OSDN Git Service

uclinux-h8/uClibc.git
12 years agoldso: merge duplicate dl-syscalls.h code to common header
Mike Frysinger [Wed, 5 Oct 2011 03:26:08 +0000 (23:26 -0400)]
ldso: merge duplicate dl-syscalls.h code to common header

Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
12 years agomicroblaze mmu/elf/shared lib support
Ryan Flux [Tue, 4 Oct 2011 00:50:49 +0000 (10:50 +1000)]
microblaze mmu/elf/shared lib support

microblaze can either be with mmu or without
If with, use elf rather than flat, and support shared libs

Signed-off-by: Ryan Flux <ryan.flux@emsolutions.com.au>
Signed-off-by: Steve Bennett <steveb@workware.net.au>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
12 years agomicroblaze linux has no SDA support
Steve Bennett [Tue, 4 Oct 2011 00:50:48 +0000 (10:50 +1000)]
microblaze linux has no SDA support

The linker does not set up the Small Data Area symbols,
_SDA_BASE_ and _SDA2_BASE_ so don't try to resolve them

Signed-off-by: Ryan Flux <ryan.flux@emsolutions.com.au>
Signed-off-by: Steve Bennett <steveb@workware.net.au>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
12 years agomicroblaze can be either big or little endian
Steve Bennett [Tue, 4 Oct 2011 00:50:47 +0000 (10:50 +1000)]
microblaze can be either big or little endian

Signed-off-by: Steve Bennett <steveb@workware.net.au>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
12 years agomicroblaze has long double support too
Ryan Flux [Tue, 4 Oct 2011 00:50:46 +0000 (10:50 +1000)]
microblaze has long double support too

Signed-off-by: Ryan Flux <ryan.flux@emsolutions.com.au>
Signed-off-by: Steve Bennett <steveb@workware.net.au>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
12 years agoUse the generic bits/select.h for microblaze
Steve Bennett [Tue, 4 Oct 2011 00:50:45 +0000 (10:50 +1000)]
Use the generic bits/select.h for microblaze

This avoids warnings about type-punned pointers

Signed-off-by: Steve Bennett <steveb@workware.net.au>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
12 years agotests: ignore new inet test
Mike Frysinger [Sat, 1 Oct 2011 19:41:06 +0000 (15:41 -0400)]
tests: ignore new inet test

Signed-off-by: Mike Frysinger <vapier@gentoo.org>
12 years agotests: add fallback for mktemp()
Mike Frysinger [Sat, 1 Oct 2011 19:39:46 +0000 (15:39 -0400)]
tests: add fallback for mktemp()

If SuSv3 legacy support is disabled, then mktemp() isn't available, and
we end up getting a lot of build errors.  So add a fallback built on top
of tempnam() since that currently cannot be disabled.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>
12 years agoldso: x86_64: delete unused variable
Mike Frysinger [Sat, 1 Oct 2011 19:12:14 +0000 (15:12 -0400)]
ldso: x86_64: delete unused variable

gcc informed me:
In file included from ldso/ldso/ldso.c:46:0:
ldso/ldso/x86_64/elfinterp.c: In function '_dl_do_lazy_reloc':
ldso/ldso/x86_64/elfinterp.c:294:6: warning:
variable 'symtab_index' set but not used [-Wunused-but-set-variable]

Signed-off-by: Mike Frysinger <vapier@gentoo.org>
12 years agoldso: constify reloc array itself
Mike Frysinger [Sat, 1 Oct 2011 19:10:39 +0000 (15:10 -0400)]
ldso: constify reloc array itself

Signed-off-by: Mike Frysinger <vapier@gentoo.org>
12 years agoDefeat compiler optimization which assumes function addresses are never NULL
Denys Vlasenko [Thu, 15 Sep 2011 14:59:21 +0000 (16:59 +0200)]
Defeat compiler optimization which assumes function addresses are never NULL

From email:
A warning for people who can be hit by the same or similar issue:
gcc 4.1.2 with -march=i486 here with -Os and even with -O2 or -O
is "optimizing away" the check
       if (_stdio_term)
in libc/stdlib/_atexit.c
which results in a "call 0" and a segfault at exit
if you do not happen to link in stdio.

Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com>
12 years agoconfig: tweak help text
Bernhard Reutner-Fischer [Mon, 12 Sep 2011 15:29:21 +0000 (17:29 +0200)]
config: tweak help text

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
12 years agoresolv: simple optimizations and style/readability fixes
Denys Vlasenko [Sat, 10 Sep 2011 01:01:33 +0000 (03:01 +0200)]
resolv: simple optimizations and style/readability fixes

    text           data     bss     dec     hex filename
-   2915              0       0    2915     b63 libc/inet/ns_name_old.o
+   2656              0       0    2656     a60 libc/inet/ns_name.o
-    958              0       0     958     3be libc/inet/ns_parse_old.o
+    898              0       0     898     382 libc/inet/ns_parse.o

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
12 years agoldso: fix build with PRELINK enabled and !TLS
Carmelo Amoroso [Fri, 26 Aug 2011 08:26:17 +0000 (10:26 +0200)]
ldso: fix build with PRELINK enabled and !TLS

Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
12 years agoldso: add missing include
Bernhard Reutner-Fischer [Fri, 26 Aug 2011 07:40:30 +0000 (09:40 +0200)]
ldso: add missing include

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
12 years agoinet: build all res_* and ns_* functions conditionally
Daniel Mack [Tue, 12 Jul 2011 22:30:56 +0000 (00:30 +0200)]
inet: build all res_* and ns_* functions conditionally

This patch changes the build system so that all DNS resolver related
functions (ie, those starting with dn_, ns_ and res_) are only built if
UCLIBC_HAS_RESOLVER_SUPPORT is set. This means that the libc will lack
support for things that existed in the library before unless this new
config option is set. On the other hand, if users really need support
for resolver functions, they now get a more complete set, and the
library is smaller than before it is deselected.

