devel

devel@uclibc-ng.org

1 participants
1164 discussions

[RFC PATCH] elf-{fdpic, dsbt}.h: avoid void pointer's subtraction
by Rahul Bedarkar 20 Sep '16

20 Sep '16

elf-fdpic.h or elf-dsbt.h is included by link.h. When C++ program includes <link.h>, we get following build failure. ../usr/include/bits/elf-fdpic.h: In function 'void* __reloc_pointer(void*, const elf32_fdpic_loadmap*)': ../usr/include/bits/elf-fdpic.h:95: error: invalid use of 'void' void pointer addition and subtraction is not allowed in C++ as it has undetermined size, however in C with language extension it is possible because sizeof void is treated as one byte. Instead of performing subtraction on void pointers, typecast it to char* first. This build failure is detected by Buildroot autobuilder. http://autobuild.buildroot.net/results/a10/a10ed48e6eb8411a3d8372f57c05fd11… Signed-off-by: Rahul Bedarkar <rahul.bedarkar(a)imgtec.com> --- Note: Compile tested only --- libc/sysdeps/linux/bfin/bits/elf-fdpic.h | 2 +- libc/sysdeps/linux/c6x/bits/elf-dsbt.h | 2 +- libc/sysdeps/linux/frv/bits/elf-fdpic.h | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/libc/sysdeps/linux/bfin/bits/elf-fdpic.h b/libc/sysdeps/linux/bfin/bits/elf-fdpic.h index dddf82c..9fcf93f 100644 --- a/libc/sysdeps/linux/bfin/bits/elf-fdpic.h +++ b/libc/sysdeps/linux/bfin/bits/elf-fdpic.h @@ -91,7 +91,7 @@ __reloc_pointer (void *p, /* This should be computed as part of the pointer comparison above, but we want to use the carry in the comparison, so we can't convert it to an integer type beforehand. */ - unsigned long offset = p - (void*)map->segs[c].p_vaddr; + unsigned long offset = (char*)p - (char*)map->segs[c].p_vaddr; /* We only check for one-past-the-end for the last segment, assumed to be the data segment, because other cases are ambiguous in the absence of padding between segments, and diff --git a/libc/sysdeps/linux/c6x/bits/elf-dsbt.h b/libc/sysdeps/linux/c6x/bits/elf-dsbt.h index a4e3e7d..0e82ec7 100644 --- a/libc/sysdeps/linux/c6x/bits/elf-dsbt.h +++ b/libc/sysdeps/linux/c6x/bits/elf-dsbt.h @@ -94,7 +94,7 @@ __reloc_pointer (void *p, for (c = 0; c < map->nsegs; c++) { - unsigned long offset = p - (void*)map->segs[c].p_vaddr; + unsigned long offset = (char*)p - (char*)map->segs[c].p_vaddr; /* We only check for one-past-the-end for the second segment, assumed to be the data segment, because other cases are ambiguous in the absence of padding between segments, and diff --git a/libc/sysdeps/linux/frv/bits/elf-fdpic.h b/libc/sysdeps/linux/frv/bits/elf-fdpic.h index dddf82c..9fcf93f 100644 --- a/libc/sysdeps/linux/frv/bits/elf-fdpic.h +++ b/libc/sysdeps/linux/frv/bits/elf-fdpic.h @@ -91,7 +91,7 @@ __reloc_pointer (void *p, /* This should be computed as part of the pointer comparison above, but we want to use the carry in the comparison, so we can't convert it to an integer type beforehand. */ - unsigned long offset = p - (void*)map->segs[c].p_vaddr; + unsigned long offset = (char*)p - (char*)map->segs[c].p_vaddr; /* We only check for one-past-the-end for the last segment, assumed to be the data segment, because other cases are ambiguous in the absence of padding between segments, and -- 2.6.2

2 1

uClibc-ng - small C library for embedded systems branch master updated. v1.0.17-12-gf1d7505
by wbx＠helium.openadk.org 20 Sep '16

20 Sep '16

1 0

RFC PATCH
by Waldemar Brodkorb 19 Sep '16

19 Sep '16

Hi, I would like to suggest following patch to have a single libc library. This would solve a lot of problems in buildroot and other buildsystems. I would love to make a release at the end of this month. What do you think about it? For buildroot people, here is a patch for testing: http://debug.openadk.org/buildroot/0001-experimental-uClibc-ng-patches.patch This will solve a lot of old autobuild failures, f.e. the libarchive static stuff. I worked for many hours on this, so don't be too harsh with your feedback ;) Thanks in advance! Waldemar P.s.: I know it is a big patch, but I see no good chances to split it in pieces.

