This reduces open coding and allows to easily add a knob to enable
it treewide, where chosen packages can still opt-out via "no-lto".
Some packages used LTO, but not the linker plugin. This unifies 'em
all to attempt to produce better code.
Quoting man gcc(1):
"This improves the quality of optimization by exposing more code to the
link-time optimizer."
Also use -flto=auto instead of -flto=jobserver, as it's not guaranteed
that every buildsystem uses +$(MAKE) correctly.
Signed-off-by: Andre Heider <a.heider@gmail.com>
This reduces open coding and allows to easily add a knob to
enable it treewide, where chosen packages can still opt-out via
"no-gc-sections".
Note: libnl, mbedtls and opkg only used the CFLAGS part without the
LDFLAGS counterpart. That doesn't help at all if the goal is to produce
smaller binaries. I consider that an accident, and this fixes it.
Note: there are also packages using only the LDFLAGS part. I didn't
touch those, as gc might have been disabled via CFLAGS intentionally.
Signed-off-by: Andre Heider <a.heider@gmail.com>
PKG_BUILD_FLAGS is a new variable for package Makefiles similar to
PKG_FLAGS. It's a whitespace separated list of flags to control various
aspects of how a package is build.
The build system and/or .config defines the default for each, but
every package has the means to override it. Using $flagname enables
a flag, no-$flagname disables it.
Start with PKG_IREMAP as "iremap". That's easy as no package here
nor in any package feed uses it. The default is unchanged: enabled.
Packages can opt-out via:
PKG_BUILD_FLAGS:=no-iremap
(Not that any should, just to illustrate how to use it)
Signed-off-by: Andre Heider <a.heider@gmail.com>
Manually rebased:
ramips/patches-5.10/810-uvc-add-iPassion-iP2970-support.patch
All other patches automatically rebased.
Signed-off-by: John Audia <therealgraysky@proton.me>
No patches needed to be rebased.
Signed-off-by: John Audia <therealgraysky@proton.me>
Tested-by: Robert Marko <robimarko@gmail.com> #ipq807x/Dynalink WRX36
Tested-by: Stefan Lippers-Hollmann <s.l-h@gmx.de> #ipq807x/ax3600, x86_64/FW-7543B, ath79/tl-wdr3600, ipq806x/g10, ipq806x/nbg6817
No patches rebased as a function of this bump.
Build system: x86_64
Build-tested: ramips/tplink_archer-a6-v3
Run-tested: ramips/tplink_archer-a6-v3
Signed-off-by: John Audia <therealgraysky@proton.me>
autoreconf wants to use the gtkdocize tool now if a configure.ac file
defines GTK_DOC_CHECk(). OpenWrt does not ship the gtkdocize tool, just
use true instead. This fixes the build of some applications like guntls.
Fixes: 030447b8f4 ("tools/autoconf: bump to 2.71")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Recent versions of Automake
have changed dependency tracking significantly
(reference commit below)
causing breakage in some package builds
when using newer Automake with packages that need autoreconf
that were bootstrapped with an old version of Automake.
Those changes cause a great inconsistency between packages over time
where some packages may or may not use this feature,
and may or may not update the .ac and .am files
to work with the new methods.
This problem might exist in many packages
where autoreconf is not currently required,
but would cause build failure if autoreconf is used.
Fortunately, this feature is practically useless
for the purposes of Openwrt and the average developer,
so we can disable it.
GNU Automake manual states in part:
"Because dependencies are only computed as a side-effect of compilation...
no dependency information exists the first time a package is built...
dependency tracking is completely useless for one-time builds..."
A nice side-effect is that build times are slightly faster.
Ref: 6a675ef17edf7109da189f5ae70e2dc6b7665896 (automake.git)
Signed-off-by: Michael Pratt <mcpratt@pm.me>
The -Xpreset option changed format and dropped the -Xe and just require
the extreme string to be passed to the -Xpreset option.
Update the LZMA_XZ_OPTIONS to reflect this change.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
The name of squashfs is confusing since in reality it's a really old
version using an old lzma library. This tools is used for old ath79
netgear target and to produde a fake squasfs3 image needed for some
specific bootloader from some OEM (AVM for example)
Rename squashfs tool to squasfs3-lzma to better describe it.
Rename the installed bin from mksquashfs-lzma to mksquashfs3-lzma.