Signed-off-by: Daniel Mack <zonque@gmail.com>
12 years agoinclude/resolv.h: provide __res_state._u even if !__UCLIBC_HAS_IPV6__
Daniel Mack [Tue, 12 Jul 2011 22:30:55 +0000 (00:30 +0200)]
include/resolv.h: provide __res_state._u even if !__UCLIBC_HAS_IPV6__

This struct is used for IPv4 as well, so only make the members
conditional that are purely used by IPv6.

Signed-off-by: Daniel Mack <zonque@gmail.com>
12 years agoinet/resolv: add res_ninit and res_nclose
Daniel Mack [Tue, 12 Jul 2011 22:30:54 +0000 (00:30 +0200)]
inet/resolv: add res_ninit and res_nclose

Some reordering of existing functions was necessary in order to provide
this functionality.

Signed-off-by: Daniel Mack <zonque@gmail.com>
12 years agoinet/resolv: add res_mkquery
Daniel Mack [Tue, 12 Jul 2011 22:30:53 +0000 (00:30 +0200)]
inet/resolv: add res_mkquery

Signed-off-by: Daniel Mack <zonque@gmail.com>
12 years agoinet/resolv: add ns_initparse, ns_parserr, ns_skiprr and ns_msg_getflag
Daniel Mack [Tue, 12 Jul 2011 22:30:52 +0000 (00:30 +0200)]
inet/resolv: add ns_initparse, ns_parserr, ns_skiprr and ns_msg_getflag

There are build 'ifdef L_ns_parse'.

Signed-off-by: Daniel Mack <zonque@gmail.com>
12 years agoinet/resolv: add dn_skipname and ns_name_skip
Daniel Mack [Tue, 12 Jul 2011 22:30:51 +0000 (00:30 +0200)]
inet/resolv: add dn_skipname and ns_name_skip

One uses the other, so add them in one go.

Signed-off-by: Daniel Mack <zonque@gmail.com>
12 years agoinet/resolv: Add dn_comp
Daniel Mack [Tue, 12 Jul 2011 22:30:50 +0000 (00:30 +0200)]
inet/resolv: Add dn_comp

Signed-off-by: Daniel Mack <zonque@gmail.com>
12 years agoinet/resolv: add ns_name_pton, ns_name_pack and ns_name_compress
Daniel Mack [Tue, 12 Jul 2011 22:30:49 +0000 (00:30 +0200)]
inet/resolv: add ns_name_pton, ns_name_pack and ns_name_compress

These are built '#ifdef L_ns_name'

Signed-off-by: Daniel Mack <zonque@gmail.com>
12 years agoinet/resolv: add ns_{put,get}{16,32}
Daniel Mack [Tue, 12 Jul 2011 22:30:48 +0000 (00:30 +0200)]
inet/resolv: add ns_{put,get}{16,32}

This patch introduces a new config directive
'UCLIBC_HAS_RESOLVER_SUPPORT' and adds the new symbols conditionally.

Signed-off-by: Daniel Mack <zonque@gmail.com>
12 years agolibc/inet/resolv.c: comment on #endif lines
Daniel Mack [Tue, 12 Jul 2011 22:30:47 +0000 (00:30 +0200)]
libc/inet/resolv.c: comment on #endif lines

While reading this file, I found it rather confusing to relate the
"endif" lines to their opening counter-part. This patch adds comments to
most of these lines to denote what the actually do.

Signed-off-by: Daniel Mack <zonque@gmail.com>
12 years agoextra/Configs/Config.in: drop "default n" lines
Daniel Mack [Tue, 12 Jul 2011 22:30:46 +0000 (00:30 +0200)]
extra/Configs/Config.in: drop "default n" lines

"default n" is the default default. No need to mention it explicitly.

Signed-off-by: Daniel Mack <zonque@gmail.com>
12 years agoldso/mips: Clean up warnings
Kevin Cernekee [Sun, 24 Jul 2011 08:19:14 +0000 (01:19 -0700)]
ldso/mips: Clean up warnings

Trivial fixes for these warnings:

  CC ldso/libdl/libdl.oS
In file included from ldso/ldso/ldso.c:46:0:
ldso/ldso/mips/elfinterp.c:88:1: warning: no previous prototype for '__dl_runtime_pltresolve'
ldso/ldso/ldso.c: In function '_dl_get_ready_to_run':
ldso/ldso/ldso.c:475:5: warning: assignment makes pointer from integer without a cast
In file included from ldso/ldso/ldso.c:1097:0:
ldso/ldso/dl-elf.c: In function '_dl_load_elf_shared_library':
ldso/ldso/dl-elf.c:811:3: warning: assignment makes pointer from integer without a cast

Signed-off-by: Kevin Cernekee <cernekee@gmail.com>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
12 years agoldso/mips: dlsym() incorrectly matches undefined symbols
Kevin Cernekee [Sun, 24 Jul 2011 08:19:13 +0000 (01:19 -0700)]
ldso/mips: dlsym() incorrectly matches undefined symbols

check_match() relies on checking for (sym->st_value == 0) to see if the
symbol is undefined.  This works reasonably well on most architectures,
such as ARM or i386:

$ readelf -s /lib32/libcap.so.2 | grep -E "\<malloc\>"
    17: 00000000     0 FUNC    GLOBAL DEFAULT  UND malloc@GLIBC_2.0 (2)

However, on MIPS, libbfd puts nonzero data in the st_value field to
facilitate resetting the symbol's GOT entry if the library that defines
the symbol gets unloaded:

$ mipsel-linux-readelf -s libfoo.so | grep -E "\<malloc\>"
    74: 00003140     0 FUNC    GLOBAL DEFAULT  UND malloc

This can cause check_match to report a false positive when examining the
external symbol reference.  Consequently dlsym() will return a bad pointer
to the caller.

