed-by: Carmelo Amoroso <carmelo.amoroso@st.com>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>

linux: pread/write: fix 64bit handling

The syscall on 64bit ports takes 4 args as there is no need to split
up the value into two args.  Add support for that to the common code.

Once we fix that, the mips code can now leverage it for its 64bit and
32bit needs.  However, we can't just drop it entirely yet because its
n32 ABI needs special handling to treat it like a 64bit port.  This
does change the existing behavior which treats the n32 like a 32bit
port, but we want to do this.

In the future, we'll probably have to introduce a define for this as
it currently affects x86_64/x32 and mips/n32.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>

buildsys: fix realclean

For anything but (toplevel) clean, CLEAN_%, we need .config
Otherwise realclean would leave stuff behind that was not intended to
survive.

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

add UCLIBC_SUSV2_LEGACY to hide valloc

valloc was marked as LEGACY in SUSv2, removed from SUSv3 and later.
TODO: Remove this (point people to memalign and it's successors?).

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

relinfo.pl: use correct interpreter

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

relinfo.pl: silence warning

$* is no longer supported at extra/scripts/relinfo.pl line 2.

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

stdlib: avoid relocation

valloc uses memalign

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

linux: pread/write: convert to SYSCALL_ALIGN_64BIT

The pread64/write64 syscalls have the 64bit register align issue for
all arches.  Use this new define so we can merge the powerc/xtensa
versions back into the common code.

SuperH is funky and also allows us to do this.

We should be able to merge the mips version too, but that'll require
someone to take a closer look as the current stuff doesn't look quite
right.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>

linux: readahead: convert to SYSCALL_ALIGN_64BIT

The readahead syscall has the 64bit register align issue for all
arches.  Only mips was handling this though.

Clean up the common readahead.c to use the SYSCALL_ALIGN_64BIT
define so that we can throw away the mips version and make this
work correctly on arm/ppc/xtensa.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>

linux: posix_fadvise: use new SYSCALL_ALIGN_64BIT

Now that we have a new SYSCALL_ALIGN_64BIT define for tracking the
64bit register shift behavior, use it.  This allows us to delete
duplicated arm/xtensa files, as well as drop a few arch ifdefs from
common code.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>

linux: use OFF64_HI_LO rather than __LONG_LONG_PAIR

This macro takes care of the shift/mask split for us, so no need
to open code this ourselves and then use __LONG_LONG_PAIR.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>

mips: only define SYSCALL_ALIGN_64BIT for O32 ABIs

The 64bit register alignment issue only affects the O32 ABI, so wrap
the define accordingly.  We don't want this being used for N32 ABIs.

This doesn't directly affect the N64 ABI since these files wouldn't
even be compiled for those targets (no need for the 32bit/64bit shim).

Signed-off-by: Mike Frysinger <vapier@gentoo.org>

libc: rename TRUNCATE64_HAS_4_ARGS to SYSCALL_ALIGN_64BIT

The reason truncate64 takes 4 args on some arches is that their ABI
requires 64bit values to be aligned on register pair boundaries.
Since this alignment affects more than just truncate64, rename the
define to properly document its purpose.  This also allows us to
expand it to the other impacted syscalls (which will be done in a
follow up commit).

Signed-off-by: Mike Frysinger <vapier@gentoo.org>

Revert "_scanf.c: Implement 'm' modifier for 'c' and '[' conversions."

This reverts commit e567c399ff86d007d8c4586f0dd5e0ca61e283ca.
since it breaks badly (e.g. busybox netstat)

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

_vfprintf.c: use 'fputws_unlocked(S, F)' instead of 'fputws(S, F)'

This eliminates a source of reproduceable freezes

Signed-off-by: Mirko Vogt <dev@nanl.de>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

inet: fix getting the nameserver from _res state after res_init.

Fixes displaying the nameserver in busybox nslookup.

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

libc: Fix common-generic stat.h for BE

Signed-off-by: Mischa Jonker <mjonker@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

Fix some fragileness in dlopen/do_dlopen wrapper & worker pair.

do_dlopen contains __builtin_return_address to determine from
whence it was called, and uses that to determine which dynamic
object's data it should use to start the search.  (In the bug I was
tracking, this related to whether the application's RPATH was used or
not.)  For that to work, it has to have been inlined into the wrapper
function.

As it happens, it wasn't being inlined.  That's an unfortunate compiler
behaviour, but it isn't wrong and shouldn't have caused dlopen to fail.

