It was tested that cache scaling currently cause instability problem.
This is probably caused by a latent misconfiguration that cause the L2
cache to be sourced from the wrong source and runs at an unstable freq
compared to the original QSDK fw.
To improve stability while the problem is bisected, disable the devfreq
drivers with minimal perf penality.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Define the kernel crash log storage ramoops/pstore feature
for C2600/AD7200 and add kmod-ramoops to default.
Tested with C2600 only.
Signed-off-by: Edward Matijevic <motolav@gmail.com>
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>
Linksys EA8500 is currently broken after the kernel 5.15 bump. Disable
compiling it by default from buildbot to prevent brick from the user.
Don't mark it as BROKEN to permit user to compile images and permit devs
to bisect the problem with the users.
The current problem with the device is that the switch is not detected
and we can't comunicate with it via MDIO.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Linksys EA8500 is currently broken after the kernel 5.15 bump. Disable
compiling it by default from buildbot to prevent brick from the user.
Don't mark it as BROKEN to permit user to compile images and permit devs
to bisect the problem with the users.
The current problem with the device is that the switch is not detected
and we can't comunicate with it via MDIO.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
This patch was wrongly dropped with the assumption that it was moved to
generic. This wasn't the case and caused the malfunction of the Asrock
G10 router.
Reintroduce it to fix Asrock G10 functionality.
Fixes: 8cc2caed58 ("ipq806x: 5:15: add testing kernel version")
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Zyxel NGB6817 is the only router that use mmc for rootfs. Upstream
kernel dtsi have mmc-ddr-1_8v enabled for sddc1. This is wrong as mmc on
ipq806x is supplied by a fixed 3.3v regulator and can't operate at 1.8v.
This cause the sddc1 to malfunction and cause kernel panic.
In old 5.15 version this was disabled but it was put in addition to many
other changes so it was dropped silently. Restore this patch to fix
working condition of such router.
Fixes: 88bf652 ("ipq806x: 5.15: replace dtsi patches with upstream version")
Fixes: #11000
Tested-by: Hendrik Koerner <koerhen@web.de>
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
In refreshing DTS to the upstream version an unwanted change slipped in
the commit. The ASRock G10 dts got converted to DSA without any support.
Revert this to swconfig driver to restore normal functionality.
Fixes: 88bf652525 ("ipq806x: 5.15: replace dtsi patches with upstream version")
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
We currently ignore ret of the nandc partition parser if unprotected
spare data is true. This is the case for ipq806x nand.
Backport patch that fix this error and correctly handle error from
partition parser.
Fixes: ae6a63bc97 ("ipq806x: 5.15: replace nandc patch with upstream version")
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Add 6.1 tag to upstream patch now that 6.1 got tagged. This permits to
track patch in a better way and directly drop them on kernel bump.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Add multiple patch for krait-cc modernization and multiple fixup for the
driver. Also modify a patch to enable the qsb fixed clock and add pxo to
krait-cc node.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
nand_pins definition is now shipped in ipq8064 dtsi. Rework the
nand_pins definition for wg2600hp3 5.15 files.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
The patch has changed implementation and now the binding has changed.
Replace the old binding with boot-partitions and reimplement the
definition with the new definition.
The new definition is:
<offset1 size1 offset2 size2 offset3 ...>
and now supports sparsed patch.
Also add missing binding in some dts and add the backup boot partition
to the boot-partitions list.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Remove useless spm patch as using the normal qcom,spm compatible is
enough to register it with no clks.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Use a new implementation by using a devfreq driver to scale the shared
cache of the krait cpu cores.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Backport devfreq new cpufreq based PASSIVE governor needed for devfreq
based fab and cache scaling.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Reorganize dtsi patches with upstream version and drop dtsi in 5.15
files.
Also add an additional upstream patch for hwspinlock support.
Refresh all the dts with needed changes.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
This doesn't cause any panic anymore and no regression are observed with
ath10k. Remove this additional patch.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Remove qcom adm Documentation patch that is not needed for the target.
Probably a leftover when the adm bus was added, now merged upstream.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
In preparation for a cleanup of 5.15 patches copy the files dir to 5.10
and 5.15 kernel version.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Removed following upstreamed patch:
* bcm53xx: 081-next-ARM_dts_BCM53015-add-mr26.patch
All other patches automagically rebased.
Signed-off-by: Petr Štetiar <ynezz@true.cz>
Add missing scaling_available_frequencies sysfs entry for dedicated
cpufreq driver.
This sysfs entry is not standard and each cpufreq driver needs to
provide it and declare it in the cpufreq driver struct attr.
