Olof Johansson [Thu, 7 Jul 2016 05:23:27 +0000 (22:23 -0700)]
Merge tag 'v4.8-rockchip-dts64-1' of git://git.kernel.org/pub/scm/linux/kernel/git/mmind/linux-rockchip into next/dt64
The rk3399 gets support for its emmc controller as well as thermal,
i2c and core io-domain nodes and some reasonable default rates
for core clocks. The rk3368 also gets io-domains for its r88 board
as well as a small fix for the gic's memory regions.
* tag 'v4.8-rockchip-dts64-1' of git://git.kernel.org/pub/scm/linux/kernel/git/mmind/linux-rockchip:
arm64: dts: rockchip: add ap_pwroff and ddrio_pwroff pins for rk3399
arm64: dts: rockchip: Provide emmcclk to PHY for rk3399
arm64: dts: rockchip: Add soc-ctl-syscon to sdhci for rk3399
arm64: dts: rockchip: fixes the gic400 2nd region size for rk3368
arm64: dts: rockchip: add i2c nodes for rk3399
arm64: dts: rockchip: add thermal nodes for rk3399 SoCs
arm64: dts: rockchip: add rk3399 io-domain core nodes
arm64: dts: rockchip: add rk3368-r88 iodomains
arm64: dts: rockchip: add rk3368 io-domain core nodes
arm64: dts: rockchip: make rk3368 grf syscons simple-mfds
arm64: dts: rockchip: enable eMMC for rk3399 EVB
arm64: dts: rockchip: add sdhci/emmc for rk3399
arm64: dts: rockchip: make rk3399's grf a "simple-mfd"
arm64: dts: rockchip: assign default rates for core rk3399 clocks
Olof Johansson [Wed, 6 Jul 2016 04:10:09 +0000 (21:10 -0700)]
Merge tag 'imx-dt64-4.8' of git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux into next/dt64
The Freescale arm64 device tree updates for 4.8:
- Update address-cells and reg properties of cpu nodes, considering
MPIDR_EL1[63:32] bits are not used for CPUs identification on ls1043a
and ls2080a
- Adds the cache nodes and next-level-cache property for ls1043a and
ls2080a to get cacheinfo work on these platforms
- Add dma-coherent for ls1043a PCI nodes to utilize the hardware
capability on data coherency
- Add dis_rxdet_inp3_quirk property for USB3 device to disable rx
detection in P3 PHY mode
* tag 'imx-dt64-4.8' of git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux:
arm64: dts: ls2080a: Add cache nodes for cacheinfo support
arm64: dts: ls1043a: Add cache nodes for cacheinfo support
arm64: dts: ls1043a: Add 'dma-coherent' for ls1043a PCI nodes
bindings: PCI: layerscape: Add 'dma-coherent' property
arm64: dts: ls1043a: Add dis_rxdet_inp3_quirk property to USB3 node
arm64: dts: ls2080a: Add dis_rxdet_inp3_quirk property to USB3 node
arm64: dts: fsl: Update address-cells and reg properties of cpu nodes
Olof Johansson [Tue, 5 Jul 2016 04:33:31 +0000 (21:33 -0700)]
Merge tag 'xgene-dts-for-v4.8-part1' of https://github.com/AppliedMicro/xgene-next into next/dt64
First part of X-Gene DTS changes queued for v4.8
The changes include:
+ 2 clean-up and style-fix patches from Bjorn
+ Correct timer interrupt polarity for X-Gene 2
+ Remove unused qmlclk node on X-Gene 1
* tag 'xgene-dts-for-v4.8-part1' of https://github.com/AppliedMicro/xgene-next:
arm64: dts: apm: Remove unused qmlclk node on X-Gene 1
arm64: dts: apm: Fix timer interrupt polarity for X-Gene 2 SoC
arm64: dts: apm: Remove leading '0x' from unit addresses
arm64: dts: apm: Use lowercase consistently for hex constants
Linus Walleij [Thu, 23 Jun 2016 23:06:04 +0000 (01:06 +0200)]
arm64: dts: hikey: name the GPIO lines
This names the GPIO lines on the HiKey board in accordance with
the 96Board Specification for especially the Low Speed External
Connector: "GPIO-A" thru "GPIO-L".