Use the special MIPS logic from glibc-ports-2.13 to avoid this situation.

Signed-off-by: Kevin Cernekee <cernekee@gmail.com>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
12 years agoldso/mips: Enable bootstrap relocations
Kevin Cernekee [Sun, 24 Jul 2011 08:19:12 +0000 (01:19 -0700)]
ldso/mips: Enable bootstrap relocations

_dl_reltypes_tab[] is an array of pointers to constant strings:

Contents of section .data:
 20000 01000000 02000000 00000000 00000000  ................
 20010 70e50000 7ce50000 88e50000 94e50000  p...|...........
       ^^^^^^^^ ^^^^^^^^ ^^^^^^^^ ^^^^^^^^

(pointers are LE)

Contents of section .rodata:
 e570 525f4d49 50535f4e 4f4e4500 525f4d49  R_MIPS_NONE.R_MI
 e580 50535f31 36000000 525f4d49 50535f33  PS_16...R_MIPS_3
 e590 32000000 525f4d49 50535f52 454c3332  2...R_MIPS_REL32

These pointers require relocation:

DYNAMIC RELOCATION RECORDS
OFFSET   TYPE              VALUE
00000000 R_MIPS_NONE       *ABS*
0001fffc R_MIPS_REL32      *ABS*
00020010 R_MIPS_REL32      *ABS*
00020014 R_MIPS_REL32      *ABS*
00020018 R_MIPS_REL32      *ABS*

On MIPS, only GOT relocations are currently handled by ldso during
startup.  The net effect is that when running with "LD_DEBUG=reloc",
ldso itself crashes before the program even starts.  This is caused
by _dl_dprintf() dereferencing an unadjusted string pointer such as
0xe570.

This patch enables the missing relocations and allows LD_DEBUG to work
as designed.

Signed-off-by: Kevin Cernekee <cernekee@gmail.com>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
12 years agolibdl: add option for controlling dl_cleanup
Richard Braun [Sat, 9 Jul 2011 12:06:52 +0000 (14:06 +0200)]
libdl: add option for controlling dl_cleanup

When debugging memory leaks with Valgrind, it is required that
dynamically loaded shared objects are not unloaded when a process exits,
otherwise symbols from those files aren't correctly resolved in
allocation traces. This patch adds the LDSO_NO_CLEANUP configuration
option to control this behaviour.

Signed-off-by: Richard Braun <rbraun@sceen.net>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
12 years agoldso: fix build error due to missing variable 'st'
Douglas Mencken [Tue, 12 Jul 2011 07:39:17 +0000 (09:39 +0200)]
ldso: fix build error due to missing variable 'st'

Fix a build error triggered when LDSO_PRELOAD_FILE_SUPPORT is enabled
due to missing definition of 'st' variable.

Signed-off-by: Douglas Mencken <dougmencken@gmail.com>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
12 years agox86_64/elfinterp.c: Protect missed debug _dl_printf with __SUPPORT_LD_DEBUG__
Khem Raj [Thu, 7 Jul 2011 16:50:17 +0000 (09:50 -0700)]
x86_64/elfinterp.c: Protect missed debug _dl_printf with __SUPPORT_LD_DEBUG__

Signed-off-by: Khem Raj <raj.khem@gmail.com>
12 years agomips/signalfd.h: SFD_NONBLOCK for mips is 0200 unlike 04000 commonly
Khem Raj [Thu, 30 Jun 2011 07:32:11 +0000 (00:32 -0700)]
mips/signalfd.h: SFD_NONBLOCK for mips is 0200 unlike 04000 commonly

Exposed by udev 171 which uses signalfd

Signed-off-by: Khem Raj <raj.khem@gmail.com>
12 years agoConfig.in.arch: Free UCLIBC_HAS_FPU setting from depending on UCLIBC_HAS_FLOATS
Khem Raj [Tue, 28 Jun 2011 00:48:18 +0000 (17:48 -0700)]
Config.in.arch: Free UCLIBC_HAS_FPU setting from depending on UCLIBC_HAS_FLOATS

UCLIBC_HAS_FLOATS and UCLIBC_HAS_FPU are denoting two different aspects.
UCLIBC_HAS_FLOATS covers the floating point operations which has nothing
to do if you have FPU or not.

Signed-off-by: Khem Raj <raj.khem@gmail.com>
12 years agoldso: arm: fix build when prelinking is enabled.
Carmelo Amoroso [Mon, 27 Jun 2011 13:26:22 +0000 (15:26 +0200)]
ldso: arm: fix build when prelinking is enabled.

This is a post merge fix to address build issue when PRELINK
support is enabled.

Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
12 years agoMerge remote-tracking branch 'origin/master' into prelink
Carmelo Amoroso [Fri, 1 Jul 2011 07:48:01 +0000 (09:48 +0200)]
Merge remote-tracking branch 'origin/master' into prelink

* origin/master:
  libc: add missing lock initialization in vswprintf
  libubacktrace: fix makefile clean target

Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
12 years agolibc: add missing lock initialization in vswprintf
Maksim Rayskiy [Fri, 17 Jun 2011 01:14:01 +0000 (18:14 -0700)]
libc: add missing lock initialization in vswprintf

Unlike vsnprintf, vswprintf does not properly initialize locking
elements of FILE structure, which in some unfortunate cases can result
in lockups in _vfwprintf_internal.
Interesting, the initialization code was removed in
2a915734a32c5aec9a6a76c13bcb074d30e64171 at the same time as it was added
to vsnprintf.

Signed-off-by: Maksim Rayskiy <mrayskiy@broadcom.com>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
12 years agolibubacktrace: fix makefile clean target
Carmelo Amoroso [Fri, 24 Jun 2011 14:27:36 +0000 (16:27 +0200)]
libubacktrace: fix makefile clean target

Fix makefile lean target removing the old reference to the non
existing libubacktrace_ARCH_OUT variable.

Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
12 years agoMerge remote-tracking branch 'origin/master' into prelink
Carmelo Amoroso [Fri, 24 Jun 2011 14:24:25 +0000 (16:24 +0200)]
Merge remote-tracking branch 'origin/master' into prelink

* origin/master: (61 commits)
  fts: fix warning due to old-style function definition
  ldso_tls: fix compiler warning due to missing cast
  resolv: fix bug in res_init with ipv6 nameservers
  config: Fix passing defconfig args
  buildsys: pt-initfini.s depends on uClibc_config.h
  libdl: search for ELF_RTYPE_CLASS_DLSYM in dlsym()
  resolv: try next server on SERVFAIL
  getaddrinfo: allow numeric service without any hints
  bump version to 0.9.33-git
  nptl/pthread: Correct path for machine specific pt-initfini.c
  ctor/dtor nptl: Fix init and fini function compilation
  Rules.mak: Rearrange appending UCLIBC_EXTRA_CFLAGS to CFLAGS
  ARM: remove EABI/OABI selection
  ARM: detect BX availibility at build time
  ARM: #include <bits/arm_asm.h> where __USE_BX__ is used
  ARM: transform the EABI/OABI choice into a boolean
  ARM: remove sub-arch/variants selection from menuconfig
  ARM: introduce blind options to select & force THUMB mode
  ARM: reorder "Use BX" option
  Fix __libc_epoll_pwait compile failure on x86
  ...

Conflicts:
ldso/libdl/libdl.c

Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
12 years agofts: fix warning due to old-style function definition
Carmelo Amoroso [Fri, 24 Jun 2011 13:05:55 +0000 (15:05 +0200)]
fts: fix warning due to old-style function definition

Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
12 years agoldso_tls: fix compiler warning due to missing cast
Carmelo Amoroso [Mon, 13 Jun 2011 17:58:22 +0000 (19:58 +0200)]
ldso_tls: fix compiler warning due to missing cast

Fix compiler warning (as below) due to missign cast

In file included from ldso/ldso/ldso.c:42:0:
ldso/ldso/dl-tls.c: In function 'init_tls':
ldso/ldso/dl-tls.c:1028:24: error: initialization makes pointer from integer without a cast

Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
12 years agoresolv: fix bug in res_init with ipv6 nameservers
Bernhard Reutner-Fischer [Fri, 17 Jun 2011 18:27:51 +0000 (20:27 +0200)]
resolv: fix bug in res_init with ipv6 nameservers

Thanks to Christian Krause <chkr plauener.de> for finding and fixing
this!

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
12 years agoconfig: Fix passing defconfig args
Bernhard Reutner-Fischer [Tue, 14 Jun 2011 16:48:19 +0000 (18:48 +0200)]
config: Fix passing defconfig args

Don't rely on argument-shuffling but use correct arguments in
the first place.

In bug#3589 Thomas Suckow writes:

The modified kconfig defines -D to be the same function as -d. In
addition -d does not take an argument and Config.in should be passed as
a non-option argument. The existing argument order appears to work on
distributions such as Fedora but chokes on Cygwin (newlib?).

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
12 years agobuildsys: pt-initfini.s depends on uClibc_config.h
Bernhard Reutner-Fischer [Tue, 14 Jun 2011 15:51:58 +0000 (17:51 +0200)]
buildsys: pt-initfini.s depends on uClibc_config.h

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
12 years agolibdl: search for ELF_RTYPE_CLASS_DLSYM in dlsym()
Bernhard Reutner-Fischer [Tue, 14 Jun 2011 14:48:51 +0000 (16:48 +0200)]
libdl: search for ELF_RTYPE_CLASS_DLSYM in dlsym()

On FDPIC platforms, functions are passed by function descriptor, not by
pointers. If you don't specify ELF_RTYPE_CLASS_DLSYM when calling
_dl_find_hash() the return value from dlsym() will be a pointer not a
function descriptor, crashing the program.

The bug was introduced when TLS support was added in
534661b91c98492995274c364c8177c45efc63db

Closes bug#3433

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
12 years agoresolv: try next server on SERVFAIL
Bernhard Reutner-Fischer [Tue, 14 Jun 2011 14:31:00 +0000 (16:31 +0200)]
resolv: try next server on SERVFAIL

in bug 3637 Andrey Kovalev aka pxe.ru writes:
getaddrinfo does NOT add domain to query when receive SERVFAIL

RFC1035 7.2 suggests that
   - If a resolver gets a server error or other bizarre response
     from a name server, it should remove it from SLIST, and may
     wish to schedule an immediate transmission to the next
     candidate server address.

