Merge branch 'akpm' (Andrew's patch-bomb)

Merge misc patches from Andrew Morton:
 "The MM tree is rather stuck while I wait to find out what the heck is
  happening with sched/numa.  Probably I'll need to route around all the
  code which was added to -next, sigh.

  So this is "everything else", or at least most of it - other small
  bits are still awaiting resolutions of various kinds."

* emailed patches from Andrew Morton <akpm@linux-foundation.org>: (180 commits)
  lib/decompress.c add __init to decompress_method and data
  kernel/resource.c: fix stack overflow in __reserve_region_with_split()
  omfs: convert to use beXX_add_cpu()
  taskstats: cgroupstats_user_cmd() may leak on error
  aoe: update aoe-internal version number to 50
  aoe: update documentation to better reflect aoe-plus-udev usage
  aoe: remove unused code
  aoe: make dynamic block minor numbers the default
  aoe: update and specify AoE address guards and error messages
  aoe: retain static block device numbers for backwards compatibility
  aoe: support more AoE addresses with dynamic block device minor numbers
  aoe: update documentation with new URL and VM settings reference
  aoe: update copyright year in touched files
  aoe: update internal version number to 49
  aoe: remove unused code and add cosmetic improvements
  aoe: increase net_device reference count while using it
  aoe: associate frames with the AoE storage target
  aoe: disallow unsupported AoE minor addresses
  aoe: do revalidation steps in order
  aoe: failover remote interface based on aoe_deadsecs parameter
  ...

Documentation/CodingStyle

@@ -454,6 +454,16 @@ The preferred style for long (multi-line) comments is:
 	 * with beginning and ending almost-blank lines.
 	 */
 
+For files in net/ and drivers/net/ the preferred style for long (multi-line)
+comments is a little different.
+
+	/* The preferred comment style for files in net/ and drivers/net
+	 * looks like this.
+	 *
+	 * It is nearly the same as the generally preferred comment style,
+	 * but there is no initial almost-blank line.
+	 */
+
 It's also important to comment data, whether they are basic types or derived
 types.  To this end, use just one data declaration per line (no commas for
 multiple data declarations).  This leaves you room for a small comment on each

Documentation/aoe/aoe.txt

@@ -1,8 +1,16 @@
-The EtherDrive (R) HOWTO for users of 2.6 kernels is found at ...
+ATA over Ethernet is a network protocol that provides simple access to
+block storage on the LAN.
 
-  http://www.coraid.com/SUPPORT/EtherDrive-HBA  
+  http://support.coraid.com/documents/AoEr11.txt
 
-  It has many tips and hints!
+The EtherDrive (R) HOWTO for 2.6 and 3.x kernels is found at ...
+
+  http://support.coraid.com/support/linux/EtherDrive-2.6-HOWTO.html
+
+It has many tips and hints!  Please see, especially, recommended
+tunings for virtual memory:
+
+  http://support.coraid.com/support/linux/EtherDrive-2.6-HOWTO-5.html#ss5.19
 
 The aoetools are userland programs that are designed to work with this
 driver.  The aoetools are on sourceforge.
@@ -23,20 +31,12 @@ CREATING DEVICE NODES
   There is a udev-install.sh script that shows how to install these
   rules on your system.
 
-  If you are not using udev, two scripts are provided in
-  Documentation/aoe as examples of static device node creation for
-  using the aoe driver.
-
-    rm -rf /dev/etherd
-    sh Documentation/aoe/mkdevs.sh /dev/etherd
-
-  ... or to make just one shelf's worth of block device nodes ...
-
-    sh Documentation/aoe/mkshelf.sh /dev/etherd 0
-
   There is also an autoload script that shows how to edit
   /etc/modprobe.d/aoe.conf to ensure that the aoe module is loaded when
-  necessary.
+  necessary.  Preloading the aoe module is preferable to autoloading,
+  however, because AoE discovery takes a few seconds.  It can be
+  confusing when an AoE device is not present the first time the a
+  command is run but appears a second later.
 
 USING DEVICE NODES
 
@@ -51,9 +51,9 @@ USING DEVICE NODES
   "echo > /dev/etherd/discover" tells the driver to find out what AoE
   devices are available.
 
-  These character devices may disappear and be replaced by sysfs
-  counterparts.  Using the commands in aoetools insulates users from
-  these implementation details.
+  In the future these character devices may disappear and be replaced
+  by sysfs counterparts.  Using the commands in aoetools insulates
+  users from these implementation details.
 
   The block devices are named like this:
 
@@ -76,8 +76,8 @@ USING SYSFS
   The netif attribute is the network interface on the localhost
   through which we are communicating with the remote AoE device.
 
-  There is a script in this directory that formats this information
-  in a convenient way.  Users with aoetools can use the aoe-stat
+  There is a script in this directory that formats this information in
+  a convenient way.  Users with aoetools should use the aoe-stat
   command.
 
   root@makki root# sh Documentation/aoe/status.sh 
@@ -121,3 +121,21 @@ DRIVER OPTIONS
   usage example for the module parameter.
 
     modprobe aoe_iflist="eth1 eth3"
+
+  The aoe_deadsecs module parameter determines the maximum number of
+  seconds that the driver will wait for an AoE device to provide a
+  response to an AoE command.  After aoe_deadsecs seconds have
+  elapsed, the AoE device will be marked as "down".
+
+  The aoe_maxout module parameter has a default of 128.  This is the
+  maximum number of unresponded packets that will be sent to an AoE
+  target at one time.
+
+  The aoe_dyndevs module parameter defaults to 1, meaning that the
+  driver will assign a block device minor number to a discovered AoE
+  target based on the order of its discovery.  With dynamic minor
+  device numbers in use, a greater range of AoE shelf and slot
+  addresses can be supported.  Users with udev will never have to
+  think about minor numbers.  Using aoe_dyndevs=0 allows device nodes
+  to be pre-created using a static minor-number scheme with the
+  aoe-mkshelf script in the aoetools.

Documentation/aoe/mkdevs.sh

@@ -1,41 +0,0 @@
-#!/bin/sh
-
-n_shelves=${n_shelves:-10}
-n_partitions=${n_partitions:-16}
-
-if test "$#" != "1"; then
-	echo "Usage: sh `basename $0` {dir}" 1>&2
-	echo "       n_partitions=16 sh `basename $0` {dir}" 1>&2
-	exit 1
-fi
-dir=$1
-
-MAJOR=152
-
-echo "Creating AoE devnode files in $dir ..."
-
-set -e
-
-mkdir -p $dir
-
-# (Status info is in sysfs.  See status.sh.)
-# rm -f $dir/stat
-# mknod -m 0400 $dir/stat c $MAJOR 1
-rm -f $dir/err
-mknod -m 0400 $dir/err c $MAJOR 2
-rm -f $dir/discover
-mknod -m 0200 $dir/discover c $MAJOR 3
-rm -f $dir/interfaces
-mknod -m 0200 $dir/interfaces c $MAJOR 4
-rm -f $dir/revalidate
-mknod -m 0200 $dir/revalidate c $MAJOR 5
-rm -f $dir/flush
-mknod -m 0200 $dir/flush c $MAJOR 6
-
-export n_partitions
-mkshelf=`echo $0 | sed 's!mkdevs!mkshelf!'`
-i=0
-while test $i -lt $n_shelves; do
-	sh -xc "sh $mkshelf $dir $i"
-	i=`expr $i + 1`
-done

Documentation/aoe/mkshelf.sh

@@ -1,28 +0,0 @@
-#! /bin/sh
-
-if test "$#" != "2"; then
-	echo "Usage: sh `basename $0` {dir} {shelfaddress}" 1>&2
-	echo "       n_partitions=16 sh `basename $0` {dir} {shelfaddress}" 1>&2
-	exit 1
-fi
-n_partitions=${n_partitions:-16}
-dir=$1
-shelf=$2
-nslots=16
-maxslot=`echo $nslots 1 - p | dc`
-MAJOR=152
-
-set -e
-
-minor=`echo $nslots \* $shelf \* $n_partitions | bc`
-endp=`echo $n_partitions - 1 | bc`
-for slot in `seq 0 $maxslot`; do
-	for part in `seq 0 $endp`; do
-		name=e$shelf.$slot
-		test "$part" != "0" && name=${name}p$part
-		rm -f $dir/$name
-		mknod -m 0660 $dir/$name b $MAJOR $minor
-
-		minor=`expr $minor + 1`
-	done
-done

Documentation/aoe/status.sh

@@ -1,5 +1,8 @@
 #! /bin/sh
 # collate and present sysfs information about AoE storage
+#
+# A more complete version of this script is aoe-stat, in the
+# aoetools.
 
 set -e
 format="%8s\t%8s\t%8s\n"

Documentation/devicetree/bindings/crypto/fsl-sec4.txt

@@ -9,6 +9,7 @@ Copyright (C) 2008-2011 Freescale Semiconductor Inc.
    -Run Time Integrity Check (RTIC) Node
    -Run Time Integrity Check (RTIC) Memory Node
    -Secure Non-Volatile Storage (SNVS) Node
+   -Secure Non-Volatile Storage (SNVS) Low Power (LP) RTC Node
    -Full Example
 
 NOTE: the SEC 4 is also known as Freescale's Cryptographic Accelerator
@@ -294,6 +295,27 @@ Secure Non-Volatile Storage (SNVS) Node
           address and length of the SEC4 configuration
           registers.
 
+   - #address-cells
+       Usage: required
+       Value type: <u32>
+       Definition: A standard property.  Defines the number of cells
+           for representing physical addresses in child nodes.  Must
+           have a value of 1.
+
+   - #size-cells
+       Usage: required
+       Value type: <u32>
+       Definition: A standard property.  Defines the number of cells
+           for representing the size of physical addresses in
+           child nodes.  Must have a value of 1.
+
+   - ranges
+       Usage: required
+       Value type: <prop-encoded-array>
+       Definition: A standard property.  Specifies the physical address
+           range of the SNVS register space.  A triplet that includes
+           the child address, parent address, & length.
+
    - interrupts
       Usage: required
       Value type: <prop_encoded-array>
@@ -314,10 +336,33 @@ EXAMPLE
 	sec_mon@314000 {
 		compatible = "fsl,sec-v4.0-mon";
 		reg = <0x314000 0x1000>;
+		ranges = <0 0x314000 0x1000>;
 		interrupt-parent = <&mpic>;
 		interrupts = <93 2>;
 	};
 
+=====================================================================
+Secure Non-Volatile Storage (SNVS) Low Power (LP) RTC Node
+
+  A SNVS child node that defines SNVS LP RTC.
+
+  - compatible
+      Usage: required
+      Value type: <string>
+      Definition: Must include "fsl,sec-v4.0-mon-rtc-lp".
+
+  - reg
+      Usage: required
+      Value type: <prop-encoded-array>
+      Definition: A standard property.  Specifies the physical
+          address and length of the SNVS LP configuration registers.
+
+EXAMPLE
+	sec_mon_rtc_lp@314000 {
+		compatible = "fsl,sec-v4.0-mon-rtc-lp";
+		reg = <0x34 0x58>;
+	};
+
 =====================================================================
 FULL EXAMPLE
 
@@ -390,8 +435,14 @@ FULL EXAMPLE
 	sec_mon: sec_mon@314000 {
 		compatible = "fsl,sec-v4.0-mon";
 		reg = <0x314000 0x1000>;
+		ranges = <0 0x314000 0x1000>;
 		interrupt-parent = <&mpic>;
 		interrupts = <93 2>;
+
+		sec_mon_rtc_lp@34 {
+			compatible = "fsl,sec-v4.0-mon-rtc-lp";
+			reg = <0x34 0x58>;
+		};
 	};
 
 =====================================================================

Documentation/devicetree/bindings/rtc/snvs-rtc.txt

@@ -0,0 +1 @@
+See Documentation/devicetree/bindings/crypto/fsl-sec4.txt for details.

Documentation/printk-formats.txt

@@ -69,6 +69,7 @@ MAC/FDDI addresses:
 	%pMR	05:04:03:02:01:00
 	%pMF	00-01-02-03-04-05
 	%pm	000102030405
+	%pmR	050403020100
 
 	For printing 6-byte MAC/FDDI addresses in hex notation. The 'M' and 'm'
 	specifiers result in a printed address with ('M') or without ('m') byte

Documentation/rtc.txt

@@ -119,8 +119,9 @@ three different userspace interfaces:
     *	/sys/class/rtc/rtcN ... sysfs attributes support readonly
 	access to some RTC attributes.
 
-    *	/proc/driver/rtc ... the first RTC (rtc0) may expose itself
-	using a procfs interface.  More information is (currently) shown
+    *	/proc/driver/rtc ... the system clock RTC may expose itself
+	using a procfs interface. If there is no RTC for the system clock,
+	rtc0 is used by default. More information is (currently) shown
 	here than through sysfs.
 
 The RTC Class framework supports a wide variety of RTCs, ranging from those

Documentation/sysctl/kernel.txt

@@ -181,6 +181,8 @@ core_pattern is used to specify a core dumpfile pattern name.
 	%p	pid
 	%u	uid
 	%g	gid
+	%d	dump mode, matches PR_SET_DUMPABLE and
+		/proc/sys/fs/suid_dumpable
 	%s	signal number
 	%t	UNIX time of dump
 	%h	hostname

MAINTAINERS

@@ -572,7 +572,7 @@ F:	drivers/net/appletalk/
 F:	net/appletalk/
 
 ARASAN COMPACT FLASH PATA CONTROLLER
-M:     Viresh Kumar <viresh.linux@gmail.com>
+M:	Viresh Kumar <viresh.linux@gmail.com>
 L:	linux-ide@vger.kernel.org
 S:	Maintained
 F:	include/linux/pata_arasan_cf_data.h
@@ -760,6 +760,7 @@ S:	Maintained
 T:	git git://git.pengutronix.de/git/imx/linux-2.6.git
 F:	arch/arm/mach-imx/
 F:	arch/arm/plat-mxc/
+F:	arch/arm/configs/imx*_defconfig
 
 ARM/FREESCALE IMX6
 M:	Shawn Guo <shawn.guo@linaro.org>
@@ -1245,7 +1246,7 @@ F:	include/linux/i2c/at24.h
 
 ATA OVER ETHERNET (AOE) DRIVER
 M:	"Ed L. Cashin" <ecashin@coraid.com>
-W:	http://www.coraid.com/support/linux
+W:	http://support.coraid.com/support/linux
 S:	Supported
 F:	Documentation/aoe/
 F:	drivers/block/aoe/
@@ -3123,6 +3124,7 @@ T:	git git://git.secretlab.ca/git/linux-2.6.git
 F:	Documentation/gpio.txt
 F:	drivers/gpio/
 F:	include/linux/gpio*
+F:	include/asm-generic/gpio.h
 
 GRE DEMULTIPLEXER DRIVER
 M:	Dmitry Kozlov <xeb@mail.ru>
@@ -5265,7 +5267,7 @@ F:	include/linux/i2c-algo-pca.h
 F:	include/linux/i2c-pca-platform.h
 
 PCDP - PRIMARY CONSOLE AND DEBUG PORT
-M:	Khalid Aziz <khalid.aziz@hp.com>
+M:	Khalid Aziz <khalid@gonehiking.org>
 S:	Maintained
 F:	drivers/firmware/pcdp.*
 
@@ -5554,7 +5556,7 @@ S:	Maintained
 W:	http://linuxptp.sourceforge.net/
 F:	Documentation/ABI/testing/sysfs-ptp
 F:	Documentation/ptp/*
-F:	drivers/net/gianfar_ptp.c
+F:	drivers/net/ethernet/freescale/gianfar_ptp.c
 F:	drivers/net/phy/dp83640*
 F:	drivers/ptp/*
 F:	include/linux/ptp_cl*
@@ -5986,7 +5988,7 @@ S:	Maintained
 F:	drivers/tty/serial
 
 SYNOPSYS DESIGNWARE DMAC DRIVER
-M:     Viresh Kumar <viresh.linux@gmail.com>
+M:	Viresh Kumar <viresh.linux@gmail.com>
 S:	Maintained
 F:	include/linux/dw_dmac.h
 F:	drivers/dma/dw_dmac_regs.h
@@ -6134,7 +6136,7 @@ S:	Maintained
 F:	drivers/mmc/host/sdhci-s3c.c
 
 SECURE DIGITAL HOST CONTROLLER INTERFACE (SDHCI) ST SPEAR DRIVER
-M:     Viresh Kumar <viresh.linux@gmail.com>
+M:	Viresh Kumar <viresh.linux@gmail.com>
 L:	spear-devel@list.st.com
 L:	linux-mmc@vger.kernel.org
 S:	Maintained
@@ -6499,7 +6501,7 @@ S:	Maintained
 F:	include/linux/compiler.h
 
 SPEAR PLATFORM SUPPORT
-M:     Viresh Kumar <viresh.linux@gmail.com>
+M:	Viresh Kumar <viresh.linux@gmail.com>
 M:	Shiraz Hashim <shiraz.hashim@st.com>
 L:	spear-devel@list.st.com
 L:	linux-arm-kernel@lists.infradead.org (moderated for non-subscribers)
@@ -6508,7 +6510,7 @@ S:	Maintained
 F:	arch/arm/plat-spear/
 
 SPEAR13XX MACHINE SUPPORT
-M:     Viresh Kumar <viresh.linux@gmail.com>
+M:	Viresh Kumar <viresh.linux@gmail.com>
 M:	Shiraz Hashim <shiraz.hashim@st.com>
 L:	spear-devel@list.st.com
 L:	linux-arm-kernel@lists.infradead.org (moderated for non-subscribers)
@@ -6517,7 +6519,7 @@ S:	Maintained
 F:	arch/arm/mach-spear13xx/
 
 SPEAR3XX MACHINE SUPPORT
-M:     Viresh Kumar <viresh.linux@gmail.com>
+M:	Viresh Kumar <viresh.linux@gmail.com>
 M:	Shiraz Hashim <shiraz.hashim@st.com>
 L:	spear-devel@list.st.com
 L:	linux-arm-kernel@lists.infradead.org (moderated for non-subscribers)
@@ -6528,7 +6530,7 @@ F:	arch/arm/mach-spear3xx/
 SPEAR6XX MACHINE SUPPORT
 M:	Rajeev Kumar <rajeev-dlh.kumar@st.com>
 M:	Shiraz Hashim <shiraz.hashim@st.com>
-M:     Viresh Kumar <viresh.linux@gmail.com>
+M:	Viresh Kumar <viresh.linux@gmail.com>
 L:	spear-devel@list.st.com
 L:	linux-arm-kernel@lists.infradead.org (moderated for non-subscribers)
 W:	http://www.st.com/spear
@@ -6536,7 +6538,7 @@ S:	Maintained
 F:	arch/arm/mach-spear6xx/
 
 SPEAR CLOCK FRAMEWORK SUPPORT
-M:     Viresh Kumar <viresh.linux@gmail.com>
+M:	Viresh Kumar <viresh.linux@gmail.com>
 L:	spear-devel@list.st.com
 L:	linux-arm-kernel@lists.infradead.org (moderated for non-subscribers)
 W:	http://www.st.com/spear

arch/arm/mach-at91/clock.c

@@ -625,7 +625,7 @@ fail:
 	return 0;
 }
 
-static struct clk *const standard_pmc_clocks[] __initdata = {
+static struct clk *const standard_pmc_clocks[] __initconst = {
 	/* four primary clocks */
 	&clk32k,
 	&main_clk,

arch/arm/mach-davinci/board-tnetv107x-evm.c

@@ -88,7 +88,7 @@ static struct davinci_mmc_config mmc_config = {
 	.version	= MMC_CTLR_VERSION_1,
 };
 
-static const short sdio1_pins[] __initdata = {
+static const short sdio1_pins[] __initconst = {
 	TNETV107X_SDIO1_CLK_1,		TNETV107X_SDIO1_CMD_1,
 	TNETV107X_SDIO1_DATA0_1,	TNETV107X_SDIO1_DATA1_1,
 	TNETV107X_SDIO1_DATA2_1,	TNETV107X_SDIO1_DATA3_1,
@@ -96,12 +96,12 @@ static const short sdio1_pins[] __initdata = {
 	-1
 };
 
-static const short uart1_pins[] __initdata = {
+static const short uart1_pins[] __initconst = {
 	TNETV107X_UART1_RD,		TNETV107X_UART1_TD,
 	-1
 };
 
-static const short ssp_pins[] __initdata = {
+static const short ssp_pins[] __initconst = {
 	TNETV107X_SSP0_0, TNETV107X_SSP0_1, TNETV107X_SSP0_2,
 	TNETV107X_SSP1_0, TNETV107X_SSP1_1, TNETV107X_SSP1_2,
 	TNETV107X_SSP1_3, -1

arch/arm/mach-davinci/da830.c

@@ -838,7 +838,7 @@ static const struct mux_config da830_pins[] = {
 #endif
 };
 
-const short da830_emif25_pins[] __initdata = {
+const short da830_emif25_pins[] __initconst = {
 	DA830_EMA_D_0, DA830_EMA_D_1, DA830_EMA_D_2, DA830_EMA_D_3,
 	DA830_EMA_D_4, DA830_EMA_D_5, DA830_EMA_D_6, DA830_EMA_D_7,
 	DA830_EMA_D_8, DA830_EMA_D_9, DA830_EMA_D_10, DA830_EMA_D_11,
@@ -853,19 +853,19 @@ const short da830_emif25_pins[] __initdata = {
 	-1
 };
 
-const short da830_spi0_pins[] __initdata = {
+const short da830_spi0_pins[] __initconst = {
 	DA830_SPI0_SOMI_0, DA830_SPI0_SIMO_0, DA830_SPI0_CLK, DA830_NSPI0_ENA,
 	DA830_NSPI0_SCS_0,
 	-1
 };
 
-const short da830_spi1_pins[] __initdata = {
+const short da830_spi1_pins[] __initconst = {
 	DA830_SPI1_SOMI_0, DA830_SPI1_SIMO_0, DA830_SPI1_CLK, DA830_NSPI1_ENA,
 	DA830_NSPI1_SCS_0,
 	-1
 };
 
-const short da830_mmc_sd_pins[] __initdata = {
+const short da830_mmc_sd_pins[] __initconst = {
 	DA830_MMCSD_DAT_0, DA830_MMCSD_DAT_1, DA830_MMCSD_DAT_2,
 	DA830_MMCSD_DAT_3, DA830_MMCSD_DAT_4, DA830_MMCSD_DAT_5,
 	DA830_MMCSD_DAT_6, DA830_MMCSD_DAT_7, DA830_MMCSD_CLK,
@@ -873,32 +873,32 @@ const short da830_mmc_sd_pins[] __initdata = {
 	-1
 };
 
-const short da830_uart0_pins[] __initdata = {
+const short da830_uart0_pins[] __initconst = {
 	DA830_NUART0_CTS, DA830_NUART0_RTS, DA830_UART0_RXD, DA830_UART0_TXD,
 	-1
 };
 
-const short da830_uart1_pins[] __initdata = {
+const short da830_uart1_pins[] __initconst = {
 	DA830_UART1_RXD, DA830_UART1_TXD,
 	-1
 };
 
-const short da830_uart2_pins[] __initdata = {
+const short da830_uart2_pins[] __initconst = {
 	DA830_UART2_RXD, DA830_UART2_TXD,
 	-1
 };
 
-const short da830_usb20_pins[] __initdata = {
+const short da830_usb20_pins[] __initconst = {
 	DA830_USB0_DRVVBUS, DA830_USB_REFCLKIN,
 	-1
 };
 
-const short da830_usb11_pins[] __initdata = {
+const short da830_usb11_pins[] __initconst = {
 	DA830_USB_REFCLKIN,
 	-1
 };
 
