GitHub/LineageOS/android_kernel_motorola_exynos9610.git
11 years agoARM: imx: imx6sl iomuxc syscon is compatible to imx6q
Shawn Guo [Fri, 18 Oct 2013 15:27:37 +0000 (23:27 +0800)]
ARM: imx: imx6sl iomuxc syscon is compatible to imx6q

The imx6sl iomuxc syscon is compatible to imx6q, so let's add
compatible string 'fsl,imx6q-iomuxc-gpr' for imx6sl iomuxc syscon node.

Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx6sl-evk: enable the SPI NOR
Huang Shijie [Fri, 18 Oct 2013 02:32:53 +0000 (10:32 +0800)]
ARM: dts: imx6sl-evk: enable the SPI NOR

enable the spi nor for imx6sl-evk boards.

Signed-off-by: Huang Shijie <b32955@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx6sl: add a pinctrl for ECSPI1
Huang Shijie [Fri, 18 Oct 2013 02:32:52 +0000 (10:32 +0800)]
ARM: dts: imx6sl: add a pinctrl for ECSPI1

add a pinctrl for ECSPI1. This pinctrl can be used in the imx6sl-evk board.

Signed-off-by: Huang Shijie <b32955@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: imx27: add missing #pwm-cells property
Steffen Trumtrar [Thu, 17 Oct 2013 13:03:16 +0000 (15:03 +0200)]
ARM: imx27: add missing #pwm-cells property

The pwm-node is missing its #pwm-cells property. The pwm-framework will
complain about this.

Add the missing property.

Signed-off-by: Steffen Trumtrar <s.trumtrar@pengutronix.de>
Acked-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx6sl: add pinctrl uhs states for usdhc
Dong Aisheng [Wed, 9 Oct 2013 11:20:07 +0000 (19:20 +0800)]
ARM: dts: imx6sl: add pinctrl uhs states for usdhc

This is needed for SD3.0 cards working on UHS mode.

Signed-off-by: Dong Aisheng <b29396@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx6qdl-sabresd: Add backlight support for lvds
Rogerio Pimentel [Fri, 11 Oct 2013 19:48:16 +0000 (16:48 -0300)]
ARM: dts: imx6qdl-sabresd: Add backlight support for lvds

This patch adds support for lvds backlight on boards
i.MX6q-SabreSD and i.MX6dl-SabreSD

Signed-off-by: Rogerio Pimentel <rogerio.pimentel@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx51-babbage: Make DVI and WVGA panel functional
Fabio Estevam [Tue, 8 Oct 2013 18:52:12 +0000 (15:52 -0300)]
ARM: dts: imx51-babbage: Make DVI and WVGA panel functional

Currently we get the following errors on imx51-babbage:

/display@di0: could not find display-timings node
/display@di0: no timings specified
/display@di1: could not find display-timings node
/display@di1: no timings specified
imx-drm imx-drm: failed to allocate buffer with size 0

Provide timing values for IPU1, which is connected to a DVI bridge and
to IPU2, which can be connected to the WVGA panel, so that both of them can
be functional.

While at it, disable the WVGA panel, so that DVI output becomes the default one.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: imx27-apf27dev: Add framebuffer support
Gwenhael Goavec-Merou [Sun, 6 Oct 2013 14:07:54 +0000 (16:07 +0200)]
ARM: imx27-apf27dev: Add framebuffer support

Signed-off-by: Gwenhael Goavec-Merou <gwenhael.goavec-merou@armadeus.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: imx51-apf51dev: Add parallel display support
Gwenhael Goavec-Merou [Sun, 6 Oct 2013 11:48:52 +0000 (13:48 +0200)]
ARM: imx51-apf51dev: Add parallel display support

Signed-off-by: Gwenhael Goavec-Merou <gwenhael.goavec-merou@armadeus.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx53-qsb: Do not use GPIO1_8 as wakeup source
Fabio Estevam [Fri, 4 Oct 2013 16:15:28 +0000 (13:15 -0300)]
ARM: dts: imx53-qsb: Do not use GPIO1_8 as wakeup source

On the mx53qsb board with mc34708 PMIC, GPIO1_8 resets the system, so better not
to use it as a wakeup source.

Use GPIO 2_14 and 2_15 for wakeup sources instead.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx53-qsb: SDHC1 does not have cd-gpios
Fabio Estevam [Fri, 4 Oct 2013 16:15:27 +0000 (13:15 -0300)]
ARM: dts: imx53-qsb: SDHC1 does not have cd-gpios

SDHC1 does not have any GPIO for reading the card detection status, so
remove 'cd-gpios'.

After this change card detection works via the internal SD controller mechanism.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx53-qsb: SDHC3 is connected in 8-bit mode
Fabio Estevam [Fri, 4 Oct 2013 16:15:26 +0000 (13:15 -0300)]
ARM: dts: imx53-qsb: SDHC3 is connected in 8-bit mode

SDHC3 is 8 bit-wide, so pass the bus-width property to reflect that.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: mxs: Add MSR M28CU3 board
Marek Vasut [Sun, 29 Sep 2013 22:41:29 +0000 (00:41 +0200)]
ARM: dts: mxs: Add MSR M28CU3 board

This board is based on the M28 SoM with custom baseboard. Supported
are LEDs, ethernet, PWM, LCD, SD slots.

Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx6qdl-wandboard: Provide phy-reset-gpios
Fabio Estevam [Fri, 27 Sep 2013 14:12:42 +0000 (11:12 -0300)]
ARM: dts: imx6qdl-wandboard: Provide phy-reset-gpios

GPIO3_29 is used to reset the ethernet phy.

Add support for it.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx6qdl-sabresd: Provide phy-reset-gpios
Fabio Estevam [Fri, 27 Sep 2013 14:12:41 +0000 (11:12 -0300)]
ARM: dts: imx6qdl-sabresd: Provide phy-reset-gpios

GPIO1_25 is used to reset the ethernet phy.

Add support for it.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx6q-sabrelite: Add ethernet phy reset pin into hog
Fabio Estevam [Fri, 27 Sep 2013 14:12:40 +0000 (11:12 -0300)]
ARM: dts: imx6q-sabrelite: Add ethernet phy reset pin into hog

MX6QDL_PAD_EIM_D23__GPIO3_IO23 pin is used to reset the ethernet phy.

Add it to the 'hog' group.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx6qdl: add pcie device node
Sean Cross [Thu, 26 Sep 2013 02:51:09 +0000 (10:51 +0800)]
ARM: dts: imx6qdl: add pcie device node

Add pcie device node for imx6qdl.

Signed-off-by: Sean Cross <xobs@kosagi.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx6q-udoo: Add initial board support
Fabio Estevam [Thu, 26 Sep 2013 20:43:52 +0000 (17:43 -0300)]
ARM: dts: imx6q-udoo: Add initial board support

For more information about the Udoo board:
http://www.udoo.org/

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: mxs: Add muxing options for the SSP2 MMC
Marek Vasut [Thu, 26 Sep 2013 11:16:16 +0000 (13:16 +0200)]
ARM: dts: mxs: Add muxing options for the SSP2 MMC

Add pinmux for 4-bit SD card connected to SSP2.

Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: add initial VF610 Cosmic/Cosmic+ board support
Matt Porter [Mon, 23 Sep 2013 15:14:44 +0000 (11:14 -0400)]
ARM: dts: add initial VF610 Cosmic/Cosmic+ board support

Add initial PHYTEC VF610 Cosmic/Cosmic+ board support with
UART and FEC enabled.

Signed-off-by: Matt Porter <matt.porter@linaro.org>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: mxs: convert usb0_id_pins_a to use symbolic pin defs
Lothar Waßmann [Mon, 23 Sep 2013 12:20:27 +0000 (14:20 +0200)]
ARM: dts: mxs: convert usb0_id_pins_a to use symbolic pin defs

This pin definition had been added after the initial patch to use
symbolic pin names in DTS files.

Signed-off-by: Lothar Waßmann <LW@KARO-electronics.de>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: tx28: restructure and update DTS file
Lothar Waßmann [Mon, 23 Sep 2013 10:20:48 +0000 (12:20 +0200)]
ARM: dts: tx28: restructure and update DTS file

Update the Ka-Ro TX28 DTS file.