So let's try the next server upon SERVFAIL even if it's not strictly
required.

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
12 years agogetaddrinfo: allow numeric service without any hints
Natanael Copa [Sun, 12 Jun 2011 12:09:04 +0000 (12:09 +0000)]
getaddrinfo: allow numeric service without any hints

This appears to correspond to what glibc does and this fixes an
issue with iptables-1.4.11 with udp and raw port numbers.

(see http://bugzilla.netfilter.org/show_bug.cgi?id=721)

This fixes #3841
https://bugs.busybox.net/show_bug.cgi?id=3841

Signed-off-by: Natanael Copa <natanael.copa@gmail.com>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
12 years agobump version to 0.9.33-git
Bernhard Reutner-Fischer [Tue, 14 Jun 2011 08:27:17 +0000 (10:27 +0200)]
bump version to 0.9.33-git

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
12 years agonptl/pthread: Correct path for machine specific pt-initfini.c
Khem Raj [Tue, 14 Jun 2011 00:13:52 +0000 (17:13 -0700)]
nptl/pthread: Correct path for machine specific pt-initfini.c

Signed-off-by: Khem Raj <raj.khem@gmail.com>
12 years agoctor/dtor nptl: Fix init and fini function compilation
Khem Raj [Mon, 13 Jun 2011 23:30:55 +0000 (16:30 -0700)]
ctor/dtor nptl: Fix init and fini function compilation

We need to define the rules for .S files so it
gets the include paths some architectures like mips
include headers

Some architectures e.g. SH have their own version
of pt-initfini.c so look for that first before resorting
to generic version of pt-initfini.c

Signed-off-by: Khem Raj <raj.khem@gmail.com>
12 years agoRules.mak: Rearrange appending UCLIBC_EXTRA_CFLAGS to CFLAGS
Khem Raj [Sat, 11 Jun 2011 07:15:02 +0000 (00:15 -0700)]
Rules.mak: Rearrange appending UCLIBC_EXTRA_CFLAGS to CFLAGS

UCLIBC_EXTRA_CFLAGS is currently added before the OPTIMIZATION flags
and OPTIMIZATION is chosen to be Os by default. But in OE we pass the optimisation
flags through UCLIBC_EXTRA_CFLAGS but they are not effective since -Os is
specified at last. So we need to change the order of these option flags

Signed-off-by: Khem Raj <raj.khem@gmail.com>
12 years agoARM: remove EABI/OABI selection
Yann E. MORIN [Sun, 9 Jan 2011 00:45:10 +0000 (01:45 +0100)]
ARM: remove EABI/OABI selection

Rely on the compiler to be properly setup for the default ABI.

When installing-headers, there are two cases:
- NPTL: no issue, a cross-compiler is already expected
- LinuxThreads: no issue, EABI/OABI has no impact on installed headers.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@anciens.enib.fr>
Cc: Khem Raj <raj.khem@gmail.com>
Cc: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Cc: Carmelo AMOROSO <carmelo.amoroso@st.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
12 years agoARM: detect BX availibility at build time
Yann E. MORIN [Sun, 9 Jan 2011 00:45:09 +0000 (01:45 +0100)]
ARM: detect BX availibility at build time

The "use BX" option is now a suggestion that BX be used if available.
Use a macro to detect if BX is available at build time. If so, and
the user requested it be used, then use it. Otherwise, error out.

Macro courtesy Khem RAJ:
  http://lists.uclibc.org/pipermail/uclibc/2009-April/042301.html

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@anciens.enib.fr>
Cc: Khem Raj <raj.khem@gmail.com>
Cc: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Cc: Carmelo AMOROSO <carmelo.amoroso@st.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
12 years agoARM: #include <bits/arm_asm.h> where __USE_BX__ is used
Yann E. MORIN [Sun, 9 Jan 2011 00:45:08 +0000 (01:45 +0100)]
ARM: #include <bits/arm_asm.h> where __USE_BX__ is used

The check for __USE_BX__ will be available in bits/arm_asm.h,
so the latter must be included wherever the former is used.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@anciens.enib.fr>
Cc: Khem Raj <raj.khem@gmail.com>
Cc: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Cc: Carmelo AMOROSO <carmelo.amoroso@st.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
12 years agoARM: transform the EABI/OABI choice into a boolean
Yann E. MORIN [Sun, 9 Jan 2011 00:45:07 +0000 (01:45 +0100)]
ARM: transform the EABI/OABI choice into a boolean

The CONFIG_ARM_OABI option is never used.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@anciens.enib.fr>
Cc: Khem Raj <raj.khem@gmail.com>
Cc: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Cc: Carmelo AMOROSO <carmelo.amoroso@st.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
12 years agoARM: remove sub-arch/variants selection from menuconfig
Yann E. MORIN [Sun, 9 Jan 2011 00:45:06 +0000 (01:45 +0100)]
ARM: remove sub-arch/variants selection from menuconfig

Rely on the compiler to be correctly set up to generate
appropriate code for the target variant.