-const short da830_uhpi_pins[] __initdata = {
+const short da830_uhpi_pins[] __initconst = {
 	DA830_UHPI_HD_0, DA830_UHPI_HD_1, DA830_UHPI_HD_2, DA830_UHPI_HD_3,
 	DA830_UHPI_HD_4, DA830_UHPI_HD_5, DA830_UHPI_HD_6, DA830_UHPI_HD_7,
 	DA830_UHPI_HD_8, DA830_UHPI_HD_9, DA830_UHPI_HD_10, DA830_UHPI_HD_11,
@@ -909,14 +909,14 @@ const short da830_uhpi_pins[] __initdata = {
 	-1
 };
 
-const short da830_cpgmac_pins[] __initdata = {
+const short da830_cpgmac_pins[] __initconst = {
 	DA830_RMII_TXD_0, DA830_RMII_TXD_1, DA830_RMII_TXEN, DA830_RMII_CRS_DV,
 	DA830_RMII_RXD_0, DA830_RMII_RXD_1, DA830_RMII_RXER, DA830_MDIO_CLK,
 	DA830_MDIO_D,
 	-1
 };
 
-const short da830_emif3c_pins[] __initdata = {
+const short da830_emif3c_pins[] __initconst = {
 	DA830_EMB_SDCKE, DA830_EMB_CLK_GLUE, DA830_EMB_CLK, DA830_NEMB_CS_0,
 	DA830_NEMB_CAS, DA830_NEMB_RAS, DA830_NEMB_WE, DA830_EMB_BA_1,
 	DA830_EMB_BA_0, DA830_EMB_A_0, DA830_EMB_A_1, DA830_EMB_A_2,
@@ -935,7 +935,7 @@ const short da830_emif3c_pins[] __initdata = {
 	-1
 };
 
-const short da830_mcasp0_pins[] __initdata = {
+const short da830_mcasp0_pins[] __initconst = {
 	DA830_AHCLKX0, DA830_ACLKX0, DA830_AFSX0,
 	DA830_AHCLKR0, DA830_ACLKR0, DA830_AFSR0, DA830_AMUTE0,
 	DA830_AXR0_0, DA830_AXR0_1, DA830_AXR0_2, DA830_AXR0_3,
@@ -945,7 +945,7 @@ const short da830_mcasp0_pins[] __initdata = {
 	-1
 };
 
-const short da830_mcasp1_pins[] __initdata = {
+const short da830_mcasp1_pins[] __initconst = {
 	DA830_AHCLKX1, DA830_ACLKX1, DA830_AFSX1,
 	DA830_AHCLKR1, DA830_ACLKR1, DA830_AFSR1, DA830_AMUTE1,
 	DA830_AXR1_0, DA830_AXR1_1, DA830_AXR1_2, DA830_AXR1_3,
@@ -954,24 +954,24 @@ const short da830_mcasp1_pins[] __initdata = {
 	-1
 };
 
-const short da830_mcasp2_pins[] __initdata = {
+const short da830_mcasp2_pins[] __initconst = {
 	DA830_AHCLKX2, DA830_ACLKX2, DA830_AFSX2,
 	DA830_AHCLKR2, DA830_ACLKR2, DA830_AFSR2, DA830_AMUTE2,
 	DA830_AXR2_0, DA830_AXR2_1, DA830_AXR2_2, DA830_AXR2_3,
 	-1
 };
 
-const short da830_i2c0_pins[] __initdata = {
+const short da830_i2c0_pins[] __initconst = {
 	DA830_I2C0_SDA, DA830_I2C0_SCL,
 	-1
 };
 
-const short da830_i2c1_pins[] __initdata = {
+const short da830_i2c1_pins[] __initconst = {
 	DA830_I2C1_SCL, DA830_I2C1_SDA,
 	-1
 };
 
-const short da830_lcdcntl_pins[] __initdata = {
+const short da830_lcdcntl_pins[] __initconst = {
 	DA830_LCD_D_0, DA830_LCD_D_1, DA830_LCD_D_2, DA830_LCD_D_3,
 	DA830_LCD_D_4, DA830_LCD_D_5, DA830_LCD_D_6, DA830_LCD_D_7,
 	DA830_LCD_D_8, DA830_LCD_D_9, DA830_LCD_D_10, DA830_LCD_D_11,
@@ -981,34 +981,34 @@ const short da830_lcdcntl_pins[] __initdata = {
 	-1
 };
 
-const short da830_pwm_pins[] __initdata = {
+const short da830_pwm_pins[] __initconst = {
 	DA830_ECAP0_APWM0, DA830_ECAP1_APWM1, DA830_EPWM0B, DA830_EPWM0A,
 	DA830_EPWMSYNCI, DA830_EPWMSYNC0, DA830_ECAP2_APWM2, DA830_EHRPWMGLUETZ,
 	DA830_EPWM2B, DA830_EPWM2A, DA830_EPWM1B, DA830_EPWM1A,
 	-1
 };
 
-const short da830_ecap0_pins[] __initdata = {
+const short da830_ecap0_pins[] __initconst = {
 	DA830_ECAP0_APWM0,
 	-1
 };
 
-const short da830_ecap1_pins[] __initdata = {
+const short da830_ecap1_pins[] __initconst = {
 	DA830_ECAP1_APWM1,
 	-1
 };
 
-const short da830_ecap2_pins[] __initdata = {
+const short da830_ecap2_pins[] __initconst = {
 	DA830_ECAP2_APWM2,
 	-1
 };
 
-const short da830_eqep0_pins[] __initdata = {
+const short da830_eqep0_pins[] __initconst = {
 	DA830_EQEP0I, DA830_EQEP0S, DA830_EQEP0A, DA830_EQEP0B,
 	-1
 };
 
-const short da830_eqep1_pins[] __initdata = {
+const short da830_eqep1_pins[] __initconst = {
 	DA830_EQEP1I, DA830_EQEP1S, DA830_EQEP1A, DA830_EQEP1B,
 	-1
 };

arch/arm/mach-davinci/da850.c

@@ -576,17 +576,17 @@ static const struct mux_config da850_pins[] = {
 #endif
 };
 
-const short da850_i2c0_pins[] __initdata = {
+const short da850_i2c0_pins[] __initconst = {
 	DA850_I2C0_SDA, DA850_I2C0_SCL,
 	-1
 };
 
-const short da850_i2c1_pins[] __initdata = {
+const short da850_i2c1_pins[] __initconst = {
 	DA850_I2C1_SCL, DA850_I2C1_SDA,
 	-1
 };
 
-const short da850_lcdcntl_pins[] __initdata = {
+const short da850_lcdcntl_pins[] __initconst = {
 	DA850_LCD_D_0, DA850_LCD_D_1, DA850_LCD_D_2, DA850_LCD_D_3,
 	DA850_LCD_D_4, DA850_LCD_D_5, DA850_LCD_D_6, DA850_LCD_D_7,
 	DA850_LCD_D_8, DA850_LCD_D_9, DA850_LCD_D_10, DA850_LCD_D_11,

arch/arm/mach-msm/board-qsd8x50.c

@@ -37,8 +37,8 @@
 #include "devices.h"
 #include "common.h"
 
-static const resource_size_t qsd8x50_surf_smc91x_base __initdata = 0x70000300;
-static const unsigned        qsd8x50_surf_smc91x_gpio __initdata = 156;
+static const resource_size_t qsd8x50_surf_smc91x_base __initconst = 0x70000300;
+static const unsigned        qsd8x50_surf_smc91x_gpio __initconst = 156;
 
 /* Leave smc91x resources empty here, as we'll fill them in
  * at run-time: they vary from board to board, and the true

arch/arm/mach-omap2/display.c

@@ -76,14 +76,14 @@ struct omap_dss_hwmod_data {
 	const int id;
 };
 
-static const struct omap_dss_hwmod_data omap2_dss_hwmod_data[] __initdata = {
+static const struct omap_dss_hwmod_data omap2_dss_hwmod_data[] __initconst = {
 	{ "dss_core", "omapdss_dss", -1 },
 	{ "dss_dispc", "omapdss_dispc", -1 },
 	{ "dss_rfbi", "omapdss_rfbi", -1 },
 	{ "dss_venc", "omapdss_venc", -1 },
 };
 
-static const struct omap_dss_hwmod_data omap3_dss_hwmod_data[] __initdata = {
+static const struct omap_dss_hwmod_data omap3_dss_hwmod_data[] __initconst = {
 	{ "dss_core", "omapdss_dss", -1 },
 	{ "dss_dispc", "omapdss_dispc", -1 },
 	{ "dss_rfbi", "omapdss_rfbi", -1 },
@@ -91,7 +91,7 @@ static const struct omap_dss_hwmod_data omap3_dss_hwmod_data[] __initdata = {
 	{ "dss_dsi1", "omapdss_dsi", 0 },
 };
 
-static const struct omap_dss_hwmod_data omap4_dss_hwmod_data[] __initdata = {
+static const struct omap_dss_hwmod_data omap4_dss_hwmod_data[] __initconst = {
 	{ "dss_core", "omapdss_dss", -1 },
 	{ "dss_dispc", "omapdss_dispc", -1 },
 	{ "dss_rfbi", "omapdss_rfbi", -1 },

arch/arm64/include/asm/compat.h

@@ -55,6 +55,7 @@ typedef s64		compat_s64;
 typedef u32		compat_uint_t;
 typedef u32		compat_ulong_t;
 typedef u64		compat_u64;
+typedef u32		compat_uptr_t;
 
 struct compat_timespec {
 	compat_time_t	tv_sec;
@@ -130,6 +131,64 @@ typedef u32		compat_old_sigset_t;
 
 typedef u32		compat_sigset_word;
 
+typedef union compat_sigval {
+	compat_int_t	sival_int;
+	compat_uptr_t	sival_ptr;
+} compat_sigval_t;
+
+typedef struct compat_siginfo {
+	int si_signo;
+	int si_errno;
+	int si_code;
+
+	union {
+		/* The padding is the same size as AArch64. */
+		int _pad[128/sizeof(int) - 3];
+
+		/* kill() */
+		struct {
+			compat_pid_t _pid;	/* sender's pid */
+			__compat_uid32_t _uid;	/* sender's uid */
+		} _kill;
+
+		/* POSIX.1b timers */
+		struct {
+			compat_timer_t _tid;	/* timer id */
+			int _overrun;		/* overrun count */
+			compat_sigval_t _sigval;	/* same as below */
+			int _sys_private;       /* not to be passed to user */
+		} _timer;
+
+		/* POSIX.1b signals */
+		struct {
+			compat_pid_t _pid;	/* sender's pid */
+			__compat_uid32_t _uid;	/* sender's uid */
+			compat_sigval_t _sigval;
+		} _rt;
+
+		/* SIGCHLD */
+		struct {
+			compat_pid_t _pid;	/* which child */
+			__compat_uid32_t _uid;	/* sender's uid */
+			int _status;		/* exit code */
+			compat_clock_t _utime;
+			compat_clock_t _stime;
+		} _sigchld;
+
+		/* SIGILL, SIGFPE, SIGSEGV, SIGBUS */
+		struct {
+			compat_uptr_t _addr; /* faulting insn/memory ref. */
+			short _addr_lsb; /* LSB of the reported address */
+		} _sigfault;
+
+		/* SIGPOLL */
+		struct {
+			compat_long_t _band;	/* POLL_IN, POLL_OUT, POLL_MSG */
+			int _fd;
+		} _sigpoll;
+	} _sifields;
+} compat_siginfo_t;
+
 #define COMPAT_OFF_T_MAX	0x7fffffff
 #define COMPAT_LOFF_T_MAX	0x7fffffffffffffffL
 
@@ -139,7 +198,6 @@ typedef u32		compat_sigset_word;
  * as pointers because the syscall entry code will have
  * appropriately converted them already.
  */
-typedef	u32		compat_uptr_t;
 
 static inline void __user *compat_ptr(compat_uptr_t uptr)
 {

arch/arm64/kernel/signal32.c

@@ -30,59 +30,6 @@
 #include <asm/uaccess.h>
 #include <asm/unistd.h>
 
-typedef struct compat_siginfo {
-	int si_signo;
-	int si_errno;
-	int si_code;
-
-	union {
-		/* The padding is the same size as AArch64. */
-		int _pad[SI_PAD_SIZE];
-
-		/* kill() */
-		struct {
-			compat_pid_t _pid;	/* sender's pid */
-			__compat_uid32_t _uid;	/* sender's uid */
-		} _kill;
-
-		/* POSIX.1b timers */
-		struct {
-			compat_timer_t _tid;	/* timer id */
-			int _overrun;		/* overrun count */
-			compat_sigval_t _sigval;	/* same as below */
-			int _sys_private;       /* not to be passed to user */
-		} _timer;
-
-		/* POSIX.1b signals */
-		struct {
-			compat_pid_t _pid;	/* sender's pid */
-			__compat_uid32_t _uid;	/* sender's uid */
-			compat_sigval_t _sigval;
-		} _rt;
-
-		/* SIGCHLD */
-		struct {
-			compat_pid_t _pid;	/* which child */
-			__compat_uid32_t _uid;	/* sender's uid */
-			int _status;		/* exit code */
-			compat_clock_t _utime;
-			compat_clock_t _stime;
-		} _sigchld;
-
-		/* SIGILL, SIGFPE, SIGSEGV, SIGBUS */
-		struct {
-			compat_uptr_t _addr; /* faulting insn/memory ref. */
-			short _addr_lsb; /* LSB of the reported address */
-		} _sigfault;
-
-		/* SIGPOLL */
-		struct {
-			compat_long_t _band;	/* POLL_IN, POLL_OUT, POLL_MSG */
-			int _fd;
-		} _sigpoll;
-	} _sifields;
-} compat_siginfo_t;
-
 struct compat_sigaction {
 	compat_uptr_t			sa_handler;
 	compat_ulong_t			sa_flags;

arch/avr32/include/asm/elf.h

@@ -102,6 +102,7 @@ typedef struct user_fpu_struct elf_fpregset_t;
 
 #define ELF_PLATFORM  (NULL)
 
-#define SET_PERSONALITY(ex) set_personality(PER_LINUX_32BIT)
+#define SET_PERSONALITY(ex) \
+	set_personality(PER_LINUX_32BIT | (current->personality & (~PER_MASK)))
 
 #endif /* __ASM_AVR32_ELF_H */

arch/blackfin/include/asm/elf.h

@@ -132,6 +132,7 @@ do {											\
 
 #define ELF_PLATFORM  (NULL)
 
-#define SET_PERSONALITY(ex) set_personality(PER_LINUX)
+#define SET_PERSONALITY(ex) \
+	set_personality(PER_LINUX | (current->personality & (~PER_MASK)))
 
 #endif

arch/c6x/include/asm/elf.h

@@ -77,7 +77,8 @@ do {								\
 
 #define ELF_PLATFORM  (NULL)
 
-#define SET_PERSONALITY(ex) set_personality(PER_LINUX)
+#define SET_PERSONALITY(ex) \
+	set_personality(PER_LINUX | (current->personality & (~PER_MASK)))
 
 /* C6X specific section types */
 #define SHT_C6000_UNWIND	0x70000001

arch/cris/include/asm/elf.h

@@ -86,6 +86,7 @@ typedef unsigned long elf_fpregset_t;
 
 #define ELF_PLATFORM  (NULL)
 
-#define SET_PERSONALITY(ex) set_personality(PER_LINUX)
+#define SET_PERSONALITY(ex) \
+	set_personality(PER_LINUX | (current->personality & (~PER_MASK)))
 
 #endif

arch/frv/include/asm/elf.h

@@ -137,6 +137,7 @@ do {											\
 
 #define ELF_PLATFORM  (NULL)
 
-#define SET_PERSONALITY(ex) set_personality(PER_LINUX)
+#define SET_PERSONALITY(ex) \
+	set_personality(PER_LINUX | (current->personality & (~PER_MASK)))
 
 #endif

arch/frv/kernel/pm.c

@@ -153,23 +153,22 @@ static int user_atoi(char __user *ubuf, size_t len)
 static int sysctl_pm_do_suspend(ctl_table *ctl, int write,
 				void __user *buffer, size_t *lenp, loff_t *fpos)
 {
-	int retval, mode;
+	int mode;
 
 	if (*lenp <= 0)
 		return -EIO;
 
 	mode = user_atoi(buffer, *lenp);
-	if ((mode != 1) && (mode != 5))
-		return -EINVAL;
+	switch (mode) {
+	case 1:
+	    return pm_do_suspend();
 
-	if (retval == 0) {
-		if (mode == 5)
-		    retval = pm_do_bus_sleep();
-		else
-		    retval = pm_do_suspend();
-	}
+	case 5:
+	    return pm_do_bus_sleep();
 
-	return retval;
+	default:
+	    return -EINVAL;
+	}
 }
 
 static int try_set_cmode(int new_cmode)

arch/frv/kernel/setup.c

@@ -184,7 +184,7 @@ static struct clock_cmode __pminitdata clock_cmodes_fr555[16] = {
 	[6]	= {	_x1,	_x1_5,	_x1_5,	_x4_5,	_x0_375	},
 };
 
-static const struct clock_cmode __pminitdata *clock_cmodes;
+static const struct clock_cmode __pminitconst *clock_cmodes;
 static int __pminitdata clock_doubled;
 
 static struct uart_port __pminitdata __frv_uart0 = {

arch/frv/mb93090-mb00/pci-irq.c

@@ -28,7 +28,7 @@
  *
  */
 
-static const uint8_t __initdata pci_bus0_irq_routing[32][4] = {
+static const uint8_t __initconst pci_bus0_irq_routing[32][4] = {
 	[0 ] = { IRQ_FPGA_MB86943_PCI_INTA },
 	[16] = { IRQ_FPGA_RTL8029_INTA },
 	[17] = { IRQ_FPGA_PCI_INTC, IRQ_FPGA_PCI_INTD, IRQ_FPGA_PCI_INTA, IRQ_FPGA_PCI_INTB },

arch/h8300/include/asm/elf.h

@@ -54,7 +54,8 @@ typedef unsigned long elf_fpregset_t;
 
 #define ELF_PLATFORM  (NULL)
 
-#define SET_PERSONALITY(ex) set_personality(PER_LINUX)
+#define SET_PERSONALITY(ex) \
+	set_personality(PER_LINUX | (current->personality & (~PER_MASK)))
 
 #define R_H8_NONE       0
 #define R_H8_DIR32      1

arch/h8300/kernel/sys_h8300.c

@@ -51,6 +51,7 @@ asmlinkage void syscall_print(void *dummy,...)
  * Do a system call from kernel instead of calling sys_execve so we
  * end up with proper pt_regs.
  */
+asmlinkage
 int kernel_execve(const char *filename,
 		  const char *const argv[],
 		  const char *const envp[])

arch/h8300/kernel/timer/itu.c

@@ -62,7 +62,7 @@ static struct irqaction itu_irq = {
 	.flags		= IRQF_DISABLED | IRQF_TIMER,
 };
 
-static const int __initdata divide_rate[] = {1, 2, 4, 8};
+static const int __initconst divide_rate[] = {1, 2, 4, 8};
 
 void __init h8300_timer_setup(void)
 {

arch/h8300/kernel/timer/timer16.c

@@ -57,7 +57,7 @@ static struct irqaction timer16_irq = {
 	.flags		= IRQF_DISABLED | IRQF_TIMER,
 };
 
-static const int __initdata divide_rate[] = {1, 2, 4, 8};
+static const int __initconst divide_rate[] = {1, 2, 4, 8};
 
 void __init h8300_timer_setup(void)
 {

arch/h8300/kernel/timer/timer8.c

@@ -77,7 +77,7 @@ static struct irqaction timer8_irq = {
 	.flags		= IRQF_DISABLED | IRQF_TIMER,
 };
 
-static const int __initdata divide_rate[] = {8, 64, 8192};
+static const int __initconst divide_rate[] = {8, 64, 8192};
 
 void __init h8300_timer_setup(void)
 {

arch/h8300/kernel/timer/tpu.c

@@ -66,7 +66,7 @@ static struct irqaction tpu_irq = {
 	.flags		= IRQF_DISABLED | IRQF_TIMER,
 };
 
-static const int __initdata divide_rate[] = {
+static const int __initconst divide_rate[] = {
 #if CONFIG_H8300_TPU_CH == 0
 	1,4,16,64,0,0,0,0,
 #elif (CONFIG_H8300_TPU_CH == 1) || (CONFIG_H8300_TPU_CH == 5)

arch/h8300/platform/h8300h/irq.c

@@ -14,14 +14,14 @@
 #include <asm/gpio-internal.h>
 #include <asm/regs306x.h>
 
-const int __initdata h8300_saved_vectors[] = {
+const int __initconst h8300_saved_vectors[] = {
 #if defined(CONFIG_GDB_DEBUG)
 	TRAP3_VEC,	/* TRAPA #3 is GDB breakpoint */
 #endif
 	-1,
 };
 
-const h8300_vector __initdata h8300_trap_table[] = {
+const h8300_vector __initconst h8300_trap_table[] = {
 	0, 0, 0, 0, 0, 0, 0, 0,
 	system_call,
 	0,

arch/h8300/platform/h8s/irq.c

@@ -18,7 +18,7 @@
 #include <asm/regs267x.h>
 
 /* saved vector list */
-const int __initdata h8300_saved_vectors[]={
+const int __initconst h8300_saved_vectors[] = {
 #if defined(CONFIG_GDB_DEBUG)
 	TRACE_VEC,
 	TRAP3_VEC,
@@ -27,7 +27,7 @@ const int __initdata h8300_saved_vectors[]={
 };
 
 /* trap entry table */
-const H8300_VECTOR __initdata h8300_trap_table[] = {
+const H8300_VECTOR __initconst h8300_trap_table[] = {
 	0,0,0,0,0,
 	trace_break,  /* TRACE */
 	0,0,

arch/hexagon/include/asm/elf.h

@@ -217,7 +217,8 @@ do {					\
 #define ELF_PLATFORM  (NULL)
 
 #ifdef __KERNEL__
-#define SET_PERSONALITY(ex) set_personality(PER_LINUX)
+#define SET_PERSONALITY(ex) \
+	set_personality(PER_LINUX | (current->personality & (~PER_MASK)))
 #endif
 
 #define ARCH_HAS_SETUP_ADDITIONAL_PAGES 1

arch/ia64/xen/irq_xen.c

@@ -433,7 +433,7 @@ xen_resend_irq(unsigned int vector)
 	(void)resend_irq_on_evtchn(vector);
 }
 
-const struct pv_irq_ops xen_irq_ops __initdata = {
+const struct pv_irq_ops xen_irq_ops __initconst = {
 	.register_ipi = xen_register_ipi,
 
 	.assign_irq_vector = xen_assign_irq_vector,

arch/ia64/xen/irq_xen.h

@@ -27,7 +27,7 @@ extern void (*late_time_init)(void);
 extern char xen_event_callback;
 void __init xen_init_IRQ(void);
 
-extern const struct pv_irq_ops xen_irq_ops __initdata;
+extern const struct pv_irq_ops xen_irq_ops __initconst;
 extern void xen_smp_intr_init(void);
 extern void xen_send_ipi(int cpu, int vec);
 

arch/m32r/include/asm/elf.h

@@ -128,6 +128,7 @@ typedef elf_fpreg_t elf_fpregset_t;
    intent than poking at uname or /proc/cpuinfo.  */
 #define ELF_PLATFORM	(NULL)
 
-#define SET_PERSONALITY(ex) set_personality(PER_LINUX)
+#define SET_PERSONALITY(ex) \
+	set_personality(PER_LINUX | (current->personality & (~PER_MASK)))
 
 #endif  /* _ASM_M32R__ELF_H */

arch/m68k/include/asm/elf.h

@@ -113,6 +113,7 @@ typedef struct user_m68kfp_struct elf_fpregset_t;
 
 #define ELF_PLATFORM  (NULL)
 
-#define SET_PERSONALITY(ex) set_personality(PER_LINUX)
+#define SET_PERSONALITY(ex) \
+	set_personality(PER_LINUX | (current->personality & (~PER_MASK)))
 
 #endif

arch/microblaze/include/asm/elf.h

@@ -116,7 +116,8 @@ do {							\
 } while (0)
 