- add Copyright header
- use label references for better readability
- sort the entries alphabetically
- add some aliases used by U-Boot to modify the DT data

Signed-off-by: Lothar Waßmann <LW@KARO-electronics.de>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx6q: add quirky select input for USB_OTG_ID
Shawn Guo [Thu, 1 Aug 2013 04:22:04 +0000 (12:22 +0800)]
ARM: dts: imx6q: add quirky select input for USB_OTG_ID

For some reason, the select input of pin function USB_OTG_ID is not
implemented via a regular select input register but using the bit
USB_OTG_ID_ SEL (shift 13) of IOMUXC_GPR1 register (offset 0x4).

As per the workaround for such quirk implemented in pinctrl driver,
we need to compose the input_val cell as below.

31     23      15      7        0
| 0xff | shift | width | select |

Thus, we have 0xff0d0100 for MX6QDL_PAD_ENET_RX_ER__USB_OTG_ID and
0xff0d0101 for MX6QDL_PAD_GPIO_1__USB_OTG_ID in input_val cell.

Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Tested-by: Peter Chen <peter.chen@freescale.com>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
11 years agoARM: dts: mxs: modify mx23/mx28 dts files to use padconfig defines
Lothar Waßmann [Sun, 22 Sep 2013 06:02:59 +0000 (14:02 +0800)]
ARM: dts: mxs: modify mx23/mx28 dts files to use padconfig defines

Convert mx23/mx28 dts files to use the padconfig defintions from
mxs-pinfunc.h.

Signed-off-by: Lothar Waßmann <LW@KARO-electronics.de>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: mxs: modify mx23/mx28 dts files to use pinctrl headers
Lothar Waßmann [Thu, 19 Sep 2013 06:59:48 +0000 (08:59 +0200)]
ARM: dts: mxs: modify mx23/mx28 dts files to use pinctrl headers

Convert mx23/mx28 dts filed to use the pinctrl header files.

NOTE: During automatic conversion of these files to use the pinconfig
definitions an inconsistency has been found in:
arch/arm/boot/dts/imx28-apx4devkit.dts

According to the comment the function for pad SSP2_SS0 should have
been MX28_PAD_SSP2_SS0__GPIO_2_19, while the given value 0x2131
represents: MX28_PAD_SSP2_SS0__AUART3_TX

I used the later (though probably wrong) definition because that's
what is actually being used in the DTB.

Signed-off-by: Lothar Waßmann <LW@KARO-electronics.de>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: mxs: add pinctrl header files
Lothar Waßmann [Thu, 19 Sep 2013 06:59:47 +0000 (08:59 +0200)]
ARM: dts: mxs: add pinctrl header files

Remove the list of possible pin configurations from the documentation
file and create header files containing those definitions.
This eliminates the need for error-prone manual lookup of those values
in the documentation and guarantees consistency between the human
readable representation of the pad function in the .dts file and the
actual binary value used in DT.

Signed-off-by: Lothar Waßmann <LW@KARO-electronics.de>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx28-evk: Enable touchscreen support
Fabio Estevam [Tue, 17 Sep 2013 19:18:38 +0000 (16:18 -0300)]
ARM: dts: imx28-evk: Enable touchscreen support

Provide 'lradc-touchscreen-wires' property to the LRADC driver, so that
touchscreen can be functional.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx6qdl-sabresd: SDHC ports are 8 bit-wide
Fabio Estevam [Tue, 17 Sep 2013 16:46:23 +0000 (13:46 -0300)]
ARM: dts: imx6qdl-sabresd: SDHC ports are 8 bit-wide

On imx6qdl-sabresd the SDHC2 and SDHC3 are 8 bit-wide, so pass the bus-width
property to reflect that.

Otherwise the mmc driver will operate with the default bus-width value of 4.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx6qdl: add uhs pinctrl state for usdhc3
Dong Aisheng [Fri, 13 Sep 2013 11:11:38 +0000 (19:11 +0800)]
ARM: dts: imx6qdl: add uhs pinctrl state for usdhc3

This is needed for supporting ultra high speed cards like SD3.0 cards.

Signed-off-by: Dong Aisheng <b29396@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx6sl: Add spi aliases
Fabio Estevam [Fri, 13 Sep 2013 21:13:00 +0000 (18:13 -0300)]
ARM: dts: imx6sl: Add spi aliases

Add spi aliases.

While at it, keep the aliases entries sorted.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: imx: mx6sl-evk: Enable USB function
Peter Chen [Tue, 10 Sep 2013 02:23:16 +0000 (10:23 +0800)]
ARM: imx: mx6sl-evk: Enable USB function

Enable USB function for OTG 1 and OTG 2 at mx6sololite evk.
Besides, fix the wrong interrupt number for OTG2 and host 1.

Signed-off-by: Peter Chen <peter.chen@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: add iomuxc-gpr device node for imx6sl
Fugang Duan [Tue, 3 Sep 2013 04:26:22 +0000 (12:26 +0800)]
ARM: dts: add iomuxc-gpr device node for imx6sl

Add iomuxc gpr device node for imx6sl.

Signed-off-by: Fugang Duan <B38611@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: imx6qdl-wandboard: Add spdif support
Fabio Estevam [Tue, 3 Sep 2013 02:51:41 +0000 (23:51 -0300)]
ARM: imx6qdl-wandboard: Add spdif support

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Reviewed-by: Nicolin Chen <b42378@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: vf610-twr: Enable DSPI0 devices and Flash at26df081a
Chao Fu [Fri, 30 Aug 2013 03:19:49 +0000 (11:19 +0800)]
ARM: dts: vf610-twr: Enable DSPI0 devices and Flash at26df081a

This patch enables DSPI0 and at26df081a flash device for Vybrid VF610
TOWER board.

Signed-off-by: Chao Fu <b44548@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: vf610: Add DSPI nodes
Chao Fu [Fri, 30 Aug 2013 03:19:48 +0000 (11:19 +0800)]
ARM: dts: vf610: Add DSPI nodes

Add Freescale DSPI node into vf610 dts.

Signed-off-by: Chao Fu <b44548@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx6q-sabrelite: Add LVDS support
Fabio Estevam [Wed, 28 Aug 2013 02:33:37 +0000 (23:33 -0300)]
ARM: dts: imx6q-sabrelite: Add LVDS support

imx6q-sabrelite board can be connected to a 1024x768 Hannstar LVDS panel.

Add support for it.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx6q-sabrelite: Put the nodes in alphabetical order
Fabio Estevam [Wed, 28 Aug 2013 02:33:36 +0000 (23:33 -0300)]
ARM: dts: imx6q-sabrelite: Put the nodes in alphabetical order

Put the nodes in alphabetical order so that further node additions can be better
organized.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx53-qsb: Allow usb peripheral mode to work
Fabio Estevam [Wed, 21 Aug 2013 13:27:04 +0000 (10:27 -0300)]
ARM: dts: imx53-qsb: Allow usb peripheral mode to work

Trying to use the usb otg port on mx53qsb results in the following error:

$ modprobe g_ether
using random self ethernet address
using random host ethernet address
usb0: HOST MAC 52:0f:8a:1e:aa:09
usb0: MAC de:f3:70:d8:6c:62
using random self ethernet address
using random host ethernet address
g_ether gadget: Ethernet Gadget, version: Memorial Day 2008
g_ether gadget: g_ether ready

(Connect the USB cable)

$ ci_hdrc ci_hdrc.0: remove, state 4
usb usb1: USB disconnect, device number 1
ci_hdrc ci_hdrc.0: USB bus 1 deregistered
ci_hdrc ci_hdrc.0: timeout waiting for 00000800 in 11

USB otg port goes to connector J3 (mini USB) and also to the USB combo (J2).

As mx53qsb does not provide a USB ID pin, pass the dr_mode as 'peripheral' so
that we can have usb device working.