This exposes the Thumb option, as it is no longer auto-selected.
The "Use BX" no longer depends on supported CPU to be selected,
so it now defaults to 'n' as it shall work by default on CPUs
that do not have BX.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@anciens.enib.fr>
Cc: Khem Raj <raj.khem@gmail.com>
Cc: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Cc: Carmelo AMOROSO <carmelo.amoroso@st.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
12 years agoARM: introduce blind options to select & force THUMB mode
Yann E. MORIN [Sun, 9 Jan 2011 00:45:05 +0000 (01:45 +0100)]
ARM: introduce blind options to select & force THUMB mode

Add three new blind options to set use of Thumb mode:
- COMPILE_IN_THUMB_MODE
  - if set, CFLAGS will contain -mthumb
  - if unset, the compiler's default is used
- HAS_THUMB
  - CPUS with Thumb instruction set can select this
  - use of BX depends on this
- FORCE_THUMB
  - CPUs that are Thumb-only must select this
  - this selects: HAS_THUMB, COMPILE_IN_THUMB_MODE and USE_BX

Also, remove leading space in Rules.mak.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@anciens.enib.fr>
Cc: Khem Raj <raj.khem@gmail.com>
Cc: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Cc: Carmelo AMOROSO <carmelo.amoroso@st.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
12 years agoARM: reorder "Use BX" option
Yann E. MORIN [Sun, 9 Jan 2011 00:45:04 +0000 (01:45 +0100)]
ARM: reorder "Use BX" option

"Use BX" is not available on all CPUs, so the option depends on
a correct CPU to be chosen . It is weird that e BX" then appears
_above_ the CPU selection, not below.

Move the "Use BX" after the CPU selection.

Signed-off-by: "Yann E. MORIN" <yann.morin.1998@anciens.enib.fr>
Cc: Khem Raj <raj.khem@gmail.com>
Cc: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
Cc: Carmelo AMOROSO <carmelo.amoroso@st.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
12 years agoFix __libc_epoll_pwait compile failure on x86
Phil Blundell [Sat, 11 Jun 2011 05:10:46 +0000 (01:10 -0400)]
Fix __libc_epoll_pwait compile failure on x86

This prevents "memory input 7 is not directly addressable" errors.

| libc/sysdeps/linux/common/epoll.c: In function '__libc_epoll_pwait':
| libc/sysdeps/linux/common/epoll.c:71:80: error: memory input 7 is not directly addressable
| libc/sysdeps/linux/common/epoll.c:75:86: error: memory input 7 is not directly addressable
| make: *** [libc/sysdeps/linux/common/epoll.o] Error 1
| make: *** Waiting for unfinished jobs....

Signed-off-by: Phil Blundell <pb@pbcl.net>
Signed-off-by: Jason Woodward <jason.woodward@timesys.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
12 years agonptl/Makefile.in: Fix the build break caused by UCLIBC_CTOR_DTOR enabling
Khem Raj [Fri, 10 Jun 2011 20:05:34 +0000 (13:05 -0700)]
nptl/Makefile.in: Fix the build break caused by UCLIBC_CTOR_DTOR enabling

Extra / somehow does not match the target and complains that
$(top_builddir)libpthread/nptl/sysdeps/pthread/crti.o has no rules to
build

Signed-off-by: Khem Raj <raj.khem@gmail.com>
12 years agobuildsys: fix pregen target (!NPTL with LOCALE)
Carmelo Amoroso [Mon, 30 May 2011 12:22:57 +0000 (14:22 +0200)]
buildsys: fix pregen target (!NPTL with LOCALE)

When NPTL is not enabled, and LOCALE support is used, compilation fails
with the following error:

-------------------------------------------------------------------------
  MKDIR include/bits
  GEN include/bits/sysnum.h
make[1]: *** No rule to make target `../..//include/bits/uClibc_ctype.h', needed by `../../extra/locale/gen_wc8bit.c'.  Stop.
make: *** [pregen] Error 2
-------------------------------------------------------------------------

The problem is that "headers" is not listed as prerequisite of pregen.
These have been changed multiple times recently:

o Austin: order headers before pregen-headers
1f5c73c9f8a98b1d2a35cea868a585c97ab0e436
o Khem: pregen: Fix the parallel build problem in pregen target
046035ecde92262d96eff2192ba3cda716f04909
o Khem: pregen-headers: Add new target which depends on headers.
ef18cfe8ebab25f5ef92e81956f50e2dc57df602
o Carmelo: build: Fix infinite loop when no threading support is enabled
ff5e4de7088fe8d34812c2a1e604bf04be713606
In all of these commits headers was actually as prerequisite

In commit 9381d622e2411a35a5fd73a5a573eb269e2dd9c9 [nptl: fix buildsys] by Bernhard, headers was removed,
causing the problem reported above.

Signed-off-by: Filippo Arcidiacono <filippo.arcidiacono@st.com>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
12 years agolibubacktrace: Fix ASNEEDED emission
Bernhard Reutner-Fischer [Tue, 31 May 2011 07:47:12 +0000 (09:47 +0200)]
libubacktrace: Fix ASNEEDED emission

In master the variable is called SHARED_LIBNAME and not
SHARED_MAJORNAME.

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
12 years agoNPTL: Fix CTOR_DTOR
Bernhard Reutner-Fischer [Tue, 31 May 2011 07:45:44 +0000 (09:45 +0200)]
NPTL: Fix CTOR_DTOR

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
12 years agolibc: make common longjmp usable with NPTL
Filippo Arcidiacono [Fri, 27 May 2011 15:36:20 +0000 (17:36 +0200)]
libc: make common longjmp usable with NPTL

Call _longjmp_unwind conditionally under NPTL config option,
making longjmp usable with NPTL.