Use tar transform to migrate the root directory in tar to the new
naming.
Drop redundant PKG_CAT variable not needed anymore.
Also update any user of this tool.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Packages in general use 4 check to trigger a recompile:
- timestamp for the build_dir
- timestamp for the staging stamp dir
- depends hash for the build_dir prepared file
- presence of package archieve in dl
If host tools are prebuilt and shipped in a container or manually
installed from an archieve, it would be ideal to skip including the
package archieve and just provide the build_dir prepared files and the
staging stamp file (and the actualy prebuilt tools).
Add some logic to skip dl download for host tools if AUTOREMOVE is
selected and checks for the presence of staging dir stamp file and build
dir stamp file.
If one of these requirements are not met, the package is redownloaded
and rebuilt.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Starting from version 2.39 binutils now warns about sections with rwx
permissions. While this is generally desirable it breaks building
ARM TrustedFirmware-A bl2 which treats warnings as errors.
Disable the warning/error for now to fix build.
Reference: 0579d9f5bc
Signed-off-by: Linhui Liu <liulinhui36@gmail.com>
To download a package the LLVM bins are not strictly needed.
Currently with an example run of make package/bridger/download V=s, the
build fail with
make[2]: Entering directory '/home/ansuel/openwrt-ansuel/openwrt/package/network/services/bridger'
bash: line 1: /home/ansuel/openwrt-ansuel/openwrt/staging_dir/host/llvm-bpf/bin/clang: No such file or directory
bash: line 1: [: : integer expression expected
/home/ansuel/openwrt-ansuel/openwrt/include/bpf.mk:71: *** ERROR: LLVM/clang version too old. Minimum required: 12, found: . Stop.
make[2]: Leaving directory '/home/ansuel/openwrt-ansuel/openwrt/package/network/services/bridger'
time: package/network/services/bridger/download#0.04#0.00#0.06
ERROR: package/network/services/bridger failed to build.
This is wrong since it may be needed to download the required packages
first and then compile them later.
Fix this by ignoring the LLVM bin check on non compile steps.
Tested-by: Robert Marko <robimarko@gmail.com>
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
libdeflate decompresses much faster than gzip.
Example:
~/d/openwrt> time gzip -dc dl/cmake-3.25.1.tar.gz > /dev/null
________________________________________________________
Executed in 1.01 secs fish external
usr time 912.61 millis 1.67 millis 910.94 millis
sys time 32.21 millis 0.25 millis 31.96 millis
~/d/openwrt> time libdeflate-gzip -dc dl/cmake-3.25.1.tar.gz > /dev/null
________________________________________________________
Executed in 523.04 millis fish external
usr time 415.48 millis 1.07 millis 414.41 millis
sys time 107.74 millis 0.15 millis 107.59 millis
~/d/openwrt> time gzip -dc dl/tessdata-4.1.0.tar.gz > /dev/null
________________________________________________________
Executed in 8.99 secs fish external
usr time 8.90 secs 530.00 micros 8.90 secs
sys time 0.07 secs 63.00 micros 0.07 secs
~/d/openwrt> time libdeflate-gzip -dc dl/tessdata-4.1.0.tar.gz > /dev/null
________________________________________________________
Executed in 2.74 secs fish external
usr time 2.38 secs 537.00 micros 2.38 secs
sys time 0.35 secs 66.00 micros 0.35 secs
Place libdeflate into tools-core as it is needed to decompress other
archives.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
Requires: tools/lz4, tools/lzop
complete the wiring so that these options work:
* `CONFIG_KERNEL_INITRAMFS_COMPRESSION_LZO`
* `CONFIG_KERNEL_INITRAMFS_COMPRESSION_LZ4`
Signed-off-by: Tony Butler <spudz76@gmail.com>
[remove blocking dependencies for separate ramdisk, fix lzop options]
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
FACTORY_SIZE is used as a device recipe variable on both the D-Link
DIR-825-B1 and Trendnet TEW-673GRU, but is not listed as a device
variable, neither globally, nor for ath79. Being assigned the same
value, this probably hasn't caused any issues.
Add FACTORY_SIZE to the global list DEFAULT_DEVICE_VARS, to ensure the
variable is reset for every device, and to allow it to be used outside
of the ath79 target.