Tested via g_ether.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx28-evk: Allow usb peripheral mode to work
Fabio Estevam [Wed, 21 Aug 2013 13:27:03 +0000 (10:27 -0300)]
ARM: dts: imx28-evk: Allow usb peripheral mode to work

Testing g_ether results in the following:

[    1.648022] using random self ethernet address
[    1.652778] using random host ethernet address
[    1.660504] usb0: HOST MAC c6:28:6a:81:dc:ff
[    1.666360] usb0: MAC 7e:81:54:16:c0:c6
[    1.670504] using random self ethernet address
[    1.675842] using random host ethernet address
[    1.682655] g_ether gadget: Ethernet Gadget, version: Memorial Day 2008
[    1.689332] g_ether gadget: g_ether ready
[    3.328974] irq 237: nobody cared (try booting with the "irqpoll" option)
[    3.335831] CPU: 0 PID: 1 Comm: swapper Not tainted 3.11.0-rc6-next-20130819 #1035
[    3.343500] [<c0014258>] (unwind_backtrace+0x0/0xf0) from [<c0011b20>] (show_stack+0x10/0x14)
[    3.352098] [<c0011b20>] (show_stack+0x10/0x14) from [<c006af4c>] (__report_bad_irq+0x20/0xc0)
[    3.360763] [<c006af4c>] (__report_bad_irq+0x20/0xc0) from [<c006b3f4>] (note_interrupt+0x1d4/0x234)
[    3.369943] [<c006b3f4>] (note_interrupt+0x1d4/0x234) from [<c0069754>] (handle_irq_event_percpu+0xc4/0x268)
[    3.379815] [<c0069754>] (handle_irq_event_percpu+0xc4/0x268) from [<c0069934>] (handle_irq_event+0x3c/0x5c)
[    3.389686] [<c0069934>] (handle_irq_event+0x3c/0x5c) from [<c006bb78>] (handle_level_irq+0x8c/0xe8)
[    3.398865] [<c006bb78>] (handle_level_irq+0x8c/0xe8) from [<c0068ff0>] (generic_handle_irq+0x20/0x30)
[    3.408213] [<c0068ff0>] (generic_handle_irq+0x20/0x30) from [<c000fc9c>] (handle_IRQ+0x30/0x84)
[    3.417040] [<c000fc9c>] (handle_IRQ+0x30/0x84) from [<c0012564>] (__irq_svc+0x44/0x54)
[    3.425094] [<c0012564>] (__irq_svc+0x44/0x54) from [<c0020244>] (__do_softirq+0x90/0x268)
[    3.433400] [<c0020244>] (__do_softirq+0x90/0x268) from [<c00207d4>] (irq_exit+0x9c/0xd8)
[    3.441619] [<c00207d4>] (irq_exit+0x9c/0xd8) from [<c000fca0>] (handle_IRQ+0x34/0x84)
[    3.449577] [<c000fca0>] (handle_IRQ+0x34/0x84) from [<c0012564>] (__irq_svc+0x44/0x54)
[    3.457629] [<c0012564>] (__irq_svc+0x44/0x54) from [<c04602b0>] (_raw_spin_unlock_irqrestore+0x34/0x44)
[    3.467164] [<c04602b0>] (_raw_spin_unlock_irqrestore+0x34/0x44) from [<c0348298>] (input_register_handler+0x98/0xb4)
[    3.477835] [<c0348298>] (input_register_handler+0x98/0xb4) from [<c064bc10>] (mousedev_init+0x30/0x60)
[    3.487275] [<c064bc10>] (mousedev_init+0x30/0x60) from [<c00088ac>] (do_one_initcall+0xe8/0x154)
[    3.496198] [<c00088ac>] (do_one_initcall+0xe8/0x154) from [<c062aabc>] (kernel_init_freeable+0xf0/0x1b4)
[    3.505810] [<c062aabc>] (kernel_init_freeable+0xf0/0x1b4) from [<c0456550>] (kernel_init+0x8/0xe4)
[    3.514899] [<c0456550>] (kernel_init+0x8/0xe4) from [<c000ee60>] (ret_from_fork+0x14/0x34)
[    3.523263] handlers:
[    3.525578] [<c0339d14>] ci_irq
[    3.528757] Disabling IRQ #237

Provide the USB0_ID pin in the usb0 node, so that g_ether can operate
correctly.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx6qdl-wandboard: Add usbotg support
Fabio Estevam [Wed, 21 Aug 2013 13:27:02 +0000 (10:27 -0300)]
ARM: dts: imx6qdl-wandboard: Add usbotg support

Tested via g_ether:

$ modprobe g_ether
using random self ethernet address
using random host ethernet address
usb0: HOST MAC 42:b5:26:a9:48:21
usb0: MAC 36:a6:85:9b:9e:13
using random self ethernet address
using random host ethernet address
g_ether gadget: Ethernet Gadget, version: Memorial Day 2008
g_ether gadget: g_ether ready
$ g_ether gadget: high-speed config #1: CDC Ethernet (ECM)
$ ifconfig usb0 10.0.0.2

Then on the PC host side:

~$ sudo ifconfig usb0 10.0.0.1
~$ ping 10.0.0.2
PING 10.0.0.2 (10.0.0.2) 56(84) bytes of data.
64 bytes from 10.0.0.2: icmp_req=1 ttl=64 time=1.26 ms
64 bytes from 10.0.0.2: icmp_req=2 ttl=64 time=0.280 ms
64 bytes from 10.0.0.2: icmp_req=3 ttl=64 time=0.297 ms

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: i.MX51: Add W1 devicetree node
Alexander Shiyan [Wed, 21 Aug 2013 07:28:25 +0000 (11:28 +0400)]
ARM: dts: i.MX51: Add W1 devicetree node

This patch adds the missing W1 (onewire) devicetree node for i.MX51 CPUs.

Signed-off-by: Alexander Shiyan <shc_work@mail.ru>
Acked-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: i.MX51: Add IRAM devicetree node
Alexander Shiyan [Wed, 21 Aug 2013 07:28:24 +0000 (11:28 +0400)]
ARM: dts: i.MX51: Add IRAM devicetree node

This patch adds the missing IRAM devicetree node for i.MX51 CPUs.

Signed-off-by: Alexander Shiyan <shc_work@mail.ru>
Acked-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: i.MX51: Separate TXD/RXD and RTS/CTS pinmux entries for UARTs
Alexander Shiyan [Wed, 21 Aug 2013 07:28:23 +0000 (11:28 +0400)]
ARM: dts: i.MX51: Separate TXD/RXD and RTS/CTS pinmux entries for UARTs

RTS/CTS pins can be used for different purposes, so create separate
definitions for these pins.

Signed-off-by: Alexander Shiyan <shc_work@mail.ru>
Acked-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoARM: dts: imx6sl reuses imx6q sdma firmware
Shawn Guo [Tue, 13 Aug 2013 00:55:02 +0000 (08:55 +0800)]
ARM: dts: imx6sl reuses imx6q sdma firmware

There is no imx6sl specific sdma firmware.  Instead, imx6sl reuses
imx6q sdma firmware.

Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
11 years agoLinux 3.12-rc2
Linus Torvalds [Mon, 23 Sep 2013 22:41:09 +0000 (15:41 -0700)]
Linux 3.12-rc2

11 years agoMerge tag 'staging-3.12-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh...
Linus Torvalds [Mon, 23 Sep 2013 19:53:07 +0000 (12:53 -0700)]
Merge tag 'staging-3.12-rc2' of git://git./linux/kernel/git/gregkh/staging

Pull staging fixes from Greg KH:
 "Here are a number of small staging tree and iio driver fixes.  Nothing
  major, just lots of little things"

* tag 'staging-3.12-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging: (34 commits)
  iio:buffer_cb: Add missing iio_buffer_init()
  iio: Prevent race between IIO chardev opening and IIO device free
  iio: fix: Keep a reference to the IIO device for open file descriptors
  iio: Stop sampling when the device is removed
  iio: Fix crash when scan_bytes is computed with active_scan_mask == NULL
  iio: Fix mcp4725 dev-to-indio_dev conversion in suspend/resume
  iio: Fix bma180 dev-to-indio_dev conversion in suspend/resume
  iio: Fix tmp006 dev-to-indio_dev conversion in suspend/resume
  iio: iio_device_add_event_sysfs() bugfix
  staging: iio: ade7854-spi: Fix return value
  staging:iio:hmc5843: Fix measurement conversion
  iio: isl29018: Fix uninitialized value
  staging:iio:dummy fix kfifo_buf kconfig dependency issue if kfifo modular and buffer enabled for built in dummy driver.
  iio: at91: fix adc_clk overflow
  staging: line6: add bounds check in snd_toneport_source_put()
  Staging: comedi: Fix dependencies for drivers misclassified as PCI
  staging: r8188eu: Adjust RX gain
  staging: r8188eu: Fix smatch warning in core/rtw_ieee80211.
  staging: r8188eu: Fix smatch error in core/rtw_mlme_ext.c
  staging: r8188eu: Fix Smatch off-by-one warning in hal/rtl8188e_hal_init.c
  ...