This will make these line names reflect through to userspace
so that they can easily be identified and used with the new
character device ABI.
Some care has been taken to name all lines, not just those used
by the external connectors, also lines that are muxed into some
other function than GPIO: these are named "[FOO]" so that users
can see with lsgpio what all lines are used for.
Cc: devicetree@vger.kernel.org Cc: John Stultz <john.stultz@linaro.org> Cc: Rob Herring <robh@kernel.org> Cc: David Mandala <david.mandala@linaro.org> Cc: Haojian Zhuang <haojian.zhuang@linaro.org> Cc: Wei Xu <xuwei5@hisilicon.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Wei Xu <xuwei5@hisilicon.com>
Douglas Anderson [Tue, 14 Jun 2016 20:21:11 +0000 (13:21 -0700)]
arm64: dts: rockchip: add ap_pwroff and ddrio_pwroff pins for rk3399
There are two sleep related pins on rk3399: ap_pwroff and ddrio_pwroff.
Let's add the definition of these two pins to rk3399's main dtsi file so
that boards can use them.
These two pins are similar to the global_pwroff and ddrio_pwroff pins in
rk3288 and are expected to be used in the same way: boards will likely
want to configure these pinctrl settings in their global pinctrl hog
list.
Note that on rk3288 there were two additional pins in the "sleep"
section: "ddr0_retention" and "ddr1_retention". On rk3288 designs these
pins appeared to actually route from rk3288 back to rk3288. Presumably
on rk3399 this is simply not needed since the pins don't appear to exist
there.
Signed-off-by: Douglas Anderson <dianders@chromium.org> Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Douglas Anderson [Mon, 20 Jun 2016 17:56:54 +0000 (10:56 -0700)]
arm64: dts: rockchip: Provide emmcclk to PHY for rk3399
Previous changes in this series allowed exposing the card clock from the
rk3399 SDHCI device and allowed consuming the card clock in the rk3399
eMMC PHY. Hook things up in the main rk3399 dtsi file.
Signed-off-by: Douglas Anderson <dianders@chromium.org> Tested-by: Heiko Stuebner <heiko@sntech.de> Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Bjorn Helgaas [Tue, 14 Jun 2016 13:00:20 +0000 (08:00 -0500)]
arm64: dts: apm: Use lowercase consistently for hex constants
The convention in these files is to use lowercase for "0x" prefixes and for
the hex constants themselves, but a few changes didn't follow that
convention, which makes the file annoying to read.
Use lowercase consistently for the hex constants. No functional change
intended.
Obviously, the region size should be greater than 0x1000.
So we should make sure to include the GICC_IDR since the kernel will access
it in some cases.
Fixes: b790c2cab5ca ("arm64: dts: add Rockchip rk3368 core dtsi and board dts for the r88 board") Signed-off-by: Caesar Wang <wxt@rock-chips.com> Reviewed-by: Shawn Lin <shawn.lin@rock-chips.com> Cc: stable@vger.kernel.org
[added Fixes and stable-cc] Signed-off-by: Heiko Stuebner <heiko@sntech.de>
David Wu [Mon, 16 May 2016 20:09:31 +0000 (13:09 -0700)]
arm64: dts: rockchip: add i2c nodes for rk3399
We've got 9 (count em!) i2c controllers on rk3399, some of which are in
the PMU power domain and some of which are normal peripherals. Add them
all to the main rk3399 dtsi file so future patches can turn them on in
the board dts files.
Note: by default we try to set the i2c clock rate to 200 MHz so that we
can achieve good i2c functional clock rates. 200 MHz gives us the
ability to make very close to 100 kHz / 400 kHz / 1 MHz rates. If
boards want to tune clock rates further they can always override.
Possibly boards could want to tune this if:
- they wanted to save an infinitesimal amount of power and they knew
their i2c bus was slow anyway. Since we gate the functional clock
when the i2c bus is not active, power savings would only be while i2c
transfers were happening and probably won't be very big anyway.