Signed-off-by: Filippo Arcidiacono <filippo.arcidiacono@st.com>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
12 years agosh: use the common implementation for longjmp
Filippo Arcidiacono [Fri, 27 May 2011 15:17:51 +0000 (17:17 +0200)]
sh: use the common implementation for longjmp

Signed-off-by: Filippo Arcidiacono <filippo.arcidiacono@st.com>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
12 years agosh: fix build in !NPTL case
Filippo Arcidiacono [Fri, 27 May 2011 15:08:36 +0000 (17:08 +0200)]
sh: fix build in !NPTL case

Fix build issue due to missing symbols in !NPTL case:
- vfork
- _longjmp_unwind

Signed-off-by: Filippo Arcidiacono <filippo.arcidiacono@st.com>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
12 years agoCorrect a bug when remapping textrel segments on nommu
Alan Davis [Thu, 26 May 2011 16:15:06 +0000 (18:15 +0200)]
Correct a bug when remapping textrel segments on nommu

From: Alan Davis <adavis@ti.com>

On C6X, when trying to execute a program that has a textrel DSO, it
fails to load. The telltale line in the LD_DEBUG output is:

_dl_get_ready_to_run:779:       file='';  needed by './a.out'

The corresponding DT_NEEDED entry has 'libc.so.0', but here the
filename is empty. This is what is happening in
_dl_elf_shared_library():

First, map all segments according to their permissions. Text gets
initially mapped read-only.

Then, parse the dynamic information. The dynamic table is in RW but
some of the tags may point to RO. For example, DT_NEEDED points to a
string in .dynstr which is in RO. These pointers get computed
according to the loadmap from the original mapping.

Then, in response to a DT_TEXTREL tag, the RO segment gets remapped,
thereby invaliding anything that points to it, in particular certain
dynamic tags such as DT_NEEDED.

The following patch re-parses the dynamic info after the remapping so
as to re-compute any invalid pointers.

Signed-off-by: Alan Davis <adavis@ti.com>
Signed-off-by: Bernd Schmidt <bernds@codesourcery.com>
13 years agounwind.h: Move to libc/sysdeps from nptl sysdeps
Khem Raj [Fri, 13 May 2011 22:31:06 +0000 (15:31 -0700)]
unwind.h: Move to libc/sysdeps from nptl sysdeps

Signed-off-by: Khem Raj <raj.khem@gmail.com>
13 years agoepoll: redo cancellation to match master
Bernhard Reutner-Fischer [Fri, 13 May 2011 10:16:29 +0000 (12:16 +0200)]
epoll: redo cancellation to match master

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agoepoll.c: add cancellation to epoll_[p]wait()
Peter S. Mazinger [Thu, 12 May 2011 23:56:08 +0000 (01:56 +0200)]
epoll.c: add cancellation to epoll_[p]wait()

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>
13 years agolibm: fix building out-of-tree
Bernhard Reutner-Fischer [Thu, 12 May 2011 10:26:41 +0000 (12:26 +0200)]
libm: fix building out-of-tree

Thanks to Thierry Reding for noticing!

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agowordexp: fix mishandled return value of pipe call
Bernhard Reutner-Fischer [Thu, 12 May 2011 10:11:14 +0000 (12:11 +0200)]
wordexp: fix mishandled return value of pipe call

and use pipe2 if available while at it.

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agolinux_specific: handle accept4 and pipe2
Bernhard Reutner-Fischer [Thu, 12 May 2011 08:37:36 +0000 (10:37 +0200)]
linux_specific: handle accept4 and pipe2

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agoNPTL: needs madvise internally
Bernhard Reutner-Fischer [Thu, 12 May 2011 08:30:48 +0000 (10:30 +0200)]
NPTL: needs madvise internally

This fixes 8d09a50a044638fde2ed3e1a1c4d3c7c5a3cce5c since NPTL also uses
madvise internally.

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agosocket: fix stubs
Bernhard Reutner-Fischer [Thu, 12 May 2011 08:07:25 +0000 (10:07 +0200)]
socket: fix stubs

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agocommentary typo fix
Bernhard Reutner-Fischer [Wed, 11 May 2011 20:22:51 +0000 (22:22 +0200)]
commentary typo fix

s/decriptor/descriptor/

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agoepoll: unify epoll.h
Bernhard Reutner-Fischer [Wed, 11 May 2011 20:17:23 +0000 (22:17 +0200)]
epoll: unify epoll.h

13 years agocommentary typo fix
Bernhard Reutner-Fischer [Wed, 11 May 2011 20:03:33 +0000 (22:03 +0200)]
commentary typo fix

use cancellation (with two 'l') uniformly.

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agoImplement epoll_create1 and epoll_pwait system calls.
Thierry Reding [Wed, 11 May 2011 07:28:13 +0000 (09:28 +0200)]
Implement epoll_create1 and epoll_pwait system calls.

Note: TODO: This lacks cancellation support.

Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agobuildsys: document UCLIBC_EXTRA_CPPFLAGS
Bernhard Reutner-Fischer [Wed, 11 May 2011 19:16:23 +0000 (21:16 +0200)]
buildsys: document UCLIBC_EXTRA_CPPFLAGS

Rename EXTRA_CPPFLAGS to UCLIBC_EXTRA_CPPFLAGS and make them override
previous flags (by appending - not prepending - them to the other flags).

13 years agolll: avoid shadow warnings
Bernhard Reutner-Fischer [Wed, 11 May 2011 18:42:18 +0000 (20:42 +0200)]
lll: avoid shadow warnings

use __ret uniformly to avoid shadow warnings as seen on x86_64.