2 1

[PATCH] ARC: Fix max ULP for cosine test
by Eugeniy Paltsev 10 Sep '16

10 Sep '16

Max ULP for "cos (M_PI_6l * 2.0) == 0.5" in "test-double" sub-test is set to "0" for ARC. It causes "test-double" test failure. I realized that this sub-test on ARM gets completely same results. But it doesn't fails on ARM because max ULP is set to "1" for ARM. I am wondering, if it was done for any special reason? Otherwise, set max ULP for "cos (M_PI_6l * 2.0) == 0.5" test to "1" instead of "0" for ARC. Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev(a)synopsys.com> --- test/math/libm-test-ulps-arc | 1 + 1 file changed, 1 insertion(+) diff --git a/test/math/libm-test-ulps-arc b/test/math/libm-test-ulps-arc index 7139447..46b2ac5 100644 --- a/test/math/libm-test-ulps-arc +++ b/test/math/libm-test-ulps-arc @@ -2,6 +2,7 @@ # cos Test "cos (M_PI_6l * 2.0) == 0.5": +double: 1 idouble: 1 Test "cos (M_PI_6l * 4.0) == -0.5": double: 2 -- 2.5.5

2 1

uClibc-ng - small C library for embedded systems branch master updated. v1.0.17-9-g8ae2b54
by wbx＠helium.openadk.org 10 Sep '16