11 years agoMerge tag 'usb-3.12-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb
Linus Torvalds [Mon, 23 Sep 2013 19:52:35 +0000 (12:52 -0700)]
Merge tag 'usb-3.12-rc2' of git://git./linux/kernel/git/gregkh/usb

Pull USB fixes from Greg KH:
 "Here are a number of small USB fixes for 3.12-rc2.

  One is a revert of a EHCI change that isn't quite ready for 3.12.
  Others are minor things, gadget fixes, Kconfig fixes, and some quirks
  and documentation updates.

  All have been in linux-next for a bit"

* tag 'usb-3.12-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb:
  USB: pl2303: distinguish between original and cloned HX chips
  USB: Faraday fotg210: fix email addresses
  USB: fix typo in usb serial simple driver Kconfig
  Revert "USB: EHCI: support running URB giveback in tasklet context"
  usb: s3c-hsotg: do not disconnect gadget when receiving ErlySusp intr
  usb: s3c-hsotg: fix unregistration function
  usb: gadget: f_mass_storage: reset endpoint driver data when disabled
  usb: host: fsl-mph-dr-of: Staticize local symbols
  usb: gadget: f_eem: Staticize eem_alloc
  usb: gadget: f_ecm: Staticize ecm_alloc
  usb: phy: omap-usb3: Fix return value
  usb: dwc3: gadget: avoid memory leak when failing to allocate all eps
  usb: dwc3: remove extcon dependency
  usb: gadget: add '__ref' for rndis_config_register() and cdc_config_register()
  usb: dwc3: pci: add support for BayTrail
  usb: gadget: cdc2: fix conversion to new interface of f_ecm
  usb: gadget: fix a bug and a WARN_ON in dummy-hcd
  usb: gadget: mv_u3d_core: fix violation of locking discipline in mv_u3d_ep_disable()

11 years agoMerge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux
Linus Torvalds [Mon, 23 Sep 2013 02:51:49 +0000 (19:51 -0700)]
Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux

Pull drm fixes from Dave Airlie:
 - some small fixes for msm and exynos
 - a regression revert affecting nouveau users with old userspace
 - intel pageflip deadlock and gpu hang fixes, hsw modesetting hangs

* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux: (22 commits)
  Revert "drm: mark context support as a legacy subsystem"
  drm/i915: Don't enable the cursor on a disable pipe
  drm/i915: do not update cursor in crtc mode set
  drm/exynos: fix return value check in lowlevel_buffer_allocate()
  drm/exynos: Fix address space warnings in exynos_drm_fbdev.c
  drm/exynos: Fix address space warning in exynos_drm_buf.c
  drm/exynos: Remove redundant OF dependency
  drm/msm: drop unnecessary set_need_resched()
  drm/i915: kill set_need_resched
  drm/msm: fix potential NULL pointer dereference
  drm/i915/dvo: set crtc timings again for panel fixed modes
  drm/i915/sdvo: Robustify the dtd<->drm_mode conversions
  drm/msm: workaround for missing irq
  drm/msm: return -EBUSY if bo still active
  drm/msm: fix return value check in ERR_PTR()
  drm/msm: fix cmdstream size check
  drm/msm: hangcheck harder
  drm/msm: handle read vs write fences
  drm/i915/sdvo: Fully translate sync flags in the dtd->mode conversion
  drm/i915: Use proper print format for debug prints
  ...

11 years agoMerge branch 'for-3.12/core' of git://git.kernel.dk/linux-block
Linus Torvalds [Sun, 22 Sep 2013 22:00:11 +0000 (15:00 -0700)]
Merge branch 'for-3.12/core' of git://git.kernel.dk/linux-block

Pull block IO fixes from Jens Axboe:
 "After merge window, no new stuff this time only a collection of neatly
  confined and simple fixes"

* 'for-3.12/core' of git://git.kernel.dk/linux-block:
  cfq: explicitly use 64bit divide operation for 64bit arguments
  block: Add nr_bios to block_rq_remap tracepoint
  If the queue is dying then we only call the rq->end_io callout. This leaves bios setup on the request, because the caller assumes when the blk_execute_rq_nowait/blk_execute_rq call has completed that the rq->bios have been cleaned up.
  bio-integrity: Fix use of bs->bio_integrity_pool after free
  blkcg: relocate root_blkg setting and clearing
  block: Convert kmalloc_node(...GFP_ZERO...) to kzalloc_node(...)
  block: trace all devices plug operation

11 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux...
Linus Torvalds [Sun, 22 Sep 2013 21:58:49 +0000 (14:58 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/mason/linux-btrfs

Pull btrfs fixes from Chris Mason:
 "These are mostly bug fixes and a two small performance fixes.  The
  most important of the bunch are Josef's fix for a snapshotting
  regression and Mark's update to fix compile problems on arm"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs: (25 commits)
  Btrfs: create the uuid tree on remount rw
  btrfs: change extent-same to copy entire argument struct
  Btrfs: dir_inode_operations should use btrfs_update_time also
  btrfs: Add btrfs: prefix to kernel log output
  btrfs: refuse to remount read-write after abort
  Btrfs: btrfs_ioctl_default_subvol: Revert back to toplevel subvolume when arg is 0
  Btrfs: don't leak transaction in btrfs_sync_file()
  Btrfs: add the missing mutex unlock in write_all_supers()
  Btrfs: iput inode on allocation failure
  Btrfs: remove space_info->reservation_progress
  Btrfs: kill delay_iput arg to the wait_ordered functions
  Btrfs: fix worst case calculator for space usage
  Revert "Btrfs: rework the overcommit logic to be based on the total size"
  Btrfs: improve replacing nocow extents
  Btrfs: drop dir i_size when adding new names on replay
  Btrfs: replay dir_index items before other items
  Btrfs: check roots last log commit when checking if an inode has been logged
  Btrfs: actually log directory we are fsync()'ing
  Btrfs: actually limit the size of delalloc range
  Btrfs: allocate the free space by the existed max extent size when ENOSPC
  ...

11 years agocfq: explicitly use 64bit divide operation for 64bit arguments
Anatol Pomozov [Sun, 22 Sep 2013 18:43:47 +0000 (12:43 -0600)]
cfq: explicitly use 64bit divide operation for 64bit arguments

'samples' is 64bit operant, but do_div() second parameter is 32.
do_div silently truncates high 32 bits and calculated result
is invalid.

In case if low 32bit of 'samples' are zeros then do_div() produces
kernel crash.

Signed-off-by: Anatol Pomozov <anatol.pomozov@gmail.com>
Acked-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
11 years agoMerge tag 'iio-fixes-for-3.12a' of git://git.kernel.org/pub/scm/linux/kernel/git...
Greg Kroah-Hartman [Sat, 21 Sep 2013 23:45:36 +0000 (16:45 -0700)]
Merge tag 'iio-fixes-for-3.12a' of git://git./linux/kernel/git/jic23/iio into staging-linus

Jonathan writes:

First round of IIO fixes for 3.12

A series of wrong 'struct dev' assumptions in suspend/resume callbacks
following on from this issue being identified in a new driver review.
One to watch out for in future.

