Curent ARC toolchain fails to build libstdc++ if -fno-plt is used.
Lots of following error messages appear:
------------------->8------------------
...
staging_dir/toolchain-arc_arc700_gcc-arc-2015.06_uClibc-1.0.9/arc-openwrt-linux-uclibc/bin/ld:
BFD (GNU Binutils) 2.23.2 assertion fail elf32-arc.c:2786
collect2: error: ld returned 1 exit status
------------------->8------------------
In newer binutils (still in development) for ARC rewritten from
scratch this seem to not happen, so once new binutils for ARC hit
the street this patch might be reverted.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Cc: Felix Fietkau <nbd@openwrt.org>
Cc: Jo-Philipp Wich <jow@openwrt.org>
Cc: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 48642
Add the basic set of kernel options to allow it from mounting a NFS root
and boot from it.
Signed-off-by: Florian Fainelli <florian@openwrt.org>
SVN-Revision: 48590
These allow the generated kernel's build metadata to be defined explicitly.
This metadata is reported, eg, at boot time and in `uname -a` on running
systems. If the variables aren't configured, the current build system username
and hostname are used as normal.
The motivation for this option is to achive reproducible (bit-for-bit
identical) kernel builds of official openwrt releases.
Signed-off-by: bryan newbold <bnewbold@robocracy.org>
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
SVN-Revision: 48541
Revision 46834 changed IPv6 support from a module to builtin. But
since the configuration of the IPv6 kernel options was left in
package/kernel/linux/modules/netsupport.mk, this means that an
empty kmod-ipv6 module was still being generated (not packaged).
This patch moves the configuration of the IPv6 kernel options to
config/Config-kernel.in to remove this last bit of the module.
Note that CONFIG_IPV6_PRIVACY was dropped (enabled by default
since Linux v3.13), so this option is no longer needed.
See 5d9efa7ee9
Signed-off-by: Arjen de Korte <arjen+openwrt@de-korte.org>
SVN-Revision: 48132
When stack protector support is disabled in libc (always the case for
!musl), gcc assumes that it needs to use __stack_chk_guard for the stack
canary.
This causes kernel build errors, because the kernel is only set up to
handle TLS stack canaries.
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
SVN-Revision: 46543
Make musl provide libssp_nonshared.a and make GCC link it unconditionally
if musl is used. This should be a no-op if SSP is disabled and seems to be
the only reliable way of dealing with SSP over all packages due to the mess
that is linkerflags handling in packages.
Signed-off-by: Steven Barth <steven@midlink.org>
SVN-Revision: 46108
Memory Resource Controller no longer depends on Resource counters since
Kernel version 4.0.
3.18 is the only still supported version needing Resource counters for
MEMCG, thus declare the dependency only for that version.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
SVN-Revision: 46024
This changes the x86 image generation to match x86_64, using the PARTUUID for
the rootfs instead of explicitly configuring the device.
It unbreaks KVM with VirtIO, which uses /dev/vda2 instead of /dev/sda2.
Tested in QEMU/KVM with VirtIO, VirtualBox and VMware.
Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
SVN-Revision: 44966
Split out kmods from ALL to make it easier to create local builds that
are compatible kmod-wise with releases.
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 44830
This is needed by many services to function properly and as
all modern distributions got it enabled, it starts to be a
de-facto standard, i.e. user-space starts to silently depend
on it.
This also pulls in EXPORTFS, however, the kernel binary size
increases only a little.
On ARM systems comes down to 800 bytes uncompressed and about
200 bytes compressed size.
On MIPS systems it's about 1.2 kB size increase of the LZMA
compressed kernel.
v2: use menuconfig option instead of just enabling the option
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
SVN-Revision: 44765
It's the eglibc packaging with a bit of spit-polishing. And testing. :-)
[blogic: merged glibc and eglibc into 1 and made eglibc a glibc variant]
Signed-off-by: Jeff Waugh <jdub@bethesignal.org>
Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 44701
Introduce configuration options to build an "hardened" OpenWRT.
Options to enable Stack-Smashing Protection, FORTIFY_SOURCE and RELRO
have been introduced.
uClibc makefile now automatically detects if SSP support is necessary.
hostapd makefile has been fixed to use "^" as sed separator since
using a comma was problematic when using "-Wl,-z,now" and the like in
TARGET_CFLAGS.
Currently enabling SSP on user space depends on enabling SSP kernel
side, this is due to the fact that TARGET_CFLAGS are used to build
kernel modules (at least). Suggestions on how to avoid this are welcome.
Using "select" instead of "depends on" doesn't seem to work with choice
entries.
Tested with a lantiq (WBMR) router, GCC 4.8, uClibc and a subset of
the available packages.
Needs to be tested with GCC 4.9 and the remaining packages.
PIE not currently included.
Signed-off-by: Alessandro Di Federico <ale+owrt@clearmind.me>
SVN-Revision: 44005
This patch completes missing help text for some options under CONFIG_DEVEL.
Provides help for BINARY_FOLDER and DOWNLOAD_FOLDER, and reduces ambiguity in
the help for BUILD_SUFFIX with an example.
Signed-off-by: Andrew McDonnell <bugs@andrewmcdonnell.net>
SVN-Revision: 42520
Non-functional changes to config/Config-*.in files, including:
* spelling mistakes
* inconsistent terminology
* grammar
* overly long lines in "help" components
Signed-off-by: Robert P. J. Day <rpjday@crashcourse.ca>
SVN-Revision: 42519
The idea is still to enable it by default at some point
I've tested all ar71xx packages (except oldpackages) using CONFIG_ALL=y
Failing packages have been marked with PKG_CHECK_FORMAT_SECURITY:=0 for now
I can test more targets but i have no idea which are the most used
Signed-off-by: Etienne CHAMPETIER <champetier.etienne@gmail.com>
SVN-Revision: 42282
Felix Fietkau