- they wanted to eek out a bit more speed by carefully tuning the source
clock to make divisions work out perfectly, accounting for the rise /
fall time measured on an actual board.
Note also that we still request 200 MHz for the PMU i2c busses even
though we expect that we won't make that exactly (currently PPLL is 676
MHz which gives us 169 MHz).
Signed-off-by: David Wu <david.wu@rock-chips.com> Signed-off-by: Jianqun Xu <jay.xu@rock-chips.com>
[dianders: wrote desc; put in assigned-clocks; reordered nodes] Signed-off-by: Douglas Anderson <dianders@chromium.org> Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Liu Gang [Tue, 7 Jun 2016 06:55:46 +0000 (14:55 +0800)]
arm64: dts: ls1043a: Add 'dma-coherent' for ls1043a PCI nodes
The 'dma-coherent' indicates that the hardware IP block can ensure
the coherency of the data transferred from/to the IP block. This
can avoid the software cache flush/invalid actions, and improve
the performance significantly.
The PCI IP block of ls1043a has this capability, so adding this
feature to improve the PCI performance.
Signed-off-by: Liu Gang <Gang.Liu@nxp.com> Signed-off-by: Shawn Guo <shawnguo@kernel.org>
The 'dma-coherent' indicates that the hardware IP block can ensure
the coherency of the data transferred from/to the IP block. This
can avoid the software cache flush/invalid actions, and improve
the performance significantly.
The PCI IP block of ls1043a has this capability, so adding
this feature to improve the PCI performance.
Signed-off-by: Liu Gang <Gang.Liu@nxp.com> Acked-by: Rob Herring <robh@kernel.org> Signed-off-by: Shawn Guo <shawnguo@kernel.org>
Masahiro Yamada [Tue, 14 Jun 2016 03:01:42 +0000 (12:01 +0900)]
arm64: dts: uniphier: change cpu-release-address
At first, 256 byte of the head of DRAM space was reserved for some
reasons. However, as the progress of development, it turned out
unnecessary, and it was never used in the end. Move the CPU release
address to leave no space.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Signed-off-by: Olof Johansson <olof@lixom.net>
Olof Johansson [Mon, 13 Jun 2016 22:29:12 +0000 (15:29 -0700)]
Merge tag 'renesas-arm64-dt-for-v4.8' of git://git.kernel.org/pub/scm/linux/kernel/git/horms/renesas into next/dt64
Renesas ARM64 Based SoC DT Updates for v4.8
* Fix W=1 dtc warnings and other cleanups
* Enable watchdog timer
* Enable DMA for I2C
* Increase the size of GIC-400 mapped registers: be nicer to hypervisors
* Support RTS/CTS hardware flow control
* tag 'renesas-arm64-dt-for-v4.8' of git://git.kernel.org/pub/scm/linux/kernel/git/horms/renesas:
arm64: dts: r8a7795: Drop 0x from unit address of gic
arm64: dts: salvator-x: Fix W=1 dtc warnings
arm64: dts: r8a7795: Fix W=1 dtc warnings
arm64: dts: r8a7795: Use SYSC "always-on" PM Domain for RWDT node
arm64: dts: salvator-x: Enable watchdog timer
arm64: dts: r8a7795: Add RWDT node
arm64: dts: r8a7795: enable DMA for I2C
arm64: dts: r8a7795: Increase the size of GIC-400 mapped registers
arm64: dts: salvator-x: SCIF1 supports RTS/CTS hardware flow control
Georgi Djakov [Thu, 4 Feb 2016 12:53:22 +0000 (14:53 +0200)]
arm64: dts: apq8016-sbc: Add DT node for the uSD SDHC interface
Add the necessary properties to enable the SD-card on db410c boards.
Signed-off-by: Georgi Djakov <georgi.djakov@linaro.org> Tested-by: Kevin Hilman <khilman@baylibre.com> Signed-off-by: Andy Gross <andy.gross@linaro.org>
Alison Wang [Mon, 9 May 2016 09:06:15 +0000 (17:06 +0800)]
arm64: dts: fsl: Update address-cells and reg properties of cpu nodes
MPIDR_EL1[63:32] value is equal to 0 for the CPUs of the LS1043A and
LS2080A SoCs. The ARM CPU binding allows #address-cells to be set to 1,
since MPIDR_EL1[63:32] bits are not used for CPUs identification. Update
the #address-cells and reg properties accordingly.