This patch changes things so the wrapper function determines the
return address, and passes it to the worker.  If the worker's inlined,
the generated code should be exactly the same as before.

Signed-off-by: Nathan Sidwell <nathan@codesourcery.com>
Signed-off-by: Bernd Schmidt <bernds@codesourcery.com>

Fix a problem with scanning wide chars.

We found that the testcase

int
main (void)
{
  wchar_t s[10];
  memset (s, 0, sizeof (s));
  int r = sscanf ("s", "%ls", s);
  printf ("%d\n", r);
  printf ("%ls\n", s);
  return 0;
}

printed
0
<blankline>

rather than the expected
1
s

The problem was the enum in _scanf.c, which has had a 'CONV_m' value
inserted. The attached patch fixes the problem in __psfs_parse_spec by
not presuming a particular displacement between the two sets of
char-like conversion values. With this patch the above program produces
the expected output.

Signed-off-by: Nathan Sidwell <nathan@codesourcery.com>
Signed-off-by: Bernd Schmidt <bernds@codesourcery.com>

inet: do not filter responses in res_query

Fixes bug #5342

res_query was silently rejecting responses against T_ANY DNS
questions.

Remove the type-filtering from res_query altogether.
__dns_lookup is supposed to return the proper stuff that you asked
for (and only that).

Signed-off-by: Timo Teräs <timo.teras@iki.fi>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

inet: Fix LT{.old,} compilation due to res_iclose

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

LT.old: add missing prototypes

to silence warning

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

rpc: silence warning

Silence warning about shadowing wait

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

test: fix non-O checking

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

test: add missing top_srcdir

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

buildsys: remove calls to mktemp

Nowadays we should not be racy in this area anymore, so simplify
generating temporary build-files.

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

test: silence warning in skeleton

Some tests use fd, rename the one in the skeleton.

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

test: silence warning in testsuite.h

silence warnings about unused parameters

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

test: update test-skeleton.c

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

libubacktrace: use libgcc_s.h

for consistency

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

libc: switch several users to unlikely()

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

inet: res_nclose: free user-buffer, not global _res

In res_iclose we were operating on the global _res even if called via
res_nclose where we are supposed to operate on the user provided
res_state.

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

test: dlopen: fix misplaced buildsys hunk

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

buildsys: remove surplus slashes from test dir paths

cosmetics

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

buildsys: tweak KCONFIG_CONFIG export

0c3eb2da578bc7ba2e74d240e3249dce62ec725e moved KCONFIG_CONFIG around and
that broke propagation of config-settings to the testsuite. Fix that.

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

test: cater for config

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

test: compile test{,_glibc} adjacent

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

metag: Add NPTL support

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

extra/locale: Mark __codeset_8_bit_t as packed for Meta

gen_ldc writes offsets into the __locale_mmap_t structure. This
relies on the host system having the same struct offsets as the
target. On Meta, the structs are always aligned to a multiple of
4 bytes, and for these targets it is required for the struct to
be packed so that it is more likely to match the host system
(typically x86 which does not align structs).

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

metag: Add support for loading shared objects into core memories.

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

libm: Add support for the Meta architecture

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

Add support for the Meta architecture

Meta cores are 32-bit, hardware multithreaded, general purpose, embedded
processors which also feature a DSP instruction set, and can be found in
many digital radios. They are capable of running different operating
systems on different hardware threads, for example a digital radio might
run RTOSes for DAB decoding and audio decoding on 3 hardware threads,
and run Linux on the 4th hardware thread to manage the user interface,
networking etc. HTPs are also capable of running SMP Linux on multiple
hardware threads.

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

ldd: Add MATCH_MACHINE definition for Meta

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

elf: Add ELF magic number and relocations for the Meta architecture

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

ldso: Fix Build warnings with common-generic ABI build

  CC ldso/ldso/ldso.oS
In file included from ./ldso/include/ldso.h:47,
                 from ldso/ldso/ldso.c:33:
./ldso/include/dl-syscall.h:35:1: warning: "S_ISUID" redefined
In file included from ./include/bits/kernel_stat.h:10,
                 from ./ldso/include/dl-syscall.h:31,
                 from ./ldso/include/ldso.h:47,
                 from ldso/ldso/ldso.c:33:
./include/sys/stat.h:163:1: warning: this is the location of the
previous definition
In file included from ./ldso/include/ldso.h:47,
                 from ldso/ldso/ldso.c:33:
./ldso/include/dl-syscall.h:36:1: warning: "S_ISGID" redefined
In file included from ./include/bits/kernel_stat.h:10,
                 from ./ldso/include/dl-syscall.h:31,
                 from ./ldso/include/ldso.h:47,
                 from ldso/ldso/ldso.c:33:
./include/sys/stat.h:164:1: warning: this is the location of the
previous definition

-----------------------------------------------------------------

Cleanly seperated the legacy ABI vs. new ABI scenario.

Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Cc: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

libpthread: linuxthreads.old: Allow building for no-legacy-syscalls ABI

Since NPTL port to an ARCH can be non-trivial, we must allow the legacy
pthread lib to be built for the common-generic syscall ABI

Note that only linuxthreads.old is allowed, since "newer" linuxthreads
relies on sysctl which the ABI no longer provides.

Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Cc: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

libc/ipc: Make IPC_64 0 for common-generic ABI

This avoids having to define __IPC_64 to 0 in each arch using the ABI

Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Cc: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

fstat64: Fix Build warnings with common-generic ABI build

  CC libc/sysdeps/linux/common/fstat64.os
In file included from libc/sysdeps/linux/common/fstat64.c:16:
libc/sysdeps/linux/common/xstatconv.h:28: warning: 'struct kernel_stat'
declared inside parameter list
libc/sysdeps/linux/common/xstatconv.h:28: warning: its scope is only
this definition or declaration, which is probably not what you want
libc/sysdeps/linux/common/fstat64.c: In function 'fstat64':
libc/sysdeps/linux/common/fstat64.c:33: warning: passing argument 2 of
'__syscall_fstat64' from incompatible pointer type
libc/sysdeps/linux/common/fstat64.c:18: note: expected 'struct stat *'
but argument is of type 'struct stat64 *'
-------------------------------------------------------------------

Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Cc: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

leek: if lseek syscall is not available, use lseek64 even for !LFS

With Busybox and uClibc - both built w/o LFS, this caused ash to be
completely broken, as lseek was simply returning error.

Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Cc: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

utimes: Export utimensat based utimes

sys/time.h has libc_hidden_proto(utimes) which generates the hidden
"__GI_utimes" symbol reference, and common/utimes.c has a
libc_hidden_def(utimes) which generates the exported "utimes" alias.

As part of no-legacy-syscall kernel ABI, Commit 80dc2ed05
"utimes: Use utimensat if arch does not have the utimes syscall"
introduced a new wrapper, but missed the corresponding libc_hidden_def,
causing Busybox (1.20.0) link to fail (for ARC).

Also don't generate a STUB, in that case.
---------------->8-------------------
....
touch.c:(.text.touch_main+0xdc): undefined reference to `utimes'
touch.c:(.text.touch_main+0x114): undefined reference to `utimes'
libbb/lib.a(copy_file.o): In function `copy_file':
copy_file.c:(.text.copy_file+0x446): undefined reference to `utimes'
collect2: ld returned 1 exit status
---------------->8-------------------

Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Cc: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

LT{,.old}: include stddef.h for size_t

A user application built with -D_POSIX_C_SOURCE=200112L and
 #include signal.h, yields following compilation error:

------------------->8-----------------------------
arc-linux-uclibc-gcc -D_POSIX_C_SOURCE=200112L hello.c
In file included from
/opt/INSTALL/lib/gcc/arc-linux-uclibc/4.4.7/../../../../arc-linux-uclibc/sys-include/signal.h:429,
                 from hello.c:1:
/opt/INSTALL/lib/gcc/arc-linux-uclibc/4.4.7/../../../../arc-linux-uclibc/sys-include/bits/pthreadtypes.h:48:
error: expected specifier-qualifier-list before 'size_t'
------------------->8-----------------------------

The fix is to make sure size_t is explicitly provided to pthreadtype.h

Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

buildsys: Add missing $(SYMBOL_PREFIX) to symbol names

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

libpthread: Remove useless .cvsignore files

Signed-off-by: Markos Chandras <hwoarang@gentoo.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

nptl/allocatestack.c: Fix stack mprotect alignment for STACK_GROWS_UP

mprotect() expects the first argument to be aligned to a page boundary.

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

libc/pause: Enable pause_nocancel for arches without __NR_pause

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

nice: fix overflow checking in int_add_no_wrap()

In C, signed integer overflow is undefined behavior.  Many compilers
optimize away checks like `a + b < a'.