10 Sep '16

This is an automated email from the git hooks/post-receive script. It was generated because a ref change was pushed to the repository containing the project "uClibc-ng - small C library for embedded systems". The branch, master has been updated via 8ae2b54ee70891ddb720bf00f8913c3716c1f67f (commit) via f9d8ce5ae0ed1e1a639ddc9b2b64726eab569c66 (commit) from 939014ddce98bd07c774815a7f0cf379c8bde38d (commit) Those revisions listed above that are new to this repository have not appeared on any other notification email; so we list those revisions in full, below. - Log ----------------------------------------------------------------- commit 8ae2b54ee70891ddb720bf00f8913c3716c1f67f Author: Waldemar Brodkorb <wbx(a)uclibc-ng.org> Date: Sat Aug 27 17:45:19 2016 +0200 remove DOMULTI support A single test with targeting ARM showed that this feature seems bit rotted. Remove DOMULTI and simplify Makefiles. commit f9d8ce5ae0ed1e1a639ddc9b2b64726eab569c66 Author: Eugeniy Paltsev <Eugeniy.Paltsev(a)synopsys.com> Date: Tue Aug 23 12:41:18 2016 +0300 ARC: Fix max ULP for cosine test Max ULP for "cos (M_PI_6l * 2.0) == 0.5" in "test-double" sub-test is set to "0" for ARC. It causes "test-double" test failure. I realized that this sub-test on ARM gets completely same results. But it doesn't fails on ARM because max ULP is set to "1" for ARM. I am wondering, if it was done for any special reason? Otherwise, set max ULP for "cos (M_PI_6l * 2.0) == 0.5" test to "1" instead of "0" for ARC. Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev(a)synopsys.com> ----------------------------------------------------------------------- Summary of changes: Makerules | 12 ------------ Rules.mak | 16 ---------------- extra/Configs/Config.in | 12 ------------ extra/Configs/defconfigs/or1k/defconfig | 1 - extra/scripts/randconfig.sh | 1 - libc/Makefile.in | 14 -------------- libc/sysdeps/linux/i386/bits/syscalls.h | 12 ------------ libcrypt/Makefile.in | 7 ------- libintl/Makefile.in | 11 ----------- libm/Makefile.in | 15 --------------- libuargp/Makefile.in | 5 ----- libubacktrace/Makefile.in | 7 ------- libutil/Makefile.in | 7 ------- test/math/libm-test-ulps-arc | 1 + 14 files changed, 1 insertion(+), 120 deletions(-) diff --git a/Makerules b/Makerules index d8cbfe4..d35e40d 100644 --- a/Makerules +++ b/Makerules @@ -365,18 +365,6 @@ define link-flat.so $(top_builddir)/lib/crtn.o endef -define linkm.so - $(Q)$(RM) $@ $@.$(2) $(dir $@)$(1) - $(do_strip) - @$(disp_ld) - $(Q)$(CC) $(LDFLAGS-$(notdir $@)) -Wl,-soname=$(notdir $(a)).$(2) \ - $(CFLAG_-nostdlib) -o $(dir $@)$(1) $(START_FILE-$(notdir $@)) \ - $^ \ - $(LIBS-$(notdir $@)) $(LIBGCC) $(END_FILE-$(notdir $@)) - $(Q)$(LN) -sf $(1) $@.$(2) - $(Q)$(LN) -sf $(1) $@ -endef - CFLAGS-.os+=$(PICFLAG) CFLAGS-.oS+=$(PICFLAG) -DSHARED diff --git a/Rules.mak b/Rules.mak index cd343f5..f226dca 100644 --- a/Rules.mak +++ b/Rules.mak @@ -677,22 +677,6 @@ else STRIPTOOL := true -Stripping_disabled endif -ifeq ($(DOMULTI),y) -# we try to compile all sources at once into an object (IMA), but -# gcc-3.3.x does not support it -# gcc-3.4.x supports it, but does not need and support --combine. though fails on many sources -# gcc-4.0.x supports it, supports the --combine flag, but does not need it -# gcc-4.1(200506xx) supports it, but needs the --combine flag, else libs are useless -ifeq ($(GCC_MAJOR_VER),3) -DOMULTI:=n -else -$(eval $(call check-gcc-var,--combine)) -CFLAGS += $(CFLAG_--combine) -endif -else -DOMULTI:=n -endif - ifneq ($(strip $(UCLIBC_EXTRA_CFLAGS)),"") CFLAGS += $(call qstrip,$(UCLIBC_EXTRA_CFLAGS)) endif diff --git a/extra/Configs/Config.in b/extra/Configs/Config.in index 1d97e78..12dc0df 100644 --- a/extra/Configs/Config.in +++ b/extra/Configs/Config.in @@ -2416,16 +2416,4 @@ config EXTRA_WARNINGS help If you wish to build with extra warnings enabled, say Y here. -config DOMULTI - bool "Compile all sources at once into an object" - help - Set this to compile all sources at once into an object (IMA). - This mode of compilation uses alot of memory but may produce - smaller binaries. - - Note that you need a very recent GCC for this to work, like - gcc >= 4.3 plus eventually some patches. - - If unsure, keep the default of N. - endmenu diff --git a/extra/Configs/defconfigs/or1k/defconfig b/extra/Configs/defconfigs/or1k/defconfig index a4658da..b1f890a 100644 --- a/extra/Configs/defconfigs/or1k/defconfig +++ b/extra/Configs/defconfigs/or1k/defconfig @@ -237,4 +237,3 @@ UCLIBC_EXTRA_CFLAGS="" # UCLIBC_HAS_BACKTRACE is not set WARNINGS="-Wall" # EXTRA_WARNINGS is not set -# DOMULTI is not set diff --git a/extra/scripts/randconfig.sh b/extra/scripts/randconfig.sh index d83731b..e2124c7 100755 --- a/extra/scripts/randconfig.sh +++ b/extra/scripts/randconfig.sh @@ -10,7 +10,6 @@ test "x$AWK" = "x" && AWK=awk test "x$ARCH" = "x" && ARCH=`uname -m` KCONFIG_ALLCONFIG=.config.allconfig (echo TARGET_$ARCH=y - echo '# DOMULTI is not set' ) > $KCONFIG_ALLCONFIG export KCONFIG_ALLCONFIG diff --git a/libc/Makefile.in b/libc/Makefile.in index 8ed8a75..05cec18 100644 --- a/libc/Makefile.in +++ b/libc/Makefile.in @@ -42,14 +42,7 @@ else libc-a-y = $(libc-y) $(libc-static-y) endif -ifeq ($(DOMULTI),n) libc-so-y = $(libc-y:.o=.os) $(libc-shared-y) -else -all_sources = $(libc-y:.o=.c) -all_sources += $(libc-shared-y:.oS=.c) -libc-multi-y = $(filter-out $(libc-nomulti-y:.o=.c),$(all_sources)) -endif - lib-a-y += $(top_builddir)lib/libc.a lib-gdb-y += $(top_builddir)lib/libc.gdb lib-so-y += $(libc.depend) @@ -57,17 +50,10 @@ objclean-y += CLEAN_libc OUTPUT_FORMAT = $(CC) $(CFLAGS) -Wl,--verbose 2>&1 | $(SED) -n '/OUTPUT_FORMAT/,/)/p' -ifeq ($(DOMULTI),n) $(libc.depend): $(libc_OUT)/libc_so.a $(LIBS-libc.so) $(call link.so,$(libc_FULL_NAME),$(ABI_VERSION)) # link for backward compatibility $(call link.so,$(libc_FULL_NAME),0) -else -$(libc.depend): $(libc_OUT)/libc.oS $(libc-nomulti-y:.o=.oS) | $(LIBS-libc.so) - $(call linkm.so,$(libc_FULL_NAME),$(ABI_VERSION)) - # link for backward compatibility - $(call linkm.so,$(libc_FULL_NAME),0) -endif $(Q)$(RM) $@ $(Q)cat $(top_srcdir)extra/scripts/format.lds > $@.tmp $(Q)$(OUTPUT_FORMAT) >> $@.tmp diff --git a/libc/sysdeps/linux/i386/bits/syscalls.h b/libc/sysdeps/linux/i386/bits/syscalls.h index 566b5ac..5e92093 100644 --- a/libc/sysdeps/linux/i386/bits/syscalls.h +++ b/libc/sysdeps/linux/i386/bits/syscalls.h @@ -44,17 +44,9 @@ /* We need some help from the assembler to generate optimal code. * We define some macros here which later will be used. */ -/* gcc>=4.6 with LTO need the same guards as IMA (a.k.a --combine) did. - * See gcc.gnu.org/PR47577 */ /* FIXME: drop