Signed-off-by: Alison Wang <alison.wang@nxp.com> Signed-off-by: Shawn Guo <shawnguo@kernel.org>
Caesar Wang [Wed, 25 May 2016 07:39:35 +0000 (15:39 +0800)]
arm64: dts: rockchip: add thermal nodes for rk3399 SoCs
This adds thermal zone and tsadc nodes to rk3399 dtsi, rk3399 thermal
data is including the cpu and gpu sensor zone node.
The thermal zone node is the node containing all the required info
for describing a thermal zone, including its cooling device bindings.
The thermal zone node must contain, apart from its own properties, one
sub-node containing trip nodes and one sub-node containing all the zone
cooling maps.
The following is the parameter is introduced:
* polling-delay:
The maximum number of milliseconds to wait between polls
* polling-delay-passive:
The maximum number of milliseconds to wait between polls when performing
passive cooling.
* trips:
A sub-node which is a container of only trip point nodes required to
describe the thermal zone.
* cooling-maps:
A sub-node which is a container of only cooling device map nodes, used to
describe the relation between trips and cooling devices.
* cooling-device:
A phandle of a cooling device with its specifier, referring to which
cooling device is used in this cooling specifier binding. In the cooling
specifier, the first cell is the minimum cooling state and the second cell
is the maximum cooling state used in this map.
Signed-off-by: Caesar Wang <wxt@rock-chips.com> Signed-off-by: Heiko Stuebner <heiko@sntech.de>
CK Hu [Fri, 3 Jun 2016 14:59:29 +0000 (16:59 +0200)]
arm64: dts: mt8173: Add display subsystem related nodes
This patch adds the device nodes for the DISP function blocks
comprising the display subsystem.
Signed-off-by: CK Hu <ck.hu@mediatek.com> Signed-off-by: Cawa Cheng <cawa.cheng@mediatek.com> Signed-off-by: Jie Qiu <jie.qiu@mediatek.com> Signed-off-by: Daniel Kurtz <djkurtz@chromium.org> Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de> Signed-off-by: Matthias Brugger <matthias.bgg@gmail.com>
Carlo Caione [Sun, 3 Apr 2016 17:14:41 +0000 (19:14 +0200)]
ARM64: dts: amlogic: Add hiu and periphs buses
Add two new buses in the DTS: hiu and periphs buses.
In the Amlogic S905/GXBB SoC several devices (clock / eth / pin
controllers, etc...) are mapped under these two buses. Add them in the
DT before starting to add new devices.
Signed-off-by: Carlo Caione <carlo@endlessm.com> Signed-off-by: Kevin Hilman <khilman@baylibre.com>
Jon Mason [Wed, 11 May 2016 22:56:08 +0000 (18:56 -0400)]
arm64: dts: NS2: Add all of the UARTs
Add all of the UARTs present on NS2 and enable them in the SVK device
tree file. Also, do some magic to make sure that uart3 is discovered as
ttyS0 (as that is the console UART).
Signed-off-by: Jon Mason <jonmason@broadcom.com> Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Heiko Stuebner [Mon, 1 Feb 2016 21:09:03 +0000 (22:09 +0100)]
arm64: dts: rockchip: make rk3368 grf syscons simple-mfds
The general register files do contain a lot of separate functions and
while some really are only registers with a lot of different 1-bit
settings, there are also a lot of them containing some bigger function
blocks. To be able to define these as sub-devices, make them simple-mfds.
Signed-off-by: Heiko Stuebner <heiko@sntech.de> Tested-by: David Wu <david.wu@rock-chips.com>
Brian Norris [Fri, 13 May 2016 22:12:04 +0000 (15:12 -0700)]
arm64: dts: rockchip: enable eMMC for rk3399 EVB
Rockchip's rk3399 evaluation board has eMMC. Let's enable the
newly-added nodes.
