Native gdb makes a ptrace (GET_THREAD_AREA) which needs to return the
TP. however when libc sets up TP reg (for main thread), it doesn't call
arc_settls syscall so kernel doesn't know of TP register details
(moreso because clone doesnt have SETTLS flag)
Note that kernel doesn't know about r25 being TP etc.
This commit got lost in merge of NPTL tools into arc-mainline-dev and
showed up again as STAR 9000919529 (native gdb can't debug threaded
apps)
------->8---------------
[ARCLinux]# gdb ./pth
Reading symbols from ./pth...(no debugging symbols found)...done.
(gdb) b main
Breakpoint 1 at 0x106f2
(gdb) r
Starting program: /pth
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/libthread_db.so.1".
thread_get_info_callback: cannot get thread info: generic error
(gdb) q
------->8---------------
Debugged-by: Anton Kolesov <akolesov@synopsys.com>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
These days IPv6 is used more and more in different software
packages. And so we're adding IPv6 support by default in uClibc
for ARC cores.
Signed-off-by: Alexey Brodkin <Alexey.Brodkin@synopsys.com>
Cc: Vineet Gupta <vgupta@synopsys.com>
Cc: Anton Kolesov <akolesov@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
- remove symbol prefix
- new startup
- new clone syscall support
Signed-off-by: Yoshinori Sato <ysato@users.sourceforge.jp>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
- New toolchain
- Add new flags
Signed-off-by: Yoshinori Sato <ysato@users.sourceforge.jp>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
uClibc doesn't define signal info for the SIGSYS signal which is issued
in case of hitting a syscall prohibited by seccomp.
This is sad as it makes debugging seccomp filter policies impossible on
some architectures (at least ARM and PowerPC, maybe also others) which
do not coincidentally set si_value.sival_int as the syscall number.
To fix this, import the definitions and macros needed from glibc.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
When attempting to build uClibc under buildroot, including building the
tests, the silly tests don't currently compile, a result of attempting
to build using a compiler that does not yet have an installed version of
uClibc available. The error is a missing header file, specifically
atomic.h.
Taking inspiration from the nptl tests, I have extended the EXTRA_CFLAGS
variable to add the required include paths. The tests can now be built
under buildroot.
Signed-off-by: Andrew Burgess <andrew.burgess@embecosm.com>
Acked-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
The current code implements the 'm' modifier only for 's'
conversions and would cause a segfault if it was used for 'c'
or '[' conversions. This patch extends the code to cover these
cases too.
The original version could write scanned data outside the passed buffer
because index i used in the '[' conversion handling block was clobbered.
Signed-off-by: Will Newton <will.newton@imgtec.com>
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
(__pthread_initialize_minimal_internal): Initialize pd->report_events
to that.
This patch helps NPTL report TD_CREATE event, so that GDB could catch the
event and update its thread_list.
Link: http://lists.uclibc.org/pipermail/uclibc/2015-April/048921.html
[shengyong:
- original patch from glibc: commit 7d9d8bd18906fdd17364f372b160d7ab896ce909
- context adjust
- update nptl_db/ChangeLog]
Signed-off-by: Roland McGrath <roland@gnu.org>
Signed-off-by: Sheng Yong <shengyong1@huawei.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Previously V=1 did print abbreviated commands and V=2 the full commands.
Kbuild-based build-systems behave in the opposite way and this is
apparently confusing or inconvenient for users so swap our V handling to
be in line with kbuild (and automake as far as V=0 / V=1 is concerned).
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
since ae9e3f46 sched.h pulls in stdlib.h (for malloc() and free()) and
thus also xlocale..
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
In file included from ./include/sys/syscall.h:33:0,
from libc/sysdeps/linux/common/sync_file_range.c:10:
libc/sysdeps/linux/common/sync_file_range.c: In function '__sync_file_range_nocancel':
./include/bits/syscalls.h:144:16: error: conflicting types for '_v3'
register int _v3 __asm__ ("v3") = _v3tmp;
^
./libc/sysdeps/linux/arm/sysdep.h:281:7: note: in expansion of macro 'LOAD_ARGS_7'
LOAD_ARGS_##nr (args) \
^
./libc/sysdeps/linux/arm/sysdep.h:324:2: note: in expansion of macro 'INTERNAL_SYSCALL_RAW'
INTERNAL_SYSCALL_RAW(SYS_ify(name), err, nr, args)
^
./libc/sysdeps/linux/arm/sysdep.h:256:40: note: in expansion of macro 'INTERNAL_SYSCALL'
({ unsigned int _inline_sys_result = INTERNAL_SYSCALL (name, , nr, args); \
^
libc/sysdeps/linux/common/sync_file_range.c:32:9: note: in expansion of macro 'INLINE_SYSCALL'
return INLINE_SYSCALL(sync_file_range, 7, fd, 0,
^
In file included from ./libpthread/nptl/sysdeps/unix/sysv/linux/arm/sysdep-cancel.h:18:0,
from ./include/cancel.h:58,
from libc/sysdeps/linux/common/sync_file_range.c:15:
./libc/sysdeps/linux/arm/sysdep.h:280:21: note: previous definition of '_v3' was here
register int *_v3 __asm__ ("v3") = _sys_buf; \
...
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
ucontext_i.h too depends on uClibc_config.h.
Update it to look like the other MANGLE generated files and tweak
prereqs
Fixes: make realclean ; make -j libc/misc/internals/__uClibc_main.i
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
We were ending up with multiple entries in LIBGCC for recursive child
make instances, avoid that.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This reverts commit 534f44d53146457b3ca686c47efb9207543b88e1.
I don't think this is wanted.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
when building uClibc with -O0 (DODEBUG build) the default sigrestorer
had some extra glue code generated for stack manipulation which was
messing up resume from signal path.
So annotate the function with -Os so that gcc would only generate the
bare min 2 instruction TRAP sequence
Reported-and-Debugged-by: Alexey Brodkin <abrodkin@synopsys.com>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
The R_ARC_NONE relocation is generated when --gc-sections, used by the
compile time linker, removes some sections. This is completely normal,
and we can see that all other targets (based on random sampling) have
support for R_*_NONE relocations (named for each target).
Handling R_ARC_NONE involves doing nothing with it, which is nice and
easy.
Signed-off-by: Andrew Burgess <andrew.burgess@embecosm.com>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
where SIMULATOR_glibc is an optional simulator to run host binaries and
SIMULATOR_uclibc is an optional simulator to run target (uClibc)
binaries.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Otherwise gcc might not understand that oom() ended control-flow and
might emit an (untaken) reference to abort() in _dl_update_slotinfo()
on e.g. SH4 which breaks linking ld-uClibc.so.
Arguably -ffreestanding should prevent GCC from emitting this
'.global abort' but alas, it does not, which is another bug..
Also mark the function cold to further lower the incoming frequency and
branch probability.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Fix arc4random exclusion for real.
Silence warning about missing prototype for external helper-function in
tst-scandir
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Note: _PATH_UTMPX == _PATH_UTMP and the utmp struct is identical to the
utmpx struct so this only changes the external API entrypoints and NOT
the underlying data source.
This saves about 500b (~1300b from previously ~1950) while at it.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Fixes SH build breakage due to attempts to define hidden defs for
unavailable long double functions.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This reverts commit 84135275cfeebc0b233c1c96eeada4d4178a0b18.
This change is said to make systemd deadlock (cannot reproduce this)
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Conflicts:
include/utmp.h
Introduce a <sys/random.h> for it.
/* FIXME: aren't there a couple of __restrict and const missing ?
*/
extern int getrandom(void *__buf, size_t count, unsigned int flags)
__nonnull ((1)) __wur;
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
nostdlib disables linking in startup files too which is not
what we want here since it needs to resolve __dso_handle
which comes from crtbeginS.o, otherwise librt has this
undefined reference to a weak undefined __dso_handle that
shows up as error (with gold linker)
when shared libraries are being built which are
linking in librt
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
atexit should only be in either uclibc_nonshared.a
shared libc case or libc.a in static build case
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
For some rarely cases(almost App bugs), calling malloc with
a very largre size, checked_request2size check will fail,set
ENOMEM, and return 0 to caller.
But this will let __malloc_lock futex locked and owned by the
caller. In multithread circumstance, other thread calling
malloc/calloc will NOT succeed and get locked.
Signed-off-by: Zhiqiang Zhang <zhangzhiqiang.zhang@huawei.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
On execution of "inet/gethost_r-align" test I noticed failure due
to unaligned access (instaed of 4-byte aligned 1-byte aligned
address was attempted to be accessed).
Further investigation confirmed this nice and helpful test failure.
Following commit removed usage of ALIGN_BUFFER_OFFSET on entry to
__read_etc_hosts_r():
http://git.uclibc.org/uClibc/commit/?id=f65e66078b9f4d2d7f0fc336dee36e78fc467c0f
So indeed if target architecture doesn't allow unaligned access
and provided tmp buffer is not word aligned (and we will deal with pointers
which means word-sized data units), then CPU will fail during execution.
In case of ARC we'll see "Unaligned access" exception like this:
--->8---
# potentially unexpected fatal signal 7.
Path: /root/uClibc/test/inet/gethost_r-align
CPU: 0 PID: 5514 Comm: gethost_r-align Not tainted 3.13.11 #2
task: 8f42a580 ti: 8f40e000 task.ti: 8f40e000
[ECR ]: 0x00230400 => Misaligned r/w from 0x5fdab341
[EFA ]: 0x5fdab341
[BLINK ]: 0x20032a18
[ERET ]: 0x20032a3c
@off 0x12a3c in [/lib/libuClibc-0.9.34-git.so]
VMA: 0x20020000 to 0x20062000
[STAT32]: 0x00000086 : U E2 E1
BTA: 0x20046014 SP: 0x5fdab260 FP: 0x00000000
LPS: 0x20046064 LPE: 0x20046068 LPC: 0x00000000
r00: 0x5fdab341 r01: 0x00000005 r02: 0x00000015
r03: 0x00000000 r04: 0x5fdab358 r05: 0x00000000
r06: 0x0a0a0a00 r07: 0x00000000 r08: 0x0000003f
r09: 0x20067050 r10: 0x00000000 r11: 0x00000014
r12: 0x00000001 r13: 0x00000000 r14: 0x20060660
r15: 0x20060661 r16: 0x00000006 r17: 0x5fdab371
r18: 0x00000018 r19: 0x5fdab2b4 r20: 0x00020000
r21: 0x00000000 r22: 0x00029068 r23: 0x5fdab371
r24: 0x00010000 r25: 0x00000000
--->8---
To fix this problem we'll re-introduce tmp buffer force alignment
before config parser invocation.
Signed-off-by: Alexey Brodkin <Alexey.Brodkin@synopsys.com>
Cc: Vineet Gupta <vgupta@synopsys.com>
Cc: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
We were not passing the correct use-ld and -m{32,64,...} when creating
the linker script. Revisit for gold later on anyway.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Otherwise we have a broken scenario with non-threading builds.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
include/bits/sigset.h:219:62: warning: right-hand operand of comma expression has no effect [-Wunused-value]
include/bits/sigset.h:210:63: warning: right-hand operand of comma expression has no effect [-Wunused-value]
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
In file included from ldso/ldso/ldso.c:87:0:
ldso/ldso/dl-startup.c: In function '_dl_start':
ldso/ldso/dl-startup.c:184:32: warning: cast from pointer to integer of differennt size [-Wpointer-to-int-cast]
auxvt[AT_BASE].a_un.a_val = (Elf32_Addr) &_begin;
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This option will modify ldso so that it will use the executables
RUNPATH/RPATH to find to find libraries even though this behavour
is not standard. Setting this option causes the uclibc dynamic linker
behavour to match the glibc dynamic linker.
Signed-off-by: Steve Ellcey <sellcey@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
perf in upstream Linux kernel 3.17 onwards expects STT_GNU_IFUNC
replicate it from glibc
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Currently UCLIBC_SYSCALL_ALIGN_64BIT is not explicitly handled. Fix that
and make sure the special handling is done for powerpc/xtensa which use
UCLIBC_SYSCALL_ALIGN_64BIT but don't use hole punched syscall handler in
kernel.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Cc: Baruch Siach <baruch@tkos.co.il>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
arm/powerpc/xtensa pass @advice as 2nd arg to syscall (vs. canonical 4th)
Current code however does this for UCLIBC_SYSCALL_ALIGN_64BIT which
powerpc/xtensa also happen to define.
This is not true for ARCv2 ISA and possibly other arch of future with
64-bit even register requirement, which uses the standard syscall
handler in kernel.
Fix that by providing 2 variants of SYSCALL_ALIGN_64BIT
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Cc: Baruch Siach <baruch@tkos.co.il>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
GNU glob is required by make.
Signed-off-by: Anton Kolesov <anton.kolesov@synopsys.com>
Cc: Vineet Gupta <Vineet.Gupta1@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Not relevant anymore since commit e8cc14e59ed3f66b84e,
"libc: rename TRUNCATE64_HAS_4_ARGS to SYSCALL_ALIGN_64BIT"
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
ARC CPU may have MMU page size of 4/8(default)/16k.
uClibc needs to have page size configured accodring to HW it will be run on.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
uClibc mainline supports NPTL which in turns depends on TLS support in
the tools (gcc/binutils), which is yet to be merged in dev branches.
However there is some non NPTL code in uClibc, added as part of NPTL
effort, which relies on certain relocations only provided by NPTL
binutils. As a result building the current upstream even for LT.old
breaks.
So conditionalize that code on tools, bu tin lack of specific versions,
we use NPTL enabling as a sign the tools are equipped to handle those
relos.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Add a configuration choice for the NaN format on MIPS (either the
standard (legacy) format or the newer IEEE 2008 format.
Change how CPU_LDFLAGS are set for MIPS. Use the same value as
CPU_CFLAGS since CC is used to do linking. This ensures consistency
between compiles and links and adds support for N32 ABI to linking.
Signed-off-by: Steve Ellcey <sellcey@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
41537a770b46f2376e74af58ba4885749339e81e erroneously stripped the
top_builddir off the per-directory flag setting which works fine for
in-tree builds but is not correct for O=
Just use the full path of the target files to apply the flags to fix
that thinko.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
The result of the calculation in register a12 is never used as the
function _dl_linux_resolver only accepts 2 arguments. Drop it.