Signed-off-by: Sander Vanheule <sander@svanheule.net>
BSD based OS have different fixup and doesn't require these header.
Limit these Header to Linux based OS.
Fixes: 36bc306ae6 ("prereq-build: add extra check for elfutils required header")
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
STAGING_DIR may be provided from command line. We currently hardcoded
STAGING_DIR_HOST and STAGING_DIR_HOSTPKG to the default location but we
currently have some relocatable patch that derive the path from
STAGING_DIR.
Fix this and correctly derive STAGING_DIR_HOST and STAGING_DIR_HOSTPKG
from STAGING_DIR.
The intention is to fix inconsistency from the relocatable patch and the
use of STAGING_DIR_HOST that is always hardcoded.
This with a wrong configuration may end up in broken state with some
host tools expecing a PATH from STAGING_DIR and others using library
from the default staging_dir/host path.
To save downstream project the original implementation is saved while
fixing the inconsistency between patch and .mk.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
While testing tools build on an alpine image it was found that with musl
libc some header are missing for elfutils tool.
Add extra prereq-build check to make sure these header are present in
the system to correctly compile host tools.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Add RequireCHeader helper that will try to compile a fake c program with
the requested header included.
This is useful to check if a specific header is present in the system
without checking for the specific path.
This is a generilized version of the current ncurses test.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Some vendors use basic gzip metadata (original filename and timestamp)
to verify valid images, along with the size of it's contents.
Also, add a new device profile variable FACTORY_IMG_NAME
which would be ideal to use with this new recipe.
Signed-off-by: Michael Pratt <mcpratt@pm.me>
Removed upstreamed:
pending-5.15/101-Use-stddefs.h-instead-of-compiler.h.patch[1]
ipq806x/patches-5.15/122-01-clk-qcom-clk-krait-fix-wrong-div2-functions.patch[2]
bcm27xx/patches-5.15/950-0198-drm-fourcc-Add-packed-10bit-YUV-4-2-0-format.patch[3]
Manually rebased:
ramips/patches-5.15/100-PCI-mt7621-Add-MediaTek-MT7621-PCIe-host-controller-.patch[4]
Added patch/backported:
ramips/patches-5.15/107-PCI-mt7621-Add-sentinel-to-quirks-table.patch[5]
All other patches automatically rebased.
1. https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=v5.15.86&id=c160505c9b574b346031fdf2c649d19e7939ca11
2. https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=v5.15.86&id=a051e10bfc6906d29dae7a31f0773f2702edfe1b
3. https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=v5.15.86&id=ec1727f89ecd6f2252c0c75e200058819f7ce47a
4. Quilt gave this output when I applied the patch to rebase it:
% quilt push -f
Applying patch platform/100-PCI-mt7621-Add-MediaTek-MT7621-PCIe-host-controller-.patch
patching file arch/mips/ralink/Kconfig
patching file drivers/pci/controller/Kconfig
patching file drivers/pci/controller/Makefile
patching file drivers/staging/Kconfig
patching file drivers/staging/Makefile
patching file drivers/staging/mt7621-pci/Kconfig
patching file drivers/staging/mt7621-pci/Makefile
patching file drivers/staging/mt7621-pci/TODO
patching file drivers/staging/mt7621-pci/mediatek,mt7621-pci.txt
patching file drivers/staging/mt7621-pci/pci-mt7621.c
Hunk #1 FAILED at 1.
Not deleting file drivers/staging/mt7621-pci/pci-mt7621.c as content differs from patch
1 out of 1 hunk FAILED -- saving rejects to file drivers/staging/mt7621-pci/pci-mt7621.c.rej
patching file drivers/pci/controller/pcie-mt7621.c
Applied patch platform/100-PCI-mt7621-Add-MediaTek-MT7621-PCIe-host-controller-.patch (forced; needs refresh)
Upon inspecting drivers/staging/mt7621-pci/pci-mt7621.c.rej, it seems that
the original patch wants to delete drivers/staging/mt7621-pci/pci-mt7621.c
but upstream's version was not an exact match. I opted to delete that
file.
5. Suggestion by hauke: 19098934f9
"This patch is in upstream kernel, but it was backported to the old
staging driver in kernel 5.15."