 #ifdef __KERNEL__
-#define SET_PERSONALITY(ex) set_personality(PER_LINUX_32BIT)
+#define SET_PERSONALITY(ex) \
+	set_personality(PER_LINUX_32BIT | (current->personality & (~PER_MASK)))
 #endif
 
 #endif /* __uClinux__ */

arch/mips/bcm63xx/boards/board_bcm963xx.c

@@ -632,7 +632,7 @@ static struct board_info __initdata board_DWVS0 = {
 /*
  * all boards
  */
-static const struct board_info __initdata *bcm963xx_boards[] = {
+static const struct board_info __initconst *bcm963xx_boards[] = {
 #ifdef CONFIG_BCM63XX_CPU_6328
 	&board_96328avng,
 #endif

arch/mips/include/asm/compat-signal.h

@@ -10,68 +10,6 @@
 
 #include <asm/uaccess.h>
 
-#define SI_PAD_SIZE32   ((SI_MAX_SIZE/sizeof(int)) - 3)
-
-typedef struct compat_siginfo {
-	int si_signo;
-	int si_code;
-	int si_errno;
-
-	union {
-		int _pad[SI_PAD_SIZE32];
-
-		/* kill() */
-		struct {
-			compat_pid_t _pid;	/* sender's pid */
-			compat_uid_t _uid;	/* sender's uid */
-		} _kill;
-
-		/* SIGCHLD */
-		struct {
-			compat_pid_t _pid;	/* which child */
-			compat_uid_t _uid;	/* sender's uid */
-			int _status;		/* exit code */
-			compat_clock_t _utime;
-			compat_clock_t _stime;
-		} _sigchld;
-
-		/* IRIX SIGCHLD */
-		struct {
-			compat_pid_t _pid;	/* which child */
-			compat_clock_t _utime;
-			int _status;		/* exit code */
-			compat_clock_t _stime;
-		} _irix_sigchld;
-
-		/* SIGILL, SIGFPE, SIGSEGV, SIGBUS */
-		struct {
-			s32 _addr; /* faulting insn/memory ref. */
-		} _sigfault;
-
-		/* SIGPOLL, SIGXFSZ (To do ...)  */
-		struct {
-			int _band;	/* POLL_IN, POLL_OUT, POLL_MSG */
-			int _fd;
-		} _sigpoll;
-
-		/* POSIX.1b timers */
-		struct {
-			timer_t _tid;		/* timer id */
-			int _overrun;		/* overrun count */
-			compat_sigval_t _sigval;/* same as below */
-			int _sys_private;       /* not to be passed to user */
-		} _timer;
-
-		/* POSIX.1b signals */
-		struct {
-			compat_pid_t _pid;	/* sender's pid */
-			compat_uid_t _uid;	/* sender's uid */
-			compat_sigval_t _sigval;
-		} _rt;
-
-	} _sifields;
-} compat_siginfo_t;
-
 static inline int __copy_conv_sigset_to_user(compat_sigset_t __user *d,
 	const sigset_t *s)
 {

arch/mips/include/asm/compat.h

@@ -43,6 +43,7 @@ typedef s64		compat_s64;
 typedef u32		compat_uint_t;
 typedef u32		compat_ulong_t;
 typedef u64		compat_u64;
+typedef u32		compat_uptr_t;
 
 struct compat_timespec {
 	compat_time_t	tv_sec;
@@ -124,6 +125,73 @@ typedef u32		compat_old_sigset_t;	/* at least 32 bits */
 
 typedef u32		compat_sigset_word;
 
+typedef union compat_sigval {
+	compat_int_t	sival_int;
+	compat_uptr_t	sival_ptr;
+} compat_sigval_t;
+
+#define SI_PAD_SIZE32	(128/sizeof(int) - 3)
+
+typedef struct compat_siginfo {
+	int si_signo;
+	int si_code;
+	int si_errno;
+
+	union {
+		int _pad[SI_PAD_SIZE32];
+
+		/* kill() */
+		struct {
+			compat_pid_t _pid;	/* sender's pid */
+			__compat_uid_t _uid;	/* sender's uid */
+		} _kill;
+
+		/* SIGCHLD */
+		struct {
+			compat_pid_t _pid;	/* which child */
+			__compat_uid_t _uid;	/* sender's uid */
+			int _status;		/* exit code */
+			compat_clock_t _utime;
+			compat_clock_t _stime;
+		} _sigchld;
+
+		/* IRIX SIGCHLD */
+		struct {
+			compat_pid_t _pid;	/* which child */
+			compat_clock_t _utime;
+			int _status;		/* exit code */
+			compat_clock_t _stime;
+		} _irix_sigchld;
+
+		/* SIGILL, SIGFPE, SIGSEGV, SIGBUS */
+		struct {
+			s32 _addr; /* faulting insn/memory ref. */
+		} _sigfault;
+
+		/* SIGPOLL, SIGXFSZ (To do ...)  */
+		struct {
+			int _band;	/* POLL_IN, POLL_OUT, POLL_MSG */
+			int _fd;
+		} _sigpoll;
+
+		/* POSIX.1b timers */
+		struct {
+			timer_t _tid;		/* timer id */
+			int _overrun;		/* overrun count */
+			compat_sigval_t _sigval;/* same as below */
+			int _sys_private;       /* not to be passed to user */
+		} _timer;
+
+		/* POSIX.1b signals */
+		struct {
+			compat_pid_t _pid;	/* sender's pid */
+			__compat_uid_t _uid;	/* sender's uid */
+			compat_sigval_t _sigval;
+		} _rt;
+
+	} _sifields;
+} compat_siginfo_t;
+
 #define COMPAT_OFF_T_MAX	0x7fffffff
 #define COMPAT_LOFF_T_MAX	0x7fffffffffffffffL
 
@@ -133,7 +201,6 @@ typedef u32		compat_sigset_word;
  * as pointers because the syscall entry code will have
  * appropriately converted them already.
  */
-typedef u32		compat_uptr_t;
 
 static inline void __user *compat_ptr(compat_uptr_t uptr)
 {

arch/mips/pci/pci-octeon.c

@@ -58,7 +58,7 @@ union octeon_pci_address {
 	} s;
 };
 
-int __initdata (*octeon_pcibios_map_irq)(const struct pci_dev *dev,
+int __initconst (*octeon_pcibios_map_irq)(const struct pci_dev *dev,
 					 u8 slot, u8 pin);
 enum octeon_dma_bar_type octeon_dma_bar_type = OCTEON_DMA_BAR_TYPE_INVALID;
 

arch/mn10300/Makefile

@@ -26,7 +26,7 @@ CHECKFLAGS	+=
 PROCESSOR	:= unset
 UNIT		:= unset
 
-KBUILD_CFLAGS	+= -mam33 -mmem-funcs -DCPU=AM33
+KBUILD_CFLAGS	+= -mam33 -DCPU=AM33 $(call cc-option,-mmem-funcs,)
 KBUILD_AFLAGS	+= -mam33 -DCPU=AM33
 
 ifeq ($(CONFIG_MN10300_CURRENT_IN_E2),y)

arch/mn10300/include/asm/elf.h

@@ -151,7 +151,8 @@ do {						\
 #define ELF_PLATFORM  (NULL)
 
 #ifdef __KERNEL__
-#define SET_PERSONALITY(ex) set_personality(PER_LINUX)
+#define SET_PERSONALITY(ex) \
+	set_personality(PER_LINUX | (current->personality & (~PER_MASK)))
 #endif
 
 #endif /* _ASM_ELF_H */

arch/openrisc/include/asm/elf.h

@@ -110,7 +110,8 @@ extern void dump_elf_thread(elf_greg_t *dest, struct pt_regs *pt);
 
 #define ELF_PLATFORM	(NULL)
 
-#define SET_PERSONALITY(ex) set_personality(PER_LINUX)
+#define SET_PERSONALITY(ex) \
+	set_personality(PER_LINUX | (current->personality & (~PER_MASK)))
 
 #endif /* __KERNEL__ */
 #endif

arch/parisc/Kconfig

@@ -13,6 +13,7 @@ config PARISC
 	select HAVE_PERF_EVENTS
 	select GENERIC_ATOMIC64 if !64BIT
 	select HAVE_GENERIC_HARDIRQS
+	select BROKEN_RODATA
 	select GENERIC_IRQ_PROBE
 	select GENERIC_PCI_IOMAP
 	select IRQ_PER_CPU

arch/parisc/include/asm/compat.h

@@ -36,6 +36,7 @@ typedef s64	compat_s64;
 typedef u32	compat_uint_t;
 typedef u32	compat_ulong_t;
 typedef u64	compat_u64;
+typedef u32	compat_uptr_t;
 
 struct compat_timespec {
 	compat_time_t		tv_sec;
@@ -127,6 +128,63 @@ typedef u32		compat_old_sigset_t;	/* at least 32 bits */
 
 typedef u32		compat_sigset_word;
 
+typedef union compat_sigval {
+	compat_int_t	sival_int;
+	compat_uptr_t	sival_ptr;
+} compat_sigval_t;
+
+typedef struct compat_siginfo {
+	int si_signo;
+	int si_errno;
+	int si_code;
+
+	union {
+		int _pad[128/sizeof(int) - 3];
+
+		/* kill() */
+		struct {
+			unsigned int _pid;      /* sender's pid */
+			unsigned int _uid;      /* sender's uid */
+		} _kill;
+
+		/* POSIX.1b timers */
+		struct {
+			compat_timer_t _tid;            /* timer id */
+			int _overrun;           /* overrun count */
+			char _pad[sizeof(unsigned int) - sizeof(int)];
+			compat_sigval_t _sigval;        /* same as below */
+			int _sys_private;       /* not to be passed to user */
+		} _timer;
+
+		/* POSIX.1b signals */
+		struct {
+			unsigned int _pid;      /* sender's pid */
+			unsigned int _uid;      /* sender's uid */
+			compat_sigval_t _sigval;
+		} _rt;
+
+		/* SIGCHLD */
+		struct {
+			unsigned int _pid;      /* which child */
+			unsigned int _uid;      /* sender's uid */
+			int _status;            /* exit code */
+			compat_clock_t _utime;
+			compat_clock_t _stime;
+		} _sigchld;
+
+		/* SIGILL, SIGFPE, SIGSEGV, SIGBUS */
+		struct {
+			unsigned int _addr;     /* faulting insn/memory ref. */
+		} _sigfault;
+
+		/* SIGPOLL */
+		struct {
+			int _band;      /* POLL_IN, POLL_OUT, POLL_MSG */
+			int _fd;
+		} _sigpoll;
+	} _sifields;
+} compat_siginfo_t;
+
 #define COMPAT_OFF_T_MAX	0x7fffffff
 #define COMPAT_LOFF_T_MAX	0x7fffffffffffffffL
 
@@ -136,7 +194,6 @@ typedef u32		compat_sigset_word;
  * as pointers because the syscall entry code will have
  * appropriately converted them already.
  */
-typedef	u32		compat_uptr_t;
 
 static inline void __user *compat_ptr(compat_uptr_t uptr)
 {

arch/parisc/kernel/signal32.h

@@ -55,58 +55,6 @@ struct k_sigaction32 {
 	struct compat_sigaction sa;
 };
 
-typedef struct compat_siginfo {
-        int si_signo;
-        int si_errno;
-        int si_code;
-
-        union {
-                int _pad[((128/sizeof(int)) - 3)];
-
-                /* kill() */
-                struct {
-                        unsigned int _pid;      /* sender's pid */
-                        unsigned int _uid;      /* sender's uid */
-                } _kill;
-
-                /* POSIX.1b timers */
-                struct {
-                        compat_timer_t _tid;            /* timer id */
-                        int _overrun;           /* overrun count */
-                        char _pad[sizeof(unsigned int) - sizeof(int)];
-                        compat_sigval_t _sigval;        /* same as below */
-                        int _sys_private;       /* not to be passed to user */
-                } _timer;
-
-                /* POSIX.1b signals */
-                struct {
-                        unsigned int _pid;      /* sender's pid */
-                        unsigned int _uid;      /* sender's uid */
-                        compat_sigval_t _sigval;
-                } _rt;
-
-                /* SIGCHLD */
-                struct {
-                        unsigned int _pid;      /* which child */
-                        unsigned int _uid;      /* sender's uid */
-                        int _status;            /* exit code */
-                        compat_clock_t _utime;
-                        compat_clock_t _stime;
-                } _sigchld;
-
-                /* SIGILL, SIGFPE, SIGSEGV, SIGBUS */
-                struct {
-                        unsigned int _addr;     /* faulting insn/memory ref. */
-                } _sigfault;
-
-                /* SIGPOLL */
-                struct {
-                        int _band;      /* POLL_IN, POLL_OUT, POLL_MSG */
-                        int _fd;
-                } _sigpoll;
-        } _sifields;
-} compat_siginfo_t;
-
 int copy_siginfo_to_user32 (compat_siginfo_t __user *to, siginfo_t *from);
 int copy_siginfo_from_user32 (siginfo_t *to, compat_siginfo_t __user *from);
 

arch/powerpc/include/asm/bitops.h

@@ -288,6 +288,16 @@ static __inline__ int test_bit_le(unsigned long nr,
 	return (tmp[nr >> 3] >> (nr & 7)) & 1;
 }
 
+static inline void set_bit_le(int nr, void *addr)
+{
+	set_bit(nr ^ BITOP_LE_SWIZZLE, addr);
+}
+
+static inline void clear_bit_le(int nr, void *addr)
+{
+	clear_bit(nr ^ BITOP_LE_SWIZZLE, addr);
+}
+
 static inline void __set_bit_le(int nr, void *addr)
 {
 	__set_bit(nr ^ BITOP_LE_SWIZZLE, addr);

arch/powerpc/include/asm/compat.h

@@ -38,6 +38,7 @@ typedef s64		compat_s64;
 typedef u32		compat_uint_t;
 typedef u32		compat_ulong_t;
 typedef u64		compat_u64;
+typedef u32		compat_uptr_t;
 
 struct compat_timespec {
 	compat_time_t	tv_sec;
@@ -114,6 +115,64 @@ typedef u32		compat_old_sigset_t;
 
 typedef u32		compat_sigset_word;
 
+typedef union compat_sigval {
+	compat_int_t	sival_int;
+	compat_uptr_t	sival_ptr;
+} compat_sigval_t;
+
+#define SI_PAD_SIZE32	(128/sizeof(int) - 3)
+
+typedef struct compat_siginfo {
+	int si_signo;
+	int si_errno;
+	int si_code;
+
+	union {
+		int _pad[SI_PAD_SIZE32];
+
+		/* kill() */
+		struct {
+			compat_pid_t _pid;		/* sender's pid */
+			__compat_uid_t _uid;		/* sender's uid */
+		} _kill;
+
+		/* POSIX.1b timers */
+		struct {
+			compat_timer_t _tid;		/* timer id */
+			int _overrun;			/* overrun count */
+			compat_sigval_t _sigval;	/* same as below */
+			int _sys_private;	/* not to be passed to user */
+		} _timer;
+
+		/* POSIX.1b signals */
+		struct {
+			compat_pid_t _pid;		/* sender's pid */
+			__compat_uid_t _uid;		/* sender's uid */
+			compat_sigval_t _sigval;
+		} _rt;
+
+		/* SIGCHLD */
+		struct {
+			compat_pid_t _pid;		/* which child */
+			__compat_uid_t _uid;		/* sender's uid */
+			int _status;			/* exit code */
+			compat_clock_t _utime;
+			compat_clock_t _stime;
+		} _sigchld;
+
+		/* SIGILL, SIGFPE, SIGSEGV, SIGBUS, SIGEMT */
+		struct {
+			unsigned int _addr; /* faulting insn/memory ref. */
+		} _sigfault;
+
+		/* SIGPOLL */
+		struct {
+			int _band;	/* POLL_IN, POLL_OUT, POLL_MSG */
+			int _fd;
+		} _sigpoll;
+	} _sifields;
+} compat_siginfo_t;
+
 #define COMPAT_OFF_T_MAX	0x7fffffff
 #define COMPAT_LOFF_T_MAX	0x7fffffffffffffffL
 
@@ -123,7 +182,6 @@ typedef u32		compat_sigset_word;
  * as pointers because the syscall entry code will have
  * appropriately converted them already.
  */
-typedef	u32		compat_uptr_t;
 
 static inline void __user *compat_ptr(compat_uptr_t uptr)
 {

arch/powerpc/include/asm/siginfo.h

@@ -10,7 +10,6 @@
 
 #ifdef __powerpc64__
 #    define __ARCH_SI_PREAMBLE_SIZE	(4 * sizeof(int))
-#    define SI_PAD_SIZE32		((SI_MAX_SIZE/sizeof(int)) - 3)
 #endif
 
 #include <asm-generic/siginfo.h>

arch/powerpc/kernel/ppc32.h

@@ -16,57 +16,6 @@
 
 /* These are here to support 32-bit syscalls on a 64-bit kernel. */
 
-typedef struct compat_siginfo {
-	int si_signo;
-	int si_errno;
-	int si_code;
-
-	union {
-		int _pad[SI_PAD_SIZE32];
-
-		/* kill() */
-		struct {
-			compat_pid_t _pid;		/* sender's pid */
-			compat_uid_t _uid;		/* sender's uid */
-		} _kill;
-
-		/* POSIX.1b timers */
-		struct {
-			compat_timer_t _tid;			/* timer id */
-			int _overrun;			/* overrun count */
-			compat_sigval_t _sigval;		/* same as below */
-			int _sys_private;		/* not to be passed to user */
-		} _timer;
-
-		/* POSIX.1b signals */
-		struct {
-			compat_pid_t _pid;		/* sender's pid */
-			compat_uid_t _uid;		/* sender's uid */
-			compat_sigval_t _sigval;
-		} _rt;
-
-		/* SIGCHLD */
-		struct {
-			compat_pid_t _pid;		/* which child */
-			compat_uid_t _uid;		/* sender's uid */
-			int _status;			/* exit code */
-			compat_clock_t _utime;
-			compat_clock_t _stime;
-		} _sigchld;
-
-		/* SIGILL, SIGFPE, SIGSEGV, SIGBUS, SIGEMT */
-		struct {
-			unsigned int _addr; /* faulting insn/memory ref. */
-		} _sigfault;
-
-		/* SIGPOLL */
-		struct {
-			int _band;	/* POLL_IN, POLL_OUT, POLL_MSG */
-			int _fd;
-		} _sigpoll;
-	} _sifields;
-} compat_siginfo_t;
-
 #define __old_sigaction32	old_sigaction32
 
 struct __old_sigaction32 {

arch/powerpc/platforms/40x/ppc40x_simple.c

@@ -50,7 +50,7 @@ machine_device_initcall(ppc40x_simple, ppc40x_device_probe);
  * Again, if your board needs to do things differently then create a
  * board.c file for it rather than adding it to this list.
  */
-static const char *board[] __initdata = {
+static const char * const board[] __initconst = {
 	"amcc,acadia",
 	"amcc,haleakala",
 	"amcc,kilauea",

arch/powerpc/platforms/512x/mpc5121_generic.c

@@ -26,7 +26,7 @@
 /*
  * list of supported boards
  */
-static const char *board[] __initdata = {
+static const char * const board[] __initconst = {
 	"prt,prtlvt",
 	NULL
 };

arch/powerpc/platforms/52xx/lite5200.c

@@ -172,7 +172,7 @@ static void __init lite5200_setup_arch(void)
 	mpc52xx_setup_pci();
 }
 
-static const char *board[] __initdata = {
+static const char * const board[] __initconst = {
 	"fsl,lite5200",
 	"fsl,lite5200b",
 	NULL,

arch/powerpc/platforms/52xx/media5200.c

@@ -232,7 +232,7 @@ static void __init media5200_setup_arch(void)
 }
 
 /* list of the supported boards */
-static const char *board[] __initdata = {
+static const char * const board[] __initconst = {
 	"fsl,media5200",
 	NULL
 };

arch/powerpc/platforms/83xx/mpc837x_rdb.c

@@ -60,7 +60,7 @@ static void __init mpc837x_rdb_setup_arch(void)
 
 machine_device_initcall(mpc837x_rdb, mpc83xx_declare_of_platform_devices);
 
-static const char *board[] __initdata = {
+static const char * const board[] __initconst = {
 	"fsl,mpc8377rdb",
 	"fsl,mpc8378rdb",
 	"fsl,mpc8379rdb",

arch/powerpc/platforms/85xx/tqm85xx.c

@@ -125,7 +125,7 @@ DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_TI, PCI_DEVICE_ID_TI_1520,
 
 machine_device_initcall(tqm85xx, mpc85xx_common_publish_devices);
 
-static const char *board[] __initdata = {
+static const char * const board[] __initconst = {
 	"tqc,tqm8540",
 	"tqc,tqm8541",
 	"tqc,tqm8548",

arch/s390/include/asm/compat.h

@@ -65,6 +65,7 @@ typedef s64		compat_s64;
 typedef u32		compat_uint_t;
 typedef u32		compat_ulong_t;
 typedef u64		compat_u64;
+typedef u32		compat_uptr_t;
 
 struct compat_timespec {
 	compat_time_t	tv_sec;
@@ -144,6 +145,79 @@ typedef u32		compat_old_sigset_t;	/* at least 32 bits */
 
 typedef u32		compat_sigset_word;
 
+typedef union compat_sigval {
+	compat_int_t	sival_int;
+	compat_uptr_t	sival_ptr;
+} compat_sigval_t;
+
+typedef struct compat_siginfo {
+	int	si_signo;
+	int	si_errno;
+	int	si_code;
+
+	union {
+		int _pad[128/sizeof(int) - 3];
+
+		/* kill() */
+		struct {
+			pid_t	_pid;	/* sender's pid */
+			uid_t	_uid;	/* sender's uid */
+		} _kill;
+
+		/* POSIX.1b timers */
+		struct {
+			compat_timer_t _tid;		/* timer id */
+			int _overrun;			/* overrun count */
+			compat_sigval_t _sigval;	/* same as below */
+			int _sys_private;	/* not to be passed to user */
+		} _timer;
+
+		/* POSIX.1b signals */
+		struct {
+			pid_t			_pid;	/* sender's pid */
+			uid_t			_uid;	/* sender's uid */
+			compat_sigval_t		_sigval;
+		} _rt;
+
+		/* SIGCHLD */
+		struct {
+			pid_t			_pid;	/* which child */
+			uid_t			_uid;	/* sender's uid */
+			int			_status;/* exit code */
+			compat_clock_t		_utime;
+			compat_clock_t		_stime;
+		} _sigchld;
+
+		/* SIGILL, SIGFPE, SIGSEGV, SIGBUS */
+		struct {
+			__u32	_addr;	/* faulting insn/memory ref. - pointer */
+		} _sigfault;
+
+		/* SIGPOLL */
+		struct {
+			int	_band;	/* POLL_IN, POLL_OUT, POLL_MSG */
+			int	_fd;
+		} _sigpoll;
+	} _sifields;
+} compat_siginfo_t;
+
+/*
+ * How these fields are to be accessed.
+ */
+#define si_pid		_sifields._kill._pid
+#define si_uid		_sifields._kill._uid
+#define si_status	_sifields._sigchld._status
+#define si_utime	_sifields._sigchld._utime
+#define si_stime	_sifields._sigchld._stime
+#define si_value	_sifields._rt._sigval
+#define si_int		_sifields._rt._sigval.sival_int
+#define si_ptr		_sifields._rt._sigval.sival_ptr
+#define si_addr		_sifields._sigfault._addr
+#define si_band		_sifields._sigpoll._band
+#define si_fd		_sifields._sigpoll._fd
+#define si_tid		_sifields._timer._tid
+#define si_overrun	_sifields._timer._overrun
+
 #define COMPAT_OFF_T_MAX	0x7fffffff
 #define COMPAT_LOFF_T_MAX	0x7fffffffffffffffL
 