these b* macros! */ __asm__ ( -#if defined __DOMULTI__ || __GNUC_PREREQ (4, 6) - /* Protect against asm macro redefinition (happens in __DOMULTI__ mode). - * Unfortunately, it ends up visible in .o files. */ - ".ifndef _BITS_SYSCALLS_ASM\n\t" - ".set _BITS_SYSCALLS_ASM,1\n\t" -#endif ".L__X'%ebx = 1\n\t" ".L__X'%ecx = 2\n\t" ".L__X'%edx = 2\n\t" @@ -95,10 +87,6 @@ __asm__ ( ".endif\n\t" ".endif\n\t" ".endm\n\t" - -#if defined __DOMULTI__ || __GNUC_PREREQ (4, 6) - ".endif\n\t" /* _BITS_SYSCALLS_ASM */ -#endif ); #define LOADARGS_0 diff --git a/libcrypt/Makefile.in b/libcrypt/Makefile.in index 94feacb..eb2d452 100644 --- a/libcrypt/Makefile.in +++ b/libcrypt/Makefile.in @@ -39,7 +39,6 @@ lib-a-$(UCLIBC_HAS_CRYPT) += $(top_builddir)lib/libcrypt.a lib-so-$(UCLIBC_HAS_CRYPT) += $(top_builddir)lib/libcrypt.so objclean-y += CLEAN_libcrypt -ifeq ($(DOMULTI),n) ifeq ($(DOPIC),y) $(top_builddir)lib/libcrypt.so: $(top_builddir)lib/libcrypt.a $(libc.depend) else @@ -48,12 +47,6 @@ endif $(call link.so,$(libcrypt_FULL_NAME),$(ABI_VERSION)) # link for backward compatibility $(call link.so,$(libcrypt_FULL_NAME),0) -else -$(top_builddir)lib/libcrypt.so: $(libcrypt_OUT)/libcrypt.oS | $(libc.depend) - $(call linkm.so,$(libcrypt_FULL_NAME),$(ABI_VERSION)) - # link for backward compatibility - $(call linkm.so,$(libcrypt_FULL_NAME),0) -endif $(libcrypt_OUT)/libcrypt_so.a: $(libcrypt-so-y) $(Q)$(RM) $@ diff --git a/libintl/Makefile.in b/libintl/Makefile.in index 39b39a5..6c3e79b 100644 --- a/libintl/Makefile.in +++ b/libintl/Makefile.in @@ -26,10 +26,6 @@ libintl_OUT := $(top_builddir)libintl libintl_MSRC := $(libintl_DIR)/intl.c libintl_MOBJ := $(patsubst %.o,$(libintl_OUT)/%.o,$(MOBJ)) -ifneq ($(DOMULTI),n) -CFLAGS-libintl += $(patsubst %,-DL_%,$(subst .o,,$(notdir $(libintl_MOBJ)))) -endif - ifeq ($(DOPIC),y) libintl-a-y := $(libintl_MOBJ:.o=.os) else @@ -41,7 +37,6 @@ lib-a-$(UCLIBC_HAS_GETTEXT_AWARENESS) += $(top_builddir)lib/libintl.a lib-so-$(UCLIBC_HAS_GETTEXT_AWARENESS) += $(top_builddir)lib/libintl.so objclean-y += CLEAN_libintl -ifeq ($(DOMULTI),n) ifeq ($(DOPIC),y) $(top_builddir)lib/libintl.so: $(top_builddir)lib/libintl.a $(libc.depend) else @@ -50,12 +45,6 @@ endif $(call link.so,$(libintl_FULL_NAME),$(ABI_VERSION)) # link for backward compatibility $(call link.so,$(libintl_FULL_NAME),0) -else -$(top_builddir)lib/libintl.so: $(libintl_OUT)/libintl.oS | $(libc.depend) - $(call linkm.so,$(libintl_FULL_NAME),$(ABI_VERSION)) - # link for backward compatibility - $(call linkm.so,$(libintl_FULL_NAME),0) -endif $(libintl_OUT)/libintl_so.a: $(libintl-so-y) $(Q)$(RM) $@ diff --git a/libm/Makefile.in b/libm/Makefile.in index 2880512..0d6ca05 100644 --- a/libm/Makefile.in +++ b/libm/Makefile.in @@ -277,14 +277,6 @@ endif libm_MSRC_FL := $(libm_DIR)/$(FL_MSRC) libm_MOBJ_FL := $(patsubst %.o,$(libm_OUT)/%.o,$(FL_MOBJ)) - -ifneq ($(DOMULTI),n) -CFLAGS-libm += $(patsubst %,-DL_%,$(subst .o,,$(notdir $(libm_MOBJ_FL)))) -ifeq ($(strip $(UCLIBC_HAS_LONG_DOUBLE_MATH)),y) -CFLAGS-libm += $(patsubst %,-DL_%,$(subst .o,,$(notdir $(libm_MOBJ_LD)))) -endif -endif - libm_OBJS := $(libm_OBJ) $(libm_MOBJ_FL) $(libm_MOBJ_LD) ifeq ($(DOPIC),y) @@ -298,7 +290,6 @@ lib-a-$(UCLIBC_HAS_FLOATS) += $(top_builddir)lib/libm.a lib-so-$(UCLIBC_HAS_FLOATS) += $(top_builddir)lib/libm.so objclean-y += CLEAN_libm -ifeq ($(DOMULTI),n) ifeq ($(DOPIC),y) $(top_builddir)lib/libm.so: $(top_builddir)lib/libm.a $(libc.depend) else @@ -307,12 +298,6 @@ endif $(call link.so,$(libm_FULL_NAME),$(ABI_VERSION)) # link for backward compatibility $(call link.so,$(libm_FULL_NAME),0) -else -$(top_builddir)lib/libm.so: $(libm_OUT)/libm.oS | $(libc.depend) - $(call linkm.so,$(libm_FULL_NAME),$(ABI_VERSION)) - # link for backward compatibility - $(call linkm.so,$(libm_FULL_NAME),0) -endif $(libm_OUT)/libm_so.a: $(libm-so-y) $(Q)$(RM) $@ diff --git a/libuargp/Makefile.in b/libuargp/Makefile.in index 1498abb..f1ccfa6 100644 --- a/libuargp/Makefile.in +++ b/libuargp/Makefile.in @@ -44,17 +44,12 @@ lib-so-$(UCLIBC_HAS_ARGP) += $(top_builddir)lib/libuargp.so objclean-y += CLEAN_libuargp -ifeq ($(DOMULTI),n) ifeq ($(DOPIC),y) $(top_builddir)lib/libuargp.so: $(top_builddir)lib/libuargp.a $(libc.depend) else $(top_builddir)lib/libuargp.so: $(libuargp_OUT)/libuargp_so.a $(libc.depend) endif $(call link.so,$(libuargp_FULL_NAME),$(MAJOR_VERSION)) -else -$(top_builddir)lib/libuargp.so: $(libuargp_OUT)/libuargp.oS $(libc.depend) - $(call linkm.so,$(libuargp_FULL_NAME),$(MAJOR_VERSION)) -endif $(libuargp_OUT)/libuargp_so.a: $(libuargp-so-y) $(Q)$(RM) $@ diff --git a/libubacktrace/Makefile.in b/libubacktrace/Makefile.in index 587eaf6..fdd73fa 100644 --- a/libubacktrace/Makefile.in +++ b/libubacktrace/Makefile.in @@ -59,17 +59,10 @@ lib-so-$(UCLIBC_HAS_BACKTRACE) += $(top_builddir)lib/libubacktrace.so objclean-y += CLEAN_libubacktrace -ifeq ($(DOMULTI),n) $(top_builddir)lib/libubacktrace.so: $(libubacktrace_OUT)/libubacktrace_so.a $(libdl.depend) $(call link.so,$(libubacktrace_FULL_NAME),$(ABI_VERSION)) # link for backward compatibility $(call link.so,$(libubacktrace_FULL_NAME),0) -else -$(top_builddir)lib/libubacktrace.so: $(libubacktrace_OUT)/libubacktrace.oS | $(libdl.depend) - $(call linkm.so,$(libubacktrace_FULL_NAME),$(ABI_VERSION)) - # link for backward compatibility - $(call linkm.so,$(libubacktrace_FULL_NAME),0) -endif $(libubacktrace_OUT)/libubacktrace_so.a: $(libubacktrace-so-y) $(Q)$(RM) $@ diff --git a/libutil/Makefile.in b/libutil/Makefile.in index 7759482..d36995f 100644 --- a/libutil/Makefile.in +++ b/libutil/Makefile.in @@ -48,7 +48,6 @@ lib-a-$(UCLIBC_HAS_LIBUTIL) += $(top_builddir)lib/libutil.a lib-so-$(UCLIBC_HAS_LIBUTIL) += $(top_builddir)lib/libutil.so objclean-y += CLEAN_libutil -ifeq ($(DOMULTI),n) ifeq ($(DOPIC),y) $(top_builddir)lib/libutil.so: $(top_builddir)lib/libutil.a $(libc.depend) else @@ -57,12 +56,6 @@ endif $(call link.so,$(libutil_FULL_NAME),$(ABI_VERSION)) # link for backward compatibility $(call link.so,$(libutil_FULL_NAME),0) -else -$(top_builddir)lib/libutil.so: $(libutil_OUT)/libutil.oS | $(libc.depend) - $(call linkm.so,$(libutil_FULL_NAME),$(ABI_VERSION)) - # link for backward compatibility - $(call linkm.so,$(libutil_FULL_NAME),0) -endif $(libutil_OUT)/libutil_so.a: $(libutil-so-y) $(Q)$(RM) $@ diff --git a/test/math/libm-test-ulps-arc b/test/math/libm-test-ulps-arc index 7139447..46b2ac5 100644 --- a/test/math/libm-test-ulps-arc +++ b/test/math/libm-test-ulps-arc @@ -2,6 +2,7 @@ # cos Test "cos (M_PI_6l * 2.0) == 0.5": +double: 1 idouble: 1 Test "cos (M_PI_6l * 4.0) == -0.5": double: 2 hooks/post-receive -- uClibc-ng - small C library for embedded systems