Build system: x86_64
Build-tested: bcm2711/RPi4B, filogic/xiaomi_redmi-router-ax6000-ubootmod
Run-tested: bcm2711/RPi4B, filogic/xiaomi_redmi-router-ax6000-ubootmod
Signed-off-by: John Audia <therealgraysky@proton.me>
Several devices provide U-Boot versions with only gzip compressed kernel
support (e.g. Realtek switches). This compression method produces larger
images than lzma. To save space on flash and avoid going the hard way with
lzma-loader we can make use of enhanced gzip tool based on libdeflate
compression library from https://github.com/ebiggers/libdeflate. It
keeps 100% deflate/gzip compatibility while improving compression ratio.
The image can be unpacked by the default inflate routines inside U-Boot.
Signed-off-by: Markus Stockhausen <markus.stockhausen@gmx.de>
[Switched to v1.15 and made it work with cmake]
Signed-off-by: Olliver Schinagl <oliver@schinagl.nl>
Reviewed-by: Robert Marko <robimarko@gmail.com>
Reviewed-by: Rosen Penev <rosenp@gmail.com>
Reviewed-by: Sander Vanheule <sander@svanheule.net>
The libraries libpthread, libdl, libutil, libanl have been integrated
into the libc library in version 2.34. it is not needed to explicitly
link them any more.
Most of the functions have been moved from the librt.so into libc.so
some time ago already.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
There are 2 different CPE IDs on the NVD website:
cpe:/a:arm:trusted_firmware-a
cpe:/o:arm:arm_trusted_firmware
The ID as currently used in trusted-firmware-a.mk does not exist. The
CPE ID using the arm_trusted_firmware product name only lists a few
records for versions 2.2 and 2.3 on the NVD site. The CPE ID using the
trusted_firmware-a product name lists many more records, and actually
has a CVE linked to it. Therefore, use the CPE ID using the
trusted_firmware-a product name.
Fixes: 104d60fe94 ("trusted-firmware-a.mk: add PKG_CPE_ID")
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
No patches affected by this update.
Build system: x86_64
Build-tested: ramips/tplink_archer-a6-v3
Run-tested: ramips/tplink_archer-a6-v3
Signed-off-by: John Audia <therealgraysky@proton.me>
Commit a01d23e75 ("image: always rebuild kernel loaders")
is a step in the right direction, but exposed some issues
and regressions in the makefile.
Some of the files made by device specific COMPILE targets
start with an "append" command (i.e. >> instead of > redirection)
and if the file already exists, the target file is the
input to itself before the first recipe-specified input.
Fixes: a01d23e75 ("image: always rebuild kernel loaders")
Fixes: a7fb589e8 ("image: always rebuild kernel loaders")
Signed-off-by: Michael Pratt <mcpratt@pm.me>
Most/all other tools use the staging dir prefix, gzip should as well.
Signed-off-by: Olliver Schinagl <oliver@schinagl.nl>
Acked-by: Christian Marangi <ansuelsmth@gmail.com>
KERNEL_MAKEOPTS will get expanded when it is used and not when it is
defined in the kernel.mk file now. This fixes problems finding dependent
kernel modules when it is used by a kernel module package.
Without this change the build of packages which depend on other out of
tree modules failed when they used KERNEL_MAKE because some symbols could
not be found. This happened because KERNEL_MAKE_FLAGS which contains a
"if $(__package_mk)" was evaluated where KERNEL_MAKEOPTS was defined
and not when the KERNEL_MAKE was used. For packages which included
kernel.mk before package.mk we saw this problem. One workaround
was to use the correct include order and the other one was to not
use KERNEL_MAKE_FLAGS, but copy its content.
Signed-off-by: Hauke Mehrtens <hmehrtens@maxlinear.com>
Starting from version 2.39 binutils now warns about sections with rwx
permissions. While this is generally desirable it breaks building
ARM TrustedFirmware-A bl2 which treats warnings as errors.
Disable the warning/error for now to fix build.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Call 'mkdir -p $(STAGING_DIR_IMAGE)' before trying to store files in
this potentially non-existing folder.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Manually rebased: ath79/patches-5.10/910-unaligned_access_hacks.patch
All other patches automatically rebased.
Signed-off-by: John Audia <therealgraysky@proton.me>
Instead of manually overriding every cmake package that uses iconv or
gettext's paths, add the prefix in here so that at least FindIconv.cmake
works. Fixes compilation with BUILD_NLS.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
Support defining a per-device loadaddress for the DTB. This is required
for devices which to not align the DTB from the bootloader correctly.