Fixes: 5dbbefcbcc ("ipq806x: introduce dedicated krait cpufreq")
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
All targets expect the malta target already activate the CONFIG_GPIOLIB
option. Move it to generic kernel configuration and also activate it for
malta.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The order of LAN ports shown in Luci is reversed compared to what is
written on the case of the device. Fix the order so that they match.
Fixes: #10275
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
This is now built-in, enable so it won't propagate on target configs.
Link: https://lkml.org/lkml/2022/1/3/168
Fixes: 79e7a2552e ("kernel: bump 5.15 to 5.15.44")
Fixes: 0ca9367069 ("kernel: bump 5.10 to 5.10.119")
Signed-off-by: Tomasz Maciej Nowak <tmn505@gmail.com>
(Link to Kernel's commit taht made it built-in,
CRYPTO_LIB_BLAKE2S[_ARM|_X86] as it's selectable, 5.10 backport)
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
Hardware specs:
SoC: Qualcomm IPQ8065 (dual core Cortex-A15)
RAM: 512 MB DDR3
Flash: 256 MB NAND, 32 MB NOR
WiFi: QCA9983 2.4 GHz, QCA9984 5 GHz
Switch: QCA8337
Ethernet: 5x 10/100/1000 Mbit/s
USB: 1x USB 3.0 Type-A
Buttons: WPS, Reset
Power: 12 VDC, 2.5 A
Ethernet ports:
1x WAN: connected to eth2
4x LAN: connected via the switch to eth0 and eth1
(eth0 is disabled in OEM firmware)
MAC addresses (OEM and OpenWrt):
fw_env @ 0x00 d4🆎82:??:??:?a LAN (eth1)
fw_env @ 0x06 d4🆎82:??:??:?b WAN (eth2)
fw_env @ 0x0c d4🆎82:??:??:?c WLAN 2.4 GHz (ath1)
fw_env @ 0x12 d4🆎82:??:??:?d WLAN 5 GHz (ath0)
fw_env @ 0x18 d4🆎82:??:??:?e OEM usage unknown (eth0 in OpenWrt)
OID d4🆎82 is registered to:
ARRIS Group, Inc., 6450 Sequence Drive, San Diego CA 92121, US
More info:
https://openwrt.org/inbox/toh/arris/tr4400_v2
IMPORTANT:
This port requires moving the 'fw_env' partition prior to first boot to
consolidate 70% of the usable space in flash into a contiguous partition.
'fw_env' contains factory-programmed MAC addresses, SSIDs, and passwords.
Its contents must be copied to 'rootfs_1' prior to booting via initramfs.
Note that the stock 'fw_env' partition will be wiped during sysupgrade.
A writable 'stock_fw_env' partition pointing to the old, stock location
is included in the port to help rolling back this change if desired.
Installation:
- Requires serial access and a TFTP server.
- Fully boot stock, press ENTER, type in:
mtd erase /dev/mtd21
dd if=/dev/mtd22 bs=128K count=1 | mtd write - /dev/mtd21
umount /config && ubidetach -m 23 && mtd erase /dev/mtd23
- Reboot and interrupt U-Boot by pressing a key, type in:
set mtdids 'nand0=nand0'
set mtdparts 'mtdparts=nand0:155M@0x6500000(mtd_ubi)'
set bootcmd 'ubi part mtd_ubi && ubi read 0x44000000 kernel && bootm'
env save
- Setup TFTP server serving initramfs image as 'recovery.bin', type in:
set ipaddr 192.168.1.1
set serverip 192.168.1.2
tftpboot recovery.bin && bootm
- Use sysupgrade to install squashfs image.
This port is based on work done by AmadeusGhost <amadeus@jmu.edu.cn>.
Signed-off-by: Rodrigo Balerdi <lanchon@gmail.com>
[add 5.15 changes for 0069-arm-boot-add-dts-files.patch]
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
Converts extraction entries from 11-ath10k-caldata into
nvmem-cells in the individual board's device-tree file.
Same as commit 2047058 ("ipq806x: utilize nvmem-cells
for pre-calibration data")
Signed-off-by: Chukun Pan <amadeus@jmu.edu.cn>
Reviewed-by: Ansuel Smith <ansuelsmth@gmail.com>
There is a mr25h256 spi flash on this machine. From the mtd backup
of the stock firmware, this spi flash is empty.
[ 3.652745] spi_qup 1a280000.spi: IN:block:16, fifo:64, OUT:block:16,
fifo:64
[ 3.653925] spi-nor spi0.0: mr25h256 (32 Kbytes)
Signed-off-by: Chukun Pan <amadeus@jmu.edu.cn>
Fix wrong CPU OPP for ipq8062. Revision of the SoC added an
extra 25mV for every pvs. Also fix the voltage min/max value
that were wrong.