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This noticeably lowers the number of mprotect calls at program startup,
e.g. for busybox: 7 calls vs 1835 calls.
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
mknod() in glibc/eglibc will check the argument, like this,
...
if (k_dev != dev) {
__set_errno (EINVAL);
return -1;
}
...
So add argument check in uclibc's mknod() too.
Signed-off-by: Wang Yufen <wangyufen@huawei.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
amending commit 49a67cf67d5a7194214566bc730ee7e28d55bbe1
could need a thumb implementation..
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
BZ #7808 asks us not to fallback to DES if the optional SHA are disabled
but requested by the user.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
mkostemp(char *template, int flags) generates a unique temporary
filename from a template. The flags parameter accepts three of
the same flags as open(2): O_APPEND, O_CLOEXEC, and O_SYNC. The
current implementation of mkostemp(3) does not respect the flags
and in fact confuses the flags with the file mode which should
always be S_IRUSR | S_IWUSR. This patch corrects this issue.
Signed-off-by: Anthony G. Basile <blueness@gentoo.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
We were relying on FE_DIVBYZERO being turned off when printing
"%f", +-.0
Avoid the whole issue by looking at the sign-bit (in a rough
approximation).
Note that we do not handle gracefully:
printf ("\n%llf\n", -0.0);
printf ("\n%llf\n", 0.0);
nor %Lf for both when NOT cast to long double. Avoiding an FPE due to
broken numbers like these does not make sense to me.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
In file included from ./include/signal.h:279:0,
from libc/sysdeps/linux/x86_64/sigaction.c:22:
./include/bits/sigaction.h:32:29: error: unknown type name 'siginfo_t'
void (*sa_sigaction)(int, siginfo_t *, void *);
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
statfs() is a Linux-specific function. When building without
UCLIBC_LINUX_SPECIFIC set, libc_hidden_proto(statfs) in
include/sys/statfs.h is unmatched with libc_hidden_def(statfs)
in libc/sysdeps/linux/common/statfs.c, leading to a compile
error. This patch fixes this, as well as adds statfs() to the
list of Linux-specific functions in extra/Configs/Config.in.
Signed-off-by: Anthony G. Basile <blueness@gentoo.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
add fmovem{x,.d,.*} ISA conditional
Fixes assembler errors for 5206.
isa_a, isa_b (or a+, don't remember) do not have emx, use em.d there.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Update sys/ptrace.h to latest glibc release for a lot of missing
definitions.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Update sys/ptrace.h to latest glibc release for a lot of missing
definitions.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Avoids circular dependency warning from make by generating the .h
directly from the .sym files, elimination intermediate files.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
We add the Linux-specific function fallocate() which allows the user to
directly manipulate allocate space for a file. fallocate() can operate
in different modes, but the default mode is equivalent to posix_fallocate()
which is specified in POSIX.1.
Recent releases of e2fsprogs 1.42.11 and above expect fallocate64() to be
available.
Signed-off-by: Anthony G. Basile <blueness@gentoo.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
The sparc64/sparcv9 code is incomplete. Furthermore there is
no real embedded hardware for sparc64 available, so better remove
it until someone comes up with a complete port.
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This test fails for non-MMU systems, because it uses fork()
Thanks to Waldemar Brodkorb!
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
ld was (erroneously) complaining that it failed to merge private bfd
data. Silence this mismatch as seen on m68k with obfd binary.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This test fails for systems without fpu.
For example m68k build for non-mmu/non-fpu systems fail.
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This breaks out of a deadlock, which occurs when testing for powerpc. (qemu)
Otherwise the test suite does not finish.
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
When bind() fails in this test, then /tmp get filled up with
temp files and upcoming tests will fail, because they need to create
temp files, too. Better exit from the loop.
Test still needs to be investigated to fix the failing bind().
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
otherwise you get compile warnings, when building for m68k.
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
When you cross-compile the testsuite and then try to run
on a system without gcc/binutils it would be nice if no compiling
or linking is tried. Refresh test/README (suggested by Khem)
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Test on Linux 3.16.1 in Qemu ARM fails with:
TEST_EXEC nptl/ tst-cancel4
ret == 1 ; expected_ret == 0
make[1]: *** [tst-cancel4.exe] Error 1
The output of failed test is:
minimum write buffer size too large
../Test.mak:89: recipe for target 'tst-cancel4.exe' failed
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
tst-spin1.c compile breaks with:
test/nptl/tst-spin1.c:34: undefined reference to `pthread_spin_lock'
pthread_spin_lock and pthread_spin_trylock is missing while
building sparc. add the meta c files here.
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
For sparc TARGET_SUBARCH is never used. And there is a typo in the
path.
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This reverts commit 1db4f8f8389f6a935ecd83aff7fcce5d1cf2c0f3.
It is unclear in which way this is breaking sparc systems.
With this change added you get following linking error:
libpthread/nptl/libpthread_so.a(pt-system.oS): In function `__libc_system':
libc/stdlib/system.c:64: undefined reference to `__wait4_nocancel'
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This reverts commit f51fb26dbcceee9e48d10facc830bd4a549f6cc2.
This fixes linking error for sparc build.
While building a sparc system you get following linking error:
libc/sysdeps/linux/sparc/pipe.S:54: undefined reference to `__GI___errno_location'
Removing the HIDDEN_JUMPTARGET fixes the issue, while using gcc 4.8.3/binutils 2.24.
The definition of __errno_location was changed recently here:
http://git.uclibc.org/uClibc/commit/?id=8f550715c2b211036fc273add3cb5219397ed312
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
When using something like this:
$ echo foo|grep foo|wc -l
with mksh shell, you get an runtime error.
Glibc and klibc does not do these extra check.
After removing this check using double pipes work fine.
Tested with Qemu 2.1.0.
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Thorsten Glaser <tg@mirbsd.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
issetugid() returns 1 if the process environment or memory address space
is considered tainted, and returns 0 otherwise. This happens, for example,
when a process's privileges are elevated by the setuid or setgid flags on
an executable belonging to root. This function first appeard in OpenBSD 2.0
and is needed for the LibreSSL.
This patch follows the same logic as the equivalent musl commit. For more
information see the commit message at
http://git.musl-libc.org/cgit/musl/commit/?id=ddddec106fd17c3aca3287005d21e92f742aa9d4
Signed-off-by: Anthony G. Basile <blueness@gentoo.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This fixes a breakage for MIPS32/64 introduced in:
http://git.uclibc.org/uClibc/commit/?id=c02bc5763ba9c8f8b464bca89fba28e091ce1ffc
Now ULP_SUFFIX is taken into account when searching for existing
libm-test-ulps-ARCH
Signed-off-by: Alexey Brodkin <Alexey.Brodkin@synopsys.com>
Cc: Waldemar Brodkorb <wbx@openadk.org>
Cc: Vineet Gupta <vgupta@synopsys.com>
Cc: Anton Kolesov <akolesov@synopsys.com>
Cc: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Commit "test: cater for config" http://git.uclibc.org/uClibc/commit/test/math/Makefile.in?id=7598eeaa1defa2884adaa890bb115c493d69cc35 inverted check from
ifeq ($(UCLIBC_HAS_LONG_DOUBLE_MATH),y)
to
ifeq ($(UCLIBC_HAS_LONG_DOUBLE_MATH),)
with first branch indeed inverted - instead of adding tests they were excluded.
But "else" branch was not moved up (remember check was inverted).
This causes compile breakage when building basic-test.c:
--->---
TEST_LINK math/ basic-test
basic-test.c: In function 'ldouble_test':
basic-test.c:111:1: warning: implicit declaration of function 'nanl' [-Wimplicit-function-declaration]
TEST_FUNC (ldouble_test, long double, nanl, LDBL_EPSILON, HUGE_VALL)
^
basic-test.o: In function `main':
basic-test.c:(.text.startup+0x834): undefined reference to `nanl'
basic-test.c:(.text.startup+0x858): undefined reference to `nanl'
basic-test.c:(.text.startup+0x880): undefined reference to `nanl'
basic-test.c:(.text.startup+0x8a8): undefined reference to `nanl'
basic-test.c:(.text.startup+0x8d4): undefined reference to `nanl'
basic-test.o:basic-test.c:(.text.startup+0x8e0): more undefined references to `nanl' follow
collect2: error: ld returned 1 exit status
make[3]: *** [basic-test] Error 1
--->---
Fix is simple - merge "old else" branch with the first current one.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Cc: Vineet Gupta <vgupta@synopsys.com>
Cc: Anton Kolesov <akolesov@synopsys.com>
Cc: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Cc: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
OBSTACK is required for native builds of binutils and GDB.
Signed-off-by: Anton Kolesov <Anton.Kolesov@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Mark non-pthread __lll_lock_wait_private weak to avoid multiple
definition error in static link mode.
Signed-off-by: Tony Wu <tung7970@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
There was a runtime error in systems without large file support. Call
fseek(fd, 4096, SEEK_SET) has been failing with EINVAL, though it was
succeeding for offset = 4092. This has been happening because llseek system
call accepts 64-bit value as an offset argument and lseek function has been
ordering 32-bits words that form this offset value, according to the
endianness. However this ordering to match endianness is not required,
because llseek doesn't accept one 64-bit offset argument, it accepts two
32-bit offset argument, then stitches them into one following its
endianness. As a result on little endian system, order of words has been
swapped two time: in libc and in kernel. Thus call to fseek with offset 4096
(0x1000) was doing a system call to llseek with offset 0x1000_0000_0000. I'm
not entirely sure why then offset = 4092 hasn't been failing then.
This patch removes malicious swap of words when calling llseek.
Signed-off-by: Anton Kolesov <Anton.Kolesov@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
commit 00571b43df2e "libc: posix_fadvise: restore implementation for xtensa"
enabled posix_fadvise() for all arches (it was just not generated
before).
However this also unearthed an issue introduced by ee84b8b400
"linux: posix_fadvise: use new SYSCALL_ALIGN_64BIT" which is to
referencing LFS'ish code (off64_t) w/o proper checks which causes build
to break for !LFS.
Fix this by calling posix_fadvise64() only for LFS case and open-code
it's equivalent for !LFS.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Cc: Mike Frysinger <vapier@gentoo.org>
Cc: Baruch Siach <baruch@tkos.co.il>
Cc: Khem Raj <raj.khem@gmail.com>
Cc: markos Chandras <markos.chandras@gmail.com>
Cc: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This commit 1e2e4ac6193ffe0900bd392fa3c596883771eb34 breaks networking
on sparc systems. In Linux the socket functions are declared, but not
implemented and must be routed through socketcall().
Tested via Qemu 2.0.0.
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
On attempt to build uClibc tests following error message appears:
=========
TEST_LINK nptl/ tst-cancel4
tst-cancel4.c: In function 'tf_sigpause':
tst-cancel4.c:765:3: warning: implicit declaration of function
'__sigpause' [-Wimplicit-function-declaration]
__sigpause (SIGCANCEL, 1);
^
tst-cancel4.o: In function `tf_sigpause':
tst-cancel4.c:(.text+0x1054): undefined reference to `__sigpause'
collect2: error: ld returned 1 exit status
=========
This commit http://git.uclibc.org/uClibc/commit/?id=cef1f9517c319d86b4b426ecc14a10d0fa5494c7
makes __sigpause static limiting its scope to "libc/signal/sigpause.c".
Switching to common implementation "sigpause()" fixes this issue.
Signed-off-by: Alexey Brodkin <Alexey.Brodkin@synopsys.com>
Suggested-by: Vineet Gupta <vgupta@synopsys.com>
Cc: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Typically kernel would set the TP register of newly created thread in
clone syscall with CLONE_SETTLS.
However this implies that:
* kernel knows of the exact TP register (which is already different for
ARCompact and ARCv2) and is a detail which becomes part of the ABI
* kernel also needs to handle the unlikely but possible case of !TLS
userspace (e.g. Android libc) case where TP reg is not reserved and
thus need not be mucked with (using ELF personality based detection)
It is better to confine this detail to userspace runtime and set TP reg
right after clone syscall returns.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Under test/math/, there is libm-test-ulps-generic used as fallback
in case that libm-test-ulps-$(TARGET_ARCH) does not exist.
This is required for arches that don't have ULPS yet.
Original patch from Jian Peng was found here:
http://lists.uclibc.org/pipermail/uclibc/2011-April/045172.html
Signed-off-by: Jian Peng <jipeng2005@gmail.com>
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Cc: Vineet Gupta <vgupta@synopsys.com>
Cc: Anton Kolesov <akolesov@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
We switched to detailed per-release Changelog some time ago, available
on the website.
Move the old Changelog for historical reasons.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Native GDB requires elf_fpregset_t to be defined, even if floating point
registers are absent. In this case we use empty structure. Additionally we
redefined prfpregset_t to use this new type.
And there has been an error: it was defined that user_regs_struct has 20
registers, while it has 40.
Signed-off-by: Anton Kolesov <Anton.Kolesov@synopsys.com>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This showed up due to longstanding test/unistd/clone failure where post
clone, the callback was not exiting and rather falling thru in into main
program.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
- Use syscalls-common.h vers of INTERNAL_SYSCALL / INLINE_SYSCALL_NOERR
- INLINE_SYSCALL takes syscall name (e.g. write)
- {INLINE,INTERNAL}_SYSCALL_NCS macros take syscall num (__NR_write)
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
For ARC builds atleast, __NR_xxx was not trickling into in open64.c
causing open64->openat a needless detour via open
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
When x86_64 cpu contains 'cpu_has_xsave' flag(in kernel),call the
pthread_cancel will have a segmentation fault. Because gcc could
not find the end of stack in user signal stack.
Signed-off-by: Zhang Pu <glen4linux@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
As Will noticed, the header this check is currently done in
is asm-only, and is not meant to be included from C code.
This breaks compilation when compiled for a Thumb2-aware CPU.
Move the BX check to its own header, and revert 7a246fd.
Reported-by: Will Newton <will.newton@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Will Newton <will.newton@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
When compiling against the 2.6.18 kernel headers, during linking of
the NPTL library it runs into a "undefined reference to PRIVATE_FUTEX"
error in pthread_cond_timedwait.S. This resolves the problem by
ensuring that the definition for PRIVATE_FUTEX is made available to
pthread_cond_timedwait.S.