@@ -153,7 +227,6 @@ typedef u32		compat_sigset_word;
  * as pointers because the syscall entry code will have
  * appropriately converted them already.
  */
-typedef	u32		compat_uptr_t;
 
 static inline void __user *compat_ptr(compat_uptr_t uptr)
 {

arch/s390/kernel/compat_linux.h

@@ -23,74 +23,6 @@ struct old_sigaction32 {
        __u32			sa_flags;
        __u32			sa_restorer;	/* Another 32 bit pointer */
 };
- 
-typedef struct compat_siginfo {
-	int	si_signo;
-	int	si_errno;
-	int	si_code;
-
-	union {
-		int _pad[((128/sizeof(int)) - 3)];
-
-		/* kill() */
-		struct {
-			pid_t	_pid;	/* sender's pid */
-			uid_t	_uid;	/* sender's uid */
-		} _kill;
-
-		/* POSIX.1b timers */
-		struct {
-			compat_timer_t _tid;		/* timer id */
-			int _overrun;		/* overrun count */
-			compat_sigval_t _sigval;	/* same as below */
-			int _sys_private;       /* not to be passed to user */
-		} _timer;
-
-		/* POSIX.1b signals */
-		struct {
-			pid_t			_pid;	/* sender's pid */
-			uid_t			_uid;	/* sender's uid */
-			compat_sigval_t		_sigval;
-		} _rt;
-
-		/* SIGCHLD */
-		struct {
-			pid_t			_pid;	/* which child */
-			uid_t			_uid;	/* sender's uid */
-			int			_status;/* exit code */
-			compat_clock_t		_utime;
-			compat_clock_t		_stime;
-		} _sigchld;
-
-		/* SIGILL, SIGFPE, SIGSEGV, SIGBUS */
-		struct {
-			__u32	_addr;	/* faulting insn/memory ref. - pointer */
-		} _sigfault;
-                          
-		/* SIGPOLL */
-		struct {
-			int	_band;	/* POLL_IN, POLL_OUT, POLL_MSG */
-			int	_fd;
-		} _sigpoll;
-	} _sifields;
-} compat_siginfo_t;
-
-/*
- * How these fields are to be accessed.
- */
-#define si_pid		_sifields._kill._pid
-#define si_uid		_sifields._kill._uid
-#define si_status	_sifields._sigchld._status
-#define si_utime	_sifields._sigchld._utime
-#define si_stime	_sifields._sigchld._stime
-#define si_value	_sifields._rt._sigval
-#define si_int		_sifields._rt._sigval.sival_int
-#define si_ptr		_sifields._rt._sigval.sival_ptr
-#define si_addr		_sifields._sigfault._addr
-#define si_band		_sifields._sigpoll._band
-#define si_fd		_sifields._sigpoll._fd    
-#define si_tid		_sifields._timer._tid
-#define si_overrun	_sifields._timer._overrun
 
 /* asm/sigcontext.h */
 typedef union

arch/score/Kconfig

@@ -5,6 +5,7 @@ config SCORE
        select HAVE_GENERIC_HARDIRQS
        select GENERIC_IRQ_SHOW
        select GENERIC_IOMAP
+       select GENERIC_ATOMIC64
        select HAVE_MEMBLOCK
        select HAVE_MEMBLOCK_NODE_MAP
        select ARCH_DISCARD_MEMBLOCK

arch/score/include/asm/elf.h

@@ -54,7 +54,7 @@ typedef elf_fpreg_t	elf_fpregset_t;
 
 #define SET_PERSONALITY(ex)					\
 do {								\
-	set_personality(PER_LINUX);				\
+	set_personality(PER_LINUX | (current->personality & (~PER_MASK))); \
 } while (0)
 
 struct task_struct;

arch/score/kernel/sys_score.c

@@ -112,6 +112,7 @@ score_execve(struct pt_regs *regs)
  * Do a system call from kernel instead of calling sys_execve so we
  * end up with proper pt_regs.
  */
+asmlinkage
 int kernel_execve(const char *filename,
 		  const char *const argv[],
 		  const char *const envp[])

arch/sh/include/asm/elf.h

@@ -183,7 +183,8 @@ do {									\
 } while (0)
 #endif
 
-#define SET_PERSONALITY(ex) set_personality(PER_LINUX_32BIT)
+#define SET_PERSONALITY(ex) \
+	set_personality(PER_LINUX_32BIT | (current->personality & (~PER_MASK)))
 
 #ifdef CONFIG_VSYSCALL
 /* vDSO has arch_setup_additional_pages */

arch/sh/include/asm/io.h

@@ -134,7 +134,7 @@ __BUILD_MEMORY_STRING(__raw_, q, u64)
  * load/store instructions. sh_io_port_base is the virtual address to
  * which all ports are being mapped.
  */
-extern const unsigned long sh_io_port_base;
+extern unsigned long sh_io_port_base;
 
 static inline void __set_io_port_base(unsigned long pbase)
 {

arch/sh/kernel/ioport.c

@@ -11,7 +11,7 @@
 #include <linux/module.h>
 #include <linux/io.h>
 
-const unsigned long sh_io_port_base __read_mostly = -1;
+unsigned long sh_io_port_base __read_mostly = -1;
 EXPORT_SYMBOL(sh_io_port_base);
 
 void __iomem *__ioport_map(unsigned long addr, unsigned int size)

arch/sparc/include/asm/compat.h

@@ -36,6 +36,7 @@ typedef s64		compat_s64;
 typedef u32		compat_uint_t;
 typedef u32		compat_ulong_t;
 typedef u64		compat_u64;
+typedef u32		compat_uptr_t;
 
 struct compat_timespec {
 	compat_time_t	tv_sec;
@@ -147,6 +148,65 @@ typedef u32		compat_old_sigset_t;
 
 typedef u32		compat_sigset_word;
 
+typedef union compat_sigval {
+	compat_int_t	sival_int;
+	compat_uptr_t	sival_ptr;
+} compat_sigval_t;
+
+#define SI_PAD_SIZE32	(128/sizeof(int) - 3)
+
+typedef struct compat_siginfo {
+	int si_signo;
+	int si_errno;
+	int si_code;
+
+	union {
+		int _pad[SI_PAD_SIZE32];
+
+		/* kill() */
+		struct {
+			compat_pid_t _pid;		/* sender's pid */
+			unsigned int _uid;		/* sender's uid */
+		} _kill;
+
+		/* POSIX.1b timers */
+		struct {
+			compat_timer_t _tid;		/* timer id */
+			int _overrun;			/* overrun count */
+			compat_sigval_t _sigval;	/* same as below */
+			int _sys_private;	/* not to be passed to user */
+		} _timer;
+
+		/* POSIX.1b signals */
+		struct {
+			compat_pid_t _pid;		/* sender's pid */
+			unsigned int _uid;		/* sender's uid */
+			compat_sigval_t _sigval;
+		} _rt;
+
+		/* SIGCHLD */
+		struct {
+			compat_pid_t _pid;		/* which child */
+			unsigned int _uid;		/* sender's uid */
+			int _status;			/* exit code */
+			compat_clock_t _utime;
+			compat_clock_t _stime;
+		} _sigchld;
+
+		/* SIGILL, SIGFPE, SIGSEGV, SIGBUS, SIGEMT */
+		struct {
+			u32 _addr; /* faulting insn/memory ref. */
+			int _trapno;
+		} _sigfault;
+
+		/* SIGPOLL */
+		struct {
+			int _band;	/* POLL_IN, POLL_OUT, POLL_MSG */
+			int _fd;
+		} _sigpoll;
+	} _sifields;
+} compat_siginfo_t;
+
 #define COMPAT_OFF_T_MAX	0x7fffffff
 #define COMPAT_LOFF_T_MAX	0x7fffffffffffffffL
 
@@ -156,7 +216,6 @@ typedef u32		compat_sigset_word;
  * as pointers because the syscall entry code will have
  * appropriately converted them already.
  */
-typedef	u32		compat_uptr_t;
 
 static inline void __user *compat_ptr(compat_uptr_t uptr)
 {

arch/sparc/include/asm/elf_32.h

@@ -128,6 +128,7 @@ typedef struct {
 
 #define ELF_PLATFORM	(NULL)
 
-#define SET_PERSONALITY(ex) set_personality(PER_LINUX)
+#define SET_PERSONALITY(ex) \
+	set_personality(PER_LINUX | (current->personality & (~PER_MASK)))
 
 #endif /* !(__ASMSPARC_ELF_H) */

arch/sparc/include/asm/siginfo.h

@@ -3,7 +3,6 @@
 
 #if defined(__sparc__) && defined(__arch64__)
 
-#define SI_PAD_SIZE32	((SI_MAX_SIZE/sizeof(int)) - 3)
 #define __ARCH_SI_PREAMBLE_SIZE	(4 * sizeof(int))
 #define __ARCH_SI_BAND_T int
 

arch/sparc/kernel/signal32.c

@@ -54,58 +54,6 @@ struct signal_frame32 {
 	/* __siginfo_rwin_t * */u32 rwin_save;
 } __attribute__((aligned(8)));
 
-typedef struct compat_siginfo{
-	int si_signo;
-	int si_errno;
-	int si_code;
-
-	union {
-		int _pad[SI_PAD_SIZE32];
-
-		/* kill() */
-		struct {
-			compat_pid_t _pid;		/* sender's pid */
-			unsigned int _uid;		/* sender's uid */
-		} _kill;
-
-		/* POSIX.1b timers */
-		struct {
-			compat_timer_t _tid;			/* timer id */
-			int _overrun;			/* overrun count */
-			compat_sigval_t _sigval;		/* same as below */
-			int _sys_private;		/* not to be passed to user */
-		} _timer;
-
-		/* POSIX.1b signals */
-		struct {
-			compat_pid_t _pid;		/* sender's pid */
-			unsigned int _uid;		/* sender's uid */
-			compat_sigval_t _sigval;
-		} _rt;
-
-		/* SIGCHLD */
-		struct {
-			compat_pid_t _pid;		/* which child */
-			unsigned int _uid;		/* sender's uid */
-			int _status;			/* exit code */
-			compat_clock_t _utime;
-			compat_clock_t _stime;
-		} _sigchld;
-
-		/* SIGILL, SIGFPE, SIGSEGV, SIGBUS, SIGEMT */
-		struct {
-			u32 _addr; /* faulting insn/memory ref. */
-			int _trapno;
-		} _sigfault;
-
-		/* SIGPOLL */
-		struct {
-			int _band;	/* POLL_IN, POLL_OUT, POLL_MSG */
-			int _fd;
-		} _sigpoll;
-	} _sifields;
-}compat_siginfo_t;
-
 struct rt_signal_frame32 {
 	struct sparc_stackf32	ss;
 	compat_siginfo_t	info;

arch/tile/include/asm/compat.h

@@ -110,6 +110,68 @@ struct compat_flock64 {
 
 typedef u32               compat_sigset_word;
 
+typedef union compat_sigval {
+	compat_int_t	sival_int;
+	compat_uptr_t	sival_ptr;
+} compat_sigval_t;
+
+#define COMPAT_SI_PAD_SIZE	(128/sizeof(int) - 3)
+
+typedef struct compat_siginfo {
+	int si_signo;
+	int si_errno;
+	int si_code;
+
+	union {
+		int _pad[COMPAT_SI_PAD_SIZE];
+
+		/* kill() */
+		struct {
+			unsigned int _pid;	/* sender's pid */
+			unsigned int _uid;	/* sender's uid */
+		} _kill;
+
+		/* POSIX.1b timers */
+		struct {
+			compat_timer_t _tid;	/* timer id */
+			int _overrun;		/* overrun count */
+			compat_sigval_t _sigval;	/* same as below */
+			int _sys_private;	/* not to be passed to user */
+			int _overrun_incr;	/* amount to add to overrun */
+		} _timer;
+
+		/* POSIX.1b signals */
+		struct {
+			unsigned int _pid;	/* sender's pid */
+			unsigned int _uid;	/* sender's uid */
+			compat_sigval_t _sigval;
+		} _rt;
+
+		/* SIGCHLD */
+		struct {
+			unsigned int _pid;	/* which child */
+			unsigned int _uid;	/* sender's uid */
+			int _status;		/* exit code */
+			compat_clock_t _utime;
+			compat_clock_t _stime;
+		} _sigchld;
+
+		/* SIGILL, SIGFPE, SIGSEGV, SIGBUS */
+		struct {
+			unsigned int _addr;	/* faulting insn/memory ref. */
+#ifdef __ARCH_SI_TRAPNO
+			int _trapno;	/* TRAP # which caused the signal */
+#endif
+		} _sigfault;
+
+		/* SIGPOLL */
+		struct {
+			int _band;	/* POLL_IN, POLL_OUT, POLL_MSG */
+			int _fd;
+		} _sigpoll;
+	} _sifields;
+} compat_siginfo_t;
+
 #define COMPAT_OFF_T_MAX	0x7fffffff
 #define COMPAT_LOFF_T_MAX	0x7fffffffffffffffL
 

arch/tile/include/asm/elf.h

@@ -156,12 +156,12 @@ extern int arch_setup_additional_pages(struct linux_binprm *bprm,
 #undef SET_PERSONALITY
 #define SET_PERSONALITY(ex) \
 do { \
-	current->personality = PER_LINUX; \
+	set_personality(PER_LINUX | (current->personality & (~PER_MASK))); \
 	current_thread_info()->status &= ~TS_COMPAT; \
 } while (0)
 #define COMPAT_SET_PERSONALITY(ex) \
 do { \
-	current->personality = PER_LINUX_32BIT; \
+	set_personality(PER_LINUX | (current->personality & (~PER_MASK))); \
 	current_thread_info()->status |= TS_COMPAT; \
 } while (0)
 

arch/tile/kernel/compat_signal.c

@@ -55,63 +55,6 @@ struct compat_ucontext {
 	sigset_t	  uc_sigmask;	/* mask last for extensibility */
 };
 
-#define COMPAT_SI_PAD_SIZE	((SI_MAX_SIZE - 3 * sizeof(int)) / sizeof(int))
-
-struct compat_siginfo {
-	int si_signo;
-	int si_errno;
-	int si_code;
-
-	union {
-		int _pad[COMPAT_SI_PAD_SIZE];
-
-		/* kill() */
-		struct {
-			unsigned int _pid;	/* sender's pid */
-			unsigned int _uid;	/* sender's uid */
-		} _kill;
-
-		/* POSIX.1b timers */
-		struct {
-			compat_timer_t _tid;	/* timer id */
-			int _overrun;		/* overrun count */
-			compat_sigval_t _sigval;	/* same as below */
-			int _sys_private;	/* not to be passed to user */
-			int _overrun_incr;	/* amount to add to overrun */
-		} _timer;
-
-		/* POSIX.1b signals */
-		struct {
-			unsigned int _pid;	/* sender's pid */
-			unsigned int _uid;	/* sender's uid */
-			compat_sigval_t _sigval;
-		} _rt;
-
-		/* SIGCHLD */
-		struct {
-			unsigned int _pid;	/* which child */
-			unsigned int _uid;	/* sender's uid */
-			int _status;		/* exit code */
-			compat_clock_t _utime;
-			compat_clock_t _stime;
-		} _sigchld;
-
-		/* SIGILL, SIGFPE, SIGSEGV, SIGBUS */
-		struct {
-			unsigned int _addr;	/* faulting insn/memory ref. */
-#ifdef __ARCH_SI_TRAPNO
-			int _trapno;	/* TRAP # which caused the signal */
-#endif
-		} _sigfault;
-
-		/* SIGPOLL */
-		struct {
-			int _band;	/* POLL_IN, POLL_OUT, POLL_MSG */
-			int _fd;
-		} _sigpoll;
-	} _sifields;
-};
-
 struct compat_rt_sigframe {
 	unsigned char save_area[C_ABI_SAVE_AREA_SIZE]; /* caller save area */
 	struct compat_siginfo info;

arch/unicore32/Kconfig

@@ -6,6 +6,7 @@ config UNICORE32
 	select HAVE_DMA_ATTRS
 	select HAVE_KERNEL_GZIP
 	select HAVE_KERNEL_BZIP2
+	select GENERIC_ATOMIC64
 	select HAVE_KERNEL_LZO
 	select HAVE_KERNEL_LZMA
 	select ARCH_HAVE_CUSTOM_GPIO_H

arch/x86/include/asm/apic.h

@@ -409,7 +409,7 @@ extern struct apic *apic;
  * to enforce the order with in them.
  */
 #define apic_driver(sym)					\
-	static struct apic *__apicdrivers_##sym __used		\
+	static const struct apic *__apicdrivers_##sym __used		\
 	__aligned(sizeof(struct apic *))			\
 	__section(.apicdrivers) = { &sym }
 

arch/x86/include/asm/compat.h

@@ -41,6 +41,7 @@ typedef s64 __attribute__((aligned(4))) compat_s64;
 typedef u32		compat_uint_t;
 typedef u32		compat_ulong_t;
 typedef u64 __attribute__((aligned(4))) compat_u64;
+typedef u32		compat_uptr_t;
 
 struct compat_timespec {
 	compat_time_t	tv_sec;
@@ -124,6 +125,78 @@ typedef u32		compat_old_sigset_t;	/* at least 32 bits */
 
 typedef u32               compat_sigset_word;
 
+typedef union compat_sigval {
+	compat_int_t	sival_int;
+	compat_uptr_t	sival_ptr;
+} compat_sigval_t;
+
+typedef struct compat_siginfo {
+	int si_signo;
+	int si_errno;
+	int si_code;
+
+	union {
+		int _pad[128/sizeof(int) - 3];
+
+		/* kill() */
+		struct {
+			unsigned int _pid;	/* sender's pid */
+			unsigned int _uid;	/* sender's uid */
+		} _kill;
+
+		/* POSIX.1b timers */
+		struct {
+			compat_timer_t _tid;	/* timer id */
+			int _overrun;		/* overrun count */
+			compat_sigval_t _sigval;	/* same as below */
+			int _sys_private;	/* not to be passed to user */
+			int _overrun_incr;	/* amount to add to overrun */
+		} _timer;
+
+		/* POSIX.1b signals */
+		struct {
+			unsigned int _pid;	/* sender's pid */
+			unsigned int _uid;	/* sender's uid */
+			compat_sigval_t _sigval;
+		} _rt;
+
+		/* SIGCHLD */
+		struct {
+			unsigned int _pid;	/* which child */
+			unsigned int _uid;	/* sender's uid */
+			int _status;		/* exit code */
+			compat_clock_t _utime;
+			compat_clock_t _stime;
+		} _sigchld;
+
+		/* SIGCHLD (x32 version) */
+		struct {
+			unsigned int _pid;	/* which child */
+			unsigned int _uid;	/* sender's uid */
+			int _status;		/* exit code */
+			compat_s64 _utime;
+			compat_s64 _stime;
+		} _sigchld_x32;
+
+		/* SIGILL, SIGFPE, SIGSEGV, SIGBUS */
+		struct {
+			unsigned int _addr;	/* faulting insn/memory ref. */
+		} _sigfault;
+
+		/* SIGPOLL */
+		struct {
+			int _band;	/* POLL_IN, POLL_OUT, POLL_MSG */
+			int _fd;
+		} _sigpoll;
+
+		struct {
+			unsigned int _call_addr; /* calling insn */
+			int _syscall;	/* triggering system call number */
+			unsigned int _arch;	/* AUDIT_ARCH_* of syscall */
+		} _sigsys;
+	} _sifields;
+} compat_siginfo_t;
+
 #define COMPAT_OFF_T_MAX	0x7fffffff
 #define COMPAT_LOFF_T_MAX	0x7fffffffffffffffL
 
@@ -209,7 +282,6 @@ typedef struct user_regs_struct32 compat_elf_gregset_t;
  * as pointers because the syscall entry code will have
  * appropriately converted them already.
  */
-typedef	u32		compat_uptr_t;
 
 static inline void __user *compat_ptr(compat_uptr_t uptr)
 {

arch/x86/include/asm/ia32.h

@@ -86,73 +86,6 @@ struct stat64 {
 	unsigned long long	st_ino;
 } __attribute__((packed));
 
-typedef struct compat_siginfo {
-	int si_signo;
-	int si_errno;
-	int si_code;
-
-	union {
-		int _pad[((128 / sizeof(int)) - 3)];
-
-		/* kill() */
-		struct {
-			unsigned int _pid;	/* sender's pid */
-			unsigned int _uid;	/* sender's uid */
-		} _kill;
-
-		/* POSIX.1b timers */
-		struct {
-			compat_timer_t _tid;	/* timer id */
-			int _overrun;		/* overrun count */
-			compat_sigval_t _sigval;	/* same as below */
-			int _sys_private;	/* not to be passed to user */
-			int _overrun_incr;	/* amount to add to overrun */
-		} _timer;
-
-		/* POSIX.1b signals */
-		struct {
-			unsigned int _pid;	/* sender's pid */
-			unsigned int _uid;	/* sender's uid */
-			compat_sigval_t _sigval;
-		} _rt;
-
-		/* SIGCHLD */
-		struct {
-			unsigned int _pid;	/* which child */
-			unsigned int _uid;	/* sender's uid */
-			int _status;		/* exit code */
-			compat_clock_t _utime;
-			compat_clock_t _stime;
-		} _sigchld;
-
-		/* SIGCHLD (x32 version) */
-		struct {
-			unsigned int _pid;	/* which child */
-			unsigned int _uid;	/* sender's uid */
-			int _status;		/* exit code */
-			compat_s64 _utime;
-			compat_s64 _stime;
-		} _sigchld_x32;
-
-		/* SIGILL, SIGFPE, SIGSEGV, SIGBUS */
-		struct {
-			unsigned int _addr;	/* faulting insn/memory ref. */
-		} _sigfault;
-
-		/* SIGPOLL */
-		struct {
-			int _band;	/* POLL_IN, POLL_OUT, POLL_MSG */
-			int _fd;
-		} _sigpoll;
-
-		struct {
-			unsigned int _call_addr; /* calling insn */
-			int _syscall;	/* triggering system call number */
-			unsigned int _arch;	/* AUDIT_ARCH_* of syscall */
-		} _sigsys;
-	} _sifields;
-} compat_siginfo_t;
-
 #define IA32_STACK_TOP IA32_PAGE_OFFSET
 
 #ifdef __KERNEL__

arch/x86/kernel/apic/apic_numachip.c

@@ -30,7 +30,7 @@
 
 static int numachip_system __read_mostly;
 
-static struct apic apic_numachip __read_mostly;
+static const struct apic apic_numachip __read_mostly;
 
 static unsigned int get_apic_id(unsigned long x)
 {
@@ -199,7 +199,7 @@ static int numachip_acpi_madt_oem_check(char *oem_id, char *oem_table_id)
 	return 0;
 }
 
-static struct apic apic_numachip __refconst = {
+static const struct apic apic_numachip __refconst = {
 
 	.name				= "NumaConnect system",
 	.probe				= numachip_probe,

arch/x86/kernel/rtc.c

@@ -225,7 +225,7 @@ static struct platform_device rtc_device = {
 static __init int add_rtc_cmos(void)
 {
 #ifdef CONFIG_PNP
-	static const char *ids[] __initconst =
+	static const char * const  const ids[] __initconst =
 	    { "PNP0b00", "PNP0b01", "PNP0b02", };
 	struct pnp_dev *dev;
 	struct pnp_id *id;

arch/xtensa/include/asm/elf.h

@@ -189,7 +189,8 @@ typedef struct {
 #endif
 } elf_xtregs_t;
 
-#define SET_PERSONALITY(ex) set_personality(PER_LINUX_32BIT)
+#define SET_PERSONALITY(ex) \
+	set_personality(PER_LINUX_32BIT | (current->personality & (~PER_MASK)))
 
 struct task_struct;
 

drivers/atm/eni.c

@@ -1567,7 +1567,7 @@ tx_complete++;
 /*--------------------------------- entries ---------------------------------*/
 
 
-static const char *media_name[] __devinitdata = {
+static char * const media_name[] __devinitconst = {
     "MMF", "SMF", "MMF", "03?", /*  0- 3 */
     "UTP", "05?", "06?", "07?", /*  4- 7 */
     "TAXI","09?", "10?", "11?", /*  8-11 */

drivers/block/aoe/aoe.h

@@ -1,5 +1,5 @@
-/* Copyright (c) 2007 Coraid, Inc.  See COPYING for GPL terms. */
-#define VERSION "47"
+/* Copyright (c) 2012 Coraid, Inc.  See COPYING for GPL terms. */
+#define VERSION "50"
 #define AOE_MAJOR 152
 #define DEVICE_NAME "aoe"
 
