Fixes compilation with both GCC 10 and 11.
Switched to AUTORELEASE for simplicity.
Removed PKG_VERSION as it's derived from PKG_SOURCE_VERSION.
Removed all patches as they are upstream backports.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
GCC 10 defaults to `-fno-common` and complains about multiple definition
of `mc_status` in restool.
Backport a patch from upstream to fix compilation with host GCC 10.
Signed-off-by: Kuan-Yi Li <kyli@abysm.org>
Update iproute2 to latest stable 5.14; for the changes see https://lwn.net/Articles/867940/
Refresh patches
Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
When doing parallel build on a fast machine with bottleneck in i/o,
m_xt.so may start linking faster than dynsyms.list gets populated,
resulting in error:
ld:dynsyms.list:0: syntax error in dynamic list
Fix this by adding dynsyms.list as make dependency to m_xt.so
Described also here:
https://bugs.openwrt.org/index.php?do=details&task_id=3353
Change from v1:
- add dynsysms.list dependancy only when shared libs are enabled
Signed-off-by: Roman Yeryomin <roman@advem.lv>
Fixes: FS#3353
Change the CONFLICTS definition from the alternative package
(ethtool-full) to the main one.
The CONFLICTS line creates a dependency to the conflicting package.
Right now, the dependency would be created in the PACKAGE_ethtool-full
symbol:
config PACKAGE_ethtool-full
depends on m || (PACKAGE_ethtool != y)
When the main package is selected by airmon-ng, it selects
PACKAGE_ethtool, *depending* on the value of PACKAGE_ethtool-full:
config PACKAGE_airmon-ng
select PACKAGE_ethtool if PACKAGE_ethtool-full<PACKAGE_airmon-ng
In the first block, the value of PACKAGE_ethtool-full depends on the
value of PACKAGE_ethtool. In the second block, the opposite is true:
the value of PACKAGE_ethtool depends on the value of
PACKAGE_ethtool-full. This is a recursive dependency.
Fix it by changing the package where the dependency is created, so that
only the value of PACKAGE_ethtool will depend on PACKAGE_ethtool-full.
Signed-off-by: Eneas U de Queiroz <cotequeiroz@gmail.com>
This increases the size of the iw_5.9-8fab0c9e-3_mips_24kc.ipk from
41166 to 41942 bytes by 776 bytes.
Signed-off-by: Jianhui Zhao <zhaojh329@gmail.com>
Netlink support is required for using the virtual cable tester
functionality.
Remove the pretty print build option and instead create a second package
variant ethtool-full. This allows users to install the full ethtool
featureset using opkg.
Signed-off-by: David Bauer <mail@david-bauer.net>
a0a0e02 iwinfo: rename hardware.txt to devices.txt
Also split common devices.txt (former hardware.txt) into a common
libiwinfo-data package to allow different libiwinfo versions to
coexist without file clashes.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Drop support for building the obsolete broadcom-wl backend and always
forcibly enable the nl82011 support. This allows us to make the package
shared again since no target specific compilation is happening anymore.
This will solve various repository coherency issues related to unavailable
libiwinfo versions in the long run.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
This script was expecting only add/remove events which has not been the
case since Kernel 4.12 (which added bind/unbind). Bind events were getting
treated as remove events which would cause hotplugged 3g modems to not
work.
More info:
https://lkml.org/lkml/2018/12/23/128https://github.com/systemd/systemd/issues/8221
Signed-off-by: Arjun AK <arjunak234@gmail.com>
aa0e3c4bbe12 iwinfo: nl80211: add support for printing the device path for a phy
dd6d6d2dec35 iwinfo: nl80211: use new path lookup function for nl80211_phy_idx_from_uci_path
268bb26d2e2a iwinfo: nl80211: support looking up phy by path=.. and macaddr=...
c0414642fead iwinfo: nl80211: fix typo
Signed-off-by: Felix Fietkau <nbd@nbd.name>
This fixes `iw dev wlan0-mesh station dump`.
