upstream commit 802b7c06adc7 ("ARM: cns3xxx: Convert PCI to use generic config accessors")
reimplemented cns3xxx_pci_read_config() using pci_generic_config_read32(),
which preserved the property of only doing 32-bit reads.
It also replaced cns3xxx_pci_write_config() with pci_generic_config_write(),
so it changed writes from always being 32 bits to being the actual size,
which works just fine.
Due to:
- The documentation does not mention that only 32 bit access is allowed.
- Writes are already executed using the actual size
- Extensive testing shows that 8b, 16b and 32b reads work as intended
It makes perfectly sense to also swap 32 bit reading in favor of actual size.
Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
Originally, cns3xxx used it's own functions for mapping, reading and writing registers.
Upstream commit 802b7c06adc7 ("ARM: cns3xxx: Convert PCI to use generic config accessors")
removed the internal PCI config write function in favor of the generic one:
cns3xxx_pci_write_config() --> pci_generic_config_write()
cns3xxx_pci_write_config() expected aligned addresses, being produced by cns3xxx_pci_map_bus()
while the generic one pci_generic_config_write() actually expects the real address
as both the function and hardware are capable of byte-aligned writes.
This currently leads to pci_generic_config_write() writing
to the wrong registers on some ocasions.
First issue seen due to this:
- driver ath9k gets loaded
- The driver wants to write value 0xA8 to register PCI_LATENCY_TIMER, located at 0x0D
- cns3xxx_pci_map_bus() aligns the address to 0x0C
- pci_generic_config_write() effectively writes 0xA8 into register 0x0C (CACHE_LINE_SIZE)
This seems to cause some slight instability when certain PCI devices are used.
Another issue example caused by this this is the PCI bus numbering,
where the primary bus is higher than the secondary, which is impossible.
Before:
00:00.0 PCI bridge: Cavium, Inc. Device 3400 (rev 01) (prog-if 00 [Normal decode])
Flags: bus master, fast devsel, latency 0, IRQ 255
Bus: primary=02, secondary=01, subordinate=ff, sec-latency=0
After fix:
00:00.0 PCI bridge: Cavium, Inc. Device 3400 (rev 01) (prog-if 00 [Normal decode])
Flags: bus master, fast devsel, latency 0, IRQ 255
Bus: primary=00, secondary=01, subordinate=02, sec-latency=0
And very likely some more ..
Fix all by omitting the alignment being done in the mapping function.
Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
The following patch was integrated upstream:
* target/linux/generic/patches-4.4/005-ext4-fix-check-to-prevent-initializing-reserved-inod.patch
This fixes tries to work around the following security problems:
* CVE-2018-3620 L1 Terminal Fault OS, SMM related aspects
* CVE-2018-3646 L1 Terminal Fault Virtualization related aspects
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The broken check would detect a newly generated root filesystem as corrupt
under certain circumstances, in some cases actually currupting the it while
trying to handle the error condition.
This is a regression introduced in kernel 4.4.140. The 4.14.y stable series
has already received this fix, while it is still pending for 4.4.y and
4.9.y.
Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
target/linux/ar71xx/patches-4.4/103-MIPS-ath79-fix-register-address-in-ath79_ddr_wb_flus.patch
has been applied upstream; the two deleted brcm2708 patches have been
useless even before (as the second one only reverted the first one).
Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
These two patches:
target/linux/ar71xx/patches-4.4/403-mtd_fix_cfi_cmdset_0002_status_check.patch
target/linux/ramips/patches-4.4/0036-mtd-fix-cfi-cmdset-0002-erase-status-check.patch
are replaced by upstream commit 242dbd2b3df ("mtd: cfi_cmdset_0002:
Change erase functions to check chip good only")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Fixes "unregister_netdevice: waiting for lo to become free. Usage count = 1"
messages which started appearing since the update to 4.4.103. That
problem was exposed by upstream commit 76da0704507bb ("ipv6: only call
ip6_route_dev_notify() once for NETDEV_UNREGISTER") backported to 4.4.x
branch in 2417da3f4d6bc.
Fixes: 2b664499cd622 ("kernel: bump 4.4 to 4.4.103 for 17.01")
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
(cherry picked from commit 58f7b5b96c301176d639540df4723c798af2a999)
This is important fix for flash parsing in some corner cases. In case
of TRX subpartition with rootfs being aligned to the flash block size it
was incorrectly registered twice. Detecting & registering it as a
standalone partition was resulting in an incorrect "firmware" partition
size and possibly broken sysupgrade.
It wasn't noticed before because "rootfs" alignment depends on a kernel
size. It can happen though - depending on the configuration and the
kernel size.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
(cherry picked from commit f5195e72c0fcf2949f7d6296a5db081eb58f8e32)
Switching from kernel 4.4.120 to 4.4.124 introduced a regression in
the genirq code. It was caused by a commit 9d0273bb1c4b6 ("genirq: Use
irqd_get_trigger_type to compare the trigger type for shared IRQs").
On bcm53xx it breaks serial console and results in a flood of:
[ 22.078829] genirq: Flags mismatch irq 18. 00000080 (serial) vs. 00000080 (gpio)
[ 22.086432] genirq: Flags mismatch irq 18. 00000080 (serial) vs. 00000080 (gpio)
[ 22.601150] genirq: Flags mismatch irq 18. 00000080 (serial) vs. 00000080 (gpio)
[ 22.608845] genirq: Flags mismatch irq 18. 00000080 (serial) vs. 00000080 (gpio)
Later in the upstream "linux-4.4.y" branch that commit was reverted and
it was followed by a 4.4.126 release. Until we switch from 4.4.124 to
4.4.126 (or newer), let's backport that reverting commit.