Signed-off-by: Brian Norris <briannorris@chromium.org> Reviewed-by: Douglas Anderson <dianders@chromium.org> Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Brian Norris [Fri, 13 May 2016 22:12:03 +0000 (15:12 -0700)]
arm64: dts: rockchip: add sdhci/emmc for rk3399
Add description for the SDHCI v5.1 eMMC controller on rk3399. Fix it to
200 MHz, to support all supported timing modes.
Note that 'rockchip,rk3399-sdhci-5.1' is not documented; we presumably
have a compliant Arasan controller, but let's have a rockchip property
as the canonical backup/precautionary measure. Per Heiko's previous
suggestion, let's not clutter the arasan doc with it.
Signed-off-by: Brian Norris <briannorris@chromium.org> Reviewed-by: Shawn Lin <shawn.lin@rock-chips.com> Reviewed-by: Douglas Anderson <dianders@chromium.org> Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Brian Norris [Fri, 13 May 2016 22:12:02 +0000 (15:12 -0700)]
arm64: dts: rockchip: make rk3399's grf a "simple-mfd"
Per the examples in
Documentation/devicetree/bindings/phy/rockchip-emmc-phy.txt, we need the
grf node to be a simple-mfd in order to properly enumerate child devices
like our eMMC PHY.
Signed-off-by: Brian Norris <briannorris@chromium.org> Reviewed-by: Douglas Anderson <dianders@chromium.org>
[directly mimic for the pmugrf, which will need the same change later
and there is no need to pollute commit history with another patch] Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Xing Zheng [Fri, 13 May 2016 20:50:18 +0000 (13:50 -0700)]
arm64: dts: rockchip: assign default rates for core rk3399 clocks
These clocks are all core clocks used by many blocks/peripherals, many
of whose drivers don't set their clock rates at all. Let's assign
reasonable default clock rates for these core clocks, so that these
peripherals get something reasonable by default, and also so that if
child devices want to select a clock rate themselves, their muxes have
some reasonable parent clock rates to branch off of (rather than just
the boot-time defaults).
This helps the eMMC PHY, for one, to get a reasonable ACLK rate.
Signed-off-by: Xing Zheng <zhengxing@rock-chips.com> Signed-off-by: Brian Norris <briannorris@chromium.org> Reviewed-by: Douglas Anderson <dianders@chromium.org> Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Warning (unit_address_vs_reg): Node /regulator@1 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /regulator@2 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /regulator@3 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /regulator@4 has a unit name, but no reg property
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be> Signed-off-by: Simon Horman <horms+renesas@verge.net.au>
Warning (unit_address_vs_reg): Node /cache-controller@0 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /cache-controller@1 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,dvc/dvc@0 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,dvc/dvc@1 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,src/src@0 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,src/src@1 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,src/src@2 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,src/src@3 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,src/src@4 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,src/src@5 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,src/src@6 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,src/src@7 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,src/src@8 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,src/src@9 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,ssi/ssi@0 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,ssi/ssi@1 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,ssi/ssi@2 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,ssi/ssi@3 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,ssi/ssi@4 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,ssi/ssi@5 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,ssi/ssi@6 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,ssi/ssi@7 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,ssi/ssi@8 has a unit name, but no reg property
Warning (unit_address_vs_reg): Node /soc/sound@ec500000/rcar_sound,ssi/ssi@9 has a unit name, but no reg property
Move the cache-controller nodes under the cpus node, and make their unit
names and reg properties match the MPIDR values.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be> Signed-off-by: Simon Horman <horms+renesas@verge.net.au>
arm64: dts: salvator-x: SCIF1 supports RTS/CTS hardware flow control
On the Salvator-X development board, the RTS and CTS pins of debug
serial-1 port SCIF1 are wired to the CP2102 Serial-USB bridge. Reflect
this in the DTS by adding the "uart-has-rtscts" property to the scif1
device node.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be> Signed-off-by: Simon Horman <horms+renesas@verge.net.au>
George Spelvin [Sun, 29 May 2016 05:26:41 +0000 (01:26 -0400)]
Rename other copy of hash_string to hashlen_string
The original name was simply hash_string(), but that conflicted with a
function with that name in drivers/base/power/trace.c, and I decided
that calling it "hashlen_" was better anyway.