Reviewed-by: Robert Marko robimarko@gmail.com
Signed-off-by: Ansuel Smith <ansuelsmth@gmail.com>
The existing device tree has incorrect definitions for usb3_0 and usb3_1
and the blocks they depend upon: their addresses and interrupts are
swapped. However, their clocks and resets are not. The result is that
the USB blocks are non-functional if only one of them is enabled.
This fix backports the definitions from mainline Linux 5.15 to
OpenWrt's 5.10 dtsi additions. See the relevant mainline code here:
https://github.com/torvalds/linux/blob/v5.17/arch/arm/boot/dts/qcom-ipq8064.dtsi#L1062-L1148
This fix does not break existing ports. But some ports may have enabled
both USB blocks even thought their board only implements one, because
enabling a single USB block would not have worked before this fix.
This means that revisiting all ports of ipq806x devices that implement
a single USB port is advised. This work must be done by maintainers that
can determine which USB block corresponds to the implemented port on
their hardware.
Note that this fix swaps the names of the hardware ports. This is
unfortunate, but will happen anyway when switching to kernel 5.15. Thus,
it is best to do this ASAP, before users get to depend on port names.
It is strongly recommended that this fix is backported to 22.03 before
its release. This will minimize the number of users affected by the port
name swap.
Signed-off-by: Rodrigo Balerdi <lanchon@gmail.com>
This commit add some enabled symbols to generic config.
LTO is only supported by clang compiler and therefore should
be disabled in the generic config instead of duplicating this
symbol in each target. CONFIG_LTO_NONE do this job.
The second group of symbols is enabled by the options available
in the generic config and is therefore added here:
* CONFIG_AF_UNIX_OOB is selected by CONFIG_NET && CONFIG_UNIX,
* CONFIG_BINARY_PRINTF is selected by CONFIG_BPF_SYSCALL,
* CONFIG_NET_SOCK_MSG is selected by CONFIG_BPF_SYSCALL && CONFIG_NET.
The other symbols are disabled and should be in the generic config.
This commit also removes these symbols from subtargets.
Signed-off-by: Aleksander Jan Bajkowski <olek2@wp.pl>
Fix dedicated cpufreq for kernel 5.15 as they changed module
order and now it can happen that cpufreq probe after cache driver.
Also add lock between cache scaling in set_target as it's now required
by opp functions.
Signed-off-by: Ansuel Smith <ansuelsmth@gmail.com>
Now that smem actually free the leaked parts, when
a rootfs partition is detected, the kernel panics as
it try to free the static space allocated for the "ubi"
name. Change the logic and fix the name at the allocate_partition
function to correctly free the space allocated by smem.
Signed-off-by: Ansuel Smith <ansuelsmth@gmail.com>
Refresh patch for 5.15
Rework tweak patch to sync with upstream ipq8064 dtsi and fix
regression introduced.
Rename nand_controller to nand in every dts.
Signed-off-by: Ansuel Smith <ansuelsmth@gmail.com>
Timo Schroeder reported:
"The TP-Link Archer VR2600v is stuck in a boot loop on written
snapshot image. It's able to boot using the snapshot uimage
though, but there ath10k firmware can't be found.
21.02.2 release version doesn't have either problem."
The VR2600v has a 512 byte header at the beginning of the
firmware that needs to be accounted for.
Fixes: f6a01d7f5c ("ipq806x: convert TP-Link Archer VR2600v to denx,uimage")
Reported-by: Timo Schroeder <der.timosch@gmail.com>
References: <https://github.com/openwrt/openwrt/issues/9467>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
Add kmod-ramoops to the default set of device packages in
R7800 and XR500, so that the ramoops kernel crash logs
are provided by default for these routers.
The capability was earlier defined by 97158fe1 and cf346dfa,
but the feature was not yet turned on by default.
The possible kernel crashes are stored into /sys/fs/pstore/*
Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>
Enable support for allocating user space page table entries in high memory [1],
for the targets which support this feature. This saves precious low memory
(permanently mapped, the only type of memory directly accessible by the kernel).
[1] https://www.kernel.org/doc/html/latest/vm/highmem.html
Signed-off-by: Rui Salvaterra <rsalvaterra@gmail.com>
Fixes following warning message during image building process:
Finalizing root filesystem...
root-ipq806x/lib/upgrade/asrock.sh: line 1: /lib/functions.sh: No such file or directory
Enabling boot
root-ipq806x/lib/upgrade/asrock.sh: line 1: /lib/functions.sh: No such file or directory
Enabling bootcount
Fixes#9350
Fixes: 98b86296e6 ("ipq806x: add support for ASRock G10")
Signed-off-by: Petr Štetiar <ynezz@true.cz>
Increase the kernel size from 3 MB to 4 MB for EA8500 and EA7500v1.