A number of driver specific fixes
1) at91 - fix a overflow in clock rate computation
2) dummy - Kconfig dependency issue
3) isl29018 - uninitialized value
4) hmc5843 - measurement conversion bug introduced by recent cleanup.
5) ade7854-spi - wrong return value.

Some IIO core fixes
1) Wrong value picked up for event code creation for a modified channel
2) A null dereference on failure to initialize a buffer after no buffer has
   been in use, when using the available_scan_masks approach.
3) Sampling not stopped when a device is removed. Effects forced removal
   such as hot unplugging.
4) Prevent device going away if a chrdev is still open in userspace.
5) Prevent race on chardev opening and device being freed.
6) Add a missing iio_buffer_init in the call back buffer.

These last few are the first part of a set from Lars-Peter Clausen who
has been taking a closer look at our removal paths and buffer handling
than anyone has for quite some time.

11 years agoMerge tag 'nfs-for-3.12-3' of git://git.linux-nfs.org/projects/trondmy/linux-nfs
Linus Torvalds [Sat, 21 Sep 2013 22:59:41 +0000 (15:59 -0700)]
Merge tag 'nfs-for-3.12-3' of git://git.linux-nfs.org/projects/trondmy/linux-nfs

Pull NFS client bugfix from Trond Myklebust:
 "Fix a regression due to incorrect sharing of gss auth caches"

* tag 'nfs-for-3.12-3' of git://git.linux-nfs.org/projects/trondmy/linux-nfs:
  RPCSEC_GSS: fix crash on destroying gss auth

11 years agoblock: Add nr_bios to block_rq_remap tracepoint
Jun'ichi Nomura [Sat, 21 Sep 2013 19:57:47 +0000 (13:57 -0600)]
block: Add nr_bios to block_rq_remap tracepoint

Adding the number of bios in a remapped request to 'block_rq_remap'
tracepoint.

Request remapper clones bios in a request to track the completion
status of each bio. So the number of bios can be useful information
for investigation.

Related discussions:
  http://www.redhat.com/archives/dm-devel/2013-August/msg00084.html
  http://www.redhat.com/archives/dm-devel/2013-September/msg00024.html

Signed-off-by: Jun'ichi Nomura <j-nomura@ce.jp.nec.com>
Acked-by: Mike Snitzer <snitzer@redhat.com>
Cc: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
11 years agoBtrfs: create the uuid tree on remount rw
Josef Bacik [Sat, 21 Sep 2013 02:33:20 +0000 (22:33 -0400)]
Btrfs: create the uuid tree on remount rw

Users have been complaining of the uuid tree stuff warning that there is no uuid
root when trying to do snapshot operations.  This is because if you mount -o ro
we will not create the uuid tree.  But then if you mount -o rw,remount we will
still not create it and then any subsequent snapshot/subvol operations you try
to do will fail gloriously.  Fix this by creating the uuid_root on remount rw if
it was not already there.  Thanks,

Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agobtrfs: change extent-same to copy entire argument struct
Mark Fasheh [Tue, 17 Sep 2013 22:43:54 +0000 (15:43 -0700)]
btrfs: change extent-same to copy entire argument struct

btrfs_ioctl_file_extent_same() uses __put_user_unaligned() to copy some data
back to it's argument struct. Unfortunately, not all architectures provide
__put_user_unaligned(), so compiles break on them if btrfs is selected.

Instead, just copy the whole struct in / out at the start and end of
operations, respectively.

Signed-off-by: Mark Fasheh <mfasheh@suse.de>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agoBtrfs: dir_inode_operations should use btrfs_update_time also
Guangyu Sun [Mon, 16 Sep 2013 17:42:03 +0000 (10:42 -0700)]
Btrfs: dir_inode_operations should use btrfs_update_time also

Commit 2bc5565286121d2a77ccd728eb3484dff2035b58 (Btrfs: don't update atime on
RO subvolumes) ensures that the access time of an inode is not updated when
the inode lives in a read-only subvolume.
However, if a directory on a read-only subvolume is accessed, the atime is
updated. This results in a write operation to a read-only subvolume. I
believe that access times should never be updated on read-only subvolumes.

To reproduce:

 # mkfs.btrfs -f /dev/dm-3
 (...)
 # mount /dev/dm-3 /mnt
 # btrfs subvol create /mnt/sub
  Create subvolume '/mnt/sub'
 # mkdir /mnt/sub/dir
 # echo "abc" > /mnt/sub/dir/file
 # btrfs subvol snapshot -r /mnt/sub /mnt/rosnap
  Create a readonly snapshot of '/mnt/sub' in '/mnt/rosnap'
 # stat /mnt/rosnap/dir
  File: `/mnt/rosnap/dir'
  Size: 8         Blocks: 0          IO Block: 4096   directory
 Device: 16h/22d    Inode: 257         Links: 1
 Access: (0755/drwxr-xr-x)  Uid: (    0/    root)   Gid: (    0/    root)
  Access: 2013-09-11 07:21:49.389157126 -0400
  Modify: 2013-09-11 07:22:02.330156079 -0400
  Change: 2013-09-11 07:22:02.330156079 -0400
 # ls /mnt/rosnap/dir
  file
 # stat /mnt/rosnap/dir
  File: `/mnt/rosnap/dir'
  Size: 8         Blocks: 0          IO Block: 4096   directory
 Device: 16h/22d    Inode: 257         Links: 1
 Access: (0755/drwxr-xr-x)  Uid: (    0/    root)   Gid: (    0/    root)
  Access: 2013-09-11 07:22:56.797151670 -0400
  Modify: 2013-09-11 07:22:02.330156079 -0400
  Change: 2013-09-11 07:22:02.330156079 -0400

Reported-by: Koen De Wit <koen.de.wit@oracle.com>
Signed-off-by: Guangyu Sun <guangyu.sun@oracle.com>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agobtrfs: Add btrfs: prefix to kernel log output
Frank Holton [Fri, 13 Sep 2013 15:46:50 +0000 (11:46 -0400)]
btrfs: Add btrfs: prefix to kernel log output

The kernel log entries for device label %s and device fsid %pU
are missing the btrfs: prefix. Add those here.

Signed-off-by: Frank Holton <fholton@gmail.com>
Reviewed-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agobtrfs: refuse to remount read-write after abort
David Sterba [Fri, 13 Sep 2013 15:41:20 +0000 (17:41 +0200)]
btrfs: refuse to remount read-write after abort

It's still possible to flip the filesystem into RW mode after it's
remounted RO due to an abort. There are lots of places that check for
the superblock error bit and will not write data, but we should not let
the filesystem appear read-write.

Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agoBtrfs: btrfs_ioctl_default_subvol: Revert back to toplevel subvolume when arg is 0
chandan [Fri, 13 Sep 2013 14:04:10 +0000 (19:34 +0530)]
Btrfs: btrfs_ioctl_default_subvol: Revert back to toplevel subvolume when arg is 0

This patch makes it possible to set BTRFS_FS_TREE_OBJECTID as the default
subvolume by passing a subvolume id of 0.

Signed-off-by: chandan <chandan@linux.vnet.ibm.com>
Reviewed-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agoBtrfs: don't leak transaction in btrfs_sync_file()
Filipe David Borba Manana [Wed, 11 Sep 2013 19:36:44 +0000 (20:36 +0100)]
Btrfs: don't leak transaction in btrfs_sync_file()

In btrfs_sync_file(), if the call to btrfs_log_dentry_safe() returns
a negative error (for e.g. -ENOMEM via btrfs_log_inode()), we would
return without ending/freeing the transaction.

Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agoBtrfs: add the missing mutex unlock in write_all_supers()
Stefan Behrens [Wed, 11 Sep 2013 07:59:22 +0000 (09:59 +0200)]
Btrfs: add the missing mutex unlock in write_all_supers()

The BUG() was replaced by btrfs_error() and return -EIO with the
patch "get rid of one BUG() in write_all_supers()", but the missing
mutex_unlock() was overlooked.