But you have to do it in two places.
[ This caused build errors for architectures that don't define
CONFIG_DCACHE_WORD_ACCESS - Linus ]
Mikulas Patocka [Tue, 24 May 2016 20:49:18 +0000 (22:49 +0200)]
hpfs: implement the show_options method
The HPFS filesystem used generic_show_options to produce string that is
displayed in /proc/mounts. However, there is a problem that the options
may disappear after remount. If we mount the filesystem with option1
and then remount it with option2, /proc/mounts should show both option1
and option2, however it only shows option2 because the whole option
string is replaced with replace_mount_options in hpfs_remount_fs.
To fix this bug, implement the hpfs_show_options function that prints
options that are currently selected.
Mikulas Patocka [Tue, 24 May 2016 20:48:33 +0000 (22:48 +0200)]
affs: fix remount failure when there are no options changed
Commit c8f33d0bec99 ("affs: kstrdup() memory handling") checks if the
kstrdup function returns NULL due to out-of-memory condition.
However, if we are remounting a filesystem with no change to
filesystem-specific options, the parameter data is NULL. In this case,
kstrdup returns NULL (because it was passed NULL parameter), although no
out of memory condition exists. The mount syscall then fails with
ENOMEM.
This patch fixes the bug. We fail with ENOMEM only if data is non-NULL.
The patch also changes the call to replace_mount_options - if we didn't
pass any filesystem-specific options, we don't call
replace_mount_options (thus we don't erase existing reported options).
Mikulas Patocka [Tue, 24 May 2016 20:47:00 +0000 (22:47 +0200)]
hpfs: fix remount failure when there are no options changed
Commit ce657611baf9 ("hpfs: kstrdup() out of memory handling") checks if
the kstrdup function returns NULL due to out-of-memory condition.
However, if we are remounting a filesystem with no change to
filesystem-specific options, the parameter data is NULL. In this case,
kstrdup returns NULL (because it was passed NULL parameter), although no
out of memory condition exists. The mount syscall then fails with
ENOMEM.
This patch fixes the bug. We fail with ENOMEM only if data is non-NULL.
The patch also changes the call to replace_mount_options - if we didn't
pass any filesystem-specific options, we don't call
replace_mount_options (thus we don't erase existing reported options).
Fixes: ce657611baf9 ("hpfs: kstrdup() out of memory handling") Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Cc: stable@vger.kernel.org Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
VDSO:
- Build microMIPS VDSO for microMIPS kernels.
- Fix aliasing warning by building with `-fno-strict-aliasing' for
debugging but also tracing them might result in recursion.
Misc:
- Add missing FROZEN hotplug notifier transitions.
- Fix clk binding example for varioius PIC32 devices.
- Fix cpu interrupt controller node-names in the DT files.
- Fix XPA CPU feature separation.
- Fix write_gc0_* macros when writing zero.
- Add inline asm encoding helpers.
- Add missing VZ accessor microMIPS encodings.
- Fix little endian microMIPS MSA encodings.
- Add 64-bit HTW fields and fix its configuration.
- Fix sigreturn via VDSO on microMIPS kernel.
- Lots of typo fixes.
- Add definitions of SegCtl registers and use them"
Linus Torvalds [Sat, 28 May 2016 23:15:25 +0000 (16:15 -0700)]
Merge branch 'hash' of git://ftp.sciencehorizons.net/linux
Pull string hash improvements from George Spelvin:
"This series does several related things:
- Makes the dcache hash (fs/namei.c) useful for general kernel use.
(Thanks to Bruce for noticing the zero-length corner case)
- Converts the string hashes in <linux/sunrpc/svcauth.h> to use the
above.
- Avoids 64-bit multiplies in hash_64() on 32-bit platforms. Two
32-bit multiplies will do well enough.
- Rids the world of the bad hash multipliers in hash_32.
This finishes the job started in commit 689de1d6ca95 ("Minimal
fix-up of bad hashing behavior of hash_64()")
The vast majority of Linux architectures have hardware support for
32x32-bit multiply and so derive no benefit from "simplified"
multipliers.