* modify the common .dtsi
* modify the kernel size in the image recipes
Define compat-version 2.0 to force factory image usage for sysupgrade.
Add explanation message. Reenable both devices.
As for 4MiB (and not more): Hannu Nyman noted that:
"We have lots of ipq806x devices with 4 MB kernel, so will
need action at that point in future in any case.
(Assuming that the bootloader did not have a 4 MB limit that
has been tested...)"
Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>
(squashed, added 4MiB notice of support in ipq806x)
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
Define the kernel crash log storage ramoops/pstore feature
for R7800 and its sister XR500.
Reference to the ramoops admin guide in upstream Linux:
https://www.kernel.org/doc/html/v5.10/admin-guide/ramoops.html
Tested with R7800.
Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>
The kernel of both images will no longer fit into
the 3072KiB / 3MiB kernel partition:
|Image Name: ARM OpenWrt Linux-5.10.100
|Created: Sat Feb 19 00:11:55 2022
|Image Type: ARM Linux Kernel Image (uncompressed)
|Data Size: 3147140 Bytes = 3073.38 KiB = 3.00 MiB
Disable both targets for now, until a solution is available.
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
Enabled `CONFIG_ALL_KMODS` and ran `make kernel_menuconfig` against
ipq806x to update defconfig.
The removed symbols are in fact present in
target/linux/generic/config-5.10. CONFIG_MDIO_DEVRES
was likely added due to this:
<https://elixir.bootlin.com/linux/v5.10.100/source/drivers/net/phy/Kconfig#L16>
Signed-off-by: John Audia <graysky@archlinux.us>
This device still had the legacy flash partitioning.
This is a problem, because neither the nvmem-cells
for mac-address and calibration. Nor the denx,uimage
mtd-splitter compatible would be picked up.
The patch also changes the node-names of the flash
and partition nodes to hopefully meet all the
current FDT trends.
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
Increase the available flash memory size in Netgear R7800
by repurposing the unused "netgear" partition that is
located after the firmware partition.
Available flash space for kernel+rootfs+overlay increases
by 68 MB from 32 MB to 100 MB.
In a typical build, overlay space increases from 15 to 85,
increasing the package installation possibilities greatly.
Reverting to the OEM firmware is still possible, as the OEM
firmware contains logic to initialise the "netgear" partition
if its contents do not match expectations. In OEM firmware,
"netgear" contains 6 UBI sub-partitions that are defined in
/etc/netgear.cfg and initialisation is done by /etc/preinit
This is based on fb8a578aa7
Signed-off-by: Mike Lothian <mike@fireburn.co.uk>
The recent device-tree modification that added pre-cal
nvmem-cells pushed the device's kernel+dtb over the
allotted 3072k KERNEL_SIZE.
> WARNING: Image file tplink_vr2600v-uImage is too big: 3147214 > 3145728
There was a previous kernel partition size upgrade:
commit 0c967d92b3 ("ipq806x: increase kernel partition size for the TP-Link Archer VR2600v")
It has been seemingly upgraded from a 2048k KERNEL_SIZE in the past.
The commit talks about using the MTD_SPLIT_TPLINK_FW. But looking at
the image make recipe, there is no code that adds a TPLINK header.
So instead the board will use "denx,umimage". This requires
MTD_SPLIT_UIMAGE_FW, but this is present thanks to some NEC devices.
(Maybe the MTD_CONFIG_ARGS can be removed as well? But it could be
there because of the padding at the beginning. This needs testing.)
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
brings back the ath10k QCA9980 wifi nodes to which
it adds ASROCK's wifi calibration data. These are
now provided by the ath10k_firmware.git's board-2.bin.
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
converts extraction entries from 11-ath10k-caldata into
nvmem-cells in the individual board's device-tree file.
The patch also moves previously existing referenced
nvmem-cells data nodes which were placed at the end
back into the partitions node. As well as removing
some duplicated properties from qcom-ipq8065-xr500.dts's
art (the included nighthawk.dtsi defines those already).
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
Commit d284e6ef0f ("treewide: convert mtd-mac-address-increment* to
generic implementation") renamed "mtd-mac-address-increment" property
to "mac-address-increment". Convert remaining usages that have been
added after that.
Fixes: f44e933458 ("ipq806x: provide WiFI mac-addresses from dts")
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
This moves bootargs-append support patch from ipq40xx and ipq806x to
generic. This way we can append additional boot arguments from DTS instead
of only being able to overwrite them.
This is a preparation for kirkwood support of ipTIME NAS1.
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
This board has 512MiB of RAM like the R7800, and the VDSL modem is
attached to the second PCIe port.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
This commit moves the device profiles within the ipq806x/generic
subtarget into their own includable .mk file, to support eventually
having subtargets other than generic.