The 0-DAY kernel build service from Intel reported the missing
unlock which was found by the coccinelle tool:

    fs/btrfs/disk-io.c:3422:2-8: preceding lock on line 3374

Signed-off-by: Stefan Behrens <sbehrens@giantdisaster.de>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agoBtrfs: iput inode on allocation failure
Josef Bacik [Tue, 17 Sep 2013 15:25:44 +0000 (11:25 -0400)]
Btrfs: iput inode on allocation failure

We don't do the iput when we fail to allocate our delayed delalloc work in
__start_delalloc_inodes, fix this.

Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agoBtrfs: remove space_info->reservation_progress
Josef Bacik [Tue, 17 Sep 2013 15:02:25 +0000 (11:02 -0400)]
Btrfs: remove space_info->reservation_progress

This isn't used for anything anymore, just remove it.

Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agoBtrfs: kill delay_iput arg to the wait_ordered functions
Josef Bacik [Tue, 17 Sep 2013 14:55:51 +0000 (10:55 -0400)]
Btrfs: kill delay_iput arg to the wait_ordered functions

This is a left over of how we used to wait for ordered extents, which was to
grab the inode and then run filemap flush on it.  However if we have an ordered
extent then we already are holding a ref on the inode, and we just use
btrfs_start_ordered_extent anyway, so there is no reason to have an extra ref on
the inode to start work on the ordered extent.  Thanks,

Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agoBtrfs: fix worst case calculator for space usage
Josef Bacik [Tue, 17 Sep 2013 14:50:06 +0000 (10:50 -0400)]
Btrfs: fix worst case calculator for space usage

Forever ago I made the worst case calculator say that we could potentially split
into 3 blocks for every level on the way down, which isn't right.  If we split
we're only going to get two new blocks, the one we originally cow'ed and the new
one we're going to split.  Thanks,

Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agoRevert "Btrfs: rework the overcommit logic to be based on the total size"
Josef Bacik [Tue, 17 Sep 2013 14:48:00 +0000 (10:48 -0400)]
Revert "Btrfs: rework the overcommit logic to be based on the total size"

This reverts commit 70afa3998c9baed4186df38988246de1abdab56d.  It is causing
performance issues and wasn't actually correct.  There were problems with the
way we flushed delalloc and that was the real cause of the early enospc.
Thanks,

Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agoBtrfs: improve replacing nocow extents
Josef Bacik [Thu, 12 Sep 2013 20:58:28 +0000 (16:58 -0400)]
Btrfs: improve replacing nocow extents

Various people have hit a deadlock when running btrfs/011.  This is because when
replacing nocow extents we will take the i_mutex to make sure nobody messes with
the file while we are replacing the extent.  The problem is we are already
holding a transaction open, which is a locking inversion, so instead we need to
save these inodes we find and then process them outside of the transaction.

Further we can't just lock the inode and assume we are good to go.  We need to
lock the extent range and then read back the extent cache for the inode to make
sure the extent really still points at the physical block we want.  If it
doesn't we don't have to copy it.  Thanks,

Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agoBtrfs: drop dir i_size when adding new names on replay
Josef Bacik [Wed, 11 Sep 2013 18:17:00 +0000 (14:17 -0400)]
Btrfs: drop dir i_size when adding new names on replay

So if we have dir_index items in the log that means we also have the inode item
as well, which means that the inode's i_size is correct.  However when we
process dir_index'es we call btrfs_add_link() which will increase the
directory's i_size for the new entry.  To fix this we need to just set the dir
items i_size to 0, and then as we find dir_index items we adjust the i_size.
btrfs_add_link() will do it for new entries, and if the entry already exists we
can just add the name_len to the i_size ourselves.  Thanks,

Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agoBtrfs: replay dir_index items before other items
Josef Bacik [Wed, 11 Sep 2013 15:57:23 +0000 (11:57 -0400)]
Btrfs: replay dir_index items before other items

A user reported a bug where his log would not replay because he was getting
-EEXIST back.  This was because he had a file moved into a directory that was
logged.  What happens is the file had a lower inode number, and so it is
processed first when replaying the log, and so we add the inode ref in for the
directory it was moved to.  But then we process the directories DIR_INDEX item
and try to add the inode ref for that inode and it fails because we already
added it when we replayed the inode.  To solve this problem we need to just
process any DIR_INDEX items we have in the log first so this all is taken care
of, and then we can replay the rest of the items.  With this patch my reproducer
can remount the file system properly instead of erroring out.  Thanks,

Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agoBtrfs: check roots last log commit when checking if an inode has been logged
Josef Bacik [Wed, 11 Sep 2013 13:55:42 +0000 (09:55 -0400)]
Btrfs: check roots last log commit when checking if an inode has been logged

Liu introduced a local copy of the last log commit for an inode to make sure we
actually log an inode even if a log commit has already taken place.  In order to
make sure we didn't relog the same inode multiple times he set this local copy
to the current trans when we log the inode, because usually we log the inode and
then sync the log.  The exception to this is during rename, we will relog an
inode if the name changed and it is already in the log.  The problem with this
is then we go to sync the inode, and our check to see if the inode has already
been logged is tripped and we don't sync the log.  To fix this we need to _also_
check against the roots last log commit, because it could be less than what is
in our local copy of the log commit.  This fixes a bug where we rename a file
into a directory and then fsync the directory and then on remount the directory
is no longer there.  Thanks,

Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agoBtrfs: actually log directory we are fsync()'ing
Josef Bacik [Wed, 11 Sep 2013 13:36:30 +0000 (09:36 -0400)]
Btrfs: actually log directory we are fsync()'ing

If you just create a directory and then fsync that directory and then pull the
power plug you will come back up and the directory will not be there.  That is
because we won't actually create directories if we've logged files inside of
them since they will be created on replay, but in this check we will set our
logged_trans of our current directory if it happens to be a directory, making us
think it doesn't need to be logged.  Fix the logic to only do this to parent
directories.  Thanks,

Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agoBtrfs: actually limit the size of delalloc range
Josef Bacik [Fri, 30 Aug 2013 18:38:49 +0000 (14:38 -0400)]
Btrfs: actually limit the size of delalloc range

So forever we have had this thing to limit the amount of delalloc pages we'll
setup to be written out to 128mb.  This is because we have to lock all the pages
in this range, so anything above this gets a bit unweildly, and also without a
limit we'll happily allocate gigantic chunks of disk space.  Turns out our check
for this wasn't quite right, we wouldn't actually limit the chunk we wanted to
write out, we'd just stop looking for more space after we went over the limit.
So if you do a giant 20gb dd on my box with lots of ram I could get 2gig
extents.  This is fine normally, except when you go to relocate these extents
and we can't find enough space to relocate these moster extents, since we have
to be able to allocate exactly the same sized extent to move it around.  So fix
this by actually enforcing the limit.  With this patch I'm no longer seeing
giant 1.5gb extents.  Thanks,

Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agoBtrfs: allocate the free space by the existed max extent size when ENOSPC
Miao Xie [Mon, 9 Sep 2013 05:19:42 +0000 (13:19 +0800)]
Btrfs: allocate the free space by the existed max extent size when ENOSPC

By the current code, if the requested size is very large, and all the extents
in the free space cache are small, we will waste lots of the cpu time to cut
the requested size in half and search the cache again and again until it gets
down to the size the allocator can return. In fact, we can know the max extent
size in the cache after the first search, so we needn't cut the size in half
repeatedly, and just use the max extent size directly. This way can save
lots of cpu time and make the performance grow up when there are only fragments
in the free space cache.

According to my test, if there are only 4KB free space extents in the fs,
and the total size of those extents are 256MB, we can reduce the execute
time of the following test from 5.4s to 1.4s.
  dd if=/dev/zero of=<testfile> bs=1MB count=1 oflag=sync

Changelog v2 -> v3:
- fix the problem that we skip the block group with the space which is
  less than we need.

Changelog v1 -> v2:
- address the problem that we return a wrong start position when searching
  the free space in a bitmap.

Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agobtrfs: add lockdep and tracing annotations for uuid tree
David Sterba [Tue, 3 Sep 2013 16:28:57 +0000 (18:28 +0200)]
btrfs: add lockdep and tracing annotations for uuid tree

Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agobtrfs: show compiled-in config features at module load time
Stefan Behrens [Tue, 3 Sep 2013 13:25:27 +0000 (15:25 +0200)]
btrfs: show compiled-in config features at module load time

We want to know if there are debugging features compiled in, this may
affect performance. The message is printed before the sanity checks.