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agox86_64: add TODO for R_X86_64_RELATIVE
Bernhard Reutner-Fischer [Wed, 11 May 2011 18:29:59 +0000 (20:29 +0200)]
x86_64: add TODO for R_X86_64_RELATIVE

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agolibubacktrace: mv unwind.h to include
Bernhard Reutner-Fischer [Wed, 11 May 2011 17:52:16 +0000 (19:52 +0200)]
libubacktrace: mv unwind.h to include

Fixes compilation if NPTL is off.
I hope this doesn't break ARM EABI but cannot check ATM.

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agoRevert "libubacktrace: fix config-symbol handling"
Bernhard Reutner-Fischer [Wed, 11 May 2011 17:16:17 +0000 (19:16 +0200)]
Revert "libubacktrace: fix config-symbol handling"

This reverts commit 7a080cd149c7b25d415d76506510d55b34819fc2.

This symbol is in fact different (as it is set no 'n').

13 years agoadd pipe2()
Bernhard Reutner-Fischer [Wed, 11 May 2011 11:12:14 +0000 (13:12 +0200)]
add pipe2()

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agosocketcalls: add stubs
Bernhard Reutner-Fischer [Wed, 11 May 2011 10:05:20 +0000 (12:05 +0200)]
socketcalls: add stubs

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agoaccept4: Implement cancellation
Bernhard Reutner-Fischer [Wed, 11 May 2011 09:31:45 +0000 (11:31 +0200)]
accept4: Implement cancellation

.. and add proper prototype, move it into it's own obj and other such
cleanups.

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agoImplement accept4 system call.
Thierry Reding [Wed, 11 May 2011 06:40:58 +0000 (08:40 +0200)]
Implement accept4 system call.

Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de>
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agolibubacktrace: fix config-symbol handling
Bernhard Reutner-Fischer [Fri, 6 May 2011 17:07:12 +0000 (19:07 +0200)]
libubacktrace: fix config-symbol handling

Config symbals that are not set are empty, not 'n'.

13 years agobuildsys: Turn on strict aliasing for everyone
Bernhard Reutner-Fischer [Wed, 11 May 2011 07:37:35 +0000 (09:37 +0200)]
buildsys: Turn on strict aliasing for everyone

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agoldso: commentary typo fix
Bernhard Reutner-Fischer [Tue, 10 May 2011 19:22:25 +0000 (21:22 +0200)]
ldso: commentary typo fix

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agogetcwd: Fix memory-leak in error path
Bernhard Reutner-Fischer [Tue, 10 May 2011 18:19:46 +0000 (20:19 +0200)]
getcwd: Fix memory-leak in error path

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years ago__uc_malloc: Fix memory-leak in error path
Bernhard Reutner-Fischer [Tue, 10 May 2011 18:09:36 +0000 (20:09 +0200)]
__uc_malloc: Fix memory-leak in error path

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agolibdl: fix local symbol's address handling in dladdr
Filippo Arcidiacono [Fri, 6 May 2011 14:49:28 +0000 (16:49 +0200)]
libdl: fix local symbol's address handling in dladdr

Fix dladdr to correctly handle local function's address so backtrace_symbols
print only the function address for these function, instead of showing the name
of nearest one.
Indeed the dladdr walk through the hash table to find the nearest symbol, that
doesn't contain local symbols.

Signed-off-by: Filippo Arcidiacono <filippo.arcidiacono@st.com>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
13 years ago*/crtn.S: Remove .size directive for _init and _fini
Bernhard Reutner-Fischer [Tue, 10 May 2011 08:03:58 +0000 (10:03 +0200)]
*/crtn.S: Remove .size directive for _init and _fini

These are split across objects so setting size does not (and never did)
work since the expression cannot be computed at assembly time.

This avoids errors from recent (> 2.21) gas.

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agoposix_fadvise64: fix x86 implementation
Timo Teräs [Wed, 20 Apr 2011 10:59:15 +0000 (13:59 +0300)]
posix_fadvise64: fix x86 implementation

Commit 73d59554144f429b1cf0d4d7fa7de42bdf59ad92 completely broke
the x86 implementation of posix_fadvise64. It moved the first
the assembly code retn instruction gets missing depending on the

Technically the file has two implementaions for posix_fadvise64,
one when __NR_fadvise64_64 is available, and second one if only
__NR_fadvise64 is there. Fix the #ifdef's to be proper for that.

Signed-off-by: Timo Teräs <timo.teras@iki.fi>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
13 years agoi386: fix .size of _init/_fini
Bernhard Reutner-Fischer [Thu, 5 May 2011 07:33:14 +0000 (09:33 +0200)]
i386: fix .size of _init/_fini

binutils-2.21 barf on .size that do not evaluate to const, so use
the section size and not a function that is not visible here.

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agox86_64: fix .size of _init/_fini
Bernhard Reutner-Fischer [Thu, 5 May 2011 07:27:15 +0000 (09:27 +0200)]
x86_64: fix .size of _init/_fini

binutils-2.21 barf on .size that do not evaluate to const, so use the
section size and not a function that is not visible here.

Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
13 years agoRevert "*/crt?.S: remove .size directives"
Bernhard Reutner-Fischer [Thu, 5 May 2011 07:19:16 +0000 (09:19 +0200)]
Revert "*/crt?.S: remove .size directives"

Removing them generally was not a good idea

This reverts commit 233c504cd940d9802226b6a3a092368b86978f5e.

13 years ago*/crt?.S: remove .size directives
Bernhard Reutner-Fischer [Wed, 4 May 2011 08:01:24 +0000 (10:01 +0200)]
*/crt?.S: remove .size directives

sed -i -e '/\.size[[:space:]]/d' $(grep -l "\.size" libc/sysdeps/linux/*/crt*.[sSc])

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