summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2015-08-03rework kernel module infrastructureWaldemar Brodkorb
Instead of maintaining mk/modules.mk which defines compilations of related kernel modules to pack together into a single package, follow an automatic approach: For every kernel module found in the modules installation directory, create a single package. There are a few caveats to cover: === Module Loading Order === Upon bootup, module loading is ordered based on the number-prefixed files in /etc/modules.d/. The correct number was previously managed in mk/modules.mk on a per-collection basis. The new approach is to have levels which modules are to be assigned to. Level 0 contains modules with no dependencies at all. Level 1 contains modules which have only level 0 dependencies, and so on. This information is determined at compile-time by make-module-ipkgs.sh. === Module Installation to Target RootFS === Since module packages are created automatically from the modules the script finds, ADK build system has no knowledge about the connection between what the user has selected in menuconfig and the actual module packages. Therefore the earlier approach to install selected packages into rootfs does not hold anymore. Instead, use wildcards to find all packages in firmware directory prefixed by 'kmod-' and install them all (hopefully doing the right thing). === Kernel Version === KERNEL_VERSION now contains KERNEL_RELEASE already By creating a localversion file, make KERNEL_RELEASE part of the kernel's version number (so KERNEL_VERSION is correct in most situations) Signed-off-by: Phil Sutter <phil@nwl.cc>
2015-08-03update firmware to latestWaldemar Brodkorb
2015-08-03simplifyWaldemar Brodkorb
2015-08-03cleanupWaldemar Brodkorb
2015-08-03for 4.1 kernel we need older gpu libWaldemar Brodkorb
2015-08-02separate serial drivers, add builtin dtb. still no shellWaldemar Brodkorb
2015-08-02remove non-lts kernelWaldemar Brodkorb
2015-08-02virtio-blk for qemu needs pci legacyWaldemar Brodkorb
2015-08-02update to uClibc-ng 1.0.5Waldemar Brodkorb
2015-08-02add uclibc.config for meta, disable gcc honour copts for this old gccWaldemar Brodkorb
2015-08-02add basic support for qemu-metagWaldemar Brodkorb
2015-08-02set suffix for OABIWaldemar Brodkorb
2015-08-02default to 4.1.3Waldemar Brodkorb
2015-08-01cleanup, add general patches for 4.1.3Waldemar Brodkorb
2015-08-01remove testkernelWaldemar Brodkorb
2015-08-01update to 4.1.3, which will be the base for stable branchWaldemar Brodkorb
2015-08-01be more quietWaldemar Brodkorb
2015-08-01remove openelec test patchWaldemar Brodkorb
2015-08-01update to latestWaldemar Brodkorb
2015-07-30added ADK_HOST_BUILD_U_BOOT to bananapro default configMartin Thomas
Signed-off-by: Martin Thomas <mthomas@conet.de>
2015-07-30fix linking problem for sure nowWaldemar Brodkorb
2015-07-30enable kuser_helper and MMC driver for banana piWaldemar Brodkorb
2015-07-30style changesWaldemar Brodkorb
2015-07-30uImage is mandatory, add LOADADDRWaldemar Brodkorb
2015-07-29update gcc/binutils to latest synopsis github versionWaldemar Brodkorb
2015-07-29update python2 to latest upstreamWaldemar Brodkorb
2015-07-29remove patch, does not apply anymoreWaldemar Brodkorb
2015-07-28enable neon for ARM and threadsWaldemar Brodkorb
2015-07-27handle git kernels in a better wayWaldemar Brodkorb
2015-07-27add new binutils versionWaldemar Brodkorb
2015-07-27add some basic support for emcraft kinetisWaldemar Brodkorb
2015-07-27fix linking error reported by mtWaldemar Brodkorb