Signed-off-by: Dan Moulding <dan.moulding@rackwareinc.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
The conditional load needs to be made part of an IT block on Thumb2
cores.
Signed-off-by: Will Newton <will.newton@linaro.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Commit e1420eca7374cd8f583e9d774c890645a205aaee fixed a bug where a
response code should mean the next server is tried. However, it tries
only the next search domain, and never skips to the next server. This
fix makes sure we try the next server on SERVFAIL.
Signed-off-by: Michel Stam <michel@reverze.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
The ULPs format in glibc has changed slightly and now deviates from
that in uClibc, so when using the files directly from glibc, for
certain types of tests the ULPs values were incorrectly defaulting to
0 rather than using the values provided. This caused some tests to
fail despite the max ULPs being correctly marked.
Regenerated and updated part of the ULPs file to fix this issue. This
also added values for exp and log.
Signed-off-by: Gregory Fong <gregory.0xf0@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
The table format is thrown off a bit because some of these names are
too long to fit the previous format so they were put on their own
lines.
Signed-off-by: Gregory Fong <gregory.0xf0@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
These should have been protected by
if (_dl_debug_reloc && _dl_debug_detail)
, but since the previous change "ldso/mips: actually print results of
each relocation" makes this redundant, just eliminate these prints
instead. I think this is the only item that is still printing
regardless of the value of LD_DEBUG when uClibc is built with
SUPPORT_LD_DEBUG=y.
Signed-off-by: Gregory Fong <gregory.0xf0@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
The patched line was outside of the body of the loop over relocations
and so would only print the results of the last relocation, fix that.
Signed-off-by: Gregory Fong <gregory.0xf0@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Apparently i made a typo when applying
278a06d7abcc8774ba9bb9c15779749c7e2d68cd
sorry..
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Patch "LT.old: Make __errno_location/__h_errno_location thread safe"
uncovered yet another bug with static linking and errno (hopefully this
is last of them all).
Currently, __errno_location is declared weak but is defined strong.
While this provides with the desired weak semantics in dso, it
is subtly broken in static links.
Quoting Joern Rennecke (ARC gcc expert):
| I think the issue is that you declare the function as weak in the
| header file. That is a rare instance where you want the reference
| use declaration that differs a bit from the definition.
| If the reference uses a weakly declared function, that creates a
| weakref, i.e. the linker won't bother to look for this symbol at
| all - if it gets linked in for some other reason, fine,
| otherwise, it stays zero.
So the solution to declare strong, define weak.
Supporting data
-----------------
orig code: ARM mmap wrapper (LT.old build + my prev patch for errno)
_mmap:
@ args = 8, pretend = 0, frame = 0
@ frame_needed = 0, uses_anonymous_args = 0
stmfd sp!, {r4, r5, r7, lr}
ldr r5, [sp, #20]
movs ip, r5, asl #20
beq .L2
bl __errno_location(PLT)
mov r3, #22
str r3, [r0, #0]
mvn r0, #0
...
...
.weak __errno_location
A statically linked hello world program which uses mmap too.
As we can see__errno_location is completely gone - which is
semantically wrong - we need functional errno.
00008274 <__GI_mmap>:
8274: e92d40b0 push {r4, r5, r7, lr}
8278: e59d5014 ldr r5, [sp, #20]
827c: e1b0ca05 lsls ip, r5, #20
8280: 0a000004 beq 8298
8284: e320f000 nop {0}
^^^^^^^^^^
8288: e3a03016 mov r3, #22
828c: e5803000 str r3, [r0]
8290: e3e00000 mvn r0, #0
This in turn is due to a fixup in ARM ld which transforms branch-to-null
into a nop. It is better than crashing but still wrong since errno
handling is removed.
With the patch, errno_location is restored back in test program.
00008274 <__GI_mmap>:
8274: e92d40b0 push {r4, r5, r7, lr}
8278: e59d5014 ldr r5, [sp, #20]
827c: e1b0ca05 lsls ip, r5, #20
8280: 0a000004 beq 8298 <__GI_mmap+0x24>
8284: eb000010 bl 82cc <__errno_location>
8288: e3a03016 mov r3, #22
828c: e5803000 str r3, [r0]
Cc: Christian Ruppert <christian.ruppert@abilis.com>
CC: Francois Bedard <Francois.Bedard@synopsys.com>
Cc: Anton Kolesov <Anton.Kolesov@synopsys.com>
Cc: Joern Rennecke <joern.rennecke@embecosm.com>
Cc: Jeremy Bennett <jeremy.bennett@embecosm.com>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Peter Korsgaard <peter@korsgaard.com>
Cc: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
WHY: errno in uClibc is not thread safe
HOW: __errno_location and it's sibling __h_errno_location need to be called
via PLT - even from within libc. That way when linked with pthread,
intra-uClibc callers will also use the thread safe version.
This is achieved by removing the GI alias for these functions,
ensuring that they get called normally (via PLT)
Verified with ARC LT.old and ARM cubieboard2 buildroot (LT.old)
NPTL is unaffected by this bug.
------------ History behind this patch ---------------------
This is a long standing bug (https://bugs.busybox.net/show_bug.cgi?id=2089)
Others have tried to fix it in past (alteast Peter Korsgaard's patch in 2010),
but somehow failed to be merged (or were backed out afterwards).
http://lists.uclibc.org/pipermail/uclibc/2010-July/044176.html
One of the causes could be side effect of atleast one more bug related to
pthreads and static link which has now been fixed.
http://lists.uclibc.org/pipermail/uclibc/2013-October/047958.html
I have solved this w/o looking at other pacthes but would like to give
credit to Peter and others for confirming that it makes sense.
Cc: Christian Ruppert <christian.ruppert@abilis.com>
CC: Francois Bedard <Francois.Bedard@synopsys.com>
Cc: Joern Rennecke <joern.rennecke@embecosm.com>
Cc: Jeremy Bennett <jeremy.bennett@embecosm.com>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Peter Korsgaard <peter@korsgaard.com>
Cc: Khem Raj <raj.khem@gmail.com>
Cc: buildroot@busybox.net
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
If the dynamic linker is built with LDSO_STANDALONE_SUPPORT=y, it can
load then run a dynamically linked program when explicitly invoked
from the command line. This is used for test and prelink purposes,
and also by other tools like PRoot to force the kernel to load the
right ELF interpreter.
This feature is currently broken on x86_64 since the number of command
line arguments (ie. argc) passed from the kernel is an "elf_addr_t"
(64 bit on x86_64), whereas it is read as an "unsigned int" (32 bit on
x86_64).
Ref: https://github.com/cedric-vincent/PRoot/issues/45
Signed-off-by: Cédric VINCENT <cedric.vincent@st.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Adds the mips and arm test ULPs from glibc.
Signed-off-by: Gregory Fong <gregory.0xf0@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
libm-test.c is a generated file needed by the other math tests, but
test rules were trying to build it as a test if trying to compile the
tests a second time.
Signed-off-by: Gregory Fong <gregory.0xf0@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Using TARGET_ARCH isn't sufficient when ISA can differ.
Signed-off-by: Gregory Fong <gregory.0xf0@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Guard against duplicate inclusion.
We should go back to just including the linux/sysinfo.h ..
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
508b20ead1710e8c295630d096960e51a65b7e5d disentangled
program_invocation_name and __progname but forgot to make __progname
conditional on __UCLIBC_HAS___PROGNAME__, resulting in getconf
attempting to re-define __progname in case __UCLIBC_HAS___PROGNAME__ was
not defined.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This patch fixes the rest of the set-but-not-used warnings in the MIPS
code by using attribute_unused and also fixes some conversion warnings
by changing __arch_compare_and_exchange_xxx_[8|16]_int. Assigning __cmp
(int type) to __prev (pointer type) generates a warning. Assigning zero
to each separately avoids this warning without having to add a cast.
Signed-off-by: Steve Ellcey <sellcey@mips.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This addition allows building uClibc with -mtext-section-literals
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Uclibc is not building for MIPS N64 because pread is trying to use the
pread/pwrite system calls instead of pread64/pwrite64. This patch fixes
the problem and was tested with LFS enabled and disabled.
Signed-off-by: Steve Ellcey <sellcey@mips.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
When explicitly running the compile target we were missing a dependency
to generate the libm-test.c.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
There is support for mips64(r1) already in uclibc, this change adds mips64r2
by adding the MIPS64R2 config variable and checking it in Rules.mak.
Signed-off-by: Steve Ellcey <sellcey@mips.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
The INLINE_SYSCALL, INTERNAL_SYSCALL*, and internal_syscall* macros
are defined for MIPS in both libc/sysdeps/linux/mips/sysdep.h and
libc/sysdeps/linux/mips/bits/syscalls.h. The macros are the same
in both cases except that syscalls.h defines internal_syscalls[567]
the same for N32 and N64 ABIs and has a different definition for O32.
I believe that is correct. The sysdep.h header uses the O32 versions
for N32 and has different definitions for N64. I think that is wrong
and that N32 and N64 should share the same definition (modulo the
type 'long' vs. 'long long' for the arguments. This setup (from
sysdep.h) now agrees with what glibc has.
I am not positive about which header (sysdep.h vs syscalls.h) is
really the right one to have these definitions in but using sysdep.h
seems to work for all my builds.
Signed-off-by: Steve Ellcey <sellcey@mips.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
The special sync_file_range handling is only needed for the O32 ABI
(regardless of whether it is on mips32 or mips64). The N32 (and N64)
ABI's should both use the standard code. This routine was using the
special code for the N32 ABI because that ABI has a word size of 32 bits
and that is wrong. This patch fixes it by checking the ABI used instead
of checking the word size.
Signed-off-by: Steve Ellcey <sellcey@mips.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
When booting a Linux system with qemu-system-mips64 the execution
of $(pwd) in the ash shell triggers a segmentation fault. Ash uses
setjmp/longjmp for exception handling.
After looking at the glibc implementation,
I found some differences, with this patch tries to resolve.
Now the system boots up fine and no segmentation faults occur.
The global pointer should be restored and the types for the
register values should be wide enough.
See:
http://www.cygwin.com/ml/libc-alpha/2003-03/msg00363.html
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This patch uses 'attribute_unused' to clean up a number of warnings
messages that are generated when doing a MIPS build. All of the
changes are in MIPS specific files and there are no code changes
other then adding the unused attribute where needed.
Signed-off-by: Steve Ellcey <sellcey@mips.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
The MIPS build is failing when UCLIBC_HAS_LFS is set to "n"
because posix_fadvise64.c is included in the build when it
should not be.
Signed-off-by: Steve Ellcey <sellcey@mips.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
The clone based version was introduced by commit
58570fc8e1fd601f15be5758ab95013d56771804
vfork: Use clone if arch does not have the vfork syscall
but the code was made unreachable by the previous commit
0a043c30ee71245dfe7c9d82d654312c5e1f6127
vfork: make all archs consistent
that adds a guard which require either __NR_vfork or __NR_fork
to be set.
Neither will be set on an arch that would need to use the
clone based version.
The double declaration of vfork was also introduced by
the merge of the two above mentioned patches.
The missing ; was present in the original patch.
CC: Markos Chandras <markos.chandras@imgtec.com>
CC: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
CC: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Stefan Kristiansson <stefan.kristiansson@saunalahti.fi>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
fork() is broken for MIPS64 N64 ABI. You can check it with a simple
C program statically linked with qemu-mips64 user emulation.
Internally fork() is using the clone system call (at least with NPTL)
with 5 arguments. See ./libpthread/nptl/sysdeps/unix/sysv/linux/i386/fork.c.
The calling conventions for MIPS N32 and N64 allow to use up to 8 registers
for that. See http://en.wikipedia.org/wiki/Calling_convention#MIPS
This is correctly implemented in libc/sysdeps/linux/mips/bits/syscalls.h,
but not in libc/sysdeps/linux/mips/sysdep.h. fork.c uses the later one.
It seems that fork() works fine for MIPS64 N32 with just using the stack like
with the O32 case. There is a user of INLINE_SYSCALL with 7 arguments in
libc/sysdeps/linux/common/sync_file_range.c for MIPS64 N32, so I decided to
only use the macros for the MIPS64 N64 case. With this patch my uClibc based
Linux system boots up fine in qemu-system-mips64.
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
We were ending up with an unrecognised CPP token (which is not diagnosed
for assembler-with-cpp, 59805)
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
crt1.S -> include sysdep.h -> include sysnum.h
In multistage gcc build, Buildroot happens to build crt1.S before
sysnum.h is generated. So break the above include chain and opencode the
ENTRY/END macros.
Reported-by: Anton Kolesov <akolesov@synopsys.com>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
It now generated a file name defconfig in current directory
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Adds support for architecture independent actions.
Cc: Chris Zankel <chris@zankel.net>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Commit ee84b8b400 (linux: posix_fadvise: use new SYSCALL_ALIGN_64BIT)
removed posix_fadvise implementation for xtensa, since xtensa does not
define __NR_fadvise64. Reuse the ARM support code to restore xtensa support.
This commit is based Mike Frysinger's suggested patch.
Cc: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Make __SIGEV_PAD_SIZE to take __WORDSIZE into account for alpha, mips
and ia64 arches.
Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Add "__pid_t _tid" member which is used for some packages, like rt-test
for instance, which fails with an error like this one:
src/cyclictest/cyclictest.c:638:9: error: 'union <anonymous>' has no
member named '_tid'
Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
sync up to 1542ea42da59018860a987f34f065cd120982e8c
(2.10 + cleaner exit status handling)
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
In a multi-threaded application where res_init() was called either
directly or implicitly, getaddrinfo() and others failed to add the
DNS search domain to hostnames.
This problem made it not possible to look up a hostname without its
domain appended.
The problem is caused by res_sync_func() overwriting the configuration
read by __open_nameservers() immediately after it is read.
The suggested solutin is to disable res_sync_func() while reading name
server configuration in res_init().
Signed-off-by: Kenneth Soerensen <kenneth.sorensen@spectralink.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Not all architectures use 'ret' as function return instruction. For example,
xtensa usually uses 'retw'. Use the ret_ERRVAL arch dependant macro instead.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
* Add a common ret_ERRVAL definition
* Remove ret_ERRVAL from architectures using the common 'ret'
* Add 'undef' to architectures that need a different return instruction
* Add '#include <common/sysdep.h>' to cris and ia64 that were missing it
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
For this port, I would like to give due credit to:
- Folks from Codito technologies (Sameer, Amit, Kanika, Ramana,...)
who did the very first port
- ARC UK from 2007-2009 (Joern, Irfan, Khurram, Phil...