8fab0c9 iw: fix ftm_request missing arguments segfault
e816fbc iw: fix mgmt dump missing arguments segfault
5d9d1b8 iw: Fix timestamp output on 32-bit architectures
4b25ae3 iw: fix pointer arithmetic in __print_he_capa
c3df363 iw: add option to print human readable event time
cd64525 iw: print ctrl port tx status event
0ba98b9 iw: use correct type in policy check for mesh
9e38dee iw: scan: fixup HE caps whitespace
17e8564 iw: scan: parse HE capabilities
5735e58 iw: util: factor out HE capability parser
6d8d507 iw: scan: add extension tag parsing
b4e1ec4 man: update wikipage URL, reformat SEE ALSO section
c56036a iw: enable 80MHz support for 6GHz band 11s mesh
fa72728 iw: handle positive error codes gracefully
7ba9093 iw: scan: add flag for scanning colocated ap
5ec60ed iw: Add 'coloc' and 'flush' options to sched_scan
f8ade75 iw: update wikipage URL
b6f2dac iw: Add support for specifying the 160MHz bandwidth when setting the channel/frequency
Signed-off-by: Andre Heider <a.heider@gmail.com>
This marks all packages which depend on a target with @TARGET nonshared.
If they are not marked nonshared they would be build by the SDK build
and if this happens with a different SDK, then the SDK from the target
the package depends on, the package would not be added to the index.
This should fix the image builder for some of these packages.
This should fix the image builder at least for bcm27xx/bcm2710 and
bcm4908/generic.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Before this commit, it was assumed that mkhash is in the PATH. While
this was fine for the normal build workflow, this led to some issues if
make TOPDIR="$(pwd)" -C "$pkgdir" compile
was called manually. In most of the cases, I just saw warnings like this:
make: Entering directory '/home/.../package/gluon-status-page'
bash: line 1: mkhash: command not found
bash: line 1: mkhash: command not found
bash: line 1: mkhash: command not found
bash: line 1: mkhash: command not found
bash: line 1: mkhash: command not found
bash: line 1: mkhash: command not found
bash: line 1: mkhash: command not found
bash: line 1: mkhash: command not found
[...]
While these were only warnings and the package still compiled sucessfully,
I also observed that some package even fail to build because of this.
After applying this commit, the variable $(MKHASH) is introduced. This
variable points to $(STAGING_DIR_HOST)/bin/mkhash, which is always the
correct path.
Signed-off-by: Leonardo Mörlein <me@irrelefant.net>
With some debug in qmi.sh using following patch, some errors are visible
in the registration step
@@ -29,6 +29,7 @@ proto_qmi_init_config() {
}
proto_qmi_setup() {
+ set -x
local interface="$1"
local dataformat connstat plmn_mode mcc mnc
local device apn auth username password pincode delay modes pdptype
@@ -224,6 +225,8 @@ proto_qmi_setup() {
fi
done
+ registration=$(uqmi -s -d "$device" --get-serving-system)
+
[ -n "$modes" ] && uqmi -s -d "$device" --set-network-modes "$modes" > /dev/null 2>&1
echo "Starting network $interface"
During the boot of the system, modem could not start automatically its
network registration.
netifd: wan (9235): + echo 'Waiting for network registration'
netifd: wan (9235): Waiting for network registration
netifd: wan (9235): + local 'registration_timeout=0'
netifd: wan (9235): + uqmi -s -d /dev/cdc-wdm1 --get-serving-system
netifd: wan (9235): + grep '"searching"'
netifd: wan (9235): + uqmi -s -d /dev/cdc-wdm1 --get-serving-system
netifd: wan (9235): + registration='{"registration":"not_registered","plmn_mcc":208,"plmn_mnc":20,"plmn_description":"","roaming":true}'
netifd: wan (9235): + '[' -n ]
netifd: wan (9235): + echo 'Starting network wan'
As the while loop checks only "searching" pattern, uqmi.sh script quits
searching loop and continues whereas the modem is not registered
Other issue, after X seconds modem stops searching.