Use safe precondition testing instead.

Signed-off-by: Xi Wang <xi@mit.edu>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

inet: remove noreturn attribute from res_nclose

Partially revert e9af4dfbd328e9f3bba235fdb2d1027dd2dbbcde

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

scripts: remove CVS header from getent

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

libc/ipc: Allow architectures to define their own __IPC_64 macro

New architectures don't define ARCH_WANT_IPC_PARSE_VERSION in their kernel.
This means that every cmd passed to semctl,msgctl and shmctl is IPC_64 by
default. For these architectures we need to define __IPC_64 as 0. Existing
architectures are not affected by this change.

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

ldso: Use newer syscalls if arch does not have the deprecated syscalls

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

open: Use openat if arch does not have the open syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

not-cancel.h: Use openat if arch does not have the open syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

open64: Use openat if arch does not have the open syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

fcntl: Use fcntl64 is arch does not have the fcntl syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

fstatfs64: Prefer fstatfs64 system call instead of __libc_fstatfs

Using __libc_fstatfs for fstatfs64 adds a small delay as it needs to
use a 32-bit data structure to get the file info and them pass them to
the 64-bit data structure which was given as a fstatfs64 argument. Using
the system call directly should make the entire process much faster.
Also fix the arguments for fstatfs64. It takes three arguments
(see fs/fstatfs.c in Linux kernel sources) so despite what the manpage
says, the size of the buffer needs to be passed as the second argument

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

fstatfs: Add __libc_fstatfs wrapper

New architectures don't have fstatfs anymore, so we use a wrapper for
__libc_fstatfs which will use fstatfs64 internally. The interface however
needs to remain the same (i.e accepting a struct statfs as a second
argument) for backwards compatibility

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

stat: Use fstatat64 if arch does not have the stat syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

ustat: Add ustat stub for arches that don't have the ustat system call

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

lstat64: Use fstatat64 if arch does not have the lstat64 syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

lstat: Use fstatat64 if arch does not have the lstat syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

stat64: Use fstatat64 if arch does not have the stat64 syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

fstatat{64}: No conversion code is needed for new architectures

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

fstatat64: Use newfstatat only for 64-bit operations

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

fstatat: Use newfstatat only for 64-bit operations

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

fstat: Use fstat64 if arch does not have the fstat syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

libc/statfs64: Skip conversion code for new architectures

Fixes the following tests in LTP

statfs02_64 ( test case #5 )

The purpose of this test is to pass an invalid pointer to the statfs64
syscall and check if the kernel returns EFAULT or not. However,
uClibc creates a new statfs struct which is then passed to the kernel.
As a result of which, the kernel returns 0 because the newly created
statfs structure is valid. But, when copying the contens of the new
pointer to the old userspace one, the uClibc segfauls because the old
pointer is invalid. Old architectures are doomed to suffer by this
problem but new architectures can use the statfs64 syscall directly
so that the userspace pointer is passed directly to the kernel and get
the correct errno at the end.

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

statfs: Use statfs64 if arch does not have the statfs syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

getdents: Use getdents64 if arch does not have the getdents syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

lseek: Use lseek64 is arch does not have the lseek syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

sendfile: Use sendfile64 if arch does not have the sendfile syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

truncate: Use truncate64 if arch does not have the truncate syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

ftruncate: Use ftruncate64 if arch does not have the ftruncate syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

fork: Use clone if arch does not have the fork syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

vfork: Use clone if arch does not have the vfork syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

utime: Use utimensat if arch does not have the utime syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

inotify_init: Use inotify_init1 if inotify_init syscall is not defined

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

epoll: Use epoll_pwait if arch does not have the epoll_wait syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

epoll: Use epoll_create1 if arch does not have the epoll_create syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

pipe: Use pipe2 if arch does not have the pipe syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

utimes: Use utimensat if arch does not have the utimes syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

symlink: Use symlinkat if arch does not have the symlink syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

readlink: Use readlinkat if arch does not have the readlink syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

rename: Use renameat if arch does not have the rename syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

mkdir: Use mkdirat if arch does not have the mkdir syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

lchown: Use fchownat if arch does not have the lchown syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

rmdir: Use unlinkat if arch does not have the rmdir syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>

chown: Use fchownat if arch does not have the chown syscall

Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>