Signed-off-by: Alex Lewontin <alex.c.lewontin@gmail.com>
Properly declare that the g10 is booting from NAND and define its
correct (larger than on other devices-) boot_pages_size, to prevent
the kernel from constantly falling over missing OOB error correction
for the bootloader.
This patch prevents a constant slew of (bogus) read errors reported
by the kernel and keeping the CPU busy and fixes:
blk_update_request: I/O error, dev mtdblock0, sector 0 op 0x0:(READ) flags 0x80700 phys_seg 4 prio class 0
blk_update_request: I/O error, dev mtdblock0, sector 8 op 0x0:(READ) flags 0x80700 phys_seg 3 prio class 0
blk_update_request: I/O error, dev mtdblock0, sector 16 op 0x0:(READ) flags 0x80700 phys_seg 2 prio class 0
blk_update_request: I/O error, dev mtdblock0, sector 24 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 0
blk_update_request: I/O error, dev mtdblock0, sector 0 op 0x0:(READ) flags 0x0 phys_seg 1 prio class 0
Buffer I/O error on dev mtdblock0, logical block 0, async page read
blk_update_request: I/O error, dev mtdblock0, sector 32 op 0x0:(READ) flags 0x80700 phys_seg 8 prio class 0
blk_update_request: I/O error, dev mtdblock0, sector 40 op 0x0:(READ) flags 0x80700 phys_seg 7 prio class 0
blk_update_request: I/O error, dev mtdblock0, sector 48 op 0x0:(READ) flags 0x80700 phys_seg 6 prio class 0
blk_update_request: I/O error, dev mtdblock0, sector 56 op 0x0:(READ) flags 0x80700 phys_seg 5 prio class 0
blk_update_request: I/O error, dev mtdblock0, sector 64 op 0x0:(READ) flags 0x80700 phys_seg 4 prio class 0
Buffer I/O error on dev mtdblock0, logical block 1, async page read
Buffer I/O error on dev mtdblock1, logical block 0, async page read
Buffer I/O error on dev mtdblock1, logical block 1, async page read
Buffer I/O error on dev mtdblock2, logical block 0, async page read
Buffer I/O error on dev mtdblock2, logical block 1, async page read
Buffer I/O error on dev mtdblock3, logical block 0, async page read
Buffer I/O error on dev mtdblock3, logical block 0, async page read
Buffer I/O error on dev mtdblock4, logical block 0, async page read
Buffer I/O error on dev mtdblock4, logical block 1, async page read
Suggested-by: Ansuel Smith <ansuelsmth@gmail.com>
Signed-off-by: Stefan Lippers-Hollmann <s.l-h@gmx.de>
Each of
- CRYPTO_AEAD2
- CRYPTO_AEAD
- CRYPTO_GF128MUL
- CRYPTO_GHASH
- CRYPTO_HASH2
- CRYPTO_HASH
- CRYPTO_MANAGER2
- CRYPTO_MANAGER
- CRYPTO_NULL2
either directly required for mac80211 crypto support, or directly
selected by such options. Support for the mac80211 crypto was enabled in
the generic config since c7182123b9 ("kernel: make cryptoapi support
needed by mac80211 built-in"). So move the above options from the target
configs to the generic config to make it clear why do we need them.
CC: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Sergey Ryazanov <ryazanov.s.a@gmail.com>
Both CLANG_VERSION and LLD_VERISON are autogenerated runtime
configuration options, so add them to the kernel configuration filter
and remove from generic and per-target configs to keep configs clean.
Signed-off-by: Sergey Ryazanov <ryazanov.s.a@gmail.com>
Revert the SDC "CLK_SET_RATE_GATE" changes to the SDC clock regulator
structures.
See https://elinux.org/images/b/b8/Elc2013_Clement.pdf
> if ((clk->flags & CLK_SET_RATE_GATE) && clk->prepare_count) {
>
> For this particular clock, setting its rate is possible only if the
> clock is ungated (not yet prepared)
This fixes the MMC failing to initialize on newer ZyXEL NBG6817
hardware revisions with Kingston MMC. Older revisions should
hopefully be unaffected.
Check MMC hardware details with:
cd /sys/block/mmcblk0/device/ && \
tail -v cid date name manfid fwrev hwrev oemid rev
Known problematic MMC names (broken before this commit):
* M62704 (dated 12/2018) via myself
* M62704 (dated 11/2018) via Drake Stefani
Known unaffected MMC names (already working without this commit):
* S10004 (dated 12/2015) via slh
Without enabling dynamic debugging, this error manifests in the kernel
hardware serial console as the following:
[ 2.746605] mmc0: error -110 whilst initialising MMC card
[…trimmed other messages…]
[ 2.877832] Waiting for root device /dev/mmcblk0p5...