netifd: wan (9213): + uqmi -s -d /dev/cdc-wdm0 --get-serving-system
netifd: wan (9213): + grep '"searching"'
netifd: wan (9213): + '[' -e /dev/cdc-wdm0 ]
netifd: wan (9213): + '[' 3 -lt 0 -o 0 '=' 0 ]
netifd: wan (9213): + let registration_timeout++
netifd: wan (9213): + sleep 1
netifd: wan (9213): + uqmi -s -d /dev/cdc-wdm0 --get-serving-system
netifd: wan (9213): + grep '"searching"'
netifd: wan (9213): + uqmi -s -d /dev/cdc-wdm0 --get-serving-system
netifd: wan (9213): + registration='{"registration":"not_registered"}'
netifd: wan (9213): + '[' -n ]
netifd: wan (9213): + echo 'Starting network wan'
netifd: wan (9213): Starting network wan
If registration_timeout is not expired, registration can be restarted
Signed-off-by: Thomas Richard <thomas.richard@kontron.com>
Tested-by: Florian Eckert <fe@dev.tdt.de>
This package depends on the lantiq target and is only build for that
target. A normal package would be build by the SDK builder probably
under a different target and then this package will not be selected.
Mark it as nonshared to build it when the lantiq target gets build.
Fixes: FS#3773, FS#3774
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Update to the latest stable upstream version.
Drop unneeded make variables to remove redundant assignments seen during
invocation of package Makefile.
Also remove the following patch now included upstream:
* 200-fix-install-param-order-on-macos.patch
Compile and run-tested on malta/mips32be, using bpftool directly and also
libbpf (linked with tc) to inspect and load simple eBPF programs.
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
Rename feature variable clang-bpf-global-var following upstream changes.
This restores the HAVE_CLANG feature override and should avoid rare build
errors where a recent host clang and BTF-enabled host kernel are present.
Fixes: 23be333401f0 ("bpftools: update to 5.10.10")
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
Skip building Makefile targets that aren't packaged: tipc, dcb, ifstat,
rtacct, lnstat, and man. Also, only compile targets needed for the current
build variant i.e. don't compile 'tc' when building an 'ip' variant and
vice versa.
These changes reduce typical build times by over 30%:
$ make package/iproute2/clean && time make -j8 package/iproute2/compile
(old)
...
real 2m24.985s
user 3m12.537s
sys 0m26.677s
(new)
...
real 1m36.945s
user 2m8.734s
sys 0m20.046s
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
Upstream iproute2 detects libbpf using a one-line $CC test-compile, which
normally ignores LDFLAGS. With NLS enabled however, LDFLAGS includes an
"rpath-link" linker option needed to resolve libintl.so. Its absence
causes both the compile and libbpf detection to fail:
ld: warning: libintl.so.8, needed by libbpf.so, not found (try using
-rpath or -rpath-link)
ld: libelf.so.1: undefined reference to `libintl_dgettext'
collect2: error: ld returned 1 exit status
Fix this by directly including $LDFLAGS in the test-compile command.
Reported-by: Ian Cooper <iancooper@hotmail.com>
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
There is no direct linking of libintl from bpftools, only secondary linking
through libelf, so remove "-lintl" from TARGET_LDFLAGS.
Fixes: 5582fbd6135f ("bpftools: support NLS, fix ppc build and update to 5.8.9")
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
This change was investigated previously [1] but not deemed necessary. With
the recent addition [2] of modern BPF loader support, however, tc gained
dependencies on libelf and libbpf, with a larger installation footprint.
Similar to ip-tiny/ip-full, split tc into tc-full and tc-tiny variants,
where the latter excludes the eBPF loader, uses a smaller executable, and
avoids libelf and libbpf package dependencies. Both variants provide the
'tc' virtual package, with tc-tiny as the default.
The previous tc package included a loadable module for iptables actions.
Separate this out into a common package, tc-mod-iptables, which both
variants depend on. Some package sizes on mips_24kc:
Before:
148343 tc_5.11.0-1_mips_24kc.ipk
After:
144833 tc-full_5.11.0-2_mips_24kc.ipk
138430 tc-tiny_5.11.0-2_mips_24kc.ipk (and no libelf or libbpf)
4115 tc-mod-iptables_5.11.0-2_mips_24kc.ipk
Also fix up some Makefile indentation.
[1] https://github.com/openwrt/openwrt/pull/1627#issuecomment-447619962
[2] b048a305a3d3 ("iproute2: update to 5.11.0")
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
The link equalizer sch_teql.ko of package kmod-sched relies on a hotplug
script historically included in iproute2's tc package. In previous
discussion [1], consensus was the hotplug script is best located together
with the module in kmod-sched, but this change was deferred at the time.
Relocate the hotplug script now. This change also simplifies adding a tc
variant for minimal size with reduced functionality.