1 0

uClibc-ng - small C library for embedded systems branch master updated. v1.0.17-7-g939014d
by wbx＠helium.openadk.org 27 Aug '16

27 Aug '16

1 0

pthread_mutex_*() functions, dynamic vs. static linking
by Thomas Petazzoni 19 Aug '16

19 Aug '16

Hello, Today I was investigating a link issue that occurs in Buildroot with the libarchive package. It turns out that the code of this package uses pthread_mutex_lock/pthread_mutex_unlock in some places to be thread-safe, but does *not* link with the libpthread library. I was originally surprised that this could even work, but I discovered that the libc intentionally provides a stub implementation of pthread_mutex_lock/unlock. The idea behind this is that a library can use pthread_mutex_lock/unlock without linking to libpthread. This way, if the application using the library is single-threaded and doesn't link with libpthread, pthread_mutex_lock/unlock are no-ops. On the other hand, if the application using the library is multi-threaded, it will link with libpthread, and therefore the libpthread versions of pthread_mutex_lock/unlock will be used. This all seems good, until you get to static linking. Indeed with uClibc-ng, the following program: #include <pthread.h> int main(void) { pthread_mutex_t lock; pthread_mutex_lock(&lock); return 0; } will link perfectly fine with dynamic linking: $ ./host/usr/bin/arm-linux-gcc -o foo foo.c $ ./host/usr/bin/arm-linux-readelf -d foo | grep NEEDED 0x00000001 (NEEDED) Shared library: [libc.so.0] but will fail to build with static linking: $ ./host/usr/bin/arm-linux-gcc -o foo foo.c -static /tmp/ccda8vkc.o: In function `main': foo.c:(.text+0x14): undefined reference to `pthread_mutex_lock' collect2: error: ld returned 1 exit status And this explains the build failures like http://autobuild.buildroot.net/results/01b/01b7088a06e7310c8773e78e8be4f6e8… that we are seeing in Buildroot. It is worth mentioning that using the musl C library, this test case works fine in both dynamic linking and static linking cases (and the libarchive error also does not occur). Thomas -- Thomas Petazzoni, CTO, Free Electrons Embedded Linux and Kernel engineering http://free-electrons.com

3 12

[uClibc] ARC: Support syscall ABI v4
by Vineet Gupta 18 Aug '16

18 Aug '16

The syscall ABI includes the gcc functional calling ABI since a syscall implies userland caller and kernel callee. The current gcc ABI (v3) for ARCv2 ISA required 64-bit data be passed in even-odd register pairs, (potentially punching reg holes when passing such values as args). This was partly driven by the fact that the double-word LDD/STD instructions in ARCv2 expect the register alignment and thus gcc forcing this avoids extra MOV at the cost of a few unused register (which we have plenty anyways). This however was rejected as part of upstreaming gcc port to HS. So the new ABI v4 doesn't enforce the even-odd reg restriction. Do note that for ARCompact ISA builds v3 and v4 are practically the same in terms of gcc code generation. This change is dormant for now (gcc 4.8.x based tools) and will only kick in with switch to gcc 6.x based tools. Signed-off-by: Vineet Gupta <vgupta(a)synopsys.com> --- libc/sysdeps/linux/arc/bits/uClibc_arch_features.h | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/libc/sysdeps/linux/arc/bits/uClibc_arch_features.h b/libc/sysdeps/linux/arc/bits/uClibc_arch_features.h index 51607240cd82..94e089d5dfa8 100755 --- a/libc/sysdeps/linux/arc/bits/uClibc_arch_features.h +++ b/libc/sysdeps/linux/arc/bits/uClibc_arch_features.h @@ -41,8 +41,14 @@ /* The default ';' is a comment on ARC. */ #define __UCLIBC_ASM_LINE_SEP__ ` -/* does your target align 64bit values in register pairs ? (32bit arches only) */ -#if defined(__A7__) +/* does your target align 64bit values in register pairs ? (32bit arches only) + * - ARC700 never had any constraint on reg pairs (even if ABI v3) + * - Inital HS ABI (v3: non upstream gcc) had 64-bit data aligned in even-odd + * reg pairs (thus allowed reg holes when passing such args to calls) + * - Upstream gcc (6.x) HS ABI doesn't have that restriction + */ + +#if defined(__A7__) || (__GNUC__ > 4) #undef __UCLIBC_SYSCALL_ALIGN_64BIT__ #else #define __UCLIBC_SYSCALL_ALIGN_64BIT__ -- 2.7.4

2 1

uClibc-ng - small C library for embedded systems branch master updated. v1.0.17-5-g3e40f96
by wbx＠helium.openadk.org 17 Aug '16

17 Aug '16

1 0

uClibc-ng - small C library for embedded systems branch master updated. v1.0.17-4-g9725c0b
by wbx＠helium.openadk.org 16 Aug '16

16 Aug '16

1 0

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

devel