The few processors that do not (68000, h8/300 and some models of
Microblaze) have arch-specific implementations added. Those
patches are last in the series.
- Overhauls the dcache hash mixing.
The patch in commit 0fed3ac866ea ("namei: Improve hash mixing if
CONFIG_DCACHE_WORD_ACCESS") was an off-the-cuff suggestion.
Replaced with a much more careful design that's simultaneously
faster and better. (My own invention, as there was noting suitable
in the literature I could find. Comments welcome!)
- Modify the hash_name() loop to skip the initial HASH_MIX(). This
would let us salt the hash if we ever wanted to.
- Sort out partial_name_hash().
The hash function is declared as using a long state, even though
it's truncated to 32 bits at the end and the extra internal state
contributes nothing to the result. And some callers do odd things:
- fs/hfs/string.c only allocates 32 bits of state
- fs/hfsplus/unicode.c uses it to hash 16-bit unicode symbols not bytes
- Modify bytemask_from_count to handle inputs of 1..sizeof(long)
rather than 0..sizeof(long)-1. This would simplify users other
than full_name_hash"
Special thanks to Bruce Fields for testing and finding bugs in v1. (I
learned some humbling lessons about "obviously correct" code.)
On the arch-specific front, the m68k assembly has been tested in a
standalone test harness, I've been in contact with the Microblaze
maintainers who mostly don't care, as the hardware multiplier is never
omitted in real-world applications, and I haven't heard anything from
the H8/300 world"
* 'hash' of git://ftp.sciencehorizons.net/linux:
h8300: Add <asm/hash.h>
microblaze: Add <asm/hash.h>
m68k: Add <asm/hash.h>
<linux/hash.h>: Add support for architecture-specific functions
fs/namei.c: Improve dcache hash function
Eliminate bad hash multipliers from hash_32() and hash_64()
Change hash_64() return value to 32 bits
<linux/sunrpc/svcauth.h>: Define hash_str() in terms of hashlen_string()
fs/namei.c: Add hashlen_string() function
Pull out string hash to <linux/stringhash.h>
George Spelvin [Wed, 25 May 2016 18:19:49 +0000 (14:19 -0400)]
h8300: Add <asm/hash.h>
This will improve the performance of hash_32() and hash_64(), but due
to complete lack of multi-bit shift instructions on H8, performance will
still be bad in surrounding code.
Designing H8-specific hash algorithms to work around that is a separate
project. (But if the maintainers would like to get in touch...)
Signed-off-by: George Spelvin <linux@sciencehorizons.net> Cc: Yoshinori Sato <ysato@users.sourceforge.jp> Cc: uclinux-h8-devel@lists.sourceforge.jp
George Spelvin [Wed, 25 May 2016 15:06:09 +0000 (11:06 -0400)]
microblaze: Add <asm/hash.h>
Microblaze is an FPGA soft core that can be configured various ways.
If it is configured without a multiplier, the standard __hash_32()
will require a call to __mulsi3, which is a slow software loop.
Instead, use a shift-and-add sequence for the constant multiply.
GCC knows how to do this, but it's not as clever as some.
Signed-off-by: George Spelvin <linux@sciencehorizons.net> Cc: Alistair Francis <alistair.francis@xilinx.com> Cc: Michal Simek <michal.simek@xilinx.com>
George Spelvin [Fri, 27 May 2016 02:11:51 +0000 (22:11 -0400)]
<linux/hash.h>: Add support for architecture-specific functions
This is just the infrastructure; there are no users yet.
This is modelled on CONFIG_ARCH_RANDOM; a CONFIG_ symbol declares
the existence of <asm/hash.h>.
That file may define its own versions of various functions, and define
HAVE_* symbols (no CONFIG_ prefix!) to suppress the generic ones.
Included is a self-test (in lib/test_hash.c) that verifies the basics.
It is NOT in general required that the arch-specific functions compute
the same thing as the generic, but if a HAVE_* symbol is defined with
the value 1, then equality is tested.