2015-07-27mikrotik-rb4xx: update kernel patchesPhil Sutter
* fix conflict in existing patch * add build fix for newer binutils Signed-off-by: Phil Sutter <phil@nwl.cc>
2015-07-26mikrotik-rb4xx: update kernel patchesPhil Sutter
- add zboot support - workaround link state bug Signed-off-by: Phil Sutter <phil@nwl.cc>
2015-07-26update u-boot to latest stable releaseWaldemar Brodkorb
2015-07-26u-boot for BananaPro pls. check target/arm/kernel/banana-pro if ↵Martin Thomas
CONFIG_SERIAL_8250=y CONFIG_SERIAL_8250_CONSOLE=y CONFIG_SERIAL_8250_SUNXI=y make sense together Signed-off-by: Martin Thomas <mthomas@hamtam.de>
2015-07-26package/asterisk: add link to /tmp for priv-callerintrosPhil Sutter
Signed-off-by: Phil Sutter <phil@nwl.cc>
2015-07-26package/asterisk: fix and improve downgrade to version 1.8Phil Sutter
Signed-off-by: Phil Sutter <phil@nwl.cc>
2015-07-26package/asterisk: update to version 1.8.28.2Phil Sutter
Signed-off-by: Phil Sutter <phil@nwl.cc>
2015-07-26Revert "update to latest upstream"Phil Sutter
This reverts commit 23c39ed2658699a56627e2609908348d4a1f9b36. Conflicts: package/asterisk/Makefile Signed-off-by: Phil Sutter <phil@nwl.cc>
2015-07-26added host build for dnsmasqJoerg Seitter
Signed-off-by: Joerg Seitter <joerg.seitter@stz-gab.de>
2015-07-26scripts/ipkg-build: fix error messagePhil Sutter
Signed-off-by: Phil Sutter <phil@nwl.cc>
2015-07-26package/libnetfilter_cttimeout: drop ltmain.sh patchPhil Sutter
The file to be patched simply does not exist.
2015-07-26package/base-files: wireless: support appending to hostapd.confPhil Sutter
This is useful for adding additional (virtual) access points. In my setup, /etc/hostapd.conf.post contains: | bss=ath6 | ssid=IPv6XS4all In /etc/network/interfaces, there is a standard entry for ath6 which defines the IP address. Maybe this could be implemented in a cleaner way by introducing a 'wireless-master' setting which points to the real wireless interface (wlan0 in my case), so ifup knows which hostapd.conf file to add the settings to.
2015-07-26package/base-files: wireless: allow setting pairwise algorithmPhil Sutter
Not perfect, since allowing WPA+WPA2 means setting wpa_pairwise as well as rsn_pairwise, but I do not want to add unnecessary complexity to interface configuration. This is relevant for my X40 with ipw2200, which bugs for CCMP pairwise cipher.
2015-07-26package/maradns: split into subpackagesPhil Sutter
For the nameserver alone, maradns is enough. fetchzone/zoneserver are only used on authoritative/secondary servers, and askmara is just a query tool probably not needed on embedded devices. Sadly, maradns itself does not support recursion anymore. For that job, deadwood was created. Note that all these daemons don't properly fork into background. Instead, the author maintains a daemonizing tool called 'duende', which itself seems to have it's flaws, either. In practice, the deadwood init script made rcS hang unless it was set to 'DAEMON' in rc.conf. Oh well. Signed-off-by: Phil Sutter <phil@nwl.cc>
2015-07-26package/maradns: update to version 2.0.09Phil Sutter
Additionally: - Also minimize necessary patching. Still a bit hacky, but a lot less intrusive at least. - Allow enabling IPv6 support. - Make sure /etc/maradns/logger exists, necessary for chroot operation Signed-off-by: Phil Sutter <phil@nwl.cc>
2015-07-26add msse for x86/x86_64 buildsWaldemar Brodkorb
2015-07-26fix build for ARM and othersWaldemar Brodkorb
2015-07-25minor cleanups, fix missing dependenciesWaldemar Brodkorb