Fixes: bed0ee7cbfaa5 ("Kernel: bump 4.4 to 4.4.124 for 17.01")
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
At least on some devices, LEDs don't work anymore since kernel 4.4.120.
Revert the broken change.
See also: https://www.spinics.net/lists/stable/msg223656.html
Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
This bumps the 4.4. kernel in LEDE 17.01 to 4.4.116.
More Meltdown & Spectre mitigation.
* Refresh patches.
* Refresh x86/config for RETPOLINE.
* Deleted 8049-PCI-layerscape-Add-fsl-ls2085a-pcie-compatible-ID.patch (accepted upstream)
* Deleted 8050-PCI-layerscape-Fix-MSG-TLP-drop-setting.patch (accepted upstream)
* 650-pppoe_header_pad.patch does not apply anymore (code was replaced).
Bumps from 4.4.113 to 4.4.115 were handled by Kevin Darbyshire-Bryant.
Compile-tested on: ar71xx, ramips/mt7621, x86/64
Run-tested on: ar71xx, ramips/mt7621, x86/64
Signed-off-by: Stijn Segers <foss@volatilesystems.org>
Backport support for raw-ip mode including all known fixes afterwards.
Newer LTE modems only tend to support this mode, which was only
introduced in kernel 4.5.
Also backport support for the Quectel EC2x LTE modem series which is
a very popular device.
No custom changes were needed in order to apply these patches.
Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
Bump 4.4 to 4.4.107 and refreshed all patches.
Made the following patch for Mediatek and Oxnas compatible with kernel 4.4.107:
0072-mtd-backport-v4.7-0day-patches-from-Boris.patch
Compile-tested: ar71xx
Run-tested: ar71xx
Signed-off-by: Etienne Haarsma <bladeoner112@gmail.com>
Reviewed-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
Tested-by: Rosen Penev <rosenp@gmail.com>
It is used by pretty much every target
Signed-off-by: Felix Fietkau <nbd@nbd.name>
(backported from commit b47fd7656336162360ebf66147326763ddae3f8d)
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
PHY core treats any positive return value as the auto-negotiation done
indication. Since we do not actually check any device register in this
callback then update it to return positive value with a neutral meaning
instead of the register flag to avoid confusing for future readers.
Signed-off-by: Sergey Ryazanov <ryazanov.s.a@gmail.com>
(cherry picked from commit 2cc61e6e8e951b1480a024fdb66bc539aa163bbc)
The Marvel 88E6060 switch has an MDIO interface, but does not emulate
regular PHY behavior for the host. The network core can not detect using
the generic code, whether the connection via the attached PHY can be
used or not. The PHY's state machine is stuck in a state of
auto-negotiation and does not go any further so the Ethernet interface
of the router stay forever in the not-runing state.
Fix this issue by implementing the aneg_done callback to be able to
inform the network core that the Ethernet interface link to which the
switch is connected can be marked as RUNNING.
Signed-off-by: Sergey Ryazanov <ryazanov.s.a@gmail.com>
(cherry picked from commit 913b2290caad437bb523d6f1ee5a70707f7f7ef4)
Keeping it in base-files was resulting in adding it to the base-files
package. This file is meant to be included manually for initramfs
images only.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
(cherry picked from commit f6433eede79db4f6276b68e815bb17a9364292c7)
RGMII RX delay setting needs to be always specified for AR8337 to
avoid port 5 RX hang on high traffic / flood conditions.
Also, the HOL registers that set per-port and per-packet-priority
buffer sizes are updated with the reduced values suggested by the
QCA switch team.
Finally, AR8327 reserved register fixups are disabled for the AR8337.
This patch is adapted from the Code Aurora QSDK, but with magic
values mapped to proper defines.
Signed-off-by: Vittorio Gambaletta <openwrt@vittgam.net>
(cherry picked from commit 967b6be118e3217e8d6a28df9c615d3255e7b1ae)
Direct-IO support has to be enabled for the release build anyway, so
this hack is not worth keeping
Signed-off-by: Felix Fietkau <nbd@nbd.name>
(backported from commit 0b7ed65cec8084bb98ae0e2758b7aca6c447cd4b)
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
This PHY requires some extra programming to work reliably with all
devices. Backport upstream fix for it.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
4.4.80+ contains 71a165f6397df07a06ce643de5c2dbae29bd3cfb, 4.9.41+ contains
6c78197e4a69c19e61dfe904fdc661b2aee8ec20 which are all backports of upstream
commit 0878fff1f42c18e448ab5b8b4f6a3eb32365b5b6 ("net: phy: Do not perform
software reset for Generic PHY").
Our local patch is no longer needed, all this patch was doing was utilizing
gen10g_soft_reset which does nothing either, so just keep the code unchanged.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
This backports fixes for setting of_node and making it possible to read
extra info from DT. This was partially fixed by:
[PATCH] leds: leds-gpio: Set of_node for created LED devices
but it didn't work during initialization.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
Refresh patches.
Minor update 704-phy-no-genphy-soft-reset.patch which was partially
accepted upstream.
Compile-tested on ar71xx.
Runtime-tested on ar71xx.
Fixes the following vulnerabilities:
- CVE-2017-7533 (4.4.80)
- CVE-2017-1000111 (4.4.82)
- CVE-2017-1000112 (4.4.82)
Signed-off-by: Kevin Darbyshire-Bryant <kevin@darbyshire-bryant.me.uk>