- Late Brendan Kehoe (may he RIP)
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
When accessing errno, a per thread variable, from _stdio_init
a SIGBUS error happens. This change fixes the wrong relocation
and debug output.
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
sync up to 55501a61dd257e941d53f175350ee52fc6ff2866
(2.9 + refactor keyword and symbol matching)
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
We run the linker through the compiler-driver, so we have to check
linker flags through the driver, too.
Otherwise we might check $LD's flags while we will run $CC's linker
and those might be different beasts.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
res_init() was not atomic, which could give undesired behaviour. Now
res_init() is completely locked under one lock and the locking is
removed from __res_vinit().
Signed-off-by: Kenneth Soerensen <kenneth.sorensen@spectralink.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Commit aab4df0fb51660300559f5f29290709db2f7bfee says that the line with
after res_init() function. Commit
7f74de5d4d6d10baafab4b37bb3d472f5c5f0e8c moves the res_init()
function below the line with #undef _res.
This commit moves res_init() back above #undef _res.
Signed-off-by: Kenneth Soerensen <kenneth.sorensen@spectralink.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
The argument names used in the prctl wrapper collide with the internal
variable names in syscall wrapper macros on some architecture. This
currently breaks xtensa. grep for '\b_a[0-9]' indicates that metag and
microblaze might also be affected.
Prefix argument names with '_prctl' to avoid collision.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Fix the following build failures:
ld: tst-tls10.o: undefined reference to symbol 'f2a'
ld: tst-tls12.o: undefined reference to symbol 'a1'
binutils ld defaults to --no-copy-dt-needed-entries since version 2.22. Add
library dependencies explicitly.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Commit e3dec33ba (powerpc: Add TLS and NPTL support) introduced a duplicate of
_dl_add_to_slotinfo and _dl_initial_error_catch_tsd declarations.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Reset CSRC-y to prevent pollution from libc/misc/glob/Makefile.in, and fix the
following error messages when UCLIBC_HAS_OBSTACK is disabled:
nm: 'libc/misc/gnu/glob-susv3.os': No such file
nm: 'libc/misc/gnu/glob64-susv3.os': No such file
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This fixes the following warning when SUPPORT_LD_DEBUG_EARLY is not enabled:
ldso/ldso/ldso.c: In function '_dl_get_ready_to_run':
ldso/ldso/ldso.c:754:16: warning: unused variable 'tmp' [-Wunused-variable]
This was introduced with commit 94cc6edb (ldso: Rework global scope handling
and symbol lookup mechanism).
Cc: Filippo Arcidiacono <filippo.arcidiacono@st.com>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
arm call to posix_fadvise simply calls posix_fadvise64
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Add __kernel_long_t and __kernel_ulong_t types for Meta. The rest of
the architectures have been fixed in:
6a76edddaa62ff06f178143b582167734cb55c18
"libc/sysdeps: add __kernel_long and __kernel_ulong"
Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
A simple statically linked hello world program was segfaulting for ARC
in linuxthreads.old configuration (although the root casue applies
cross-arch for NPTL as well as linuxthreads.old as described)
The crash was due to branch to NULL in _stdio_init
0001026c <_stdio_init>:
1026c: push_s blink
1026e: st.a r13,[sp,-8]
10272: bl.d 0 --> supposed call to __errno_location
The call was NOT getting patched to libc internal only alias
__GI___errno_location, because it was weak while it's exported cousin,
__errno_location was strong/normal.
arc-linux-uclibc-nm libc/misc/internals/__errno_location.os
00000000 W __GI___errno_location
00000000 T __errno_location
This is exactly opposite to what is expected.
Quoting Peter S. Mazinger, commit 87936cd013041 "errno and *_init cleanup"
| The rule adopted:
| for enabled threads we make in libc the __GI_x() variants strong, x() weak
| and (should) provide another strong x() in libpthread.
| If threads are disabled, even the __GI_x() variants are weak.
With the fix, we see the right settings as below
00000000 T __GI___errno_location
00000000 W __errno_location
Note that problem won't show up in a static busybox build as it references
errno and that seems to elide the issue.
I can confirm the same/more issues with latest ARM buildroot builds w/o
my fix.
(1). linuxthreads.old (broken just like ARC)
arm-linux-nm uclibc-snapshot/libc/misc/internals/__errno_location.os
00000000 W __GI___errno_location
00000000 T __errno_location
But presumably the issue there is NOT catestrophic because ARM linker is
likely smarter and patches a NOP instead of NULL branch.
00008388 <_stdio_init>:
8388: e92d4038 push {r3, r4, r5, lr}
838c: e320f000 nop {0}
(2) NPTL build (exported version is not weak)
00000000 T __GI___errno_location
00000000 T __errno_location
This causes a static link with libpthread and test program
referencing errno to fail to link.
#include <errno.h>
int main(void)
{
printf("%d\n", errno);
}
arm-linux-gcc -static -pthread -o tst tst.o
arm-buildroot-linux-uclibcgnueabi/sysroot/usr/lib/libc.a(__errno_location.os):
In function `__errno_location': __errno_location.c:(.text+0x0):
multiple definition of `__errno_location'
arm-buildroot-linux-uclibcgnueabi/sysroot/usr/lib/libpthread.a
(errno_location.os):errno_location.c:(.text+0x0): first defined here
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Cc: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
atomic_decrement_if_positive() returns the old value of &mem, not the
(sometimes undefined) value of __tmp.
Fixes the uClibc nptl/tst-sem3 test.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Chris Zankel <chris@zankel.net>
Before the patch, the query itself was sent via UDP
(the query contained correct protocol ID).
The fix is taken from glibc.
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
isfdtype(int fd, int fdtype) check whether a file descriptor fd is
of type fdtype, where the types are defined in stat(2). It is
supported in glibc and BSD, and used by utilities such as acpid.
Signed-off-by: Anthony G. Basile <blueness@gentoo.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
UCLIBC_MALLOC_DEBUGGING is useless without assert().
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Previous patch only fixed the build. This patch fixes the behavior at
run-time as well.
It fixes "Value too large for defined data type" messages caused by
fstat syscall wrapper returning -EOVERFLOW.
Signed-off-by: Mischa Jonker <Mischa.Jonker@synopsys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
fix breakage from 603af30d
Removing the whitespace from findstring for 64 bit architectures has
bad consequences since powerpc would be a match in powerpc64 and sparc
would also be a match in sparc64.
That doesn't make them 64 bits in reality causing general breakage.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Reviewed-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
res_mkquery() takes out __resolv_lock to copy _res.options to function local
_res_options on line 4204, but later unsafely accesses _res.options without a
lock, instead of its local copy _res_options, on line 4221. Looks like a
period / underscore typo.
Signed-off-by: Vanya Sergeev <vsergeev@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
gcc (eg 4.7.3) hardcodes the MIPS64 interpreters like this:
(see gcc/config/linux.h and gcc/config/mips/linux64.h)
o32: UCLIBC_DYNAMIC_LINKER "/lib/ld-uClibc.so.0"
n32: UCLIBC_DYNAMIC_LINKERN32 "/lib32/ld-uClibc.so.0"
n64: UCLIBC_DYNAMIC_LINKER64 "/lib/ld64-uClibc.so.0"
The existing check for MIPS64 in uClibc is wrong because it does
not respect the selected ABI
We fix this by explicitely checking the selected ABI instead of the
selected MIPS variant.
Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Cc: Anthony G. Basile <blueness@gentoo.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Sync up with the changes from C6X entering the kernel.
Signed-off-by: Luis Machado <lgustavo@codesourcery.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Commit a83ea57a50 introduces code for falling back on sendfile64
when the sendfile syscall is not available. However, as written,
that code leads to sendfile64 being globally hidden in situations
where __NR_sendfile and __USE_LARGEFILE64 are defined. In this case,
the flow executes strong_alias_untyped(sendfile,sendfile64) in
libc/sysdeps/linux/common/sendfile.c and libc_hidden_proto(sendfile64)
in include/sys/sendfile.h, but nowhere does it do libc_hidden_def(sendfile64).
This patch adds the needed libc_hidden_def(sendfile64).
This breaks, for example, xfsprogs on x86_64. Reported by mardok4
in IRC.
Signed-off-by: Anthony G. Basile <blueness@gentoo.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Since version 2.5 glibc allows prev to be a NULL pointer in insque, whereas
uClibc segfaults in this case. This fixes the issue and makes insque
initialize q_forw and q_back with NULLs if prev == NULL.
Signed-off-by: Bartosz Golaszewski <bartekgola@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Static applications that call pthread_exit on the main thread segfault.
This is because after a thread terminates __uClibc_main decrements
__nptl_nthreads which is only defined in pthread_create.
Therefore the right solution is to add a requirement to pthread_create
from pthread_exit.
This patch has been backported from glibc tree:
http://sourceware.org/git/?p=glibc.git;a=commitdiff;h=e1f0b2cfa14b058ab033431b70428ccda8ec6ed0
For further detail see also glibc bug at:
http://sourceware.org/bugzilla/show_bug.cgi?id=12310
Signed-off-by: Filippo Arcidiacono <filippo.arcidiacono@st.com>
Cc: Vladimir Nikulichev <v.nikulichev@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
uclibc is compiled with elf/dwarf and secondly HAVE_CPP_ASM_DEBUGINFO
is not defined so we generate bogus .stab sections needlessly
Some of post processing tools get confused when they see both
dwarf and stab sections
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Prompted by lkml discussion of a MIPS bug where sending
signal 128 was found to be able to crash the machine :/
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
975bca165c3e10e74c05c0384fd58f45a7025a3c avoided relocations to
_obstack_newchunk used by obstack_vprintf but missed to enable the
corresponding public symbol which is used by the obstack macros.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This adds a straight forward implementation for obstack_printf and
obstack_vprintf on uClibc's already existing obstack_grow and
vasprintf. It does not attempt to port over glibc's implementation
in terms of _IO_* structs and functions.
Signed-off-by: Anthony G. Basile <blueness@gentoo.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
As in timerfd.h, eventfd.h needs arch-specific definition files.
alpha, mips and sparc needs separate file, all the other arch
will use common definition.
This problem is already fixed in glibc.
Also sanitize and provide bits for hppa.
Make sure not to install the new bits/eventfd unless eventfd support is
enabled.
Signed-off-by: Hiroaki KAWAI <kawai@stratosphere.co.jp>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
On MIPS, several relocations that were original only resolved by the
dynamic linker were reused as static relocations. Consequently the
macros TLS_DTPREL_VALUE and TLS_TPREL_VALUE defined in
libpthread/nptl/sysdeps/mips/dl-tls.h need to be available even for
!SHARED.
Relevant: http://www.linux-mips.org/wiki/NPTL#History
Original patch by Vincent Wen <wenvincent@gmail.com>:
http://lists.uclibc.org/pipermail/uclibc/2013-April/047707.html
When build statically linked applications for MIPS platform,
sometimes the linker fails with following errors:
undefined reference to TLS_DTPREL_VALUE
undefined reference to TLS_TPREL_VALUE
The include of dl-tls.h is only in code guarded by SHARED,
Removing the SHARED compilation option to cover static link too.
Signed-off-by: Vincent Wen <wenvincent90@gmail.com>
Signed-off-by: Gregory Fong <gregory.0xf0@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Partially reverts "parser_config.[ch]: remove duplicated hidden functions"
There's a namespace collision for config_{open,close,read} functions
between uClibc and Busybox. Callers in uClibc and busybox need to call
their local variants. In case of static linking, both the variants need
to be pulled in the final executable.
Turns out that the existing hidden attribute is not enough for static
linked Busybox where the linker treats these symbols as duplicate and
pulls in only the local (Busybox's own) variants in the final link.
This obviously causes the uClibc callers to crap out at runtime.
This can be fixed by restoring the libc_hidden_proto() annotation
which causes the libc variants to be aliased with __GI_ prefix thus
making both the symbols distinct from linker's perspective.
This issue was found with ARC gcc 4.8 + uClibc trunk.
I can confirm the same thing with ARM uClibc buildroot build (when
switching to uClibc daily snapshot).
All the credit for debugging the issue goes to Anton,Joern and Jeremy -
I simply found the cure :-)
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Cc: Anton Kolesov <akolesov@synopsys.com>
Cc: Jeremy Bennett <jeremy.bennett@embecosm.com>
Cc: Joern Rennecke <joern.rennecke@embecosm.com>
Cc: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Cc: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Commit 71c10c484e7dc113396cccb7e503befb759c6346 broke libubactrace build
due to some typos, so that backtrace.o[sS] were not built.
This definetively fixes this problem.
Signed-off-by: Filippo Arcidiacono <filippo.arcidiacono@st.com>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
This reverts commit e567c399ff86d007d8c4586f0dd5e0ca61e283ca.
since it breaks badly (e.g. busybox netstat)
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
New architectures, which don't define ARCH_HAS_DEPRECATED syscalls, need
to use the new interfaces from libc/sysdeps/linux/generic-syscalls/bits/
Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Linuxthreads(old and new) need deprecated syscalls to build. Existing
architectures support these system calls but new architectures don't.
This symbol has no functional change apart from hidding the Linuxthreads
symbols from arches that don't support them.
Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Fixes the following compilation error:
_lfs_64.h: # error Do not include this header in files not built when LFS is disabled
Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
regressions introduced in 1b3025b7352f5e432ffa1c7adc57085ac9092b77
"buildsys: switch libc to kbuild-style"
unistd: Handle !UCLIBC_HAS_GET{, SUB}OPT cases
libc/libc_so.a(getsubopt-susv3.os): In function getsubopt':
getsubopt-susv3.c:(.text+0x0): multiple definition of getsubopt'
libc/libc_so.a(getsubopt.os):getsubopt.c:(.text+0x0):
first defined here.
