Linux 4.1 is also affected from the dma issue, so remove the dma
proprties there as well.
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 46599
Make the device name accessible for any build commands instead of
forcing targets to define it themselves.
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 46597
Instead of letting each target define it themselves, create on in
include/image.mk and let the targets use it.
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 46596
Instead of each target defining it the same, move the KDIR_TMP
definition to include/image.mk. In addition Image/Build/SysupgradeNAND
already requires KDIR_TMP to be set, so it makes sense to have it
globally defined.
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 46592
Previous patch 6f2905eeb6ce5ddec8d12d677e1f377a940b537b enabled ADM in
the kernel, which causes a kernel panic when accessing the SPI flash.
As a workaround, We'll disable DMA for the flash for now. It was not
enabled previously anyway so we'll just leave it as is.
Reported-by: Jonas Gorski <jogo@openwrt.org>
Signed-off-by: Mathieu Olivari <mathieu@codeaurora.org>
SVN-Revision: 46587
We got check for nvram_len placed too early and the code never tried
reading CFE variables used on WGT634U. This is a regression introduced
in r45942 and reported in #20291.
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
SVN-Revision: 46584
Previous patch set backported the recently posted NAND flash driver to
3.18 and 4.1 kernel. This patch now enables it in the kernel config.
There is no change to the partition layout and init yet. But the NAND
flash can be seen in the mtd list on an AP148:
root@OpenWrt:/# cat /proc/mtd
dev: size erasesize name
mtd0: 10000000 00020000 "qcom-nandc"
...
Signed-off-by: Mathieu Olivari <mathieu@codeaurora.org>
SVN-Revision: 46569
These patches add support for ipq806x NAND flash controller. Most of
these are cherry-picked & backported from LKML:
*https://lkml.org/lkml/2015/8/3/16
This patch just modifies the kernel code, but doesn't change the config.
It should be harmless.
Signed-off-by: Mathieu Olivari <mathieu@codeaurora.org>
SVN-Revision: 46568
These are cherry-picked & backported from LKML:
*https://lkml.org/lkml/2015/3/17/19
They are enabled on both 3.18 and 4.1 kernel. Patches 150 to 154 are
applying changes merged since 3.18; they enable mechanisms used by the
ADM driver.
ADM engine is used by the NAND controller, so it is necessary to
bring-up NAND flash support.
Signed-off-by: Mathieu Olivari <mathieu@codeaurora.org>
SVN-Revision: 46567
Variables dependend on JFFS2_BLOCKSIZE and NANDBLOCK_SIZE are used
for template generation, so need to be present before inclusion of
image.mk in target image Makefiles.
So move all declarations to before any includes.
Fixes: r42878 ("image.mk: clean up and parallelize mkfs calls")
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 46564
It looks like RAC flushes cause random corruption(?) when the
second thread is set as default.
Fixes#20160.
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 46563
BCM6318: add support for Plusnet / Sagem 2704N (V1)
Signed-off-by: Matt Goring <matt.goring@googlemail.com>
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 46562
We forgot to enable the stmmac driver for 4.1 kernel, so ethernet
interfaces don't show-up on this kernel.
Signed-off-by: Mathieu Olivari <mathieu@codeaurora.org>
SVN-Revision: 46559
Old bootloader (same ones which have DT disabled) don't perform any PCIe
initialization. The consequence is a freeze during PCIe bring-up on
these old u-boot. Same kernel with a newer bootloaders works fine as
they contain the corresponding PCIe init code.
In this change, we'll add the missing init and make sure the kernel
doesn't rely on some preexisting init to get PCIe to work. That includes
the following changes:
*GPIOs: set function & drive strength
*Clocks: add init code for aux & ref clocks
*PCIe driver: additional init of the hardware controller
Tested 3.18 and 4.1 on an AP148 with bootloader branch 0.0.1
Signed-off-by: Mathieu Olivari <mathieu@codeaurora.org>
SVN-Revision: 46557
PCIe controller nodes are numbers 0/1/2 in the chipset dtsi file, but
the pinmux nodes are numbers 1/2/3. We'll make it consistent by changing
the pinmux numbering to match the controller's one.
Signed-off-by: Mathieu Olivari <mathieu@codeaurora.org>
SVN-Revision: 46556
Certain AP148 platforms (and derivative) use bootloaders which did not
have DT enabled.
In order to support these old platforms, we'll now make the following
modifications:
*explicitely add the memory node in the AP148 DT: this used to be added
by new u-boot through a run-time patch mechanism. We'll now add it
explicitely so it works on boots which don't support that feature. New
boots will have the node twice, the second one will be ignored.
*add the zImage generation next to the FIT image for AP148.
Other platforms using non-DT enabled bootloaders may want to leverage
this zImage code to generate their own firmare as well.
Signed-off-by: Mathieu Olivari <mathieu@codeaurora.org>
SVN-Revision: 46555
Fixes a 100% cpu usage issue if using dhcp-script.
Signed-off-by: Kevin Darbyshire-Bryant <kevin@darbyshire-bryant.me.uk>
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 46550
The generic kernel cmdline parser ignores argv[0], this caused a
regression for all lzma-loader based boards with linux 4.1
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
SVN-Revision: 46544
When stack protector support is disabled in libc (always the case for
!musl), gcc assumes that it needs to use __stack_chk_guard for the stack
canary.
This causes kernel build errors, because the kernel is only set up to
handle TLS stack canaries.
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
SVN-Revision: 46543