[1] https://github.com/openwrt/openwrt/pull/1627#issuecomment-447923636
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
This patch has been submitted upstream to fix an error reported by a few
users. One instance seen using gcc 10.2.0, binutils 2.35.1 and musl 1.1.24:
bpf_glue.c: In function 'get_libbpf_version':
bpf_glue.c:46:11: error: 'PATH_MAX' undeclared (first use in this function);
did you mean 'AF_MAX'?
46 | char buf[PATH_MAX], *s;
| ^~~~~~~~
| AF_MAX
Reported-by: Rui Salvaterra <rsalvaterra@gmail.com>
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
Fix: bpftools 5.11.2 does not compile on macOS, because the -m option
was placed between src and dst. Corrected by moving -m 644 before src.
Signed-off-by: Georgi Valkov <gvalkov@abv.bg>
The pkgconfig file hardcodes a host library directory which cannot be
overridden by OpenWrt during builds. Use SED to fix this and potential
include directory problems, as is done with several other packages.
This fixes a strange issue intermittently seen building iproute2 on the
oxnas target:
iptables modules directory: /usr/lib/iptables
libc has setns: yes
SELinux support: no
libbpf support: no
libbpf version 0.3.0 is too low, please update it to at least 0.1.0
LIBBPF_FORCE=on set, but couldn't find a usable libbpf
Fixes: 2f0d672088c3 ("bpftools: add utility and library packages
supporting eBPF usage")
Reported-by: Russell Senior <russell@personaltelco.net>
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
This will use the new application led trigger backend. For now this is
the only package that uses leds trigger in user space to configure the
kernel led triggers.
The callback script only emmits a message for now, so that this LED is now
managed by the rssileds service. Until now a generic warning was emitted that
this LED trigger is not supported. But that is not true.
-> Skipping trigger 'rssileds' for led '<name>' due to missing kernel module
I think this callback should be changed in the future to restart the
service.
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
Compile and run-tested on malta/mip32be, using bpftool directly and also
libbpf (linked with tc) to inspect and load simple eBPF programs.
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
The latest iproute2 version brings various improvements and fixes:
https://git.kernel.org/pub/scm/network/iproute2/iproute2.git/log/?qt=range&q=v5.10.0..v5.11.0
In particular, ip and tc now use libbpf as the standard way to load BPF
programs, rather than the old, limited custom loader. This allows more
consistent and featureful BPF program handling e.g. support for global
initialized variables.
Also fix a longstanding problem with iproute2 builds where unneeded DSO
dependencies are added to most utilities, bloating their installation
footprint. From research and testing, explicitly using a "--as-needed"
linker flag avoids the issue. Update accordingly and drop extra package
dependencies from Makefile.
Additional build and packaging updates include:
- install missing development header to iproute2/bpf_elf.h
- propagate OpenWrt verbose flag during build
- update and refresh patches
Compile and run tested: QEMU/malta-mips32be on kernels 5.4 & 5.10.
All iproute2 packages were built and installed to the test image. Some
regression testing using ip-full and tc was successfully performed to
exercise several kmods, tc modules, and simple BPF programs.
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
To the vast majority of the users, wireguard-tools are not useful
without the underlying kernel module. The cornercase of only generating
keys and not using the secure tunnel is something that won't be done on
an embedded OpenWrt system often. On the other hand, maintaining a
separate meta-package only for this use case introduces extra
complexity. WireGuard changes for Linux 5.10 remove the meta-package.
So let's make wireguard-tools depend on kmod-wireguard
to make WireGuard easier to use without having to install multiple
packages.
Fixes: ea980fb9 ("wireguard: bump to 20191226")
Signed-off-by: Ilya Lipnitskiy <ilya.lipnitskiy@gmail.com>
After the ABI version rework, packages need to be declared in the order of
their dependencies, so that dependent packages will use the right ABI version
Signed-off-by: Felix Fietkau <nbd@nbd.name>
The removed patches were applied upstream.
This adapts ath10k-ct and mt76 to changed APIs.
nl80211.h in iw is updated to match the version from backports.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
5a2dd18 iwinfo: add hardware description for MediaTek MT7622
4a32b33 iwinfo: add PCI ID for MediaTek MT7613BE
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