@@ -10,9 +10,6 @@
 #define AOE_PARTITIONS (16)
 #endif
 
-#define SYSMINOR(aoemajor, aoeminor) ((aoemajor) * NPERSHELF + (aoeminor))
-#define AOEMAJOR(sysminor) ((sysminor) / NPERSHELF)
-#define AOEMINOR(sysminor) ((sysminor) % NPERSHELF)
 #define WHITESPACE " \t\v\f\n"
 
 enum {
@@ -75,72 +72,67 @@ enum {
 	DEVFL_UP = 1,	/* device is installed in system and ready for AoE->ATA commands */
 	DEVFL_TKILL = (1<<1),	/* flag for timer to know when to kill self */
 	DEVFL_EXT = (1<<2),	/* device accepts lba48 commands */
-	DEVFL_CLOSEWAIT = (1<<3), /* device is waiting for all closes to revalidate */
-	DEVFL_GDALLOC = (1<<4),	/* need to alloc gendisk */
-	DEVFL_KICKME = (1<<5),	/* slow polling network card catch */
-	DEVFL_NEWSIZE = (1<<6),	/* need to update dev size in block layer */
-
-	BUFFL_FAIL = 1,
+	DEVFL_GDALLOC = (1<<3),	/* need to alloc gendisk */
+	DEVFL_KICKME = (1<<4),	/* slow polling network card catch */
+	DEVFL_NEWSIZE = (1<<5),	/* need to update dev size in block layer */
 };
 
 enum {
 	DEFAULTBCNT = 2 * 512,	/* 2 sectors */
-	NPERSHELF = 16,		/* number of slots per shelf address */
-	FREETAG = -1,
 	MIN_BUFS = 16,
 	NTARGETS = 8,
 	NAOEIFS = 8,
-	NSKBPOOLMAX = 128,
+	NSKBPOOLMAX = 256,
+	NFACTIVE = 61,
 
 	TIMERTICK = HZ / 10,
 	MINTIMER = HZ >> 2,
 	MAXTIMER = HZ << 1,
-	HELPWAIT = 20,
 };
 
 struct buf {
-	struct list_head bufs;
-	ulong stime;	/* for disk stats */
-	ulong flags;
 	ulong nframesout;
 	ulong resid;
 	ulong bv_resid;
-	ulong bv_off;
 	sector_t sector;
 	struct bio *bio;
 	struct bio_vec *bv;
+	struct request *rq;
 };
 
 struct frame {
-	int tag;
+	struct list_head head;
+	u32 tag;
 	ulong waited;
+	struct aoetgt *t;		/* parent target I belong to */
+	sector_t lba;
+	struct sk_buff *skb;		/* command skb freed on module exit */
+	struct sk_buff *r_skb;		/* response skb for async processing */
 	struct buf *buf;
-	char *bufaddr;
+	struct bio_vec *bv;
 	ulong bcnt;
-	sector_t lba;
-	struct sk_buff *skb;
+	ulong bv_off;
 };
 
 struct aoeif {
 	struct net_device *nd;
-	unsigned char lost;
-	unsigned char lostjumbo;
-	ushort maxbcnt;
+	ulong lost;
+	int bcnt;
 };
 
 struct aoetgt {
 	unsigned char addr[6];
 	ushort nframes;
-	struct frame *frames;
+	struct aoedev *d;			/* parent device I belong to */
+	struct list_head ffree;			/* list of free frames */
 	struct aoeif ifs[NAOEIFS];
 	struct aoeif *ifp;	/* current aoeif in use */
 	ushort nout;
 	ushort maxout;
-	u16 lasttag;		/* last tag sent */
-	u16 useme;
+	ulong falloc;
 	ulong lastwadj;		/* last window adjustment */
+	int minbcnt;
 	int wpkts, rpkts;
-	int dataref;
 };
 
 struct aoedev {
@@ -153,6 +145,9 @@ struct aoedev {
 	u16 rttavg;		/* round trip average of requests/responses */
 	u16 mintimer;
 	u16 fw_ver;		/* version of blade's firmware */
+	u16 lasttag;		/* last tag sent */
+	u16 useme;
+	ulong ref;
 	struct work_struct work;/* disk create work struct */
 	struct gendisk *gd;
 	struct request_queue *blkq;
@@ -160,16 +155,31 @@ struct aoedev {
 	sector_t ssize;
 	struct timer_list timer;
 	spinlock_t lock;
-	struct sk_buff_head sendq;
 	struct sk_buff_head skbpool;
 	mempool_t *bufpool;	/* for deadlock-free Buf allocation */
-	struct list_head bufq;	/* queue of bios to work on */
-	struct buf *inprocess;	/* the one we're currently working on */
+	struct {		/* pointers to work in progress */
+		struct buf *buf;
+		struct bio *nxbio;
+		struct request *rq;
+	} ip;
+	ulong maxbcnt;
+	struct list_head factive[NFACTIVE];	/* hash of active frames */
 	struct aoetgt *targets[NTARGETS];
 	struct aoetgt **tgt;	/* target in use when working */
-	struct aoetgt **htgt;	/* target needing rexmit assistance */
+	struct aoetgt *htgt;	/* target needing rexmit assistance */
+	ulong ntargets;
+	ulong kicked;
 };
 
+/* kthread tracking */
+struct ktstate {
+	struct completion rendez;
+	struct task_struct *task;
+	wait_queue_head_t *waitq;
+	int (*fn) (void);
+	char *name;
+	spinlock_t *lock;
+};
 
 int aoeblk_init(void);
 void aoeblk_exit(void);
@@ -182,22 +192,29 @@ void aoechr_error(char *);
 
 void aoecmd_work(struct aoedev *d);
 void aoecmd_cfg(ushort aoemajor, unsigned char aoeminor);
-void aoecmd_ata_rsp(struct sk_buff *);
+struct sk_buff *aoecmd_ata_rsp(struct sk_buff *);
 void aoecmd_cfg_rsp(struct sk_buff *);
 void aoecmd_sleepwork(struct work_struct *);
 void aoecmd_cleanslate(struct aoedev *);
+void aoecmd_exit(void);
+int aoecmd_init(void);
 struct sk_buff *aoecmd_ata_id(struct aoedev *);
+void aoe_freetframe(struct frame *);
+void aoe_flush_iocq(void);
+void aoe_end_request(struct aoedev *, struct request *, int);
+int aoe_ktstart(struct ktstate *k);
+void aoe_ktstop(struct ktstate *k);
 
 int aoedev_init(void);
 void aoedev_exit(void);
-struct aoedev *aoedev_by_aoeaddr(int maj, int min);
-struct aoedev *aoedev_by_sysminor_m(ulong sysminor);
+struct aoedev *aoedev_by_aoeaddr(ulong maj, int min, int do_alloc);
 void aoedev_downdev(struct aoedev *d);
 int aoedev_flush(const char __user *str, size_t size);
+void aoe_failbuf(struct aoedev *, struct buf *);
+void aoedev_put(struct aoedev *);
 
 int aoenet_init(void);
 void aoenet_exit(void);
 void aoenet_xmit(struct sk_buff_head *);
 int is_aoe_netif(struct net_device *ifp);
 int set_aoe_iflist(const char __user *str, size_t size);
-

drivers/block/aoe/aoeblk.c

@@ -1,4 +1,4 @@
-/* Copyright (c) 2007 Coraid, Inc.  See COPYING for GPL terms. */
+/* Copyright (c) 2012 Coraid, Inc.  See COPYING for GPL terms. */
 /*
  * aoeblk.c
  * block device routines
@@ -161,68 +161,22 @@ aoeblk_release(struct gendisk *disk, fmode_t mode)
 }
 
 static void
-aoeblk_make_request(struct request_queue *q, struct bio *bio)
+aoeblk_request(struct request_queue *q)
 {
-	struct sk_buff_head queue;
 	struct aoedev *d;
-	struct buf *buf;
-	ulong flags;
-
-	blk_queue_bounce(q, &bio);
-
-	if (bio == NULL) {
-		printk(KERN_ERR "aoe: bio is NULL\n");
-		BUG();
-		return;
-	}
-	d = bio->bi_bdev->bd_disk->private_data;
-	if (d == NULL) {
-		printk(KERN_ERR "aoe: bd_disk->private_data is NULL\n");
-		BUG();
-		bio_endio(bio, -ENXIO);
-		return;
-	} else if (bio->bi_io_vec == NULL) {
-		printk(KERN_ERR "aoe: bi_io_vec is NULL\n");
-		BUG();
-		bio_endio(bio, -ENXIO);
-		return;
-	}
-	buf = mempool_alloc(d->bufpool, GFP_NOIO);
-	if (buf == NULL) {
-		printk(KERN_INFO "aoe: buf allocation failure\n");
-		bio_endio(bio, -ENOMEM);
-		return;
-	}
-	memset(buf, 0, sizeof(*buf));
-	INIT_LIST_HEAD(&buf->bufs);
-	buf->stime = jiffies;
-	buf->bio = bio;
-	buf->resid = bio->bi_size;
-	buf->sector = bio->bi_sector;
-	buf->bv = &bio->bi_io_vec[bio->bi_idx];
-	buf->bv_resid = buf->bv->bv_len;
-	WARN_ON(buf->bv_resid == 0);
-	buf->bv_off = buf->bv->bv_offset;
-
-	spin_lock_irqsave(&d->lock, flags);
+	struct request *rq;
 
+	d = q->queuedata;
 	if ((d->flags & DEVFL_UP) == 0) {
 		pr_info_ratelimited("aoe: device %ld.%d is not up\n",
 			d->aoemajor, d->aoeminor);
-		spin_unlock_irqrestore(&d->lock, flags);
-		mempool_free(buf, d->bufpool);
-		bio_endio(bio, -ENXIO);
+		while ((rq = blk_peek_request(q))) {
+			blk_start_request(rq);
+			aoe_end_request(d, rq, 1);
+		}
 		return;
 	}
-
-	list_add_tail(&buf->bufs, &d->bufq);
-
 	aoecmd_work(d);
-	__skb_queue_head_init(&queue);
-	skb_queue_splice_init(&d->sendq, &queue);
-
-	spin_unlock_irqrestore(&d->lock, flags);
-	aoenet_xmit(&queue);
 }
 
 static int
@@ -254,41 +208,54 @@ aoeblk_gdalloc(void *vp)
 {
 	struct aoedev *d = vp;
 	struct gendisk *gd;
+	mempool_t *mp;
+	struct request_queue *q;
+	enum { KB = 1024, MB = KB * KB, READ_AHEAD = 2 * MB, };
 	ulong flags;
 
 	gd = alloc_disk(AOE_PARTITIONS);
 	if (gd == NULL) {
-		printk(KERN_ERR
-			"aoe: cannot allocate disk structure for %ld.%d\n",
+		pr_err("aoe: cannot allocate disk structure for %ld.%d\n",
 			d->aoemajor, d->aoeminor);
 		goto err;
 	}
 
-	d->bufpool = mempool_create_slab_pool(MIN_BUFS, buf_pool_cache);
-	if (d->bufpool == NULL) {
+	mp = mempool_create(MIN_BUFS, mempool_alloc_slab, mempool_free_slab,
+		buf_pool_cache);
+	if (mp == NULL) {
 		printk(KERN_ERR "aoe: cannot allocate bufpool for %ld.%d\n",
 			d->aoemajor, d->aoeminor);
 		goto err_disk;
 	}
+	q = blk_init_queue(aoeblk_request, &d->lock);
+	if (q == NULL) {
+		pr_err("aoe: cannot allocate block queue for %ld.%d\n",
+			d->aoemajor, d->aoeminor);
+		mempool_destroy(mp);
+		goto err_disk;
+	}
 
 	d->blkq = blk_alloc_queue(GFP_KERNEL);
 	if (!d->blkq)
 		goto err_mempool;
-	blk_queue_make_request(d->blkq, aoeblk_make_request);
 	d->blkq->backing_dev_info.name = "aoe";
 	if (bdi_init(&d->blkq->backing_dev_info))
 		goto err_blkq;
 	spin_lock_irqsave(&d->lock, flags);
+	blk_queue_max_hw_sectors(d->blkq, BLK_DEF_MAX_SECTORS);
+	q->backing_dev_info.ra_pages = READ_AHEAD / PAGE_CACHE_SIZE;
+	d->bufpool = mp;
+	d->blkq = gd->queue = q;
+	q->queuedata = d;
+	d->gd = gd;
 	gd->major = AOE_MAJOR;
-	gd->first_minor = d->sysminor * AOE_PARTITIONS;
+	gd->first_minor = d->sysminor;
 	gd->fops = &aoe_bdops;
 	gd->private_data = d;
 	set_capacity(gd, d->ssize);
 	snprintf(gd->disk_name, sizeof gd->disk_name, "etherd/e%ld.%d",
 		d->aoemajor, d->aoeminor);
 
-	gd->queue = d->blkq;
-	d->gd = gd;
 	d->flags &= ~DEVFL_GDALLOC;
 	d->flags |= DEVFL_UP;
 

drivers/block/aoe/aoechr.c

@@ -1,4 +1,4 @@
-/* Copyright (c) 2007 Coraid, Inc.  See COPYING for GPL terms. */
+/* Copyright (c) 2012 Coraid, Inc.  See COPYING for GPL terms. */
 /*
  * aoechr.c
  * AoE character device driver
@@ -86,34 +86,34 @@ revalidate(const char __user *str, size_t size)
 	if (copy_from_user(buf, str, size))
 		return -EFAULT;
 
-	/* should be e%d.%d format */
 	n = sscanf(buf, "e%d.%d", &major, &minor);
 	if (n != 2) {
-		printk(KERN_ERR "aoe: invalid device specification\n");
+		pr_err("aoe: invalid device specification %s\n", buf);
 		return -EINVAL;
 	}
-	d = aoedev_by_aoeaddr(major, minor);
+	d = aoedev_by_aoeaddr(major, minor, 0);
 	if (!d)
 		return -EINVAL;
 	spin_lock_irqsave(&d->lock, flags);
 	aoecmd_cleanslate(d);
+	aoecmd_cfg(major, minor);
 loop:
 	skb = aoecmd_ata_id(d);
 	spin_unlock_irqrestore(&d->lock, flags);
 	/* try again if we are able to sleep a bit,
 	 * otherwise give up this revalidation
 	 */
-	if (!skb && !msleep_interruptible(200)) {
+	if (!skb && !msleep_interruptible(250)) {
 		spin_lock_irqsave(&d->lock, flags);
 		goto loop;
 	}
+	aoedev_put(d);
 	if (skb) {
 		struct sk_buff_head queue;
 		__skb_queue_head_init(&queue);
 		__skb_queue_tail(&queue, skb);
 		aoenet_xmit(&queue);
 	}
-	aoecmd_cfg(major, minor);
 	return 0;
 }
 
@@ -174,6 +174,7 @@ aoechr_write(struct file *filp, const char __user *buf, size_t cnt, loff_t *offp
 		break;
 	case MINOR_FLUSH:
 		ret = aoedev_flush(buf, cnt);
+		break;
 	}
 	if (ret == 0)
 		ret = cnt;

drivers/block/aoe/aoecmd.c

@@ -1,4 +1,4 @@
-/* Copyright (c) 2007 Coraid, Inc.  See COPYING for GPL terms. */
+/* Copyright (c) 2012 Coraid, Inc.  See COPYING for GPL terms. */
 /*
  * aoecmd.c
  * Filesystem request handling methods
@@ -12,10 +12,19 @@
 #include <linux/netdevice.h>
 #include <linux/genhd.h>
 #include <linux/moduleparam.h>
+#include <linux/workqueue.h>
+#include <linux/kthread.h>
 #include <net/net_namespace.h>
 #include <asm/unaligned.h>
+#include <linux/uio.h>
 #include "aoe.h"
 
+#define MAXIOC (8192)	/* default meant to avoid most soft lockups */
+
+static void ktcomplete(struct frame *, struct sk_buff *);
+
+static struct buf *nextbuf(struct aoedev *);
+
 static int aoe_deadsecs = 60 * 3;
 module_param(aoe_deadsecs, int, 0644);
 MODULE_PARM_DESC(aoe_deadsecs, "After aoe_deadsecs seconds, give up and fail dev.");
@@ -25,6 +34,15 @@ module_param(aoe_maxout, int, 0644);
 MODULE_PARM_DESC(aoe_maxout,
 	"Only aoe_maxout outstanding packets for every MAC on eX.Y.");
 
+static wait_queue_head_t ktiowq;
+static struct ktstate kts;
+
+/* io completion queue */
+static struct {
+	struct list_head head;
+	spinlock_t lock;
+} iocq;
+
 static struct sk_buff *
 new_skb(ulong len)
 {
@@ -41,15 +59,21 @@ new_skb(ulong len)
 }
 
 static struct frame *
-getframe(struct aoetgt *t, int tag)
+getframe(struct aoedev *d, u32 tag)
 {
-	struct frame *f, *e;
+	struct frame *f;
+	struct list_head *head, *pos, *nx;
+	u32 n;
 
-	f = t->frames;
-	e = f + t->nframes;
-	for (; f<e; f++)
-		if (f->tag == tag)
+	n = tag % NFACTIVE;
+	head = &d->factive[n];
+	list_for_each_safe(pos, nx, head) {
+		f = list_entry(pos, struct frame, head);
+		if (f->tag == tag) {
+			list_del(pos);
 			return f;
+		}
+	}
 	return NULL;
 }
 
@@ -59,18 +83,18 @@ getframe(struct aoetgt *t, int tag)
  * This driver reserves tag -1 to mean "unused frame."
  */
 static int
-newtag(struct aoetgt *t)
+newtag(struct aoedev *d)
 {
 	register ulong n;
 
 	n = jiffies & 0xffff;
-	return n |= (++t->lasttag & 0x7fff) << 16;
+	return n |= (++d->lasttag & 0x7fff) << 16;
 }
 
-static int
+static u32
 aoehdr_atainit(struct aoedev *d, struct aoetgt *t, struct aoe_hdr *h)
 {
-	u32 host_tag = newtag(t);
+	u32 host_tag = newtag(d);
 
 	memcpy(h->src, t->ifp->nd->dev_addr, sizeof h->src);
 	memcpy(h->dst, t->addr, sizeof h->dst);
@@ -95,16 +119,18 @@ put_lba(struct aoe_atahdr *ah, sector_t lba)
 	ah->lba5 = lba >>= 8;
 }
 
-static void
+static struct aoeif *
 ifrotate(struct aoetgt *t)
 {
-	t->ifp++;
-	if (t->ifp >= &t->ifs[NAOEIFS] || t->ifp->nd == NULL)
-		t->ifp = t->ifs;
-	if (t->ifp->nd == NULL) {
-		printk(KERN_INFO "aoe: no interface to rotate to\n");
-		BUG();
-	}
+	struct aoeif *ifp;
+
+	ifp = t->ifp;
+	ifp++;
+	if (ifp >= &t->ifs[NAOEIFS] || ifp->nd == NULL)
+		ifp = t->ifs;
+	if (ifp->nd == NULL)
+		return NULL;
+	return t->ifp = ifp;
 }
 
 static void
@@ -129,78 +155,128 @@ skb_pool_get(struct aoedev *d)
 	return NULL;
 }
 
-/* freeframe is where we do our load balancing so it's a little hairy. */
+void
+aoe_freetframe(struct frame *f)
+{
+	struct aoetgt *t;
+
+	t = f->t;
+	f->buf = NULL;
+	f->bv = NULL;
+	f->r_skb = NULL;
+	list_add(&f->head, &t->ffree);
+}
+
 static struct frame *
-freeframe(struct aoedev *d)
+newtframe(struct aoedev *d, struct aoetgt *t)
 {
-	struct frame *f, *e, *rf;
-	struct aoetgt **t;
+	struct frame *f;
 	struct sk_buff *skb;
+	struct list_head *pos;
+
+	if (list_empty(&t->ffree)) {
+		if (t->falloc >= NSKBPOOLMAX*2)
+			return NULL;
+		f = kcalloc(1, sizeof(*f), GFP_ATOMIC);
+		if (f == NULL)
+			return NULL;
+		t->falloc++;
+		f->t = t;
+	} else {
+		pos = t->ffree.next;
+		list_del(pos);
+		f = list_entry(pos, struct frame, head);
+	}
+
+	skb = f->skb;
+	if (skb == NULL) {
+		f->skb = skb = new_skb(ETH_ZLEN);
+		if (!skb) {
+bail:			aoe_freetframe(f);
+			return NULL;
+		}
+	}
+
+	if (atomic_read(&skb_shinfo(skb)->dataref) != 1) {
+		skb = skb_pool_get(d);
+		if (skb == NULL)
+			goto bail;
+		skb_pool_put(d, f->skb);
+		f->skb = skb;
+	}
+
+	skb->truesize -= skb->data_len;
+	skb_shinfo(skb)->nr_frags = skb->data_len = 0;
+	skb_trim(skb, 0);
+	return f;
+}
+
+static struct frame *
+newframe(struct aoedev *d)
+{
+	struct frame *f;
+	struct aoetgt *t, **tt;
+	int totout = 0;
 
 	if (d->targets[0] == NULL) {	/* shouldn't happen, but I'm paranoid */
 		printk(KERN_ERR "aoe: NULL TARGETS!\n");
 		return NULL;
 	}
-	t = d->tgt;
-	t++;
-	if (t >= &d->targets[NTARGETS] || !*t)
-		t = d->targets;
+	tt = d->tgt;	/* last used target */
 	for (;;) {
-		if ((*t)->nout < (*t)->maxout
+		tt++;
+		if (tt >= &d->targets[NTARGETS] || !*tt)
+			tt = d->targets;
+		t = *tt;
+		totout += t->nout;
+		if (t->nout < t->maxout
 		&& t != d->htgt
-		&& (*t)->ifp->nd) {
-			rf = NULL;
-			f = (*t)->frames;
-			e = f + (*t)->nframes;
-			for (; f < e; f++) {
-				if (f->tag != FREETAG)
-					continue;
-				skb = f->skb;
-				if (!skb
-				&& !(f->skb = skb = new_skb(ETH_ZLEN)))
-					continue;
-				if (atomic_read(&skb_shinfo(skb)->dataref)
-					!= 1) {
-					if (!rf)
-						rf = f;
-					continue;
-				}
-gotone:				skb_shinfo(skb)->nr_frags = skb->data_len = 0;
-				skb_trim(skb, 0);
-				d->tgt = t;
-				ifrotate(*t);
+		&& t->ifp->nd) {
+			f = newtframe(d, t);
+			if (f) {
+				ifrotate(t);
+				d->tgt = tt;
 				return f;
 			}
-			/* Work can be done, but the network layer is
-			   holding our precious packets.  Try to grab
-			   one from the pool. */
-			f = rf;
-			if (f == NULL) {	/* more paranoia */
-				printk(KERN_ERR
-					"aoe: freeframe: %s.\n",
-					"unexpected null rf");
-				d->flags |= DEVFL_KICKME;
-				return NULL;
-			}
-			skb = skb_pool_get(d);
-			if (skb) {
-				skb_pool_put(d, f->skb);
-				f->skb = skb;
-				goto gotone;
-			}
-			(*t)->dataref++;
-			if ((*t)->nout == 0)
-				d->flags |= DEVFL_KICKME;
 		}
-		if (t == d->tgt)	/* we've looped and found nada */
+		if (tt == d->tgt)	/* we've looped and found nada */
 			break;
-		t++;
-		if (t >= &d->targets[NTARGETS] || !*t)
-			t = d->targets;
+	}
+	if (totout == 0) {
+		d->kicked++;
+		d->flags |= DEVFL_KICKME;
 	}
 	return NULL;
 }
 