(This commit message is a copy of David Sterba's commit message when
he introduced btrfs_print_info()).

Signed-off-by: Stefan Behrens <sbehrens@giantdisaster.de>
Reviewed-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agoBtrfs: more efficient inode tree replace operation
Filipe David Borba Manana [Mon, 2 Sep 2013 11:19:13 +0000 (12:19 +0100)]
Btrfs: more efficient inode tree replace operation

Instead of removing the current inode from the red black tree
and then add the new one, just use the red black tree replace
operation, which is more efficient.

Signed-off-by: Filipe David Borba Manana <fdmanana@gmail.com>
Reviewed-by: Zach Brown <zab@redhat.com>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agoBtrfs: do not add replace target to the alloc_list
Ilya Dryomov [Sun, 1 Sep 2013 15:56:44 +0000 (18:56 +0300)]
Btrfs: do not add replace target to the alloc_list

If replace was suspended by the umount, replace target device is added
to the fs_devices->alloc_list during a later mount.  This is obviously
wrong.  ->is_tgtdev_for_dev_replace is supposed to guard against that,
but ->is_tgtdev_for_dev_replace is (and can only ever be) initialized
*after* everything is opened and fs_devices lists are populated.  Fix
this by checking the devid instead: for replace targets it's always
equal to BTRFS_DEV_REPLACE_DEVID.

Cc: Stefan Behrens <sbehrens@giantdisaster.de>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Reviewed-by: Stefan Behrens <sbehrens@giantdisaster.de>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agoBtrfs: fixup error handling in btrfs_reloc_cow
Josef Bacik [Fri, 30 Aug 2013 19:09:51 +0000 (15:09 -0400)]
Btrfs: fixup error handling in btrfs_reloc_cow

If we failed to actually allocate the correct size of the extent to relocate we
will end up in an infinite loop because we won't return an error, we'll just
move on to the next extent.  So fix this up by returning an error, and then fix
all the callers to return an error up the stack rather than BUG_ON()'ing.
Thanks,

Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
11 years agoMerge tag 'v3.11' into for-linus
Chris Mason [Sat, 21 Sep 2013 14:44:55 +0000 (10:44 -0400)]
Merge tag 'v3.11' into for-linus

Linux 3.11

11 years agoiio:buffer_cb: Add missing iio_buffer_init()
Lars-Peter Clausen [Wed, 18 Sep 2013 20:02:00 +0000 (21:02 +0100)]
iio:buffer_cb: Add missing iio_buffer_init()

Make sure to properly initialize the IIO buffer data structure.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
11 years agoiio: Prevent race between IIO chardev opening and IIO device free
Lars-Peter Clausen [Wed, 18 Sep 2013 20:02:00 +0000 (21:02 +0100)]
iio: Prevent race between IIO chardev opening and IIO device free

Set the IIO device as the parent for the character device

We need to make sure that the IIO device is not freed while the character device
exists, otherwise the freeing of the IIO device might race against the file open
callback. Do this by setting the character device's parent to the IIO device,
this will cause the character device to grab a reference to the IIO device and
only release it once the character device itself has been removed.

Also move the registration of the character device before the registration of
the IIO device to avoid the (rather theoretical case) that the IIO device is
already freed again before we can add the character device and grab a reference
to the IIO device.

We also need to move the call to cdev_del() from iio_dev_release() to
iio_device_unregister() (where it should have been in the first place anyway) to
avoid a reference cycle. As iio_dev_release() is only called once all reference
are dropped, but the character device holds a reference to the IIO device.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
11 years agoiio: fix: Keep a reference to the IIO device for open file descriptors
Lars-Peter Clausen [Wed, 18 Sep 2013 20:02:00 +0000 (21:02 +0100)]
iio: fix: Keep a reference to the IIO device for open file descriptors

Make sure that the IIO device is not freed while we still have file descriptors
for it.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
11 years agoiio: Stop sampling when the device is removed
Lars-Peter Clausen [Wed, 18 Sep 2013 20:02:00 +0000 (21:02 +0100)]
iio: Stop sampling when the device is removed

Make sure to stop sampling when the device is removed, otherwise it will
continue to sample forever.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
11 years agoiio: Fix crash when scan_bytes is computed with active_scan_mask == NULL
Peter Meerwald [Wed, 18 Sep 2013 21:10:00 +0000 (22:10 +0100)]
iio: Fix crash when scan_bytes is computed with active_scan_mask == NULL

if device has available_scan_masks set and the buffer is enabled without
any scan_elements enabled, in a NULL pointer is dereferenced in iio_compute_scan_bytes()

[   18.993713] Unable to handle kernel NULL pointer dereference at virtual address 00000000
[   19.002593] pgd = debd4000
[   19.005432] [00000000] *pgd=9ebc0831, *pte=00000000, *ppte=00000000
[   19.012329] Internal error: Oops: 17 [#1] PREEMPT ARM
[   19.017639] Modules linked in:
[   19.020843] CPU: 0    Not tainted  (3.9.11-00036-g75c888a-dirty #207)
[   19.027587] PC is at _find_first_bit_le+0xc/0x2c
[   19.032440] LR is at iio_compute_scan_bytes+0x2c/0xf4
[   19.037719] pc : [<c021dc60>]    lr : [<c03198d0>]    psr: 200d0013
[   19.037719] sp : debd9ed0  ip : 00000000  fp : 000802bc
[   19.049713] r10: 00000000  r9 : 00000000  r8 : deb67250
[   19.055206] r7 : 00000000  r6 : 00000000  r5 : 00000000  r4 : deb67000
[   19.062011] r3 : de96ec00  r2 : 00000000  r1 : 00000004  r0 : 00000000
[   19.068847] Flags: nzCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment user
[   19.076324] Control: 10c5387d  Table: 9ebd4019  DAC: 00000015

problem is the rollback code in iio_update_buffers(), old_mask may be NULL (e.g. on first
call)

I'm not too confident about the fix; works for me...

Signed-off-by: Peter Meerwald <pmeerw@pmeerw.net>
Reviewed-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
11 years agoiio: Fix mcp4725 dev-to-indio_dev conversion in suspend/resume
Peter Meerwald [Wed, 18 Sep 2013 21:47:00 +0000 (22:47 +0100)]
iio: Fix mcp4725 dev-to-indio_dev conversion in suspend/resume

dev_to_iio_dev() is a false friend

Signed-off-by: Peter Meerwald <pmeerw@pmeerw.net>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
11 years agoiio: Fix bma180 dev-to-indio_dev conversion in suspend/resume
Peter Meerwald [Wed, 18 Sep 2013 21:47:00 +0000 (22:47 +0100)]
iio: Fix bma180 dev-to-indio_dev conversion in suspend/resume

dev_to_iio_dev() is a false friend

Signed-off-by: Peter Meerwald <pmeerw@pmeerw.net>
Cc: Oleksandr Kravchenko <o.v.kravchenko@globallogic.com>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
11 years agoiio: Fix tmp006 dev-to-indio_dev conversion in suspend/resume
Peter Meerwald [Wed, 18 Sep 2013 21:47:00 +0000 (22:47 +0100)]
iio: Fix tmp006 dev-to-indio_dev conversion in suspend/resume

dev_to_iio_dev() is a false friend

Signed-off-by: Peter Meerwald <pmeerw@pmeerw.net>
Signed-off-by: Jonathan Cameron <jic23@kernel.org>
11 years agoMerge tag 'pm+acpi-3.12-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael...
Linus Torvalds [Fri, 20 Sep 2013 22:17:14 +0000 (15:17 -0700)]
Merge tag 'pm+acpi-3.12-rc2' of git://git./linux/kernel/git/rafael/linux-pm

Pull ACPI and power management fixes from Rafael Wysocki:

 1) Four fixes for cpufreq regressions introduced by the changes that
    removed Device Tree parsing for CPU device nodes from cpufreq
    drivers from Sudeep KarkadaNagesha.