Signed-off-by: David Bauer <mail@david-bauer.net>
unetd always includes $(INCLUDE_DIR)/bpf.mk. This file always checks if
the LLVM version is supported in CLANG_VER_VALID. unetd only needs bpf
when UNETD_VXLAN_SUPPORT is set. It fails when UNETD_VXLAN_SUPPORT is
not set and llvm is not installed.
Fix it by only checking the LLVM version when a LLVM toolchain is
available.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Manually rebased:
bcm53xx/patches-5.10/180-usb-xhci-add-support-for-performing-fake-doorbell.patch
All patches automatically rebased.
Signed-off-by: John Audia <therealgraysky@proton.me>
[Move gro_skip in 680-NET-skip-GRO-for-foreign-MAC-addresses.patch to old position]
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
c3e31b6a9b and 5f8e587240 disable stack validation when the build
host is not running Linux, as the objtool kernel build tool required for
stack validation is not portable to other build host environments. This
was achieved by setting CONFIG_STACK_VALIDATION= in KERNEL_MAKEOPTS, and
by setting SKIP_STACK_VALIDATION=1 in the environment. KERNEL_MAKEOPTS
only has effect for the kernel build, not for external module builds,
but through kernel 5.14, SKIP_STACK_VALIDATION worked to disable this
feature too, so stack validation was disabled for external module builds
as well. Since kernel 0d989ac2c90b, the kernel build no longer considers
SKIP_STACK_VALIDATION, so the feature will be disabled for the kernel
build, but not for external module builds.
When building OpenWrt on a non-Linux build host targeting x86 (the only
target architecture for which OpenWrt enables the kernel
CONFIG_STACK_VALIDATION) and using kernel 5.15 (such as via
CONFIG_TESTING_KERNEL), this caused a build failure during any external
module build, such as kmod-button-hotplug. This manifested as build
errors such as:
make[4]: *** No rule to make target
'.../build_dir/target-x86_64_musl/linux-x86_64/button-hotplug/button-hotplug.o',
needed by
'.../build_dir/target-x86_64_musl/linux-x86_64/button-hotplug/button-hotplug.mod'.
Stop.
Although button-hotplug.c was present, the implicit rule to make
$(obj)/%.o from $(src)/%.c in the kernel's scripts/Makefile.build could
not be satisfied in this case, as it also depends on $(objtool_dep),
non-empty as a result of the failure to propagate disabling of stack
validation to external module builds, in a configuration where it is not
possible to build objtool.
KERNEL_MAKEOPTS is used for just the kernel build itself, while
KERNEL_MAKE_FLAGS is used for both the kernel build and for external
module builds. This restores the ability to build OpenWrt in such
configurations by moving the CONFIG_STACK_VALIDATION= make argument from
KERNEL_MAKEOPTS to KERNEL_MAKE_FLAGS where it is able to affect external
module builds properly.
Note that the kernel's objtool and related configuration have seen a
major overhaul since kernel 5.15, and may need more attention again
after 22922deae13f, in kernel 5.19.
Signed-off-by: Mark Mentovai <mark@mentovai.com>
Introduce a new option in the "Advanced configuration options" to
configure a custom download tool.
By declaring a string in "Use custom download tool" an user can force
what command to use to download package. With the string empty the
default tool used is curl, with wget as a fallback if not available.
download.pl supports 3 tools officially aria2c, curl and wget.
If one of the tool is used in this config, download.pl will use the
default args to make use of them.
If the provided string is different than aria2c, curl or wget, the command
is used as is and the download url will be appended at the end of such command.
While at it also tweak the tool selection logic and chose the tool only
once when the script is called and move aria2c specific variables in the
relevant section.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
To use from the following devices in ath79 target, move edimax-header to
image-commands.mk.
- ELECOM WRC-300GHBK2-I
- ELECOM WRC-1750GHBK2-I/C
Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>
Kernel loaders like the lzma-loader currently don't track changes to
their sources. This can lead to an old version of a loader to be used
when a build tree is not clean between builds.
As the loaders are tiny and the build times are insignificant, simply
force rebuilding them on every build to avoid this problem.
Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