+static void
+skb_fillup(struct sk_buff *skb, struct bio_vec *bv, ulong off, ulong cnt)
+{
+	int frag = 0;
+	ulong fcnt;
+loop:
+	fcnt = bv->bv_len - (off - bv->bv_offset);
+	if (fcnt > cnt)
+		fcnt = cnt;
+	skb_fill_page_desc(skb, frag++, bv->bv_page, off, fcnt);
+	cnt -= fcnt;
+	if (cnt <= 0)
+		return;
+	bv++;
+	off = bv->bv_offset;
+	goto loop;
+}
+
+static void
+fhash(struct frame *f)
+{
+	struct aoedev *d = f->t->d;
+	u32 n;
+
+	n = f->tag % NFACTIVE;
+	list_add_tail(&f->head, &d->factive[n]);
+}
+
 static int
 aoecmd_ata_rw(struct aoedev *d)
 {
@@ -208,26 +284,47 @@ aoecmd_ata_rw(struct aoedev *d)
 	struct aoe_hdr *h;
 	struct aoe_atahdr *ah;
 	struct buf *buf;
-	struct bio_vec *bv;
 	struct aoetgt *t;
 	struct sk_buff *skb;
-	ulong bcnt;
+	struct sk_buff_head queue;
+	ulong bcnt, fbcnt;
 	char writebit, extbit;
 
 	writebit = 0x10;
 	extbit = 0x4;
 
-	f = freeframe(d);
+	buf = nextbuf(d);
+	if (buf == NULL)
+		return 0;
+	f = newframe(d);
 	if (f == NULL)
 		return 0;
 	t = *d->tgt;
-	buf = d->inprocess;
-	bv = buf->bv;
-	bcnt = t->ifp->maxbcnt;
+	bcnt = d->maxbcnt;
 	if (bcnt == 0)
 		bcnt = DEFAULTBCNT;
-	if (bcnt > buf->bv_resid)
-		bcnt = buf->bv_resid;
+	if (bcnt > buf->resid)
+		bcnt = buf->resid;
+	fbcnt = bcnt;
+	f->bv = buf->bv;
+	f->bv_off = f->bv->bv_offset + (f->bv->bv_len - buf->bv_resid);
+	do {
+		if (fbcnt < buf->bv_resid) {
+			buf->bv_resid -= fbcnt;
+			buf->resid -= fbcnt;
+			break;
+		}
+		fbcnt -= buf->bv_resid;
+		buf->resid -= buf->bv_resid;
+		if (buf->resid == 0) {
+			d->ip.buf = NULL;
+			break;
+		}
+		buf->bv++;
+		buf->bv_resid = buf->bv->bv_len;
+		WARN_ON(buf->bv_resid == 0);
+	} while (fbcnt);
+
 	/* initialize the headers & frame */
 	skb = f->skb;
 	h = (struct aoe_hdr *) skb_mac_header(skb);
@@ -235,10 +332,10 @@ aoecmd_ata_rw(struct aoedev *d)
 	skb_put(skb, sizeof *h + sizeof *ah);
 	memset(h, 0, skb->len);
 	f->tag = aoehdr_atainit(d, t, h);
+	fhash(f);
 	t->nout++;
 	f->waited = 0;
 	f->buf = buf;
-	f->bufaddr = page_address(bv->bv_page) + buf->bv_off;
 	f->bcnt = bcnt;
 	f->lba = buf->sector;
 
@@ -253,10 +350,11 @@ aoecmd_ata_rw(struct aoedev *d)
 		ah->lba3 |= 0xe0;	/* LBA bit + obsolete 0xa0 */
 	}
 	if (bio_data_dir(buf->bio) == WRITE) {
-		skb_fill_page_desc(skb, 0, bv->bv_page, buf->bv_off, bcnt);
+		skb_fillup(skb, f->bv, f->bv_off, bcnt);
 		ah->aflags |= AOEAFL_WRITE;
 		skb->len += bcnt;
 		skb->data_len = bcnt;
+		skb->truesize += bcnt;
 		t->wpkts++;
 	} else {
 		t->rpkts++;
@@ -267,23 +365,15 @@ aoecmd_ata_rw(struct aoedev *d)
 
 	/* mark all tracking fields and load out */
 	buf->nframesout += 1;
-	buf->bv_off += bcnt;
-	buf->bv_resid -= bcnt;
-	buf->resid -= bcnt;
 	buf->sector += bcnt >> 9;
-	if (buf->resid == 0) {
-		d->inprocess = NULL;
-	} else if (buf->bv_resid == 0) {
-		buf->bv = ++bv;
-		buf->bv_resid = bv->bv_len;
-		WARN_ON(buf->bv_resid == 0);
-		buf->bv_off = bv->bv_offset;
-	}
 
 	skb->dev = t->ifp->nd;
 	skb = skb_clone(skb, GFP_ATOMIC);
-	if (skb)
-		__skb_queue_tail(&d->sendq, skb);
+	if (skb) {
+		__skb_queue_head_init(&queue);
+		__skb_queue_tail(&queue, skb);
+		aoenet_xmit(&queue);
+	}
 	return 1;
 }
 
@@ -330,17 +420,25 @@ cont:
 }
 
 static void
-resend(struct aoedev *d, struct aoetgt *t, struct frame *f)
+resend(struct aoedev *d, struct frame *f)
 {
 	struct sk_buff *skb;
+	struct sk_buff_head queue;
 	struct aoe_hdr *h;
 	struct aoe_atahdr *ah;
+	struct aoetgt *t;
 	char buf[128];
 	u32 n;
 
-	ifrotate(t);
-	n = newtag(t);
+	t = f->t;
+	n = newtag(d);
 	skb = f->skb;
+	if (ifrotate(t) == NULL) {
+		/* probably can't happen, but set it up to fail anyway */
+		pr_info("aoe: resend: no interfaces to rotate to.\n");
+		ktcomplete(f, NULL);
+		return;
+	}
 	h = (struct aoe_hdr *) skb_mac_header(skb);
 	ah = (struct aoe_atahdr *) (h+1);
 
@@ -351,39 +449,22 @@ resend(struct aoedev *d, struct aoetgt *t, struct frame *f)
 	aoechr_error(buf);
 
 	f->tag = n;
+	fhash(f);
 	h->tag = cpu_to_be32(n);
 	memcpy(h->dst, t->addr, sizeof h->dst);
 	memcpy(h->src, t->ifp->nd->dev_addr, sizeof h->src);
 
-	switch (ah->cmdstat) {
-	default:
-		break;
-	case ATA_CMD_PIO_READ:
-	case ATA_CMD_PIO_READ_EXT:
-	case ATA_CMD_PIO_WRITE:
-	case ATA_CMD_PIO_WRITE_EXT:
-		put_lba(ah, f->lba);
-
-		n = f->bcnt;
-		if (n > DEFAULTBCNT)
-			n = DEFAULTBCNT;
-		ah->scnt = n >> 9;
-		if (ah->aflags & AOEAFL_WRITE) {
-			skb_fill_page_desc(skb, 0, virt_to_page(f->bufaddr),
-				offset_in_page(f->bufaddr), n);
-			skb->len = sizeof *h + sizeof *ah + n;
-			skb->data_len = n;
-		}
-	}
 	skb->dev = t->ifp->nd;
 	skb = skb_clone(skb, GFP_ATOMIC);
 	if (skb == NULL)
 		return;
-	__skb_queue_tail(&d->sendq, skb);
+	__skb_queue_head_init(&queue);
+	__skb_queue_tail(&queue, skb);
+	aoenet_xmit(&queue);
 }
 
 static int
-tsince(int tag)
+tsince(u32 tag)
 {
 	int n;
 
@@ -407,58 +488,65 @@ getif(struct aoetgt *t, struct net_device *nd)
 	return NULL;
 }
 
-static struct aoeif *
-addif(struct aoetgt *t, struct net_device *nd)
-{
-	struct aoeif *p;
-
-	p = getif(t, NULL);
-	if (!p)
-		return NULL;
-	p->nd = nd;
-	p->maxbcnt = DEFAULTBCNT;
-	p->lost = 0;
-	p->lostjumbo = 0;
-	return p;
-}
-
 static void
 ejectif(struct aoetgt *t, struct aoeif *ifp)
 {
 	struct aoeif *e;
+	struct net_device *nd;
 	ulong n;
 
+	nd = ifp->nd;
 	e = t->ifs + NAOEIFS - 1;
 	n = (e - ifp) * sizeof *ifp;
 	memmove(ifp, ifp+1, n);
 	e->nd = NULL;
+	dev_put(nd);
 }
 
 static int
 sthtith(struct aoedev *d)
 {
-	struct frame *f, *e, *nf;
+	struct frame *f, *nf;
+	struct list_head *nx, *pos, *head;
 	struct sk_buff *skb;
-	struct aoetgt *ht = *d->htgt;
-
-	f = ht->frames;
-	e = f + ht->nframes;
-	for (; f < e; f++) {
-		if (f->tag == FREETAG)
-			continue;
-		nf = freeframe(d);
-		if (!nf)
-			return 0;
-		skb = nf->skb;
-		*nf = *f;
-		f->skb = skb;
-		f->tag = FREETAG;
-		nf->waited = 0;
-		ht->nout--;
-		(*d->tgt)->nout++;
-		resend(d, *d->tgt, nf);
+	struct aoetgt *ht = d->htgt;
+	int i;
+
+	for (i = 0; i < NFACTIVE; i++) {
+		head = &d->factive[i];
+		list_for_each_safe(pos, nx, head) {
+			f = list_entry(pos, struct frame, head);
+			if (f->t != ht)
+				continue;
+
+			nf = newframe(d);
+			if (!nf)
+				return 0;
+
+			/* remove frame from active list */
+			list_del(pos);
+
+			/* reassign all pertinent bits to new outbound frame */
+			skb = nf->skb;
+			nf->skb = f->skb;
+			nf->buf = f->buf;
+			nf->bcnt = f->bcnt;
+			nf->lba = f->lba;
+			nf->bv = f->bv;
+			nf->bv_off = f->bv_off;
+			nf->waited = 0;
+			f->skb = skb;
+			aoe_freetframe(f);
+			ht->nout--;
+			nf->t->nout++;
+			resend(d, nf);
+		}
 	}
-	/* he's clean, he's useless.  take away his interfaces */
+	/* We've cleaned up the outstanding so take away his
+	 * interfaces so he won't be used.  We should remove him from
+	 * the target array here, but cleaning up a target is
+	 * involved.  PUNT!
+	 */
 	memset(ht->ifs, 0, sizeof ht->ifs);
 	d->htgt = NULL;
 	return 1;
@@ -477,13 +565,15 @@ ata_scnt(unsigned char *packet) {
 static void
 rexmit_timer(ulong vp)
 {
-	struct sk_buff_head queue;
 	struct aoedev *d;
 	struct aoetgt *t, **tt, **te;
 	struct aoeif *ifp;
-	struct frame *f, *e;
+	struct frame *f;
+	struct list_head *head, *pos, *nx;
+	LIST_HEAD(flist);
 	register long timeout;
 	ulong flags, n;
+	int i;
 
 	d = (struct aoedev *) vp;
 
@@ -497,58 +587,22 @@ rexmit_timer(ulong vp)
 		spin_unlock_irqrestore(&d->lock, flags);
 		return;
 	}
-	tt = d->targets;
-	te = tt + NTARGETS;
-	for (; tt < te && *tt; tt++) {
-		t = *tt;
-		f = t->frames;
-		e = f + t->nframes;
-		for (; f < e; f++) {
-			if (f->tag == FREETAG
-			|| tsince(f->tag) < timeout)
-				continue;
-			n = f->waited += timeout;
-			n /= HZ;
-			if (n > aoe_deadsecs) {
-				/* waited too long.  device failure. */
-				aoedev_downdev(d);
-				break;
-			}
-
-			if (n > HELPWAIT /* see if another target can help */
-			&& (tt != d->targets || d->targets[1]))
-				d->htgt = tt;
-
-			if (t->nout == t->maxout) {
-				if (t->maxout > 1)
-					t->maxout--;
-				t->lastwadj = jiffies;
-			}
-
-			ifp = getif(t, f->skb->dev);
-			if (ifp && ++ifp->lost > (t->nframes << 1)
-			&& (ifp != t->ifs || t->ifs[1].nd)) {
-				ejectif(t, ifp);
-				ifp = NULL;
-			}
 
-			if (ata_scnt(skb_mac_header(f->skb)) > DEFAULTBCNT / 512
-			&& ifp && ++ifp->lostjumbo > (t->nframes << 1)
-			&& ifp->maxbcnt != DEFAULTBCNT) {
-				printk(KERN_INFO
-					"aoe: e%ld.%d: "
-					"too many lost jumbo on "
-					"%s:%pm - "
-					"falling back to %d frames.\n",
-					d->aoemajor, d->aoeminor,
-					ifp->nd->name, t->addr,
-					DEFAULTBCNT);
-				ifp->maxbcnt = 0;
-			}
-			resend(d, t, f);
+	/* collect all frames to rexmit into flist */
+	for (i = 0; i < NFACTIVE; i++) {
+		head = &d->factive[i];
+		list_for_each_safe(pos, nx, head) {
+			f = list_entry(pos, struct frame, head);
+			if (tsince(f->tag) < timeout)
+				break;	/* end of expired frames */
+			/* move to flist for later processing */
+			list_move_tail(pos, &flist);
 		}
-
-		/* window check */
+	}
+	/* window check */
+	tt = d->targets;
+	te = tt + d->ntargets;
+	for (; tt < te && (t = *tt); tt++) {
 		if (t->nout == t->maxout
 		&& t->maxout < t->nframes
 		&& (jiffies - t->lastwadj)/HZ > 10) {
@@ -557,45 +611,173 @@ rexmit_timer(ulong vp)
 		}
 	}
 
-	if (!skb_queue_empty(&d->sendq)) {
+	if (!list_empty(&flist)) {	/* retransmissions necessary */
 		n = d->rttavg <<= 1;
 		if (n > MAXTIMER)
 			d->rttavg = MAXTIMER;
 	}
 
-	if (d->flags & DEVFL_KICKME || d->htgt) {
-		d->flags &= ~DEVFL_KICKME;
-		aoecmd_work(d);
+	/* process expired frames */
+	while (!list_empty(&flist)) {
+		pos = flist.next;
+		f = list_entry(pos, struct frame, head);
+		n = f->waited += timeout;
+		n /= HZ;
+		if (n > aoe_deadsecs) {
+			/* Waited too long.  Device failure.
+			 * Hang all frames on first hash bucket for downdev
+			 * to clean up.
+			 */
+			list_splice(&flist, &d->factive[0]);
+			aoedev_downdev(d);
+			break;
+		}
+		list_del(pos);
+
+		t = f->t;
+		if (n > aoe_deadsecs/2)
+			d->htgt = t; /* see if another target can help */
+
+		if (t->nout == t->maxout) {
+			if (t->maxout > 1)
+				t->maxout--;
+			t->lastwadj = jiffies;
+		}
+
+		ifp = getif(t, f->skb->dev);
+		if (ifp && ++ifp->lost > (t->nframes << 1)
+		&& (ifp != t->ifs || t->ifs[1].nd)) {
+			ejectif(t, ifp);
+			ifp = NULL;
+		}
+		resend(d, f);
 	}
 
-	__skb_queue_head_init(&queue);
-	skb_queue_splice_init(&d->sendq, &queue);
+	if ((d->flags & DEVFL_KICKME || d->htgt) && d->blkq) {
+		d->flags &= ~DEVFL_KICKME;
+		d->blkq->request_fn(d->blkq);
+	}
 
 	d->timer.expires = jiffies + TIMERTICK;
 	add_timer(&d->timer);
 
 	spin_unlock_irqrestore(&d->lock, flags);
+}
 
-	aoenet_xmit(&queue);
+static unsigned long
+rqbiocnt(struct request *r)
+{
+	struct bio *bio;
+	unsigned long n = 0;
+
+	__rq_for_each_bio(bio, r)
+		n++;
+	return n;
+}
+
+/* This can be removed if we are certain that no users of the block
+ * layer will ever use zero-count pages in bios.  Otherwise we have to
+ * protect against the put_page sometimes done by the network layer.
+ *
+ * See http://oss.sgi.com/archives/xfs/2007-01/msg00594.html for
+ * discussion.
+ *
+ * We cannot use get_page in the workaround, because it insists on a
+ * positive page count as a precondition.  So we use _count directly.
+ */
+static void
+bio_pageinc(struct bio *bio)
+{
+	struct bio_vec *bv;
+	struct page *page;
+	int i;
+
+	bio_for_each_segment(bv, bio, i) {
+		page = bv->bv_page;
+		/* Non-zero page count for non-head members of
+		 * compound pages is no longer allowed by the kernel,
+		 * but this has never been seen here.
+		 */
+		if (unlikely(PageCompound(page)))
+			if (compound_trans_head(page) != page) {
+				pr_crit("page tail used for block I/O\n");
+				BUG();
+			}
+		atomic_inc(&page->_count);
+	}
+}
+
+static void
+bio_pagedec(struct bio *bio)
+{
+	struct bio_vec *bv;
+	int i;
+
+	bio_for_each_segment(bv, bio, i)
+		atomic_dec(&bv->bv_page->_count);
+}
+
+static void
+bufinit(struct buf *buf, struct request *rq, struct bio *bio)
+{
+	struct bio_vec *bv;
+
+	memset(buf, 0, sizeof(*buf));
+	buf->rq = rq;
+	buf->bio = bio;
+	buf->resid = bio->bi_size;
+	buf->sector = bio->bi_sector;
+	bio_pageinc(bio);
+	buf->bv = bv = &bio->bi_io_vec[bio->bi_idx];
+	buf->bv_resid = bv->bv_len;
+	WARN_ON(buf->bv_resid == 0);
+}
+
+static struct buf *
+nextbuf(struct aoedev *d)
+{
+	struct request *rq;
+	struct request_queue *q;
+	struct buf *buf;
+	struct bio *bio;
+
+	q = d->blkq;
+	if (q == NULL)
+		return NULL;	/* initializing */
+	if (d->ip.buf)
+		return d->ip.buf;
+	rq = d->ip.rq;
+	if (rq == NULL) {
+		rq = blk_peek_request(q);
+		if (rq == NULL)
+			return NULL;
+		blk_start_request(rq);
+		d->ip.rq = rq;
+		d->ip.nxbio = rq->bio;
+		rq->special = (void *) rqbiocnt(rq);
+	}
+	buf = mempool_alloc(d->bufpool, GFP_ATOMIC);
+	if (buf == NULL) {
+		pr_err("aoe: nextbuf: unable to mempool_alloc!\n");
+		return NULL;
+	}
+	bio = d->ip.nxbio;
+	bufinit(buf, rq, bio);
+	bio = bio->bi_next;
+	d->ip.nxbio = bio;
+	if (bio == NULL)
+		d->ip.rq = NULL;
+	return d->ip.buf = buf;
 }
 
 /* enters with d->lock held */
 void
 aoecmd_work(struct aoedev *d)
 {
-	struct buf *buf;
-loop:
 	if (d->htgt && !sthtith(d))
 		return;
-	if (d->inprocess == NULL) {
-		if (list_empty(&d->bufq))
-			return;
-		buf = container_of(d->bufq.next, struct buf, bufs);
-		list_del(d->bufq.next);
-		d->inprocess = buf;
-	}
-	if (aoecmd_ata_rw(d))
-		goto loop;
+	while (aoecmd_ata_rw(d))
+		;
 }
 
 /* this function performs work that has been deferred until sleeping is OK
@@ -604,28 +786,25 @@ void
 aoecmd_sleepwork(struct work_struct *work)
 {
 	struct aoedev *d = container_of(work, struct aoedev, work);
+	struct block_device *bd;
+	u64 ssize;
 
 	if (d->flags & DEVFL_GDALLOC)
 		aoeblk_gdalloc(d);
 
 	if (d->flags & DEVFL_NEWSIZE) {
-		struct block_device *bd;
-		unsigned long flags;
-		u64 ssize;
-
 		ssize = get_capacity(d->gd);
 		bd = bdget_disk(d->gd, 0);
-
 		if (bd) {
 			mutex_lock(&bd->bd_inode->i_mutex);
 			i_size_write(bd->bd_inode, (loff_t)ssize<<9);
 			mutex_unlock(&bd->bd_inode->i_mutex);
 			bdput(bd);
 		}
-		spin_lock_irqsave(&d->lock, flags);
+		spin_lock_irq(&d->lock);
 		d->flags |= DEVFL_UP;
 		d->flags &= ~DEVFL_NEWSIZE;
-		spin_unlock_irqrestore(&d->lock, flags);
+		spin_unlock_irq(&d->lock);
 	}
 }
 
@@ -718,163 +897,299 @@ gettgt(struct aoedev *d, char *addr)
 	return NULL;
 }
 
-static inline void
-diskstats(struct gendisk *disk, struct bio *bio, ulong duration, sector_t sector)
+static void
+bvcpy(struct bio_vec *bv, ulong off, struct sk_buff *skb, long cnt)
+{
+	ulong fcnt;
+	char *p;
+	int soff = 0;
+loop:
+	fcnt = bv->bv_len - (off - bv->bv_offset);
+	if (fcnt > cnt)
+		fcnt = cnt;
+	p = page_address(bv->bv_page) + off;
+	skb_copy_bits(skb, soff, p, fcnt);
+	soff += fcnt;
+	cnt -= fcnt;
+	if (cnt <= 0)
+		return;
+	bv++;
+	off = bv->bv_offset;
+	goto loop;
+}
+
+void
+aoe_end_request(struct aoedev *d, struct request *rq, int fastfail)
+{
+	struct bio *bio;
+	int bok;
+	struct request_queue *q;
+
+	q = d->blkq;
+	if (rq == d->ip.rq)
+		d->ip.rq = NULL;
+	do {
+		bio = rq->bio;
+		bok = !fastfail && test_bit(BIO_UPTODATE, &bio->bi_flags);
+	} while (__blk_end_request(rq, bok ? 0 : -EIO, bio->bi_size));
+
+	/* cf. http://lkml.org/lkml/2006/10/31/28 */
+	if (!fastfail)
+		q->request_fn(q);
+}
+
+static void
+aoe_end_buf(struct aoedev *d, struct buf *buf)
+{
+	struct request *rq;
+	unsigned long n;
+
+	if (buf == d->ip.buf)
+		d->ip.buf = NULL;
+	rq = buf->rq;
+	bio_pagedec(buf->bio);
+	mempool_free(buf, d->bufpool);
+	n = (unsigned long) rq->special;
+	rq->special = (void *) --n;
+	if (n == 0)
+		aoe_end_request(d, rq, 0);
+}
+
+static void
+ktiocomplete(struct frame *f)
 {
-	unsigned long n_sect = bio->bi_size >> 9;
-	const int rw = bio_data_dir(bio);
-	struct hd_struct *part;
-	int cpu;
+	struct aoe_hdr *hin, *hout;
+	struct aoe_atahdr *ahin, *ahout;
+	struct buf *buf;
+	struct sk_buff *skb;
+	struct aoetgt *t;
+	struct aoeif *ifp;
+	struct aoedev *d;
+	long n;
+
+	if (f == NULL)
+		return;
+
+	t = f->t;
+	d = t->d;
+
+	hout = (struct aoe_hdr *) skb_mac_header(f->skb);
+	ahout = (struct aoe_atahdr *) (hout+1);
+	buf = f->buf;
+	skb = f->r_skb;
+	if (skb == NULL)
+		goto noskb;	/* just fail the buf. */
+
+	hin = (struct aoe_hdr *) skb->data;
+	skb_pull(skb, sizeof(*hin));
+	ahin = (struct aoe_atahdr *) skb->data;
+	skb_pull(skb, sizeof(*ahin));
+	if (ahin->cmdstat & 0xa9) {	/* these bits cleared on success */
+		pr_err("aoe: ata error cmd=%2.2Xh stat=%2.2Xh from e%ld.%d\n",
+			ahout->cmdstat, ahin->cmdstat,
+			d->aoemajor, d->aoeminor);
+noskb:	if (buf)
+			clear_bit(BIO_UPTODATE, &buf->bio->bi_flags);
+		goto badrsp;
+	}
 