Enabling Linux dynamic kernel debugging provides additional messages.
For guidance, see the Linux kernel documentation:
https://www.kernel.org/doc/html/latest/admin-guide/dynamic-debug-howto.html
First, enable dynamic debugging in OpenWRT's configuration:
1. Run "make menuconfig"
2. Select "Global build settings --->"
3. Select "Kernel build options --->"
4. Enable "Compile the kernel with dynamic printk" via spacebar
5. Save and exit (arrow key to "Exit" until prompted to save, save)
Alternatively, set "CONFIG_KERNEL_DYNAMIC_DEBUG=y" in your .config.
Then, turn on dynamic debugging at boot:
Modify bootargs in
target/linux/ipq806x/files/arch/arm/boot/dts/qcom-ipq8065-nbg6817.dts
to add…
bootargs = "[…existing bootargs…] dyndbg=\"file drivers/mmc/* +p\" dynamic_debug.verbose=1 loglevel=8";
For example:
chosen {
- bootargs = "rootfstype=squashfs,ext4 rootwait noinitrd fstools_ignore_partname=1";
+ bootargs = "rootfstype=squashfs,ext4 rootwait noinitrd fstools_ignore_partname=1 dyndbg=\"file drivers/mmc/* +p\" dynamic_debug.verbose=1 loglevel=8";
append-rootblock = "root=/dev/mmcblk0p";
Then, compile and flash the resulting build. If you are testing
before this commit on newer MMC hardware, be prepared to recover!
NOTE: If you have hardware serial console access, you don't need to
use TFTP recovery to change the active boot partition.
Reboot to working alternative partition via serial console:
1. Connect to hardware serial console
* See https://openwrt.org/toh/zyxel/nbg6817#serial
2. Interrupt boot at "Hit any key to stop autoboot:"
3. Run "ATSE NBG6817"
4. Copy the result (e.g. "001976FE4B04")
* Changes with **every boot** - can't reuse this
5. On your local system, run
"./zyxel-uboot-password-tool.sh <copied value here>"
* Example: "./zyxel-uboot-password-tool.sh 001976FE4B04"
6. Run the command provided by the password tool
* Example: "ATEN 1,910F129B"
* Changes with **every boot** - can't reuse this
7. Run "ATGU"
* You now have full u-boot shell until next boot - unlocking is
not remembered
8. Run either "run boot_mmc" (for booting partition set "FF") or
"run boot_mmc_1" (for booting partition set "01")
* These commands are not affected by dual-boot partition flags
NOTE: This will NOT set the dual-boot partition flag. You'll need to
fix that manually. The "nbg6817-dualboot" script may help:
https://github.com/pkgadd/nbg6817/blob/master/nbg6817-dualboot
zyxel-uboot-password-tool.sh - sourced from
commit 459c8c9ef8:
ror32() {
echo $(( ($1 >> $2) | (($1 << (32 - $2) & (2**32-1)) ) ))
}
v="0x$1"
a="0x${v:2:6}"
b=$(( a + 0x10F0A563))
c=$(( 0x${v:12:14} & 7 ))
p=$(( $(ror32 $b $c) ^ a ))
printf "ATEN 1,%X\n" $p
Kernel serial console log BEFORE commit with dynamic debug enabled:
[…trimmed…]
[ 3.171343] mmci-pl18x 12400000.sdcc: designer ID = 0x51
[ 3.171397] mmci-pl18x 12400000.sdcc: revision = 0x0
[ 3.175811] mmci-pl18x 12400000.sdcc: clocking block at 96000000 Hz
[ 3.181134] mmci-pl18x 12400000.sdcc: No vqmmc regulator found
[ 3.186788] mmci-pl18x 12400000.sdcc: mmc0: PL180 manf 51 rev0 at 0x12400000 irq 41,0 (pio)
[ 3.192902] mmci-pl18x 12400000.sdcc: DMA channels RX dma1chan1, TX dma1chan2
[ 3.215609] mmc0: clock 0Hz busmode 2 powermode 1 cs 0 Vdd 21 width 1 timing 0
[ 3.227532] mmci-pl18x 12400000.sdcc: Initial signal voltage of 3.3v
[ 3.247518] mmc0: clock 52000000Hz busmode 2 powermode 2 cs 0 Vdd 21 width 1 timing 0
[…trimmed…]
[ 3.997725] mmc0: req done (CMD2): -110: 00000000 00000000 00000000 00000000
[ 4.003631] mmci-pl18x 12400000.sdcc: irq0 (data+cmd) 00000000
[ 4.003659] mmc0: error -110 whilst initialising MMC card
[ 4.016481] mmc0: clock 0Hz busmode 2 powermode 0 cs 0 Vdd 0 width 1 timing 0
Notice how the initial clock is 52 MHz, which is incorrect - MMC
requires negotiation to enable higher speeds.