The solution is to filter out the correct files when
UCLIBC_HAS_GETOPT and UCLIBC_HAS_GETSUBOPT are not defined.
inet: Move INET_RPC_{DIR, OUT} before CSRC_ALL
libc/libc_so.a(rpc_thread.oS): In function __rpc_thread_svc_max_pollfd':
rpc_thread.c:(.text+0x148): undefined reference to svc_max_pollfd'
libc/libc_so.a(rpc_thread.oS): In function __rpc_thread_svc_pollfd':
rpc_thread.c:(.text+0x184): undefined reference to svc_pollfd'
The CSRC_ALL variable uses the INET_RPC_DIR variable which is
initialized later on. We fix this problem by moving the
INET_RCP_{DIR,OUT} definitions before we use them in CSRC_ALL
Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
TARGET_SUBARCH can only ever be set if TARGET_ARCH is set.
Only include SUBARCH includes if SUBARCH is non-empty.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Continuosly dlopen and dlclose of shared object will cause a memory leak
in atexit function. This fix reuse free slots at the end of the list.
For further detail see https://bugs.busybox.net/show_bug.cgi?id=2455
Signed-off-by: Ronald Wahl <ronald.wahl@raritan.com>
Tested-by: Filippo Arcidiacono <filippo.arcidiacono@st.com>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
Fixes a build error:
In file included from libc/inet/recv.c:8:0:
libc/inet/socketcalls.c: In function '__recv_nocancel':
libc/inet/socketcalls.c:203:57: error: 'NULL' undeclared (first use in this function)
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
If we end up doing '#undef __NR_xxx', we don't want to leave the
corresponding SYS_xxx symbol defined. So undef it too.
For example, with the ARM EABI layer, we have a bunch of legacy
syscalls that we define early on and then later undefine (such
as __NR_utime). But we left SYS_utime defined so code that tests
for that define before using it would be broken (since it'd be
defined to a non-existent symbol).
URL: https://bugs.gentoo.org/425006
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Setting KERNEL_HEADERS up to the exact path all the time is a pita,
especially when the toolchain is often times already configured to
do the right thing. So if the user has set that to "", don't force
any specific paths.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
with -O0 we (e.g. lockf) might end up with references to
_Unwind_Resume, so pull in gcc_eh in this case..
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
As noted by Florian Fainelli:
LD libpthread-0.9.34-git.so
libpthread/linuxthreads.old/libpthread_so.a(wrapsyscall.oS): In function `fcntl64':
wrapsyscall.c:(.text+0xd8): undefined reference to `__libc_fcntl64'
libpthread/linuxthreads.old/libpthread_so.a(wrapsyscall.oS): In function `lseek64':
wrapsyscall.c:(.text+0x190): undefined reference to `__libc_lseek64'
libpthread/linuxthreads.old/libpthread_so.a(wrapsyscall.oS): In function `pread64':
wrapsyscall.c:(.text+0x39c): undefined reference to `__libc_pread64'
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Add the obsolescent SUSv3 family of user context manipulating functions
for arm, i386, mips, x86_64.
Signed-off-by: Timon ter Braak <timonterbraak@gmail.com>
Signed-off-by: Florian Fainelli <florian@openwrt.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Signed-off-by: Natanael Copa <ncopa@alpinelinux.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
The current code for dlsym() when invoked with RTLD_NEXT lookup
searches for the module where it's being called from, and executes the
_dl_find_hash only for the next module in the chain. However, if the
looked symbol is not there, the rest of the modules are not checked.
Generally this is not a problem as symbols are merged for the parent
modules; so this affects only RTLD_NEXT.
This patch adds a loop iterating through all the following modules.
Signed-off-by: Timo Teräs <timo.teras@iki.fi>
Reviewed-by: Filippo ARCIDIACONO <filippo.arcidiacono@st.com>
Tested-by: Florian Fainelli <florian@openwrt.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
If the syscall returns with an error the stack pointer and r4 register
are not restored because the instruction 'ldmnefd sp!, {r4}' is executed
after branching to '__error' label.
This bug has been spotted out by running './utstest clone 5' from LTP
built with -fstack-protector-all compiler flag as log below:
root@cortex-a9:/usr/tests/ltp/testcases/bin# ./utstest clone 5
stack smashing detected: ./utstest terminated()
Regression introduced by commit e58798e107d652644629a1daaa95d76430808d53
Signed-off-by: Filippo Arcidiacono <filippo.arcidiacono@st.com>
Signed-off-by: Giuseppe Di Giore <giuseppe.di-giore@st.com>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
Setjmp was only saving the lower order of registers (a0-a3) correctly, but
not the higher ones (a4-a8/a12).
The change also includes additional information, and renames many of the
registers, so that setjmp and longjmp look more like the inverse of
each other.
Signed-off-by: Chris Zankel <chris@zankel.net>
Update libc/sysdeps/linux/common/sys/ptrace.h to latest from glibc's
sysdeps/unix/sysv/linux/sys/ptrace.h.
This adds definitions for operations:
- PTRACE_GETREGSET
- PTRACE_SETREGSET
- PTRACE_SEIZE
- PTRACE_INTERRUPT
- PTRACE_LISTEN
And adds flags:
- PTRACE_SEIZE_DEVEL
And adds event codes:
- PTRACE_EVENT_SECCOMP
This is to allow access to the generic interface for accessing
architecture specific regsets using the corresponding NT_* types,
required for new Linux kernel architecture ports.
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
When ASSUME_DEVPTS is disabled, we end up exporting __unix_grantpt and
get a reloc to it from grantpt. Mark it static to fix all of that.
Reported-by: Michael Deutschmann <michael@talamasca.ocis.net>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
This drops __signed, __volatile, and __const. Only the latter was
used in the code base, and for uClibc, not consistently. Much of
the code used plain "const" which meant "__const" was useless.
Really, the point of this is to stay in sync with what glibc did.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Current builds show the warning:
CC libpthread/nptl/pthread_create.os
libpthread/nptl/pthread_create.c: In function 'start_thread':
libpthread/nptl/pthread_create.c:418:3: warning: function declared
'noreturn' has a 'return' statement [enabled by default]
libpthread/nptl/pthread_create.c:418:3: warning: 'noreturn' function
does return [enabled by default]
Looking at glibc, it does not mark this function as noreturn. Testing
code size, gcc is smart enough to detect the code never actually returns
regardless of the attribute. So drop it to keep in sync with glibc and
to avoid the warnings.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Current builds show the warning:
CC libpthread/nptl/sysdeps/unix/sysv/linux/mq_notify.os
libpthread/nptl/sysdeps/unix/sysv/linux/mq_notify.c: In function 'helper_thread':
libpthread/nptl/sysdeps/unix/sysv/linux/mq_notify.c:142:3: warning:
function declared 'noreturn' has a 'return' statement [enabled by default]
Looking at glibc, it does not mark this function as noreturn. Testing
code size, gcc is smart enough to detect the while(1){} never actually
returns regardless of the attribute. So drop it to keep in sync with
glibc and to avoid the warning.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Building with NPTL enabled and shared library support disabled we hit:
In file included from libpthread/nptl/sysdeps/generic/dl-tls.c:30:0:
./ldso/include/dl-elf.h: In function '__dl_parse_dynamic_info':
./ldso/include/dl-elf.h:173:20: error: 'RTLD_NODELETE' undeclared (first use in this function)
./ldso/include/dl-elf.h:173:20: note: each undeclared identifier is reported only once for each function it appears in
make: *** [libpthread/nptl/sysdeps/generic/dl-tls.os] Error 1
A previous commit (f26c5f6952ce9bf8edec9c1571c47addb1bcc442) touched
on a similar issue, but added the include to the incorrect location.
Reported-by: Christophe Lyon <christophe.lyon@st.com> [arm nommu]
Reported-by: Daniel Beecham <daniel@lunix.se> [static x86_64]
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
ELFOSABI_GNU replaces ELFOSABI_LINUX, the latter is kept as a compatibility
alias, and ELFOSABI_HURD is removed. See the table on
<http://www.sco.com/developers/gabi/latest/ch4.eheader.html#osabi> for
reference.
Signed-off-by: Thomas Schwinge <thomas@codesourcery.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Fix the asm-optimised memcpy and memmove so they
work for little-endian as well as big-endian.
Testing has shown no issues, but I am not a microblaze
asm expert so YMMV.
Signed-off-by: Steve Bennett <steveb@workware.net.au>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Add more atomic intrinsics. These are mostly non 32-bit versions, which
are not support by Xtensa. This file needs some more clean-up and
consolidation.
Signed-off-by: Chris Zankel <chris@zankel.net>
The Xtensa version of sysdep.h didn't include sys/syscall.h. This header
file is, for example, needed by not-cancel.h, which uses the
INTERNAL_SYSCALL macro.
Signed-off-by: Chris Zankel <chris@zankel.net>
Xtensa only supports the ELF binary format, so the user.h file isn't
needed. However, it's included by other files, so provide a stub user.h.
Signed-off-by: Chris Zankel <chris@zankel.net>
The current code implements the 'm' modifier only for 's'
conversions and would cause a segfault if it was used for 'c'
or '[' conversions. This patch extends the code to cover these
cases too.
Signed-off-by: Will Newton <will.newton@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Linux 3.4 added __kernel_long_t and __kernel_ulong_t and various
exported header files were updated to use these new types. Add the
definitions for __kernel_long_t and __kernel_ulong_t to the relevant
kernel_types.h headers.
This change was automated with the following scriptlet
git grep --name-only 'typedef.*__kernel_old_dev_t' \
| xargs sed -i '/typedef.*__kernel_old_dev_t/ a\
typedef long\t\t__kernel_long_t;\
typedef unsigned long\t__kernel_ulong_t;'
Whitespace in arm, avr32, hppa, sparc was then manually fixed up.
Signed-off-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
--
Here's a cleaned up patch which should get the whitespace right. I'm a
bit iffy about the sparc changes they make sense to me but it's not a
platform I have access to.
I can break this up per arch or per maintainer if requested.
libc/sysdeps/linux/alpha/bits/kernel_types.h | 2 ++
libc/sysdeps/linux/arm/bits/kernel_types.h | 2 ++
libc/sysdeps/linux/avr32/bits/kernel_types.h | 2 ++
libc/sysdeps/linux/bfin/bits/kernel_types.h | 2 ++
libc/sysdeps/linux/c6x/bits/kernel_types.h | 2 ++
libc/sysdeps/linux/cris/bits/kernel_types.h | 2 ++
libc/sysdeps/linux/e1/bits/kernel_types.h | 2 ++
libc/sysdeps/linux/h8300/bits/kernel_types.h | 2 ++
libc/sysdeps/linux/hppa/bits/kernel_types.h | 2 ++
libc/sysdeps/linux/i386/bits/kernel_types.h | 2 ++
libc/sysdeps/linux/ia64/bits/kernel_types.h | 2 ++
libc/sysdeps/linux/m68k/bits/kernel_types.h | 2 ++
libc/sysdeps/linux/microblaze/bits/kernel_types.h | 2 ++
libc/sysdeps/linux/mips/bits/kernel_types.h | 4 ++++
libc/sysdeps/linux/nios2/bits/kernel_types.h | 2 ++
libc/sysdeps/linux/powerpc/bits/kernel_types.h | 4 ++++
libc/sysdeps/linux/sh/bits/kernel_types.h | 2 ++
libc/sysdeps/linux/sh64/bits/kernel_types.h | 2 ++
libc/sysdeps/linux/sparc/bits/kernel_types.h | 4 ++++
libc/sysdeps/linux/v850/bits/kernel_types.h | 2 ++
libc/sysdeps/linux/x86_64/bits/kernel_types.h | 2 ++
libc/sysdeps/linux/xtensa/bits/kernel_types.h | 2 ++
22 files changed, 50 insertions(+)
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Indeed if the libpthread is before the libc in the library look up
the SIGCANCEL is removed from the list of the blocked signal by
sigfillset func, this can produce the handler not properly called.
This commit revert what Denys modified in commit
162cfaea20d807f0ae329efe39292a9b22593b41.
Signed-off-by: Filippo Arcidiacono <filippo.arcidiacono@st.com>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
Allow use of immedate values as the 6th syscall argument. Otherwise we must
store the arg on memory. This gives gcc more options to optimize better.
This also works around an issue with posix_fallocate.
Signed-off-by: Natanael Copa <ncopa@alpinelinux.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
CLOCK_MONOTONIC_RAW is available since 2.6.28
(2d42244ae71d: clocksource: introduce CLOCK_MONOTONIC_RAW), and
CLOCK_*_COARSE since 2.6.32 (da15cfdae033: time: Introduce
CLOCK_REALTIME_COARSE).
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
All other TLS data in uClibc is explicitly using the IE model,
so use it for __resp too. This generates smaller and faster
code.
Signed-off-by: Will Newton <will.newton@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
The definition of __resp has been hidden for some years (commit
22de495d). Remove the comment suggesting that it is not.
Signed-off-by: Will Newton <will.newton@imgtec.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
In llseek.c, we already allow both variants; the same needs to be done here
to avoid duplicate definitions of lseek64_nocancel.
Signed-off-by: Bernd Schmidt <bernds@codesourcery.com>
pthread_initialize is a static function and should not be mentioned in a
header. The #pragma weak for it appears to confuse gcc-4.7.
Signed-off-by: Bernd Schmidt <bernds@codesourcery.com
Make a local copy of the tid value to avoid a race condition,
as the value could have been changed to 0, thus using a pointer
it would have been passed to the lll_futex_wait modified.
Signed-off-by: Stas Sergeev <stsp@users.sourceforge.net>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
temporarily work around circular dependency in linking that was
introduced by 01c125c3bd8f949f8e5711e09152859eecd1b004
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Remove hardcoded path from xtensa, we have sysdep.h in path
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Add config option to provide arc4random without device access.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
While there, fix epoll_pwait syscall, it takes 6 arguments
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
based on a patch from Kevin Cernekee <cernekee@gmail.com>
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
While there, guard __clone[2] with _LIBC
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
rcmd.c uses sigblock and sigsetmask, make the function static and include them directly until
a rewrite is done.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
replace USE_TLS, HAVE___THREAD and USE___THREAD with __UCLIBC_HAS_TLS__
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Use __UCLIBC_HAS_TLS__ instead of USE___THREADS (although this LT implementation
will never support TLS).
Disable unused/unneeded members of the pthread_functions structure.
No redirect/forward for _pthread_cleanup_push/pop, it would have not worked anyway
due to the fact, that the structure member was not initialized. Disable it's possible
internal use in libc-lock.h.
Avoid using internals.h in libc_pthread_init.c (moving a prototype to pthread-functions.h).