-	cpu = part_stat_lock();
-	part = disk_map_sector_rcu(disk, sector);
+	n = ahout->scnt << 9;
+	switch (ahout->cmdstat) {
+	case ATA_CMD_PIO_READ:
+	case ATA_CMD_PIO_READ_EXT:
+		if (skb->len < n) {
+			pr_err("aoe: runt data size in read.  skb->len=%d need=%ld\n",
+				skb->len, n);
+			clear_bit(BIO_UPTODATE, &buf->bio->bi_flags);
+			break;
+		}
+		bvcpy(f->bv, f->bv_off, skb, n);
+	case ATA_CMD_PIO_WRITE:
+	case ATA_CMD_PIO_WRITE_EXT:
+		spin_lock_irq(&d->lock);
+		ifp = getif(t, skb->dev);
+		if (ifp)
+			ifp->lost = 0;
+		if (d->htgt == t) /* I'll help myself, thank you. */
+			d->htgt = NULL;
+		spin_unlock_irq(&d->lock);
+		break;
+	case ATA_CMD_ID_ATA:
+		if (skb->len < 512) {
+			pr_info("aoe: runt data size in ataid.  skb->len=%d\n",
+				skb->len);
+			break;
+		}
+		if (skb_linearize(skb))
+			break;
+		spin_lock_irq(&d->lock);
+		ataid_complete(d, t, skb->data);
+		spin_unlock_irq(&d->lock);
+		break;
+	default:
+		pr_info("aoe: unrecognized ata command %2.2Xh for %d.%d\n",
+			ahout->cmdstat,
+			be16_to_cpu(get_unaligned(&hin->major)),
+			hin->minor);
+	}
+badrsp:
+	spin_lock_irq(&d->lock);
+
+	aoe_freetframe(f);
+
+	if (buf && --buf->nframesout == 0 && buf->resid == 0)
+		aoe_end_buf(d, buf);
+
+	aoecmd_work(d);
+
+	spin_unlock_irq(&d->lock);
+	aoedev_put(d);
+	dev_kfree_skb(skb);
+}
+
+/* Enters with iocq.lock held.
+ * Returns true iff responses needing processing remain.
+ */
+static int
+ktio(void)
+{
+	struct frame *f;
+	struct list_head *pos;
+	int i;
 
-	part_stat_inc(cpu, part, ios[rw]);
-	part_stat_add(cpu, part, ticks[rw], duration);
-	part_stat_add(cpu, part, sectors[rw], n_sect);
-	part_stat_add(cpu, part, io_ticks, duration);
+	for (i = 0; ; ++i) {
+		if (i == MAXIOC)
+			return 1;
+		if (list_empty(&iocq.head))
+			return 0;
+		pos = iocq.head.next;
+		list_del(pos);
+		spin_unlock_irq(&iocq.lock);
+		f = list_entry(pos, struct frame, head);
+		ktiocomplete(f);
+		spin_lock_irq(&iocq.lock);
+	}
+}
 
-	part_stat_unlock();
+static int
+kthread(void *vp)
+{
+	struct ktstate *k;
+	DECLARE_WAITQUEUE(wait, current);
+	int more;
+
+	k = vp;
+	current->flags |= PF_NOFREEZE;
+	set_user_nice(current, -10);
+	complete(&k->rendez);	/* tell spawner we're running */
+	do {
+		spin_lock_irq(k->lock);
+		more = k->fn();
+		if (!more) {
+			add_wait_queue(k->waitq, &wait);
+			__set_current_state(TASK_INTERRUPTIBLE);
+		}
+		spin_unlock_irq(k->lock);
+		if (!more) {
+			schedule();
+			remove_wait_queue(k->waitq, &wait);
+		} else
+			cond_resched();
+	} while (!kthread_should_stop());
+	complete(&k->rendez);	/* tell spawner we're stopping */
+	return 0;
 }
 
 void
+aoe_ktstop(struct ktstate *k)
+{
+	kthread_stop(k->task);
+	wait_for_completion(&k->rendez);
+}
+
+int
+aoe_ktstart(struct ktstate *k)
+{
+	struct task_struct *task;
+
+	init_completion(&k->rendez);
+	task = kthread_run(kthread, k, k->name);
+	if (task == NULL || IS_ERR(task))
+		return -ENOMEM;
+	k->task = task;
+	wait_for_completion(&k->rendez); /* allow kthread to start */
+	init_completion(&k->rendez);	/* for waiting for exit later */
+	return 0;
+}
+
+/* pass it off to kthreads for processing */
+static void
+ktcomplete(struct frame *f, struct sk_buff *skb)
+{
+	ulong flags;
+
+	f->r_skb = skb;
+	spin_lock_irqsave(&iocq.lock, flags);
+	list_add_tail(&f->head, &iocq.head);
+	spin_unlock_irqrestore(&iocq.lock, flags);
+	wake_up(&ktiowq);
+}
+
+struct sk_buff *
 aoecmd_ata_rsp(struct sk_buff *skb)
 {
-	struct sk_buff_head queue;
 	struct aoedev *d;
-	struct aoe_hdr *hin, *hout;
-	struct aoe_atahdr *ahin, *ahout;
+	struct aoe_hdr *h;
 	struct frame *f;
-	struct buf *buf;
 	struct aoetgt *t;
-	struct aoeif *ifp;
-	register long n;
+	u32 n;
 	ulong flags;
 	char ebuf[128];
 	u16 aoemajor;
 
-	hin = (struct aoe_hdr *) skb_mac_header(skb);
-	aoemajor = get_unaligned_be16(&hin->major);
-	d = aoedev_by_aoeaddr(aoemajor, hin->minor);
+	h = (struct aoe_hdr *) skb->data;
+	aoemajor = be16_to_cpu(get_unaligned(&h->major));
+	d = aoedev_by_aoeaddr(aoemajor, h->minor, 0);
 	if (d == NULL) {
 		snprintf(ebuf, sizeof ebuf, "aoecmd_ata_rsp: ata response "
 			"for unknown device %d.%d\n",
-			 aoemajor, hin->minor);
+			aoemajor, h->minor);
 		aoechr_error(ebuf);
-		return;
+		return skb;
 	}
 
 	spin_lock_irqsave(&d->lock, flags);
 
-	n = get_unaligned_be32(&hin->tag);
-	t = gettgt(d, hin->src);
-	if (t == NULL) {
-		printk(KERN_INFO "aoe: can't find target e%ld.%d:%pm\n",
-			d->aoemajor, d->aoeminor, hin->src);
-		spin_unlock_irqrestore(&d->lock, flags);
-		return;
-	}
-	f = getframe(t, n);
+	n = be32_to_cpu(get_unaligned(&h->tag));
+	f = getframe(d, n);
 	if (f == NULL) {
 		calc_rttavg(d, -tsince(n));
 		spin_unlock_irqrestore(&d->lock, flags);
+		aoedev_put(d);
 		snprintf(ebuf, sizeof ebuf,
 			"%15s e%d.%d    tag=%08x@%08lx\n",
 			"unexpected rsp",
-			get_unaligned_be16(&hin->major),
-			hin->minor,
-			get_unaligned_be32(&hin->tag),
+			get_unaligned_be16(&h->major),
+			h->minor,
+			get_unaligned_be32(&h->tag),
 			jiffies);
 		aoechr_error(ebuf);
-		return;
+		return skb;
 	}
-
+	t = f->t;
 	calc_rttavg(d, tsince(f->tag));
-
-	ahin = (struct aoe_atahdr *) (hin+1);
-	hout = (struct aoe_hdr *) skb_mac_header(f->skb);
-	ahout = (struct aoe_atahdr *) (hout+1);
-	buf = f->buf;
-
-	if (ahin->cmdstat & 0xa9) {	/* these bits cleared on success */
-		printk(KERN_ERR
-			"aoe: ata error cmd=%2.2Xh stat=%2.2Xh from e%ld.%d\n",
-			ahout->cmdstat, ahin->cmdstat,
-			d->aoemajor, d->aoeminor);
-		if (buf)
-			buf->flags |= BUFFL_FAIL;
-	} else {
-		if (d->htgt && t == *d->htgt) /* I'll help myself, thank you. */
-			d->htgt = NULL;
-		n = ahout->scnt << 9;
-		switch (ahout->cmdstat) {
-		case ATA_CMD_PIO_READ:
-		case ATA_CMD_PIO_READ_EXT:
-			if (skb->len - sizeof *hin - sizeof *ahin < n) {
-				printk(KERN_ERR
-					"aoe: %s.  skb->len=%d need=%ld\n",
-					"runt data size in read", skb->len, n);
-				/* fail frame f?  just returning will rexmit. */
-				spin_unlock_irqrestore(&d->lock, flags);
-				return;
-			}
-			memcpy(f->bufaddr, ahin+1, n);
-		case ATA_CMD_PIO_WRITE:
-		case ATA_CMD_PIO_WRITE_EXT:
-			ifp = getif(t, skb->dev);
-			if (ifp) {
-				ifp->lost = 0;
-				if (n > DEFAULTBCNT)
-					ifp->lostjumbo = 0;
-			}
-			if (f->bcnt -= n) {
-				f->lba += n >> 9;
-				f->bufaddr += n;
-				resend(d, t, f);
-				goto xmit;
-			}
-			break;
-		case ATA_CMD_ID_ATA:
-			if (skb->len - sizeof *hin - sizeof *ahin < 512) {
-				printk(KERN_INFO
-					"aoe: runt data size in ataid.  skb->len=%d\n",
-					skb->len);
-				spin_unlock_irqrestore(&d->lock, flags);
-				return;
-			}
-			ataid_complete(d, t, (char *) (ahin+1));
-			break;
-		default:
-			printk(KERN_INFO
-				"aoe: unrecognized ata command %2.2Xh for %d.%d\n",
-				ahout->cmdstat,
-				get_unaligned_be16(&hin->major),
-				hin->minor);
-		}
-	}
-
-	if (buf && --buf->nframesout == 0 && buf->resid == 0) {
-		diskstats(d->gd, buf->bio, jiffies - buf->stime, buf->sector);
-		if (buf->flags & BUFFL_FAIL)
-			bio_endio(buf->bio, -EIO);
-		else {
-			bio_flush_dcache_pages(buf->bio);
-			bio_endio(buf->bio, 0);
-		}
-		mempool_free(buf, d->bufpool);
-	}
-
-	f->buf = NULL;
-	f->tag = FREETAG;
 	t->nout--;
-
 	aoecmd_work(d);
-xmit:
-	__skb_queue_head_init(&queue);
-	skb_queue_splice_init(&d->sendq, &queue);
 
 	spin_unlock_irqrestore(&d->lock, flags);
-	aoenet_xmit(&queue);
+
+	ktcomplete(f, skb);
+
+	/*
+	 * Note here that we do not perform an aoedev_put, as we are
+	 * leaving this reference for the ktio to release.
+	 */
+	return NULL;
 }
 
 void
@@ -896,7 +1211,7 @@ aoecmd_ata_id(struct aoedev *d)
 	struct sk_buff *skb;
 	struct aoetgt *t;
 
-	f = freeframe(d);
+	f = newframe(d);
 	if (f == NULL)
 		return NULL;
 
@@ -909,6 +1224,7 @@ aoecmd_ata_id(struct aoedev *d)
 	skb_put(skb, sizeof *h + sizeof *ah);
 	memset(h, 0, skb->len);
 	f->tag = aoehdr_atainit(d, t, h);
+	fhash(f);
 	t->nout++;
 	f->waited = 0;
 
@@ -929,7 +1245,6 @@ static struct aoetgt *
 addtgt(struct aoedev *d, char *addr, ulong nframes)
 {
 	struct aoetgt *t, **tt, **te;
-	struct frame *f, *e;
 
 	tt = d->targets;
 	te = tt + NTARGETS;
@@ -941,26 +1256,73 @@ addtgt(struct aoedev *d, char *addr, ulong nframes)
 			"aoe: device addtgt failure; too many targets\n");
 		return NULL;
 	}
-	t = kcalloc(1, sizeof *t, GFP_ATOMIC);
-	f = kcalloc(nframes, sizeof *f, GFP_ATOMIC);
-	if (!t || !f) {
-		kfree(f);
-		kfree(t);
+	t = kzalloc(sizeof(*t), GFP_ATOMIC);
+	if (!t) {
 		printk(KERN_INFO "aoe: cannot allocate memory to add target\n");
 		return NULL;
 	}
 
+	d->ntargets++;
 	t->nframes = nframes;
-	t->frames = f;
-	e = f + nframes;
-	for (; f < e; f++)
-		f->tag = FREETAG;
+	t->d = d;
 	memcpy(t->addr, addr, sizeof t->addr);
 	t->ifp = t->ifs;
 	t->maxout = t->nframes;
+	INIT_LIST_HEAD(&t->ffree);
 	return *tt = t;
 }
 
+static void
+setdbcnt(struct aoedev *d)
+{
+	struct aoetgt **t, **e;
+	int bcnt = 0;
+
+	t = d->targets;
+	e = t + NTARGETS;
+	for (; t < e && *t; t++)
+		if (bcnt == 0 || bcnt > (*t)->minbcnt)
+			bcnt = (*t)->minbcnt;
+	if (bcnt != d->maxbcnt) {
+		d->maxbcnt = bcnt;
+		pr_info("aoe: e%ld.%d: setting %d byte data frames\n",
+			d->aoemajor, d->aoeminor, bcnt);
+	}
+}
+
+static void
+setifbcnt(struct aoetgt *t, struct net_device *nd, int bcnt)
+{
+	struct aoedev *d;
+	struct aoeif *p, *e;
+	int minbcnt;
+
+	d = t->d;
+	minbcnt = bcnt;
+	p = t->ifs;
+	e = p + NAOEIFS;
+	for (; p < e; p++) {
+		if (p->nd == NULL)
+			break;		/* end of the valid interfaces */
+		if (p->nd == nd) {
+			p->bcnt = bcnt;	/* we're updating */
+			nd = NULL;
+		} else if (minbcnt > p->bcnt)
+			minbcnt = p->bcnt; /* find the min interface */
+	}
+	if (nd) {
+		if (p == e) {
+			pr_err("aoe: device setifbcnt failure; too many interfaces.\n");
+			return;
+		}
+		dev_hold(nd);
+		p->nd = nd;
+		p->bcnt = bcnt;
+	}
+	t->minbcnt = minbcnt;
+	setdbcnt(d);
+}
+
 void
 aoecmd_cfg_rsp(struct sk_buff *skb)
 {
@@ -968,11 +1330,12 @@ aoecmd_cfg_rsp(struct sk_buff *skb)
 	struct aoe_hdr *h;
 	struct aoe_cfghdr *ch;
 	struct aoetgt *t;
-	struct aoeif *ifp;
-	ulong flags, sysminor, aoemajor;
+	ulong flags, aoemajor;
 	struct sk_buff *sl;
+	struct sk_buff_head queue;
 	u16 n;
 
+	sl = NULL;
 	h = (struct aoe_hdr *) skb_mac_header(skb);
 	ch = (struct aoe_cfghdr *) (h+1);
 
@@ -986,10 +1349,13 @@ aoecmd_cfg_rsp(struct sk_buff *skb)
 			"Check shelf dip switches.\n");
 		return;
 	}
-
-	sysminor = SYSMINOR(aoemajor, h->minor);
-	if (sysminor * AOE_PARTITIONS + AOE_PARTITIONS > MINORMASK) {
-		printk(KERN_INFO "aoe: e%ld.%d: minor number too large\n",
+	if (aoemajor == 0xffff) {
+		pr_info("aoe: e%ld.%d: broadcast shelf number invalid\n",
+			aoemajor, (int) h->minor);
+		return;
+	}
+	if (h->minor == 0xff) {
+		pr_info("aoe: e%ld.%d: broadcast slot number invalid\n",
 			aoemajor, (int) h->minor);
 		return;
 	}
@@ -998,9 +1364,9 @@ aoecmd_cfg_rsp(struct sk_buff *skb)
 	if (n > aoe_maxout)	/* keep it reasonable */
 		n = aoe_maxout;
 
-	d = aoedev_by_sysminor_m(sysminor);
+	d = aoedev_by_aoeaddr(aoemajor, h->minor, 1);
 	if (d == NULL) {
-		printk(KERN_INFO "aoe: device sysminor_m failure\n");
+		pr_info("aoe: device allocation failure\n");
 		return;
 	}
 
@@ -1009,52 +1375,26 @@ aoecmd_cfg_rsp(struct sk_buff *skb)
 	t = gettgt(d, h->src);
 	if (!t) {
 		t = addtgt(d, h->src, n);
-		if (!t) {
-			spin_unlock_irqrestore(&d->lock, flags);
-			return;
-		}
-	}
-	ifp = getif(t, skb->dev);
-	if (!ifp) {
-		ifp = addif(t, skb->dev);
-		if (!ifp) {
-			printk(KERN_INFO
-				"aoe: device addif failure; "
-				"too many interfaces?\n");
-			spin_unlock_irqrestore(&d->lock, flags);
-			return;
-		}
-	}
-	if (ifp->maxbcnt) {
-		n = ifp->nd->mtu;
-		n -= sizeof (struct aoe_hdr) + sizeof (struct aoe_atahdr);
-		n /= 512;
-		if (n > ch->scnt)
-			n = ch->scnt;
-		n = n ? n * 512 : DEFAULTBCNT;
-		if (n != ifp->maxbcnt) {
-			printk(KERN_INFO
-				"aoe: e%ld.%d: setting %d%s%s:%pm\n",
-				d->aoemajor, d->aoeminor, n,
-				" byte data frames on ", ifp->nd->name,
-				t->addr);
-			ifp->maxbcnt = n;
-		}
+		if (!t)
+			goto bail;
 	}
+	n = skb->dev->mtu;
+	n -= sizeof(struct aoe_hdr) + sizeof(struct aoe_atahdr);
+	n /= 512;
+	if (n > ch->scnt)
+		n = ch->scnt;
+	n = n ? n * 512 : DEFAULTBCNT;
+	setifbcnt(t, skb->dev, n);
 
 	/* don't change users' perspective */
-	if (d->nopen) {
-		spin_unlock_irqrestore(&d->lock, flags);
-		return;
+	if (d->nopen == 0) {
+		d->fw_ver = be16_to_cpu(ch->fwver);
+		sl = aoecmd_ata_id(d);
 	}
-	d->fw_ver = be16_to_cpu(ch->fwver);
-
-	sl = aoecmd_ata_id(d);
-
+bail:
 	spin_unlock_irqrestore(&d->lock, flags);
-
+	aoedev_put(d);
 	if (sl) {
-		struct sk_buff_head queue;
 		__skb_queue_head_init(&queue);
 		__skb_queue_tail(&queue, sl);
 		aoenet_xmit(&queue);
@@ -1065,20 +1405,74 @@ void
 aoecmd_cleanslate(struct aoedev *d)
 {
 	struct aoetgt **t, **te;
-	struct aoeif *p, *e;
 
 	d->mintimer = MINTIMER;
+	d->maxbcnt = 0;
 
 	t = d->targets;
 	te = t + NTARGETS;
-	for (; t < te && *t; t++) {
+	for (; t < te && *t; t++)
 		(*t)->maxout = (*t)->nframes;
-		p = (*t)->ifs;
-		e = p + NAOEIFS;
-		for (; p < e; p++) {
-			p->lostjumbo = 0;
-			p->lost = 0;
-			p->maxbcnt = DEFAULTBCNT;
+}
+
+void
+aoe_failbuf(struct aoedev *d, struct buf *buf)
+{
+	if (buf == NULL)
+		return;
+	buf->resid = 0;
+	clear_bit(BIO_UPTODATE, &buf->bio->bi_flags);
+	if (buf->nframesout == 0)
+		aoe_end_buf(d, buf);
+}
+
+void
+aoe_flush_iocq(void)
+{
+	struct frame *f;
+	struct aoedev *d;
+	LIST_HEAD(flist);
+	struct list_head *pos;
+	struct sk_buff *skb;
+	ulong flags;
+
+	spin_lock_irqsave(&iocq.lock, flags);
+	list_splice_init(&iocq.head, &flist);
+	spin_unlock_irqrestore(&iocq.lock, flags);
+	while (!list_empty(&flist)) {
+		pos = flist.next;
+		list_del(pos);
+		f = list_entry(pos, struct frame, head);
+		d = f->t->d;
+		skb = f->r_skb;
+		spin_lock_irqsave(&d->lock, flags);
+		if (f->buf) {
+			f->buf->nframesout--;
+			aoe_failbuf(d, f->buf);
 		}
+		aoe_freetframe(f);
+		spin_unlock_irqrestore(&d->lock, flags);
+		dev_kfree_skb(skb);
+		aoedev_put(d);
 	}
 }
+
+int __init
+aoecmd_init(void)
+{
+	INIT_LIST_HEAD(&iocq.head);
+	spin_lock_init(&iocq.lock);
+	init_waitqueue_head(&ktiowq);
+	kts.name = "aoe_ktio";
+	kts.fn = ktio;
+	kts.waitq = &ktiowq;
+	kts.lock = &iocq.lock;
+	return aoe_ktstart(&kts);
+}
+
+void
+aoecmd_exit(void)
+{
+	aoe_ktstop(&kts);
+	aoe_flush_iocq();
+}

drivers/block/aoe/aoedev.c

@@ -1,4 +1,4 @@
-/* Copyright (c) 2007 Coraid, Inc.  See COPYING for GPL terms. */
+/* Copyright (c) 2012 Coraid, Inc.  See COPYING for GPL terms. */
 /*
  * aoedev.c
  * AoE device utility functions; maintains device list.
@@ -9,6 +9,9 @@
 #include <linux/netdevice.h>
 #include <linux/delay.h>
 #include <linux/slab.h>
+#include <linux/bitmap.h>
+#include <linux/kdev_t.h>
+#include <linux/moduleparam.h>
 #include "aoe.h"
 
 static void dummy_timer(ulong);
@@ -16,23 +19,121 @@ static void aoedev_freedev(struct aoedev *);
 static void freetgt(struct aoedev *d, struct aoetgt *t);
 static void skbpoolfree(struct aoedev *d);
 
+static int aoe_dyndevs = 1;
+module_param(aoe_dyndevs, int, 0644);
+MODULE_PARM_DESC(aoe_dyndevs, "Use dynamic minor numbers for devices.");
+
 static struct aoedev *devlist;
 static DEFINE_SPINLOCK(devlist_lock);
 
-struct aoedev *
-aoedev_by_aoeaddr(int maj, int min)
+/* Because some systems will have one, many, or no
+ *   - partitions,
+ *   - slots per shelf,
+ *   - or shelves,
+ * we need some flexibility in the way the minor numbers
+ * are allocated.  So they are dynamic.
+ */
+#define N_DEVS ((1U<<MINORBITS)/AOE_PARTITIONS)
+
+static DEFINE_SPINLOCK(used_minors_lock);
+static DECLARE_BITMAP(used_minors, N_DEVS);
+
+static int
+minor_get_dyn(ulong *sysminor)
 {
-	struct aoedev *d;
 	ulong flags;
+	ulong n;
+	int error = 0;
+
+	spin_lock_irqsave(&used_minors_lock, flags);
+	n = find_first_zero_bit(used_minors, N_DEVS);
+	if (n < N_DEVS)
+		set_bit(n, used_minors);
+	else
+		error = -1;
+	spin_unlock_irqrestore(&used_minors_lock, flags);
+
+	*sysminor = n * AOE_PARTITIONS;
+	return error;
+}
 