Kernel serial console log AFTER commit with dynamic debug enabled:
[…trimmed…]
[ 3.168996] mmci-pl18x 12400000.sdcc: designer ID = 0x51
[ 3.169051] mmci-pl18x 12400000.sdcc: revision = 0x0
[ 3.173492] mmci-pl18x 12400000.sdcc: clocking block at 96000000 Hz
[ 3.178808] mmci-pl18x 12400000.sdcc: No vqmmc regulator found
[ 3.184702] mmci-pl18x 12400000.sdcc: mmc0: PL180 manf 51 rev0 at 0x12400000 irq 41,0 (pio)
[ 3.190573] mmci-pl18x 12400000.sdcc: DMA channels RX dma1chan1, TX dma1chan2
[ 3.217873] mmc0: clock 0Hz busmode 2 powermode 1 cs 0 Vdd 21 width 1 timing 0
[ 3.229250] mmci-pl18x 12400000.sdcc: Initial signal voltage of 3.3v
[ 3.249111] mmc0: clock 400000Hz busmode 2 powermode 2 cs 0 Vdd 21 width 1 timing 0
[…trimmed…]
[ 4.392652] mmci-pl18x 12400000.sdcc: irq0 (data+cmd) 00000000
[ 4.392785] mmc0: clock 52000000Hz busmode 2 powermode 2 cs 0 Vdd 21 width 1 timing 1
[ 4.406554] mmc0: starting CMD6 arg 03b70201 flags 0000049d
[…trimmed…]
Now, the MMC properly initializes and later switches to high speed.
Thanks to:
* Ansuel for maintaining/help with the IPQ806x platform, kernel code
* slh for additional debugging and suggestions
* dwfreed for confirming newer MMC details, clock frequency
* robimarko for device driver debug printing help, clock debugging
* Drake for testing and confirmation with their own newer NBG6817
...and anyone else I missed!
Signed-off-by: Shane Synan <digitalcircuit36939@gmail.com>
Tested-by: Shane Synan <digitalcircuit36939@gmail.com>
The purpose of this code seems to be to avoid issues caused
by partially overwriting an existing UBI partition, where some
of the erase counters would be reset but not the unmodified
ones. This problem has been solved in a more generic way by
the UBI EOF marker. This ensures that any old PEBs after the
marker are properly initialized. It is therefore unnecessary
to erase the whole partition before flashing a new OpenWrt
factory image.
Signed-off-by: Bjørn Mork <bjorn@mork.no>
The MR42 and MR52 are two similar IPQ806x based devices from the Cisco
Meraki "Cryptid" series.
MR42 main features:
- IPQ8068 1.4GHz
- 512MB RAM
- 128MB NAND
- 2x QCA9992 (2.4 & 5GHz)
- 1x QCA9889 (2.4 & 5GHz)
- 1x AR8033 PHY
- PoE/AC power
MR52 main features:
- IPQ8068 1.4GHz
- 512MB RAM
- 128MB NAND
- 2x QCA9994 (2.4 & 5GHz)
- 1x QCA9889 (2.4 & 5GHz)
- 2x AR8033 PHYs
- PoE/AC power
(MR42 Only) Installation via diagnostic mode:
If you can successfully complete step 1 then you can continue to install
via this method without having to open the device. Otherwise please use
the standard UART method. Please note that when booting via TFTP, some
Ethernet devices, in particular those on laptops, will not connect in
time, resulting in TFTP boot not succeeding. In this instance it is
advised to connect via a switch.
1. Hold down reset at power on and keep holding, after around 10 seconds
if the orange LED changes behaviour to begin flashing, proceed to
release reset, then press reset two times. Ensure that the LED has
turned blue. Note that flashing will occur on some devices, but it
will not be possible to change the LED colour using the reset button.
In this case it will still be possible to continue with this install
method.
2. Set your IP to 192.168.1.250. Set up a TFTP server serving
mr42_u-boot.mbn and
openwrt-ipq806x-generic-meraki_mr42-initramfs-fit-uImage.itb, obtained
from [1].
3. Use telnet and connect to 192.168.1.1. Run the following commands to
install u-boot. Note that all these commands are critical, an error
will likely render the device unusable.
Option 3.1:
If you are sure you have set up the TFTP server correctly you can
run this script on the device. This will download and flash the
u-boot image immediately:
`/etc/update_uboot.sh 192.168.1.250 mr42_u-boot.mbn`
Once completed successfully, power off the device.
