This was found while investigating why luarocks does not work. It was
traced to a quite old lnum patch for 5.1.3. I compared against the
latest 5.1.4 patch - https://github.com/LuaDist/lualnum and discovered
the lessthan/lessequal evaluation was not falling through to the
call_orderTM (tag methods).
I have tested LuCI (simple tests) and used the following lua code to
validate the patch (both host and target patches supplied): -
> local my_mt = {
> __eq = function(v1, v2)
> print("__eq")
> return false
> end,
> __lt = function(v1, v2)
> print("__lt")
> return false
> end,
> __le = function(v1, v2)
> print("__le")
> return false
> end
> }
>
> function get_my(vstring)
> local my = {}
> my.string = vstring;
> setmetatable(my, my_mt);
> return my;
> end
>
> local a = get_my("1.0")
> local b = get_my("1.0")
>
> local eq_works = a == b;
> local lt_works = a < b;
> local gt_works = a > b;
>
> local lte_works = a <= b;
> local gte_works = a >= b;
Without the patch the following error will be presented: -
“attempt to compare two table values”
Signed-off-by: David Thornley <david.thornley@touchstargroup.com>
Add nonshared flag to package depending on specific targets or subtargets as
there's no guarantee otherwise that they'll be available in the shared repo.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
The settimeofday() syscall wrapper provided by musl filters out the timezone
argument, breaking the ability to set the kernel timezone through the function.
Adjust busybox patch to issue the syscall directly in order to circumvent the
problem.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Change mirror to Github (Gentoo repo) and drop the gzip compression.
Worst case there's about 4kbyte increase in size but most images ends up
beign somewhere between 4-100kbyte smaller due to the lzma compression.
Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
SVN-Revision: 49246
The following patches were merged upstream:
* 0001-switch_root-improve-statfs-f_type-portability.patch
* 0002-lib-colors-use-static-buffers-when-parse-scheme.patch
* 002-mkostemp.patch
The following patch is not needed any more because all libc
implementations support alloc in sscanf:
* 001-no-printf-alloc.patch
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 49191
Updates USB IDs list to snapshot 2015-12-17 and changes copyright to
2016.
Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
SVN-Revision: 49112
One of the host patches introduces the new header file lnum_config.h
included by luaconf.h, but doesn't install it.
Install it to allow building C modules for the host Lua.
Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
SVN-Revision: 48907
This will be used to create a diff between the Lantiq annex A and the
annex B firmware.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 48810
Changeset r47775 introduced a wrong default value "n" for the ip applets
config search path, breaking the handling of named routing tables and other
objects having alias configurations.
Fix the issue by providing a proper default value aligned with the value
used by full iproute2.
Signed-off-by: Jo-Philipp Wich <jow@openwrt.org>
SVN-Revision: 48716
This will be needed to handle sysupgrade on NAND in a smart way. We'll
need to extract kernel our of provided Seama container, put in in a new
Seama entity and flash.
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
SVN-Revision: 48552
This reverts r47245.
Linking to shared libraries under $STAGING_DIR_HOST{,/usr}/lib is harmful,
as these directories aren't added to LD_LIBRARY_PATH (see r47103 for an
explanation why LD_LIBRARY_PATH is not used).
Revert to static linking; in particular, this fixes the build of the python
bz2 module on OpenSUSE and Fedora (which in turn broke the build of
nodejs).
Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
SVN-Revision: 47950
r47288 updated to Busybox 1.24.1 but did not update the configuration.
The configuration is updated by running
cd config
../convert_menuconfig.pl .../build_dir/target-*/busybox-1.24.1
cd ..
./convert_defaults.pl < .../build_dir/target-*/busybox-1.24.1/.config > \
Config-defaults.in
Signed-off-by: Mark Mentovai <mark@moxienet.com>
SVN-Revision: 47775
It looks like the bzip2 package does not install any shared libs
and has no build rules to install any shared libs.
So, for the host build we're installing the libbz2 shared libs manually
so that other modules can link against them.
Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
SVN-Revision: 47245
due to ordering PKG_SOURCE_VERSION is not defined leading
to a filename "ugps-.tar.bz2"
This errors out when an older version is in the dl/ dir (or LOCALMIRROR)
fix order and use uhttpd file naming scheme to visibly include date
Signed-off-by: Dirk Neukirchen <dirkneukirchen@web.de>
SVN-Revision: 46868
lock -n is similiar to flock -n. If the lock was already taken,
fail with exit code = 1 and write error message to stderr.
example:
if ! lock -n /tmp/foo ; then
echo lock exits.
else
echo lock was free. But is locked now.
fi
> lock was free. But is locked now.
> lock exists.
Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
SVN-Revision: 46836
added: ip addr, ip route, ip link, traceroute6
removed: hostid, devmem, vconfig, arping
deprecated (to be removed): ifconfig, route
Signed-off-by: Steven Barth <steven@midlink.org>
SVN-Revision: 46829
This enables passworldless login for root via SSH whenever no root
password is set (e.g. after reset, flashing without keeping config
or in failsafe) and removes telnet support alltogether.
Signed-off-by: Steven Barth <steven@midlink.org>
SVN-Revision: 46809
This controls the old Broadcom switch driver which was removed over a
year ago, since then this is useless. Use swconfig instead it works
with more (Broadcom) switches.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 46373
caddr_t is not defined everywhere any more, cast it the the real
pointer type of ifr_data.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 46371
This reverts commit r45579.
With the latest change r46251 ("nvram: fix "Segmentation fault" caused
by setting memory out of buffer") nvram utility shouldn't crash anymore.
It was tested on 3 brcm47xx devices:
1) Unknown with 0x10000 NVRAM size (0x8000 offset)
2) Linksys E1000 V2.1 with 0x10000 (0x8000 offset)
3) Linksys WRT300N V1 with 0x10000 (0x8000 offset)
And 3 bcm53xx devices:
1) Buffalo WZR-600DHP2 with 0x160000 NVRAM size
2) Buffalo WZR-1750DHP with 0x10000 NVRAM size
3) Netgear R6250 V1 with 0x180000 NVRAM size
(all using 0 offset)
This is an important change as it allows reading whole NVRAM. This may
critical when reading some basic configuration (e.g. switch ports).
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
SVN-Revision: 46253
Some MTD partitions with NVRAM have content starting in the middle. In
such case offset is set and nvram_header returns pointer to the middle.
It means we have to respect offset when calculating remaining space.
By the way use real MTD partition size (nvram_part_size variable) as we
may want to bump NVRAM_SPACE in the (very near) future.
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
SVN-Revision: 46251
Without this the condition
if __BYTE_ORDER == __BIG_ENDIAN
was always true.
Thanks Szabolcs.
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
SVN-Revision: 46105
So far we got only one generic function accessing this table, but
implementing optimizations will require calculating crc32 in other code
parts as well.
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
SVN-Revision: 46077
bzip2 is required by elfutils, itself required by perf. So we'll move
this package from packages.git and make it part of the core
distribution.
Signed-off-by: Mathieu Olivari <mathieu@codeaurora.org>
SVN-Revision: 45968
This reverts commit ff84c27a281bc19df19bc62ee8688cca5586f6e3.
This tool has really broken size handling (many values hardcoded), it
crashes right now in case of NVRAM not filling whole MTD partition.
Conflicts:
package/utils/nvram/src/nvram.h
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
SVN-Revision: 45579
For years Broadcom devices use 64 KiB NVRAM partition size and some of
them indeed have it filled in more than 50%. This change allows handling
whole NVRAM e.g. on Netgear WNDR4500 and Netgear R8000.
The same fix was applied to kernel in upstream commit 6ab7c29.
Reported-by: Hante Meuleman <meuleman@broadcom.com>
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
SVN-Revision: 45566
We don't have broadcom-diag for months or years now and the correct
solution is to simply don't have "nvram" partition on WGT634U anyway.
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
SVN-Revision: 45564
Sytax of /proc/mtd is following:
dev: size erasesize name
which means that sscanf "mtd%d: %08x" reads size, not erasesize.
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
SVN-Revision: 45563