Signed-off-by: George Spelvin <linux@sciencehorizons.net> Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Greg Ungerer <gerg@linux-m68k.org> Cc: Andreas Schwab <schwab@linux-m68k.org> Cc: Philippe De Muyter <phdm@macq.eu> Cc: linux-m68k@lists.linux-m68k.org Cc: Alistair Francis <alistai@xilinx.com> Cc: Michal Simek <michal.simek@xilinx.com> Cc: Yoshinori Sato <ysato@users.sourceforge.jp> Cc: uclinux-h8-devel@lists.sourceforge.jp
George Spelvin [Mon, 23 May 2016 11:43:58 +0000 (07:43 -0400)]
fs/namei.c: Improve dcache hash function
Patch 0fed3ac866 improved the hash mixing, but the function is slower
than necessary; there's a 7-instruction dependency chain (10 on x86)
each loop iteration.
Word-at-a-time access is a very tight loop (which is good, because
link_path_walk() is one of the hottest code paths in the entire kernel),
and the hash mixing function must not have a longer latency to avoid
slowing it down.
There do not appear to be any published fast hash functions that:
1) Operate on the input a word at a time, and
2) Don't need to know the length of the input beforehand, and
3) Have a single iterated mixing function, not needing conditional
branches or unrolling to distinguish different loop iterations.
One of the algorithms which comes closest is Yann Collet's xxHash, but
that's two dependent multiplies per word, which is too much.
The key insights in this design are:
1) Barring expensive ops like multiplies, to diffuse one input bit
across 64 bits of hash state takes at least log2(64) = 6 sequentially
dependent instructions. That is more cycles than we'd like.
2) An operation like "hash ^= hash << 13" requires a second temporary
register anyway, and on a 2-operand machine like x86, it's three
instructions.
3) A better use of a second register is to hold a two-word hash state.
With careful design, no temporaries are needed at all, so it doesn't
increase register pressure. And this gets rid of register copying
on 2-operand machines, so the code is smaller and faster.
4) Using two words of state weakens the requirement for one-round mixing;
we now have two rounds of mixing before cancellation is possible.
5) A two-word hash state also allows operations on both halves to be
done in parallel, so on a superscalar processor we get more mixing
in fewer cycles.
I ended up using a mixing function inspired by the ChaCha and Speck
round functions. It is 6 simple instructions and 3 cycles per iteration
(assuming multiply by 9 can be done by an "lea" instruction):
x ^= *input++;
y ^= x; x = ROL(x, K1);
x += y; y = ROL(y, K2);
y *= 9;
Not only is this reversible, two consecutive rounds are reversible:
if you are given the initial and final states, but not the intermediate
state, it is possible to compute both input words. This means that at
least 3 words of input are required to create a collision.
(It also has the property, used by hash_name() to avoid a branch, that
it hashes all-zero to all-zero.)
The rotate constants K1 and K2 were found by experiment. The search took
a sample of random initial states (I used 1023) and considered the effect
of flipping each of the 64 input bits on each of the 128 output bits two
rounds later. Each of the 8192 pairs can be considered a biased coin, and
adding up the Shannon entropy of all of them produces a score.
The best-scoring shifts also did well in other tests (flipping bits in y,
trying 3 or 4 rounds of mixing, flipping all 64*63/2 pairs of input bits),
so the choice was made with the additional constraint that the sum of the
shifts is odd and not too close to the word size.
The final state is then folded into a 32-bit hash value by a less carefully
optimized multiply-based scheme. This also has to be fast, as pathname
components tend to be short (the most common case is one iteration!), but
there's some room for latency, as there is a fair bit of intervening logic
before the hash value is used for anything.
(Performance verified with "bonnie++ -s 0 -n 1536:-2" on tmpfs. I need
a better benchmark; the numbers seem to show a slight dip in performance
between 4.6.0 and this patch, but they're too noisy to quote.)
Special thanks to Bruce fields for diligent testing which uncovered a
nasty fencepost error in an earlier version of this patch.
[checkpatch.pl formatting complaints noted and respectfully disagreed with.]
Signed-off-by: George Spelvin <linux@sciencehorizons.net> Tested-by: J. Bruce Fields <bfields@redhat.com>
projects / linux-beck.git / log