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This file needs anyway some massage (style update), I'll leave the fix
to the committer.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Use sizeof(info) instead of sizeof(siginfo_t).
stubs.c: add sigqueue stub.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Since mips has different kernel sigset_t, use the new macro in common
code that is used by mips as well.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Remove the related headers.
Adapt sh's type to the one used everywhere and add prototype for it.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
__libc_X names are intended to be used in other uClibc provided libs,
but not in libc.
Mark the syscall __always_inline.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Do not export __sigpause.
Provide __bsd_sigpause (unprototyped) if anyone may need it.
Adapt cancellation to use cancel.h.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
__socketcall.c: remove guard, we definitely need this syscall
Re-added guard plus added comment (Bernhard)
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
If an arch does not support some option, that has to be coded in Config.in
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
xstatconv.h needs features.h to really "see" config options
xstatconv.c does not need sys/syscall.h
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Update file (same as in NPTL).
Disable unused parts, the remaining inlines will be fixed later.
Copy it to include to be sure that we use this one independently of
chosen threads.
Provide write,close and waitpid_not_cancel for LT new.
Use wait4 syscall instead of waitpid syscall (kernel says so).
Provide open_not_cancel and close_not_cancel for librt's spawn.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
No idea why it was removed when NPTL was added,
but glibc uses it as well.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
text data bss dec hex filename
272224 2298 17576 292098 47502 lib/libuClibc-0.9.32-rc3-git.so.old
271806 2298 17576 291680 47360 lib/libuClibc-0.9.32-rc3-git.so
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Having them defined in both places was causing errors with static linking
Signed-off-by: Austin Foxley <austinf@cetoncorp.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Prepare the file to be used in all arch specific files
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
the generic open, close, waitpid, read, write are good for NPTL too
no good reason to have the same sleep in both libc and libpthread (NPTL)
no good reason to add sigwaitinfo, sigtimedwait and sigwait to libpthread (NPTL)
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
make __getpid static
provide getppid alias if needed
remove unneeded libc_hidden_proto
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
use __SYSCALL_SIGSET_T_SIZE
use non-cancellable select in fallback
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
sigtimedwait:
- provide __sigtimedwait_nocancel
- use __SYSCALL_SIGSET_T_SIZE instead of _NSIG / 8
- do not provide __sigtimedwait
- guard a section to avoid failure on archs if SI_TKILL/SI_USER are not defined
sigwaitinfo:
- simply use sigtimedwait since that handles cancellation already
sigwait:
- use non-cancellable functions (sigtimedwait, sigsuspend)
- get rid of code already done in __sigtimedwait_nocancel
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
LT_OLD provides cancellable versions, do it for all THREADS.
llseek.c: use newly added macros for offset handling.
Add a comment about endianness issue around offset.
Compile llseek.c only on 32bit archs.
Provide aliases for 64bit archs or if syscall is not available.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
signal.h: add __SYSCALL_SIGSET_T_SIZE, since kernel sigset_t is different
on mips and use that instead of _NSIG / 8
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Use __select_nocancel instead of select
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Guard changed from NPTL to __NEW_THREADS to cover LT new as well.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Use __wait4_nocancel, since wait4 is not available all the time
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
fcntl64 missed cancellation.
Guard fcntl64 for 32bit archs.
Reuse as much code as possible in __syscall_fcntl.c.
Provide alias fcntl64 if that syscall is not available.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
and use __waitpid_nocancel in fallback code
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Use __wait4_nocancel since wait4 is not available all the time.
Add a comment about not using waitpid syscall.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
close.c: add function __close_nocancel_no_status to be used internally
in libc avoiding inlining it everywhere.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Correct wait4 guard (it is only __USE_BSD)
wait3, system: use __wait4_nocancel
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
__vfork is hidden.
__GI_vfork is strong (not weak as it was).
vfork is weak.
e1 can use the generic version.
Note: libc_hidden_def(x) has different meaning in .c and .S files.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
utime.c: fix a cast, tv_sec is of type time_t
Add stub for utimes.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Add hidden umount2 to be used in umount eventually.
Rework umount to either use directly umount2 syscall or function.
docs say, that mount and umount are also Linux specific
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Rework a bit for less code.
always_inline syscall used only once, do not inline syscall used twice.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Add hidden stime for possible use in settimeofday.
Add stubs for both functions if none of the syscalls is present.
Avoid circular dependency.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Remove syscall.h as well, not needed anymore.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Fix arg type (const missing).
Reorganize a bit.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Guard as GNU extension.
Reorganize a bit.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Use param.h instead of MIN.
Use stddef.h instead of offsetof.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
_lfs_64.h makes the compile fail, if LFS is not enabled, no need for the guard.
Reorganize to include only the minimal necessary headers.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
No need for hidden openat64
_lfs_64.h does not allow to compile *64.c if LFS is not enabled, remove guard
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Use __sigemptyset instead of sigemptyset.
Remove hidden sigemptyset.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Touch signals only if DODEBUG is enabled.
Make the signal selection dependent on DODEBUG, as last resort use SIGKILL.
Use internal functions with less arguments, some savings.
Fix a warning about unused argument.
Do not use openlog/closelog, while there remove their hidden versions.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
It is only used within the same file.
Guard the visible version according to header.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Include only what is necessary.
Mark __syscall_getgroups __always_inline.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Warn if the arch has no __UCLIBC_ABORT_INSTRUCTION__.
Run only one syscall, exit_group is not an exclusivity, use it if THREADS are enabled.
Guard according header.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Prefer static inline instead of hidden (__bsd_getpt)
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Remove all versions, that can be covered by the common one.
Drop incorrectly implemented fallback support, if __NR_fadvise64_64 is not available.
To support 64 bit ppc, move ppc support into common.
Cleanup including headers.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Makefile.arch: include it only if ADVANCED_REALTIME is defined.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
They can't be compiled if LFS is disabled, they are using off64_t.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
The header had to be removed under the guard, else asm/cachectl.h might
not be available at all.
While there, disable __cachectl prototype.
Use untyped strong_alias for cacheflush because char * != void *.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Add support for %p to _dl_dprintf for later corrections in debug messages.
Disable _dl_debug_file if LD debugging is not enabled and change it's use to 2.
The use of dprintf in elfinterp.c will spit out warnings, will be fixed in the
reworked ldso.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
While there, add some defines to help build, cleanup
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Include stddef.h instead of defining NULL
Adapt guard to file name.
Reorder includes a bit.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
dl-sysdep.h has to come before the fallbacks in dl-defs.h,
else the wrong macros are used.
While there adapt the guard to the file name.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
While there change the guard of the file
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
dlinfo is a GNU extension, it should be of the same type
as on glibc.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
The guard is changed to allow to be used the file in libc as well.
Include string.h (although already included by ldso.h).
Guard the rest of the file according to config options.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
powerpc guard depends on the compiler options, __powerpc__ should
work all the time.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This can't happen, __mips__ was already checked earlier
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Add a comment to uClibc_stdio.h to keep macros in sync with uClibc_mutex.h
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Proposed-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Move h_errno related stuff to separate file.
Do not hide errno and h_errno for non-TLS, else it keeps being 0.
Make __[h_]errno_location weak for non-TLS.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Remove unneeded headers.
Guard inclusion of fcntl.h.
While there, remove an obsoleted comment.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Rename PAGE_SIZE to __PAGE_SIZE to be sure that we do not get the
definition from uClibc_page.h.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Move weak_function and weak_const_function under HAVE_WEAK_SYMBOLS guard
and provide fallbacks.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Use the one from libc.
While there, remove unused _dl_strncmp() and superfluos static forward declarations.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
While there, remove unused __strtok.
Use err() and warn() if uClibc provides them (if ldconfig is not static).
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Host iconv needs only L_iconv_main defined and __iconv_codesets
has to included, as host system does not provide it.
Keep __iconv_codesets out of iconv linked against uClibc.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Define a common view of __errno_location, __h_errno_location in common header
and use that everywhere, __uClibc_main.c is no special.
The rule adopted:
for enabled threads we make in libc the __GI_x() variants strong, x() weak
and (should) provide another strong x() in libpthread.
If threads are disabled, even the __GI_x() variants are weak.
_stdio_init,_stdio_term,_locale_init: make all hidden weak in common header
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
If ADVANCED_REALTIME is disabled, these prototypes are missing and
librt_hidden_proto() fails.
Makefile.in: added a comment, we build mq_timedreceive/mq_timedsend
on NPTL even if ADVANCED_REALTIME is disabled.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Unused option (and is in !ARCH_HAS_MMU guarded section, but depends on ARCH_HAS_MMU)
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
My intention was to get rid of uClibc_mutex.h completely on install, but it
introduces other dependencies. It affects anyway only the installed case.
Note: since NPTL was added, the FILE structure changes between the build
and the installed version. I wanted to fix this by the earlier commit.
Reverting the behaviour to IMHO a wrong one.
While there, remove libc-lock.h and stdio-lock.h on install (even though
glibc has them -wrongly- installed).
Reported-by: Tony Wu <tung7970@gmail.com>
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
The linker can cope with it if it is hidden.
No need for the __GI___uClibc_init.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Do not include unneeded headers.
Move headers to the needed place.
Fix a comment typo.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Even if glibc reuses the syscall (bloat), we can reuse the function.
Make sure, that we use it only if it is provided by the syscall
and the function is really around (UCLIBC_HAS_REALTIME disables it),
else we get into a loop.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
No need for internal _string.h.
For now include string.h, although the correct header would be strings.h.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Rely completely on the configuration options chosen,
in this case on UCLIBC_HAS_FLOATS.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
It shows that the internal functions differ...
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2 earlier (2006 and 2009) commits killed the proper return: -1 for failure, 0 on success
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
do not include pthread.h as the locking macro is provided by stdio.h
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
move __libc_free_aligned prototype up to a common place.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Use one common prototype for consistency.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Other compilers do not expect absolute path after -include,
they want to find the included file by path search (-I).
While there, prefer $(TARGET_ARCH) before common path.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This should come properly from features.h or the build.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Added attribute_noreturn even if it has return at the end of function
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Change CANCELLATION_P macro and adapt usage.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
psiginfo() is disabled for now.
Take care syncing this file, there are 3 bugs in glibc
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Provide __UCLIBC_IO_MUTEX for the installed version
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Redefining stuff is no good for consistency, anyway,
Config.in takes care of this case, it can't happen.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Try to have one and only view of settings, no special handling for stdio code.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Move __stdio_mutex_init,_stdio_openlist_{use,del}_count to _stdio.h.
Make _stdio_validate_FILE hidden.
Make _stdio_openlist_{add,del}_lock hidden for non-NPTL.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Remove all others, only avr32 needs to be kept.
Define __UCLIBC_ARCH_HAS_6_ARGS__ where needed.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
make it clear, that the code is disabled
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
stdio.h provides __STDIO* locking related macros
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
libc-internal.h: this has to provide the reversed case of what string.h defines.
Accomodate for later possibility to compile without _GNU_SOURCE defined
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Since we reuse the functions in the same file, provide static __X()
and use strong_alias to provide the visible functions.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Comments in header were used by vda on a similar commit
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Change uint32_t to u_int32_t and uint8_t to u_int8_t, removing
completely the dependency on stdint.h.
Based on patch by Timo Teraes <timo.teras@iki.fi>. His comment:
This also fixes a major bug that stdlib.h includes stdint.h. Things
might go very wrong because stdint.h has conditional defines and
if stdlib.h is included before #define's for stdint.h we end up
missing things and breaking builds (e.g. openjdk).
Signed-off-by: Timo Teraes <timo.teras@iki.fi>
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
while there add one for future (currently in unused code)
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
sigreturn is an internal linux specific function and is not exported
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
raise.c: only LT new provided gsignal, disable it
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Since the sources differ only minimally, use only scandir.c,readdir[_t].c
and redefine the used functions accordingly.
Use a strong_alias instead of compiling *64.c if __WORDSIZE = 64
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
getdents[64].c: forgot to include dirent.h
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Guard some UCLIBC specific parts.
Add comment about bits/getopt.h.
open_memstream.c: remove __restrict according to SuSv4.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
While there, guard everywhere ioperm and iopl prototypes with __UCLIBC_LINUX_SPECIFIC__
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
before the change (threads enabled)
text data bss dec hex filename
1107 8 408 1523 5f3 libc/misc/utmp/utent.os
152 0 0 152 98 libc/misc/utmp/wtent.os
240 0 0 240 f0 libc/misc/utmp/utxent.os
after the change
text data bss dec hex filename
1072 8 408 1488 5d0 libc/misc/utmp/utent.os
157 0 0 157 9d libc/misc/utmp/wtent.os
200 0 0 200 c8 libc/misc/utmp/utxent.os
The smaller size on utent.os is unexpected, could be due to better inlining/not inlining
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
__X() {...} and strong_alias(__X,X) keeps size the same as X() {...}
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
use fputwc instead of fputc
add hidden fputwc to avoid jump relocation
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Keep only fork (and it's hidden variant) and __libc_fork.
Common __libc_fork prototype in unistd.h.
Get rid of __fork, by changing HIDDEN_JUMPTARGET(__fork) to use fork
I can't see the reason to have fork() in libpthread, but kept it for now
making __fork static in these files.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
LT old and new provided SEM_VALUE_MAX in semaphore.h, NPTL in local_lim.h
Make sure to get the same definitions, independently of the fact, that
semaphore.h is included or not.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
relevant only for gcc < 3.3 or non-gcc compiler
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
add uClibc_posix_opt.h to handle uClibc specific options/features
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
avoid the possibility to use the wrong bits/local_lim.h
uClibc_local_lim.h fix it be correct for LT
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
cleanup bits/setjmp.h and avr32's setjmp.S
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Make sure we use the new jmpbuf-*.h headers
While there, adapt the offset on microblaze and the _JMPBUF_UNWINDS macro on bfin,
it seems, these were changed in the meantime.