Option 3.2:
If you are unsure the TFTP server is correctly set up you can
obtain the image and flash manually:
3.2.1. `cd /tmp`
3.2.2. `tftp-hpa 192.168.1.250 -m binary -c get mr42_u-boot.mbn`
3.2.3. Confirm file has downloaded correctly by comparing the
md5sum:
`md5sum mr42_u-boot.mbn`
3.2.4. The following are the required commands to write the image.
`echo 1 > /sys/devices/platform/msm_nand/boot_layout
mtd erase /dev/mtd1
nandwrite -pam /dev/mtd1 mr42_u-boot.mbn
echo 0 > /sys/devices/platform/msm_nand/boot_layout`
Important: You must observe the output of the `nandwrite`
command. Look for the following to verify writing is occurring:
`Writing data to block 0 at offset 0x0
Writing data to block 1 at offset 0x20000
Writing data to block 2 at offset 0x40000`
If you do not see this then do not power off the device. Check
your previous commands and that mr42_u-boot.mbn was downloaded
correctly. Once you are sure the image has been written you
can proceed to power off the device.
4. Hold the reset button and power on the device. This will immediately
begin downloading the appropriate initramfs image and boot into it.
Note: If the device does not download the initramfs, this is likely
due to the interface not being brought up in time. Changing Ethernet
source to a router or switch will likely resolve this. You can also
try manually setting the link speed to 10Mb/s Half-Duplex.
5. Once a solid white LED is displayed on the device, continue to the
UART installation method, step 6.
Standard installation via UART - MR42 & MR52
1. Disassemble the device and connect a UART header. The header pinout
is as follows:
1 - 3.3v
2 - TXD
3 - RXD
4 - GND
Important: You should only connect TXD, RXD and GND. Connecting
3.3v may damage the device.
2. Set your IP to 192.168.1.250. Set up a TFTP server serving
openwrt-ipq806x-generic-meraki_(mr42|mr52)-initramfs-fit-uImage.itb.
Separately obtain the respective sysupgrade image.
3. Run the following commands, preferably from a Linux host. The
mentioned files, including ubootwrite.py and u-boot images, can be
obtained from [1].
`python ubootwrite.py --write=(mr42|mr52)_u-boot.bin`
The default for "--serial" option is /dev/ttyUSB0.
4. Power on the device. The ubootwrite script will upload the image to
the device and launch it. The second stage u-boot will in turn load
the initramfs image by TFTP, provided the TFTP server is running
correctly. This process will take about 13 minutes. Once a solid
white LED is displayed, the image has successfully finished
loading. Note: If the image does not load via TFTP, try again with
the Ethernet link to 10Mb/s Half-Duplex.
5. (MR42 only) Do not connect over the network. Instead connect over
the UART using minicom or similar tool. To replace u-boot with
the network enabled version, please run the following commands.
Note that in the provided initramfs images, the u-boot.mbn file
is located in /root:
If you have not used the provided initramfs, you must ensure you
are using an image with "boot_layout" ECC configuration enabled in
the Kernel. This will be version 5.10 or higher. If you do not do
this correctly the device will be bricked.
`insmod mtd-rw i_want_a_brick=1
mtd erase /dev/mtd8
nandwrite -pam /dev/mtd8 /root/mr42_u-boot.mbn`
After running nandwrite, ensure you observe the following output:
`Writing data to block 0 at offset 0x0
Writing data to block 1 at offset 0x20000
Writing data to block 2 at offset 0x40000`
6. (Optional) If you have no further use for the Meraki OS, you can
remove all other UBI volumes on ubi0 (mtd11), including diagnostic1,
part.old, storage and part.safe. You must not remove the ubi1 ART
partition (mtd12).
`for i in diagnostic1 part.old storage part.safe ; do
ubirmvol /dev/ubi0 -N $i
done`
7. Proceed to flash the sysupgrade image via luci, or else download or
scp the image to /tmp and use the sysupgrade command.
[1] The mentioned images and ubootwrite.py script can be found in this repo:
https://github.com/clayface/openwrt-cryptid
[2] The modified u-boot sources for the MR42 and MR52 are available:
https://github.com/clayface/U-boot-MR52-20200629
Signed-off-by: Matthew Hagan <mnhagan88@gmail.com>
Add backports of the following patches:
"net: stmmac: explicitly deassert GMAC_AHB_RESET" and
"ARM: dts: qcom: add ahb reset to ipq806x-gmac"
Required for Meraki MR42/MR52.
Signed-off-by: Matthew Hagan <mnhagan88@gmail.com>
Rather than having separate patches for each GSBI node added, this patch
consolidates the existing GSBI1 patch into
083-ipq8064-dtsi-additions.patch. In addition, GSBI6 and GSBI7 I2C nodes,
required for the MR42 and MR52 respectively, are added.
Signed-off-by: Matthew Hagan <mnhagan88@gmail.com>