 2) Two fixes for recent cpufreq regressions introduced by changes
    related to the preservation of sysfs attributes over system
    suspend/resume cycles from Viresh Kumar.

 3) Fix for ACPI-based wakeup signaling in the PCI subsystem that
    fails to stop PME polling for devices put into the D3cold power
    state from Rafael J Wysocki.

 4) Fix for bad interactions between cpufreq and udev on systems
    supporting intel_pstate where acpi-cpufreq is available as well
    from Yinghai Lu.

* tag 'pm+acpi-3.12-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  cpufreq: return EEXIST instead of EBUSY for second registering
  PCI / ACPI / PM: Clear pme_poll for devices in D3cold on wakeup
  ARM: shmobile: change dev_id to cpu0 while registering cpu clock
  ARM: i.MX: change dev_id to cpu0 while registering cpu clock
  cpufreq: imx6q-cpufreq: assign cpu_dev correctly to cpu0 device
  cpufreq: cpufreq-cpu0: assign cpu_dev correctly to cpu0 device
  cpufreq: unlock correct rwsem while updating policy->cpu
  cpufreq: Clear policy->cpus bits in __cpufreq_remove_dev_finish()

11 years agoMerge tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost
Linus Torvalds [Fri, 20 Sep 2013 22:16:15 +0000 (15:16 -0700)]
Merge tag 'for_linus' of git://git./linux/kernel/git/mst/vhost

Pull vhost updates from Michael Tsirkin:
 "vhost: minor changes on top of 3.12-rc1

  This fixes module loading for vhost-scsi, and tweaks locking in vhost
  core a bit.  Both of these are not exactly release blockers but it's
  early in the cycle so I think it's a good idea to apply them now"

* tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost:
  vhost-scsi: whitespace tweak
  vhost/scsi: use vmalloc for order-10 allocation
  vhost: wake up worker outside spin_lock

11 years agoCacheFiles: Don't try to dump the index key if the cookie has been cleared
David Howells [Fri, 20 Sep 2013 13:18:00 +0000 (14:18 +0100)]
CacheFiles: Don't try to dump the index key if the cookie has been cleared

Don't try to dump the index key that distinguishes an object if netfs
data in the cookie the object refers to has been cleared (ie.  the
cookie has passed most of the way through
__fscache_relinquish_cookie()).

Since the netfs holds the index key, we can't get at it once the ->def
and ->netfs_data pointers have been cleared - and a NULL pointer
exception will ensue, usually just after a:

CacheFiles: Error: Unexpected object collision

error is reported.

Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
11 years agoCacheFiles: Fix memory leak in cachefiles_check_auxdata error paths
Josh Boyer [Fri, 20 Sep 2013 13:17:52 +0000 (14:17 +0100)]
CacheFiles: Fix memory leak in cachefiles_check_auxdata error paths

In cachefiles_check_auxdata(), we allocate auxbuf but fail to free it if
we determine there's an error or that the data is stale.

Further, assigning the output of vfs_getxattr() to auxbuf->len gives
problems with checking for errors as auxbuf->len is a u16.  We don't
actually need to set auxbuf->len, so keep the length in a variable for
now.  We shouldn't need to check the upper limit of the buffer as an
overflow there should be indicated by -ERANGE.

While we're at it, fscache_check_aux() returns an enum value, not an
int, so assign it to an appropriately typed variable rather than to ret.

Signed-off-by: Josh Boyer <jwboyer@fedoraproject.org>
Signed-off-by: David Howells <dhowells@redhat.com>
cc: Hongyi Jia <jiayisuse@gmail.com>
cc: Milosz Tanski <milosz@adfin.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
11 years agolockref: use cmpxchg64 explicitly for lockless updates
Will Deacon [Thu, 19 Sep 2013 18:06:46 +0000 (19:06 +0100)]
lockref: use cmpxchg64 explicitly for lockless updates

The cmpxchg() function tends not to support 64-bit arguments on 32-bit
architectures.  This could be either due to use of unsigned long
arguments (like on ARM) or lack of instruction support (cmpxchgq on
x86).  However, these architectures may implement a specific cmpxchg64()
function to provide 64-bit cmpxchg support instead.

Since the lockref code requires a 64-bit cmpxchg and relies on the
architecture selecting ARCH_USE_CMPXCHG_LOCKREF, move to using cmpxchg64
instead of cmpxchg and allow 32-bit architectures to make use of the
lockless lockref implementation.

Cc: Waiman Long <Waiman.Long@hp.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
11 years agoMerge branch 'pm-cpufreq'
Rafael J. Wysocki [Fri, 20 Sep 2013 13:40:41 +0000 (15:40 +0200)]
Merge branch 'pm-cpufreq'

* pm-cpufreq:
  cpufreq: return EEXIST instead of EBUSY for second registering
  ARM: shmobile: change dev_id to cpu0 while registering cpu clock
  ARM: i.MX: change dev_id to cpu0 while registering cpu clock
  cpufreq: imx6q-cpufreq: assign cpu_dev correctly to cpu0 device
  cpufreq: cpufreq-cpu0: assign cpu_dev correctly to cpu0 device
  cpufreq: unlock correct rwsem while updating policy->cpu
  cpufreq: Clear policy->cpus bits in __cpufreq_remove_dev_finish()

11 years agoMerge branch 'acpi-pci'
Rafael J. Wysocki [Fri, 20 Sep 2013 13:40:30 +0000 (15:40 +0200)]
Merge branch 'acpi-pci'

* acpi-pci:
  PCI / ACPI / PM: Clear pme_poll for devices in D3cold on wakeup

11 years agoMerge tag 'arm64-stable' of git://git.kernel.org/pub/scm/linux/kernel/git/cmarinas...
Linus Torvalds [Fri, 20 Sep 2013 13:18:51 +0000 (08:18 -0500)]
Merge tag 'arm64-stable' of git://git./linux/kernel/git/cmarinas/linux-aarch64

Pull ARM64 fixes from Catalin Marinas:
 - Compat register fault reporting fix
 - Documentation clarification on tagged pointers
 - hwcap widened to 64-bit (user space already reading it as 64-bit)

* tag 'arm64-stable' of git://git.kernel.org/pub/scm/linux/kernel/git/cmarinas/linux-aarch64:
  arm64: Widen hwcap to be 64 bit
  arm64: Correctly report LR and SP for compat tasks
  arm64: documentation: tighten up tagged pointer documentation
  arm64: Make do_bad_area() function static

11 years agoarm64: Widen hwcap to be 64 bit
Steve Capper [Wed, 18 Sep 2013 15:14:28 +0000 (16:14 +0100)]
arm64: Widen hwcap to be 64 bit

Under arm64 elf_hwcap is a 32 bit quantity, but it is stored in
a 64 bit auxiliary ELF field and glibc reads hwcap as 64 bit.

This patch widens elf_hwcap to be 64 bit.

Signed-off-by: Steve Capper <steve.capper@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
11 years agoarm64: Correctly report LR and SP for compat tasks
Catalin Marinas [Tue, 17 Sep 2013 17:49:46 +0000 (18:49 +0100)]
arm64: Correctly report LR and SP for compat tasks

When a task crashes and we print debugging information, ensure that
compat tasks show the actual AArch32 LR and SP registers rather than the
AArch64 ones.

Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
11 years agoarm64: documentation: tighten up tagged pointer documentation
Will Deacon [Tue, 17 Sep 2013 10:46:23 +0000 (11:46 +0100)]
arm64: documentation: tighten up tagged pointer documentation

Commit d50240a5f6ce ("arm64: mm: permit use of tagged pointers at EL0")
added support for tagged pointers in userspace, but the corresponding
update to Documentation/ contained some imprecise statements.

This patch fixes up some minor ambiguities in the text, hopefully making
it more clear about exactly what the kernel expects from user virtual
addresses.

Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
11 years agoarm64: Make do_bad_area() function static
Catalin Marinas [Mon, 16 Sep 2013 14:18:28 +0000 (15:18 +0100)]
arm64: Make do_bad_area() function static

This function is only called from arch/arm64/mm/fault.c.

Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>