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Provide common prototypes for __longjmp, __libc_longjmp, __libc_siglongjmp
_longjmp_unwind, __libc_unwind_longjmp in setjmp.h in preparation for use in LT new
Add __longjmp to h8300 and i960
Make common longjmp.c good for NPTL
Guard _longjmp_unwind use in sh's longjmp.c for NPTL (I think sh could use the common one)
Remove unneeded attribute_noreturn, prototype provides it already
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
clnt_perror.c: change _(x) to N_(x) to allow compile with the change in libintl.h
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
_LIBC is undefined here, no need to specially guard with __UCLIBC__
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
remove gettext workarounds and use libintl.h's _(x)
use strong_alias instead of weak_alias
change how strong_alias and libc_hidden_def are disabled
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
add restrict keyword to wcsstr/wcswcs
remove restrict keyword from wmemcmp
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
use __sigwaitinfo in sigwait.c, since that is already hidden
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
remove unneeded attribute_hidden too, as the prototype provides it
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
solve inconsistency between archs
NPTL should not use __syscall_rt_sigaction at all, for now it needs to be visible
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
remove _LIBC part from rpc headers
include libintl.h to rpc_private.h and use that instead of rpc.h
include libintl.h anywhere else needed replacing _(x) and N_(x)
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
need another guard, else rpc/rpc.h might not be included
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
change the guard from LOCALE to XLOCALE for uselocale/newlocale...
move locale_t to xlocale.h
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
guard internal constant with _LIBC
strftime_l is SuSv4
reorder a bit to ease syncing with glibc
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
glibc does not provide these prototypes
libtirpc provides them in rpc_msg.h, copied comments from there
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
it is reused within the same file, solve it through a static function
Signed-off-by: Peter S. Mazinger <ps.m@gmx.net>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Fixed typo in ucontext header, NFPREG wrongly defined twice instead of
NGREG for general registers and NFPREG for floating point register.
Signed-off-by: Filippo Arcidiacono <filippo.arcidiacono@st.com>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
Currently, the installed libubacktrace linker script entry always uses
HARDWIRED_ABSPATH semantics, even if HARDWIRED_ABSPATH is disabled:
$ grep GROUP $PREFIX/mipsel-linux-uclibc/sys-root/usr/lib/libc.so
GROUP ( libc.so.0 uclibc_nonshared.a AS_NEEDED ( ld-uClibc.so.0 ) )
GROUP ( AS_NEEDED ( /lib/libubacktrace.so.0 ) )
This causes problems when building a non-sysroot toolchain.
Move the AS_NEEDED insertion into libc/Makefile.in, and add a case to
the HARDWIRED_ABSPATH substitution command so libubacktrace is handled
the same way as the other libraries listed in the script.
Signed-off-by: Kevin Cernekee <cernekee@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This patch updates the C6X support to work with latest uClibc code and
uses reworked DSBT support to allow using kernel FDPIC loader.
Signed-off-by: Mark Salter <msalter@redhat.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
eventfd: evntfd assumes to take two arguments instead it
should be one evntfd expects two therefore implement both syscalls with
correct parameters
Thanks Eugene Rudoy for reporting it and also providing the patch
Signed-off-by: Khem Raj <raj.khem@gmail.com>
$a0 is saved across _dl_runtime_resolve(); $v0 is not. Unfortunately,
__syscall_error() uses $v0 for its argument, not $a0 as is the MIPS ABI
standard. This means that if lazy binding was used for __syscall_error(),
the errno value in $v0 could get corrupted.
The problem can be easily seen in testcases where syscalls in librt fail;
when librt tries to call __syscall_error() in libc, the argument gets
lost and errno gets set to a bogus value:
# ./tst-mqueue1 ; echo $?
mq_receive on O_WRONLY mqd_t did not fail with EBADF: Unknown error 2004684208
1
# ./tst-mqueue2 ; echo $?
mq_timedreceive with too small msg_len did not fail with EMSGSIZE: Unknown error 1997360560
1
# ./tst-mqueue4 ; echo $?
mq_timedsend did not fail with ETIMEDOUT: Unknown error 2008747440
1
When _dl_runtime_resolve() was taken out of the equation, the same test
cases passed:
# LD_BIND_NOW=y ./tst-mqueue1 ; echo $?
0
# LD_BIND_NOW=y ./tst-mqueue2 ; echo $?
0
# LD_BIND_NOW=y ./tst-mqueue4 ; echo $?
0
Changing __syscall_error() to look at $a0 instead of $v0 fixed the
problem.
(Note that there is also a "__syscall_error.c" file which presumably
uses the standard C calling conventions, but I do not think it is used
on MIPS.)
Signed-off-by: Kevin Cernekee <cernekee@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Some callers passed the first argument in $v0, while others used $a0.
Change the callers to use $a0 consistently.
Signed-off-by: Kevin Cernekee <cernekee@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
For set-user-ID/set-group-ID ELF binaries, only libraries in the standard
search directories that are also set-user-ID must be loaded.
This patch fixes existing logic according to the above statement.
Furthermore if either library setuid bit isn't set or it cannot be found in
the standard search path, the library won't be preloaded but execution go
ahead with default one.
Signed-off-by: Filippo Arcidiacono <filippo.arcidiacono@st.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
The existing DSBT support relies on the kernel to provide DSBT info
as part of the load maps passed to user space. The problem with this
approach is that the DSBT info is in the dynamic section, so the
kernel must access a userspace mapping of the dynamic section (or
separately read a copy for the kernel) in order to retrieve the
information needed by userspace.
This patch reworks the DSBT support to remove the reliance on DSBT
info coming from the kernel. Instead, ldso reads the info itself from
the dynamic section. One other benefit of this is that it allows the
existing kernel FDPIC loader to also load DSBT binaries.
Signed-off-by: Mark Salter <msalter@redhat.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Linux has provided these functions since the 2.1.x era, so no need to
keep these around. We'd rather find out when things are missing (and
fix that) than fall back to the unsafe hacks.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Some arches got this fix, but many did not. So copy the ifdef logic to
the ones that missed it to fix behavior in linux-2.6+.
URL: https://bugs.busybox.net/show_bug.cgi?id=5258
Reported-by: David Laight <david.laight@aculab.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Peter Mazinger says in [1]:
"I can revert it to the old behaviour in master (the change s://:/: should stay though)."
[1] http://lists.uclibc.org/pipermail/uclibc/2011-March/045049.html
Signed-off-by: Sedat Dilek <sedat.dilek@gmail.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
This reverts commit 5dffed7dd1a413f3965af702fa7ecd79809d1988.
This breaks ldd when HARDWIRED_ABSPATH=n.
[1] http://lists.uclibc.org/pipermail/uclibc/2011-March/045048.html
From: Timo Teräs <timo.teras@iki.fi>
Your commit 5dffed7dd1a413f3965af702fa7ecd79809d1988 removed absolute
path from the interpreter binary name embedded in ELF files.
This makes ldd seg.fault as it seems to assume absolute paths.
gdb sayeth:
(gdb) where
#0 0x00111ae3 in find_elf_interpreter (ehdr=0xb7a74000) at
../utils/ldd.c:556
#1 0x00111f2a in find_dependencies (filename=0x114910
"/home/fabled/aports/main/libc0.9.32/src/uClibc-0.9.32-rc3/lib/librt.so.0.9.32")
at ../utils/ldd.c:676
#2 0x0011228e in main (argc=1, argv=0xbffff6e8) at ../utils/ldd.c:777
(gdb) p tmp
tmp = NULL
(gdb) p interp_dir
$1 = 0x116660 "ld-uClibc.so.0.9.32"
[2] http://lists.uclibc.org/pipermail/uclibc/2011-March/045054.html
From: Mike Frysinger <vapier@gentoo.org>
On Wed, Mar 23, 2011 at 7:18 AM, Peter Mazinger wrote:
>> Your commit 5dffed7dd1a413f3965af702fa7ecd79809d1988 removed absolute
>> path from the interpreter binary name embedded in ELF files.
>
> do you consider my patch wrong?
yes. the absolute path to the ldso must be encoded in binaries. the
point of HARDWIRED_ABSPATH is purely for sysroot/non-sysroot
toolchain, and the interp path isnt involved with that.
are you sure this even works at runtime ? i'm pretty sure the kernel
doesnt do any path lookups on the interp string.
Signed-off-by: Sedat Dilek <sedat.dilek@gmail.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Fix the implementation of mmap based on the mmap2 system call, to
construct pgoffset from offset with an unsigned shift rather than a
signed (off_t) shift. The mmap2 test in the testsuite catches this case
by mmap'ing with a large offset (with the sign bit set). The signed
shift repeats the sign bit making the page shift way out of range. This
is already fixed similarly in mmap64().
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
The latest POSIX spec introduces a "m" character to allocate buffers for
the user when using scanf type functions. This is like the old glibc "a"
flag, but now standardized. With packages starting to use these, we need
to implement it.
for example:
char *s;
sscanf("foo", "%ms", &s);
printf("%s\n", s);
free(s);
This will automatically allocate storage for "s", read in "foo" to it,
and then display it.
I'm not terribly familiar with the stdio layer, so this could be wrong.
But it seems to work for me.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Handle _dl_skip_args in the asm part of the dynamic linker startup,
to skip the ldso arguments, so we can keep this symbol hidden as other archs do.
Signed-off-by: Filippo Arcidiacono <filippo.arcidiacono@st.com>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
This reverts commit 5e0dbdb9e5acdcf7dea29335a5db065f48c58766.
CROSS is only a compat thing and should not be advertised
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Only really old systems (<=linux-2.0) lack a dedicated vfork system call.
The code that is in place to support them is causing issues with newer
arches that also don't provide a vfork system call -- instead, they do
vfork by calling clone in userspace.
If anyone cares about these really old systems, they can submit a patch
to make the system work with them while not breaking newer systems.
Signed-off-by: Mark Salter <msalter@redhat.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
The throw markings on the inline func trigger errors when building with
C++. Since this inline version doesn't throw any exceptions, just drop
the markings completely and be done with it.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Since rpmatch will match ^[Yy] and ^[Nn] regardless of locale, this
lets us create a simple inline version. Newer procps versions use
this function.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Since glibc has split out arch-specific stuff to bits/epoll.h, we can
pull that sanity back into our tree and drop the arch ifdefs.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Older kernels use __NR__llseek for a 64-bit lseek syscall. Newer
kernels use __NR_llseek. This patch fixes the implementation of
lseek64 to use the __NR_llseek syscall if it exists.
Signed-off-by: Mark Salter <msalter@redhat.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
The test for generating a stub for getpgrp was wrong and would
result in duplicate symbols when building without __NR_getpgrp,
but with __NR_getpgid and __NR_getpid.
A closer look at the getpgrp implementation using getpgid showed
that getpid was being called to pass the current pid to getpgid.
This is unnecessary because passing 0 to getpgid returns the
pgid of the current process.
This patch cleans up the getpgrp implementation and the stub test.
Signed-off-by: Mark Salter <msalter@redhat.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Some arches check the size in INLINE_SYSCALL() and barf if it's
too big (i.e. a 64bit value getting truncated to 32bit).
Satisfy error-check on ppc32.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Should be no functional changes; just fixing:
libpthread/linuxthreads.old/manager.c: In function 'pthread_free':
libpthread/linuxthreads.old/manager.c:707:9: warning:
variable 'h_bottom_save' set but not used [-Wunused-but-set-variable]
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
This matches the newer linuxthreads code, and fixes the build warning:
libpthread/linuxthreads.old/manager.c: In function 'pthread_handle_create':
libpthread/linuxthreads.old/manager.c:487:7: warning:
variable 'saved_errno' set but not used [-Wunused-but-set-variable]
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Move spin_count down to the code block where it's used to fix the warning:
libpthread/linuxthreads.old/spinlock.c: In function '__pthread_lock':
libpthread/linuxthreads.old/spinlock.c:70:7: warning:
variable 'spin_count' set but not used [-Wunused-but-set-variable]
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
The normal error() func implicitly appends a newline, so make sure we do
the same. Otherwise getconf outputs an error message like so:
prompt$ getconf -h
Unrecognized variable `-h'prompt$
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
These stubs get used in libc as forwarding funcs, so we don't want them
labeled as hidden. Otherwise, attempting to use them in other libraries
(like libdl) will result in references to symbols that only libpthread
provides. Such as:
/usr/bin/python2.6: can't resolve symbol '__pthread_mutex_lock' in lib '/lib/libdl.so.0'.
This also brings the old code in line with the newer linuxthreads.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
PR 4964
Eric Nadler writes:
It looks to me like the base structure is not fully initialized
with 0 (only a subset of the structure is initialized).
This later results in a free of base->collate.index2weight in
_locale_set_l.
This free tries to free uninitialized memory and segfaults.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Fix incorrect output of BOM when converting charactersets by name. Only
affective when iconv/locale enabled.
Signed-off-by: Ed Wildgoose <lists@wildgooses.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
PR networking/4916
Reserve space for the terminating 0 alias and zero out the
scratch-buffer so the last entry of the alias list is empty.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
__check_pf() function is called from getaddrinfo() and it calls
getifaddrs(), which is too much overhead especially if RSBAC-Net
is enabled. So with this patch __check_pf() is being called only
when AI_ADDRCONFIG hint flag is specified - just when we really
need that check.
Signed-off-by: Alexander Komyagin <komyagin@altell.ru>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Posix says that canonname should use the text representation of an IP address
when a numerical nodename given
See: http://pubs.opengroup.org/onlinepubs/9699919799/functions/getaddrinfo.html
Signed-off-by: Ed Wildgoose <lists@wildgooses.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Some targets include this implicitly, but pull it in explicitly for those
which don't to fix building for them.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
The current code will sometimes cast the offset to an unsigned long.
On 32bit systems, that doesn't make much sense since we always know
that the off64_t type is 64bits. So drop the ifdef logic, cast it
to an unsigned 64bit value, and then do the shift.
If the system is 32bits, and the address is still too large, then
there's really nothing we could do about it anyways, so the extra
ifdef logic wasn't helping.
URL: https://bugs.busybox.net/show_bug.cgi?id=3853
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Otherwise we get:
/tmp/cc14nHjG.s: Assembler messages:
/tmp/cc14nHjG.s: Error: .size expression for _init does not evaluate to a constant
/tmp/cc14nHjG.s: Error: .size expression for _fini does not evaluate to a constant
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
glibc has split off the arch-specific flag definitions (like
SFD_NONBLOCK) into <arch>/bits/signalfd.h, so that there are no longer
multiple redundant copies of the entire header file. We will adopt the
same scheme in uClibc.
Special cases are included for: alpha mips sparc
hppa was omitted because it has not been updated in glibc.
