staging/noltari/config, branch master

kernel: set default for KERNEL_BPF_KPROBE_OVERRIDE

2023-03-05T16:36:01Z

Without the default value this still causes a missing symbol. Disable by default as it depends on FUNCTION_ERROR_INJECTION, which is disabled in the generic config and we don't have a build symbol to enable that. Fixes: 500c37c56ff6 ("kernel: add missing symbol") Signed-off-by: Stijn Tintel

build: fix for sourcing targets image config installed via feeds

2023-02-26T21:22:48Z

Sourcing of image/Config.in will not happen When a target is installed from target/linux/feeds/ Signed-off-by: Prasun Maiti Acked-by: Petr Štetiar

kernel: set default values for ARM low level debugging symbols

2023-02-08T15:21:05Z

Set default values for KERNEL_DEBUG_LL and KERNEL_DEBUG_LL_UART_NONE again as both of these symbols are non visible if KERNEL_EARLY_PRINTK is not selected and KConfig wont write their value to .config. This usually is the intended behaviour, but in OpenWrt we are relying on the KConfig to set these and disable the debug console settings that multiple targets like mvebu have set in their kernel config. This was the behaviour before removing all of the "default n" settings as KConfig by default considers symbols disabled but they are not visible anymore and thus their value is not set in .config and build system then later does not override the values from target kernel config. So, to restore the behaviour to the previous one lets a default value for KERNEL_DEBUG_LL and KERNEL_DEBUG_LL_UART_NONE. Fixes: 8bc72ea7be39 ("treewide: strip useless default n Kconfig lines") Tested-by: Georgi Valkov Signed-off-by: Robert Marko

treewide: strip useless `default n` Kconfig lines

2023-02-03T11:50:15Z

Kconfig docs say: > The default value deliberately defaults to 'n' in order to avoid > bloating the build. Apply this rule everywhere, to avoid more cloning of bad examples Signed-off-by: Tony Butler

kernel: make it possible for packages to select page pool support

2023-01-29T09:08:21Z

Will be used by the next mt76 update Signed-off-by: Felix Fietkau

build: fix incomplete initramfs compression options

2023-01-17T00:16:35Z

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 [remove blocking dependencies for separate ramdisk, fix lzop options] Signed-off-by: Daniel Golle

ipq807x: add Qualcomm Atheros IPQ807x target

2023-01-16T11:42:23Z

Qualcomm Atheros IPQ807x is a modern WiSoC featuring: * Quad Core ARMv8 Cortex A-53 * @ 2.2 GHz (IPQ8072A/4A/6A/8A) Codename Hawkeye * @ 1.4 GHz (IPQ8070A/1A) Codename Acorn * Dual Band simultaneaous IEEE 802.11ax * 5G: 8x8/80 or 4x4/160MHz (IPQ8074A/8A) * 5G: 4x4/80 or 2x2/160MHz (IPQ8071A/2A/6A) * 5G: 2x2/80MHz (IPQ8070A) * 2G: 4x4/40MHz (IPQ8072A/4A/6A/8A) * 2G: 2x2/40MHz (IPQ8070A/1A) * 1x PSGMII via QCA8072/5 (Max 5x 1GbE ports) * 2x SGMII/USXGMII (1/2.5/5/10 GbE) on Hawkeye * 2x SGMII/USXGMII (1/2.5/5 GbE) on Acorn * DDR3L/4 32/16 bit up to 2400MT/s * SDIO 3.0/SD card 3.0/eMMC 5.1 * Dual USB 3.0 * One PCIe Gen2.1 and one PCIe Gen3.0 port (Single lane) * Parallel NAND (ONFI)/LCD * 6x QUP BLSP SPI/I2C/UART * I2S, PCM, and TDMA * HW PWM * 1.8V configurable GPIO * Companion PMP8074 PMIC via SPMI (GPIOS, RTC etc) Note that only v2 SOC models aka the ones ending with A suffix are supported, v1 models do not comply to the final 802.11ax and have lower clocks, lack the Gen3 PCIe etc. SoC itself has two UBI32 cores for the NSS offloading system, however currently no offloading is supported. Signed-off-by: Robert Marko

build: images: squashfs: add help, fix description

2022-12-06T22:12:14Z

add help text for `TARGET_SQUASHFS_BLOCK_SIZE` to match the only valid settings accepted by `mksquashfs4` ("block size not power of two or not between 4096 and 1Mbyte") thus for this setting in "KB", the set: `4, 8, 16, 32, 64, 128, 256, 512, 1024` replace `squashfs-lzma` with `squashfs` in the description for `TARGET_ROOTFS_SQUASHFS` because it has various compressions, and not just lzma as it did in the past cosmetic change with no functional effect Signed-off-by: Tony Butler

build: config: cleanup impossible dependency logic

2022-12-03T19:45:55Z

some config `depends on` lines contained outdated kernel version checks that can no longer happen and had become non-operational; clean them up cosmetic change with no functional effect Signed-off-by: Tony Butler

Revert "build: fix for sourcing target image config a when target is being replaced"

2022-11-09T21:28:53Z

This reverts commit ce1346a8fc0ce9640a4ecbc37bc1686a25c1165d. Seems to cause buildbot compilation to fail and require more testing. Signed-off-by: Christian Marangi