All others use the common definition.
Signed-off-by: Kevin Cernekee <cernekee@gmail.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
glibc has split off the arch-specific flag definitions (like
TFD_NONBLOCK) into <arch>/bits/timerfd.h, so that there are no longer
multiple redundant copies of the entire header file. We will adopt the
same scheme in uClibc.
Special cases are included for: alpha mips sparc
hppa was omitted because it has not been updated in glibc.
All others use the common definition.
Signed-off-by: Kevin Cernekee <cernekee@gmail.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
glibc has split off the arch-specific flag definitions (like
IN_NONBLOCK) into <arch>/bits/inotify.h, so that there are no longer
multiple redundant copies of the entire header file. We will adopt the
same scheme in uClibc.
Special cases are included for: alpha mips sparc
hppa was omitted because it has not been updated in glibc.
All others use the common definition.
Signed-off-by: Kevin Cernekee <cernekee@gmail.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
2009-01-30 Ulrich Drepper <drepper@redhat.com>
[BZ #7040]
* sysdeps/unix/sysv/linux/sys/inotify.h: Second parameter of
inotify_rm_watch should have type int.
This change allows us to copy inotify.h verbatim from glibc into uClibc.
Signed-off-by: Kevin Cernekee <cernekee@gmail.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
The fdpic support has been broken since the prelink support was added,
because it didn't take into account DL_LOADADDR_TYPE could be a different
type of ElfW(Addr).
Signed-off-by: Filippo Arcidiacono <filippo.arcidiacono@st.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
We can't merge pattern and normal rules as newer make barfs:
*** mixed implicit and normal rules. Stop.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
On Blackfin systems, we have underscore prefixes in our ABI. So instead
of _start, we use __start. However, the makefile explicitly sets _start
as the entry point to support toolchains that implicitly use other symbols
as the default entry. It's easy enough to add a hidden alias in the
Blackfin code for free to support this.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
There's no point in reformatting the code style when we're sourcing
most of it from glibc. Further, the changed style breaks with older
compilers, and pollutes the namespace with non-double leading under-
scores.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
No functional changes here. We don't need line continuations as gcc
is smart enough to merge the multiple strings for us.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
ARM always provides its own userspace funcs for ioperm/iopl.
They've never had syscalls. So disable the stubs on ARM to
avoid multiple definition errors.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
This reverts commit 3bcd031f97d61a8f732d865a0f4248aed2d191ab.
The arm code is supposed to provide its own userspace ioperm/iopl
functions. They've never had system calls for these funcs.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Future branch extends stubs with ret_enosys_stub.
Resolve conflicts, and use superset of future and current master.
Adjust posix_fadvise* accordingly.
Upon future merge, pick this master version as the final state.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
This fixes the following link failure:
LD libuClibc-0.9.33.so
libc/libc_so.a(stubs.os): In function `sync_file_range':
stubs.c:(.text+0x0): multiple definition of `ioperm'
libc/libc_so.a(ioperm.os):ioperm.c:(.text+0x260): first defined here
libc/libc_so.a(stubs.os): In function `sync_file_range':
stubs.c:(.text+0x0): multiple definition of `iopl'
libc/libc_so.a(iopl.os):iopl.c:(.text+0x0): first defined here
collect2: ld returned 1 exit status
make[1]: *** [lib/libc.so] Error 1
Signed-off-by: Jason Woodward <jason.woodward@timesys.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
libpthread/nptl/sysdeps/pthread/pt-longjmp.c: In function 'longjmp':
libpthread/nptl/sysdeps/pthread/pt-longjmp.c:27:3: warning: implicit declaration of function '__libc_longjmp' [-Wimplicit-function-declaration]
libpthread/nptl/sysdeps/pthread/pt-longjmp.c:28:1: warning: 'noreturn' function does return [enabled by default]
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
gcc 4.6.x raises useful warnings due to unused-but-set-variable,
This patch fixes the following one by using properly preprocessor
guard
libpthread/nptl/sysdeps/pthread/pthread_cond_timedwait.c: In function '__pthread_cond_timedwait':
libpthread/nptl/sysdeps/pthread/pthread_cond_timedwait.c:103:6: warning: variable 'ret' set but not used [-Wunused-but-set-variable]
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
__libc_once is not available / needed when multithreading support
is not enabled, so authnone_create() calls authnone_create_once()
directly.
When LT.{old,new} is used instead of NPTL, it needs to explicitly
include <bits/libc-lock.h> to get __libc_once to be visible.
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
They expand into the same code, but using the ELF_xxx()/ElfW() macros
makes it much easier to spot similarities between code bases.
Acked-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Acked-by: Carmelo Amoroso <carmelo.amoroso@st.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Some test results:
The longer patch posted at Sun 14:46:24 +0100 made my target system
unbootable. I did not attempt to troubleshoot it, as we are focusing
our efforts on the shorter patch now.
The shorter patch posted at Mon 01:50:27 +0100 is a good start, but it
didn't completely fix the problem for me. I am posting an updated
version with a few changes at the end of this message; the patched
uClibc 0.9.32.1 tree passes both of my test cases.
My changes:
1) Need to break out of the loop on "hard" errors. Otherwise the
library call never returns:
open("/dev/null", O_RDONLY) = 4
dup2(4, 1) = 1
write(1, "hello world\n", 12) = -1 EBADF (Bad file descriptor)
write(1, "hello world\n", 12) = -1 EBADF (Bad file descriptor)
write(1, "hello world\n", 12) = -1 EBADF (Bad file descriptor)
write(1, "hello world\n", 12) = -1 EBADF (Bad file descriptor)
...
2) Move all of the error handling logic back into the "else" clause. In
particular, I believe we do not want to be checking errno unless
__WRITE() had indicated a failure, since the value may be undefined:
if (errno == EINTR
|| errno == EAGAIN
/* do we have other "soft" errors? */
) {
3) Whitespace/indentation consistency.
-- 8< --
From: Denys Vlasenko <vda.linux@googlemail.com>
Currently, uclibc retains buffered data on stdio write errors,
and subsequent fclose and fflush will try to write it out again
(in most cases, in vain).
Which results in something like this:
On Wednesday 26 January 2011 13:21, Baruch Siach wrote:
> Hi busybox list,
>
> I'm running the following command under strace (thanks Rob):
>
> echo 56 > /sys/class/gpio/export
>
> and I see the following output:
>
> write(1, "56\n", 3) = -1 EBUSY (Device or resource busy)
> write(1, "5", 1) = 1
>
> The first EBUSY is OK, since GPIO 56 is already requested. But the second
> write() attempt seems strange, and leads to an unwanted outcome. GPIO 5 gets
> exported.
This patch prevents that.
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
Signed-off-by: Kevin Cernekee <cernekee@gmail.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
A microblaze-specific sys/user.h is needed for (e.g.) strace,
so use this version from glibc.
Signed-off-by: Steve Bennett <steveb@workware.net.au>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
gcc 4.6.x raises useful warnings due to unused-but-set-variable,
This patch fixes the following one by using properly preprocessor
guard
In file included from ldso/libdl/libdl.c:129:0:
./ldso/ldso/sh/elfinterp.c: In function '_dl_do_reloc':
./ldso/ldso/sh/elfinterp.c:163:22: warning: variable 'tls_tpnt' set but not used [-Wunused-but-set-variable]
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
Pulled assignments out of ifs.
do {...} while (1); is a weird method of writing "loop forever",
thus rewrote it.
No code changes: objdump output is the same.
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
Should make adding more knob control simpler as people only need
to modify one variable now (librt_filter_SRC).
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Rename libm-test-ulps-sh4 to libm-test-ulps-sh as the ULPS file
is expected to match pattern libm-test-ulps-$(TARGET_ARCH).
Indeed "sh4" is the TARGET_SUBARCH.
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
Test case to trigger an issue raised by the new symbol scope design,
that was erroneously removing local symbol scope from the global one
too early while dl-closing a shared library.
Based on original test-case by Khem Raj
Signed-off-by: Filippo Arcidiacono <filippo.arcidiacono@st.com>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
Reported-by: Khem Raj <raj.khem@gmail.com>
Since commit a899b909, variable LDFLAGS_GNUHASH has been replaced
by CFLAG_-Wl--hash-style=gnu. This change has not been reported to
the tests buildsys.
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
This is needed for stat'ing loop devices > 255
since otherwise kernel returns EOVERFLOW becasue
it needs st_rdev/st_dev to be larger than 16bits but
in kernel it uses __old_kernel_stat for stat
syscall which has st_rdev/st_dev as unsigned short
Add a testcase
Signed-off-by: Khem Raj <raj.khem@gmail.com>
pthread_getconcurrency / pthread_setconcurrency is OBSOLETE in SUSv4
so disable pthread debug support too
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Do not pollute libraries with its own implementation of __aeabi_unwind_cpp_pr1 code.
Just add it to uclibc_nonshared.a archive and share it.
This fixes build issue with static linking of application due to multiple symbol
definition in libgcc_eh-a and libpthread.a or librt.a
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
The double slash variants are not documented (doc bug?) so better use
the single slash variants like we usually do.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
When built without NPTL support (or for a sparc target), the system()
function doesn't conform to its specification. Namely, it uses signal()
to install/save/restore signal handlers, which may break applications
using custom handlers installed with sigaction(). In addition, it resets
the SIGCHLD handler to SIG_DFL instead of blocking the signal, which may
result in "lost" signals if a custom handler was installed.
Fix system() by replacing uses of signal() with appropriate calls to
sigaction() and sigprocmask().
Signed-off-by: Richard Braun <rbraun@sceen.net>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Fill properly the sym_ref fields when invoking _dl_find_hash to lookup
symbols
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Defer removal of the local scope of a dl-opened library after
all the destructors (of itself and related dependencies) are actually
get unloaded, otherwise any function registered via atexit()
won't be resolved.
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Filippo Arcidiacono <filippo.arcidiacono@st.com>
In some case where CROSS is defined to be empty
we define CROSS_COMPILE ?= CROSS so at this point
it will be defined but will be empty so check
for the same
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Arm has a different mechanism of getting
_Unwind_GetIP. Therefore we provide arch
specific backtrace file.
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Without this the relocations for the current shared object
are not resolved since the scope is not added to map yet
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Avoids warning from -Wunused-function about the alias target that is
only used at link-time.
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
modularisation added too much bloat for no benefit, undo.
$ ../busybox/scripts/bloat-o-meter .lib.05/libcrypt-0.9.33-rc1-git.so lib/libcrypt-0.9.33-rc1-git.so
function old new delta
crypt 130 46 -84
.rodata 2704 - -2704
------------------------------------------------------------------------------
(add/remove: 0/1 grow/shrink: 0/1 up/down: 0/-2788) Total: -2788 bytes
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
The spaces aren't handled correctly (currently) when outputting the
CFLAGS version of the linker flag.
Reported-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Need a little indirection/delayed evaluation to handle variables with
equal signs and commas in them.
Reported-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
We call open() with O_CREAT, so make sure we pass in the mode else
we might get build errors with fortify source.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
As suggested by Bernhard, there is no point in evaluating the compiler's
flag availability when cleaning, so skip things in that case. If there
are variables that change targets based on the flags, then things are
already broken and need fixing independently.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Since we want to clean out all the headers we symlinked into include/,
just use `find` to locate all the symlinks for us. This simplifies it
greatly, and actually fixes bugs where we build for one arch, switch
to another, and then do a clean but the previous arch headers are left
behind.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
The MAKEFLAGS variable contains all the flags without the leading dash.
From the GNU make manual:
This variable is set up automatically by make to contain the flag
letters that make received. Thus, if you do ‘make -ks’ then MAKEFLAGS
gets the value ‘ks’.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
If backtrace support is turned on, build raise() and abort() with
unwind info (fixup 4c9b7f3c21ff21c199e54bfad2fdf3445fa4573d).
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Invalid signals have no handlers so when trying to restore the old
handler to a bad signal a SIGSEGV occurs. This is because the library
tries to store the old handler to an invalid memory area where it things
the bad signal lives.
URL: https://bugs.busybox.net/show_bug.cgi?id=4640
Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Running `make clean` atm takes like 20 seconds because every subdir
re-evaluates all the toolchain flags. Add some helpers to automate
the process of setting up variables to cache the result of tests,
as well as the checking of an already set flag.
Now `make clean` takes like 2 seconds on my system.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
From gcc documentation, we can read:
" ...
-fasynchronous-unwind-tables
Generate unwind table in dwarf2 format, if supported by target
machine. The table is exact at each instruction boundary, so it can be
used for stack unwinding from asynchronous events (such as debugger or
garbage collector)
..."
So it seems better rather than using -funwind-tables (glibc seems to prefer
-fasynchronous-unwind-tables).
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
Acked-by: Mike Frysinger <vapier@gentoo.org>
Build raise and abort function with dwarf2 info by using
-fasynchronous-unwind-tables to make backtrace() working across a
call to abort.
A scenario where it could be useful is within a signal handler that
wants to dump a backtrace when catching some signal (i.e SIGABRT).
Without having abort & raise built with DWARF2 information, it is not
possible for the libgcc stack unwinder to walk through the call stack
out of the abort/raise function.
Impacts in terms of size are really limited
Current
text data bss dec hex filename
164 0 0 164 a4 ./libpthread/nptl/sysdeps/unix/sysv/linux/raise.os
164 0 0 164 a4 ./libpthread/nptl/sysdeps/unix/sysv/linux/raise.oS
220 24 4 248 f8 ./libc/stdlib/abort.os
With (-fasynchronous-unwind-tables)
text data bss dec hex filename
216 0 0 216 d8 ./libpthread/nptl/sysdeps/unix/sysv/linux/raise.os
216 0 0 216 d8 ./libpthread/nptl/sysdeps/unix/sysv/linux/raise.oS
280 24 4 308 134 ./libc/stdlib/abort.os
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
Acked-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
If people use an interp path such as "ld.so", then there is no "/" found,
and we end up dereferencing a NULL pointer. Simplify the logic by having
a relative interp path like that be the same as "" (which the code later
on interprets as $PWD).
While we're here, document some of the nuances of this code.
Reported-by: Ignacy Gawędzki <uclibc@qult.net>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
The setup logic is duplicated, so unify it in a local func.
Mark the variable const while we're doing this, and add missing
ifdef protection to the header that declares it availability.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