-	spin_lock_irqsave(&devlist_lock, flags);
+static int
+minor_get_static(ulong *sysminor, ulong aoemaj, int aoemin)
+{
+	ulong flags;
+	ulong n;
+	int error = 0;
+	enum {
+		/* for backwards compatibility when !aoe_dyndevs,
+		 * a static number of supported slots per shelf */
+		NPERSHELF = 16,
+	};
+
+	n = aoemaj * NPERSHELF + aoemin;
+	if (aoemin >= NPERSHELF || n >= N_DEVS) {
+		pr_err("aoe: %s with e%ld.%d\n",
+			"cannot use static minor device numbers",
+			aoemaj, aoemin);
+		error = -1;
+	} else {
+		spin_lock_irqsave(&used_minors_lock, flags);
+		if (test_bit(n, used_minors)) {
+			pr_err("aoe: %s %lu\n",
+				"existing device already has static minor number",
+				n);
+			error = -1;
+		} else
+			set_bit(n, used_minors);
+		spin_unlock_irqrestore(&used_minors_lock, flags);
+	}
 
-	for (d=devlist; d; d=d->next)
-		if (d->aoemajor == maj && d->aoeminor == min)
-			break;
+	*sysminor = n;
+	return error;
+}
+
+static int
+minor_get(ulong *sysminor, ulong aoemaj, int aoemin)
+{
+	if (aoe_dyndevs)
+		return minor_get_dyn(sysminor);
+	else
+		return minor_get_static(sysminor, aoemaj, aoemin);
+}
+
+static void
+minor_free(ulong minor)
+{
+	ulong flags;
+
+	minor /= AOE_PARTITIONS;
+	BUG_ON(minor >= N_DEVS);
+
+	spin_lock_irqsave(&used_minors_lock, flags);
+	BUG_ON(!test_bit(minor, used_minors));
+	clear_bit(minor, used_minors);
+	spin_unlock_irqrestore(&used_minors_lock, flags);
+}
+
+/*
+ * Users who grab a pointer to the device with aoedev_by_aoeaddr
+ * automatically get a reference count and must be responsible
+ * for performing a aoedev_put.  With the addition of async
+ * kthread processing I'm no longer confident that we can
+ * guarantee consistency in the face of device flushes.
+ *
+ * For the time being, we only bother to add extra references for
+ * frames sitting on the iocq.  When the kthreads finish processing
+ * these frames, they will aoedev_put the device.
+ */
+
+void
+aoedev_put(struct aoedev *d)
+{
+	ulong flags;
 
+	spin_lock_irqsave(&devlist_lock, flags);
+	d->ref--;
 	spin_unlock_irqrestore(&devlist_lock, flags);
-	return d;
 }
 
 static void
@@ -47,54 +148,74 @@ dummy_timer(ulong vp)
 	add_timer(&d->timer);
 }
 
+static void
+aoe_failip(struct aoedev *d)
+{
+	struct request *rq;
+	struct bio *bio;
+	unsigned long n;
+
+	aoe_failbuf(d, d->ip.buf);
+
+	rq = d->ip.rq;
+	if (rq == NULL)
+		return;
+	while ((bio = d->ip.nxbio)) {
+		clear_bit(BIO_UPTODATE, &bio->bi_flags);
+		d->ip.nxbio = bio->bi_next;
+		n = (unsigned long) rq->special;
+		rq->special = (void *) --n;
+	}
+	if ((unsigned long) rq->special == 0)
+		aoe_end_request(d, rq, 0);
+}
+
 void
 aoedev_downdev(struct aoedev *d)
 {
-	struct aoetgt **t, **te;
-	struct frame *f, *e;
-	struct buf *buf;
-	struct bio *bio;
+	struct aoetgt *t, **tt, **te;
+	struct frame *f;
+	struct list_head *head, *pos, *nx;
+	struct request *rq;
+	int i;
 
-	t = d->targets;
-	te = t + NTARGETS;
-	for (; t < te && *t; t++) {
-		f = (*t)->frames;
-		e = f + (*t)->nframes;
-		for (; f < e; f->tag = FREETAG, f->buf = NULL, f++) {
-			if (f->tag == FREETAG || f->buf == NULL)
-				continue;
-			buf = f->buf;
-			bio = buf->bio;
-			if (--buf->nframesout == 0
-			&& buf != d->inprocess) {
-				mempool_free(buf, d->bufpool);
-				bio_endio(bio, -EIO);
+	d->flags &= ~DEVFL_UP;
+
+	/* clean out active buffers */
+	for (i = 0; i < NFACTIVE; i++) {
+		head = &d->factive[i];
+		list_for_each_safe(pos, nx, head) {
+			f = list_entry(pos, struct frame, head);
+			list_del(pos);
+			if (f->buf) {
+				f->buf->nframesout--;
+				aoe_failbuf(d, f->buf);
 			}
+			aoe_freetframe(f);
 		}
-		(*t)->maxout = (*t)->nframes;
-		(*t)->nout = 0;
 	}
-	buf = d->inprocess;
-	if (buf) {
-		bio = buf->bio;
-		mempool_free(buf, d->bufpool);
-		bio_endio(bio, -EIO);
+	/* reset window dressings */
+	tt = d->targets;
+	te = tt + NTARGETS;
+	for (; tt < te && (t = *tt); tt++) {
+		t->maxout = t->nframes;
+		t->nout = 0;
 	}
-	d->inprocess = NULL;
+
+	/* clean out the in-process request (if any) */
+	aoe_failip(d);
 	d->htgt = NULL;
 
-	while (!list_empty(&d->bufq)) {
-		buf = container_of(d->bufq.next, struct buf, bufs);
-		list_del(d->bufq.next);
-		bio = buf->bio;
-		mempool_free(buf, d->bufpool);
-		bio_endio(bio, -EIO);
+	/* fast fail all pending I/O */
+	if (d->blkq) {
+		while ((rq = blk_peek_request(d->blkq))) {
+			blk_start_request(rq);
+			aoe_end_request(d, rq, 1);
+		}
 	}
 
 	if (d->gd)
 		set_capacity(d->gd, 0);
-
-	d->flags &= ~DEVFL_UP;
 }
 
 static void
@@ -107,6 +228,7 @@ aoedev_freedev(struct aoedev *d)
 		aoedisk_rm_sysfs(d);
 		del_gendisk(d->gd);
 		put_disk(d->gd);
+		blk_cleanup_queue(d->blkq);
 	}
 	t = d->targets;
 	e = t + NTARGETS;
@@ -115,7 +237,7 @@ aoedev_freedev(struct aoedev *d)
 	if (d->bufpool)
 		mempool_destroy(d->bufpool);
 	skbpoolfree(d);
-	blk_cleanup_queue(d->blkq);
+	minor_free(d->sysminor);
 	kfree(d);
 }
 
@@ -142,7 +264,8 @@ aoedev_flush(const char __user *str, size_t cnt)
 		spin_lock(&d->lock);
 		if ((!all && (d->flags & DEVFL_UP))
 		|| (d->flags & (DEVFL_GDALLOC|DEVFL_NEWSIZE))
-		|| d->nopen) {
+		|| d->nopen
+		|| d->ref) {
 			spin_unlock(&d->lock);
 			dd = &d->next;
 			continue;
@@ -163,12 +286,15 @@ aoedev_flush(const char __user *str, size_t cnt)
 	return 0;
 }
 
-/* I'm not really sure that this is a realistic problem, but if the
-network driver goes gonzo let's just leak memory after complaining. */
+/* This has been confirmed to occur once with Tms=3*1000 due to the
+ * driver changing link and not processing its transmit ring.  The
+ * problem is hard enough to solve by returning an error that I'm
+ * still punting on "solving" this.
+ */
 static void
 skbfree(struct sk_buff *skb)
 {
-	enum { Sms = 100, Tms = 3*1000};
+	enum { Sms = 250, Tms = 30 * 1000};
 	int i = Tms / Sms;
 
 	if (skb == NULL)
@@ -182,6 +308,7 @@ skbfree(struct sk_buff *skb)
 			"cannot free skb -- memory leaked.");
 		return;
 	}
+	skb->truesize -= skb->data_len;
 	skb_shinfo(skb)->nr_frags = skb->data_len = 0;
 	skb_trim(skb, 0);
 	dev_kfree_skb(skb);
@@ -198,26 +325,29 @@ skbpoolfree(struct aoedev *d)
 	__skb_queue_head_init(&d->skbpool);
 }
 
-/* find it or malloc it */
+/* find it or allocate it */
 struct aoedev *
-aoedev_by_sysminor_m(ulong sysminor)
+aoedev_by_aoeaddr(ulong maj, int min, int do_alloc)
 {
 	struct aoedev *d;
+	int i;
 	ulong flags;
+	ulong sysminor;
 
 	spin_lock_irqsave(&devlist_lock, flags);
 
 	for (d=devlist; d; d=d->next)
-		if (d->sysminor == sysminor)
+		if (d->aoemajor == maj && d->aoeminor == min) {
+			d->ref++;
 			break;
-	if (d)
+		}
+	if (d || !do_alloc || minor_get(&sysminor, maj, min) < 0)
 		goto out;
 	d = kcalloc(1, sizeof *d, GFP_ATOMIC);
 	if (!d)
 		goto out;
 	INIT_WORK(&d->work, aoecmd_sleepwork);
 	spin_lock_init(&d->lock);
-	skb_queue_head_init(&d->sendq);
 	skb_queue_head_init(&d->skbpool);
 	init_timer(&d->timer);
 	d->timer.data = (ulong) d;
@@ -226,10 +356,12 @@ aoedev_by_sysminor_m(ulong sysminor)
 	add_timer(&d->timer);
 	d->bufpool = NULL;	/* defer to aoeblk_gdalloc */
 	d->tgt = d->targets;
-	INIT_LIST_HEAD(&d->bufq);
+	d->ref = 1;
+	for (i = 0; i < NFACTIVE; i++)
+		INIT_LIST_HEAD(&d->factive[i]);
 	d->sysminor = sysminor;
-	d->aoemajor = AOEMAJOR(sysminor);
-	d->aoeminor = AOEMINOR(sysminor);
+	d->aoemajor = maj;
+	d->aoeminor = min;
 	d->mintimer = MINTIMER;
 	d->next = devlist;
 	devlist = d;
@@ -241,13 +373,23 @@ aoedev_by_sysminor_m(ulong sysminor)
 static void
 freetgt(struct aoedev *d, struct aoetgt *t)
 {
-	struct frame *f, *e;
+	struct frame *f;
+	struct list_head *pos, *nx, *head;
+	struct aoeif *ifp;
 
-	f = t->frames;
-	e = f + t->nframes;
-	for (; f < e; f++)
+	for (ifp = t->ifs; ifp < &t->ifs[NAOEIFS]; ++ifp) {
+		if (!ifp->nd)
+			break;
+		dev_put(ifp->nd);
+	}
+
+	head = &t->ffree;
+	list_for_each_safe(pos, nx, head) {
+		list_del(pos);
+		f = list_entry(pos, struct frame, head);
 		skbfree(f->skb);
-	kfree(t->frames);
+		kfree(f);
+	}
 	kfree(t);
 }
 
@@ -257,6 +399,7 @@ aoedev_exit(void)
 	struct aoedev *d;
 	ulong flags;
 
+	aoe_flush_iocq();
 	while ((d = devlist)) {
 		devlist = d->next;
 

drivers/block/aoe/aoemain.c

@@ -1,4 +1,4 @@
-/* Copyright (c) 2007 Coraid, Inc.  See COPYING for GPL terms. */
+/* Copyright (c) 2012 Coraid, Inc.  See COPYING for GPL terms. */
 /*
  * aoemain.c
  * Module initialization routines, discover timer
@@ -61,6 +61,7 @@ aoe_exit(void)
 
 	aoenet_exit();
 	unregister_blkdev(AOE_MAJOR, DEVICE_NAME);
+	aoecmd_exit();
 	aoechr_exit();
 	aoedev_exit();
 	aoeblk_exit();		/* free cache after de-allocating bufs */
@@ -83,17 +84,20 @@ aoe_init(void)
 	ret = aoenet_init();
 	if (ret)
 		goto net_fail;
+	ret = aoecmd_init();
+	if (ret)
+		goto cmd_fail;
 	ret = register_blkdev(AOE_MAJOR, DEVICE_NAME);
 	if (ret < 0) {
 		printk(KERN_ERR "aoe: can't register major\n");
 		goto blkreg_fail;
 	}
-
 	printk(KERN_INFO "aoe: AoE v%s initialised.\n", VERSION);
 	discover_timer(TINIT);
 	return 0;
-
  blkreg_fail:
+	aoecmd_exit();
+ cmd_fail:
 	aoenet_exit();
  net_fail:
 	aoeblk_exit();

drivers/block/aoe/aoenet.c

@@ -1,4 +1,4 @@
-/* Copyright (c) 2007 Coraid, Inc.  See COPYING for GPL terms. */
+/* Copyright (c) 2012 Coraid, Inc.  See COPYING for GPL terms. */
 /*
  * aoenet.c
  * Ethernet portion of AoE driver
@@ -33,6 +33,9 @@ static char aoe_iflist[IFLISTSZ];
 module_param_string(aoe_iflist, aoe_iflist, IFLISTSZ, 0600);
 MODULE_PARM_DESC(aoe_iflist, "aoe_iflist=\"dev1 [dev2 ...]\"");
 
+static wait_queue_head_t txwq;
+static struct ktstate kts;
+
 #ifndef MODULE
 static int __init aoe_iflist_setup(char *str)
 {
@@ -44,6 +47,23 @@ static int __init aoe_iflist_setup(char *str)
 __setup("aoe_iflist=", aoe_iflist_setup);
 #endif
 
+static spinlock_t txlock;
+static struct sk_buff_head skbtxq;
+
+/* enters with txlock held */
+static int
+tx(void)
+{
+	struct sk_buff *skb;
+
+	while ((skb = skb_dequeue(&skbtxq))) {
+		spin_unlock_irq(&txlock);
+		dev_queue_xmit(skb);
+		spin_lock_irq(&txlock);
+	}
+	return 0;
+}
+
 int
 is_aoe_netif(struct net_device *ifp)
 {
@@ -88,10 +108,14 @@ void
 aoenet_xmit(struct sk_buff_head *queue)
 {
 	struct sk_buff *skb, *tmp;
+	ulong flags;
 
 	skb_queue_walk_safe(queue, skb, tmp) {
 		__skb_unlink(skb, queue);
-		dev_queue_xmit(skb);
+		spin_lock_irqsave(&txlock, flags);
+		skb_queue_tail(&skbtxq, skb);
+		spin_unlock_irqrestore(&txlock, flags);
+		wake_up(&txwq);
 	}
 }
 
@@ -102,7 +126,9 @@ static int
 aoenet_rcv(struct sk_buff *skb, struct net_device *ifp, struct packet_type *pt, struct net_device *orig_dev)
 {
 	struct aoe_hdr *h;
+	struct aoe_atahdr *ah;
 	u32 n;
+	int sn;
 
 	if (dev_net(ifp) != &init_net)
 		goto exit;
@@ -110,13 +136,16 @@ aoenet_rcv(struct sk_buff *skb, struct net_device *ifp, struct packet_type *pt,
 	skb = skb_share_check(skb, GFP_ATOMIC);
 	if (skb == NULL)
 		return 0;
-	if (skb_linearize(skb))
-		goto exit;
 	if (!is_aoe_netif(ifp))
 		goto exit;
 	skb_push(skb, ETH_HLEN);	/* (1) */
-
-	h = (struct aoe_hdr *) skb_mac_header(skb);
+	sn = sizeof(*h) + sizeof(*ah);
+	if (skb->len >= sn) {
+		sn -= skb_headlen(skb);
+		if (sn > 0 && !__pskb_pull_tail(skb, sn))
+			goto exit;
+	}
+	h = (struct aoe_hdr *) skb->data;
 	n = get_unaligned_be32(&h->tag);
 	if ((h->verfl & AOEFL_RSP) == 0 || (n & 1<<31))
 		goto exit;
@@ -137,7 +166,8 @@ aoenet_rcv(struct sk_buff *skb, struct net_device *ifp, struct packet_type *pt,
 
 	switch (h->cmd) {
 	case AOECMD_ATA:
-		aoecmd_ata_rsp(skb);
+		/* ata_rsp may keep skb for later processing or give it back */
+		skb = aoecmd_ata_rsp(skb);
 		break;
 	case AOECMD_CFG:
 		aoecmd_cfg_rsp(skb);
@@ -145,8 +175,12 @@ aoenet_rcv(struct sk_buff *skb, struct net_device *ifp, struct packet_type *pt,
 	default:
 		if (h->cmd >= AOECMD_VEND_MIN)
 			break;	/* don't complain about vendor commands */
-		printk(KERN_INFO "aoe: unknown cmd %d\n", h->cmd);
+		pr_info("aoe: unknown AoE command type 0x%02x\n", h->cmd);
+		break;
 	}
+
+	if (!skb)
+		return 0;
 exit:
 	dev_kfree_skb(skb);
 	return 0;
@@ -160,6 +194,15 @@ static struct packet_type aoe_pt __read_mostly = {
 int __init
 aoenet_init(void)
 {
+	skb_queue_head_init(&skbtxq);
+	init_waitqueue_head(&txwq);
+	spin_lock_init(&txlock);
+	kts.lock = &txlock;
+	kts.fn = tx;
+	kts.waitq = &txwq;
+	kts.name = "aoe_tx";
+	if (aoe_ktstart(&kts))
+		return -EAGAIN;
 	dev_add_pack(&aoe_pt);
 	return 0;
 }
@@ -167,6 +210,8 @@ aoenet_init(void)
 void
 aoenet_exit(void)
 {
+	aoe_ktstop(&kts);
+	skb_queue_purge(&skbtxq);
 	dev_remove_pack(&aoe_pt);
 }
 

drivers/block/nbd.c

@@ -78,6 +78,8 @@ static const char *ioctl_cmd_to_ascii(int cmd)
 	case NBD_SET_SOCK: return "set-sock";
 	case NBD_SET_BLKSIZE: return "set-blksize";
 	case NBD_SET_SIZE: return "set-size";
+	case NBD_SET_TIMEOUT: return "set-timeout";
+	case NBD_SET_FLAGS: return "set-flags";
 	case NBD_DO_IT: return "do-it";
 	case NBD_CLEAR_SOCK: return "clear-sock";
 	case NBD_CLEAR_QUE: return "clear-que";
@@ -96,6 +98,7 @@ static const char *nbdcmd_to_ascii(int cmd)
 	case  NBD_CMD_READ: return "read";
 	case NBD_CMD_WRITE: return "write";
 	case  NBD_CMD_DISC: return "disconnect";
+	case  NBD_CMD_TRIM: return "trim/discard";
 	}
 	return "invalid";
 }
@@ -467,8 +470,12 @@ static void nbd_handle_req(struct nbd_device *nbd, struct request *req)
 
 	nbd_cmd(req) = NBD_CMD_READ;
 	if (rq_data_dir(req) == WRITE) {
-		nbd_cmd(req) = NBD_CMD_WRITE;
-		if (nbd->flags & NBD_READ_ONLY) {
+		if ((req->cmd_flags & REQ_DISCARD)) {
+			WARN_ON(!(nbd->flags & NBD_FLAG_SEND_TRIM));
+			nbd_cmd(req) = NBD_CMD_TRIM;
+		} else
+			nbd_cmd(req) = NBD_CMD_WRITE;
+		if (nbd->flags & NBD_FLAG_READ_ONLY) {
 			dev_err(disk_to_dev(nbd->disk),
 				"Write on read-only\n");
 			goto error_out;
@@ -651,6 +658,10 @@ static int __nbd_ioctl(struct block_device *bdev, struct nbd_device *nbd,
 		nbd->xmit_timeout = arg * HZ;
 		return 0;
 
+	case NBD_SET_FLAGS:
+		nbd->flags = arg;
+		return 0;
+
 	case NBD_SET_SIZE_BLOCKS:
 		nbd->bytesize = ((u64) arg) * nbd->blksize;
 		bdev->bd_inode->i_size = nbd->bytesize;
@@ -670,6 +681,10 @@ static int __nbd_ioctl(struct block_device *bdev, struct nbd_device *nbd,
 
 		mutex_unlock(&nbd->tx_lock);
 
+		if (nbd->flags & NBD_FLAG_SEND_TRIM)
+			queue_flag_set_unlocked(QUEUE_FLAG_DISCARD,
+				nbd->disk->queue);
+
 		thread = kthread_create(nbd_thread, nbd, nbd->disk->disk_name);
 		if (IS_ERR(thread)) {
 			mutex_lock(&nbd->tx_lock);
@@ -687,6 +702,7 @@ static int __nbd_ioctl(struct block_device *bdev, struct nbd_device *nbd,
 		nbd->file = NULL;
 		nbd_clear_que(nbd);
 		dev_warn(disk_to_dev(nbd->disk), "queue cleared\n");
+		queue_flag_clear_unlocked(QUEUE_FLAG_DISCARD, nbd->disk->queue);
 		if (file)
 			fput(file);
 		nbd->bytesize = 0;
@@ -805,6 +821,9 @@ static int __init nbd_init(void)
 		 * Tell the block layer that we are not a rotational device
 		 */
 		queue_flag_set_unlocked(QUEUE_FLAG_NONROT, disk->queue);
+		disk->queue->limits.discard_granularity = 512;
+		disk->queue->limits.max_discard_sectors = UINT_MAX;
+		disk->queue->limits.discard_zeroes_data = 0;
 	}
 
 	if (register_blkdev(NBD_MAJOR, "nbd")) {

drivers/char/mbcs.c

@@ -799,7 +799,7 @@ static int mbcs_remove(struct cx_dev *dev)
 	return 0;
 }
 
-static const struct cx_device_id __devinitdata mbcs_id_table[] = {
+static const struct cx_device_id __devinitconst mbcs_id_table[] = {
 	{
 	 .part_num = MBCS_PART_NUM,
 	 .mfg_num = MBCS_MFG_NUM,

drivers/dma/dmaengine.c

@@ -582,7 +582,7 @@ void dmaengine_get(void)
 				list_del_rcu(&device->global_node);
 				break;
 			} else if (err)
-				pr_err("%s: failed to get %s: (%d)\n",
+				pr_debug("%s: failed to get %s: (%d)\n",
 				       __func__, dma_chan_name(chan), err);
 		}
 	}

drivers/i2c/i2c-core.c

@@ -982,7 +982,7 @@ int i2c_add_numbered_adapter(struct i2c_adapter *adap)
 
 	if (adap->nr == -1) /* -1 means dynamically assign bus id */
 		return i2c_add_adapter(adap);
-	if (adap->nr & ~MAX_ID_MASK)
+	if (adap->nr & ~MAX_IDR_MASK)
 		return -EINVAL;
 
 retry:

drivers/ide/aec62xx.c

@@ -181,7 +181,7 @@ static const struct ide_port_ops atp86x_port_ops = {
 	.cable_detect		= atp86x_cable_detect,
 };
 
-static const struct ide_port_info aec62xx_chipsets[] __devinitdata = {
+static const struct ide_port_info aec62xx_chipsets[] __devinitconst = {
 	{	/* 0: AEC6210 */
 		.name		= DRV_NAME,
 		.init_chipset	= init_chipset_aec62xx,

drivers/ide/ali14xx.c

@@ -52,13 +52,13 @@
 
 /* port addresses for auto-detection */
 #define ALI_NUM_PORTS 4
-static const int ports[ALI_NUM_PORTS] __initdata =
+static const int ports[ALI_NUM_PORTS] __initconst =
 	{ 0x074, 0x0f4, 0x034, 0x0e4 };
 
 /* register initialization data */
 typedef struct { u8 reg, data; } RegInitializer;
 
-static const RegInitializer initData[] __initdata = {
+static const RegInitializer initData[] __initconst = {
 	{0x01, 0x0f}, {0x02, 0x00}, {0x03, 0x00}, {0x04, 0x00},
 	{0x05, 0x00}, {0x06, 0x00}, {0x07, 0x2b}, {0x0a, 0x0f},
 	{0x25, 0x00}, {0x26, 0x00}, {0x27, 0x00}, {0x28, 0x00},

drivers/ide/alim15x3.c

@@ -512,7 +512,7 @@ static const struct ide_dma_ops ali_dma_ops = {
 	.dma_sff_read_status	= ide_dma_sff_read_status,
 };
 
-static const struct ide_port_info ali15x3_chipset __devinitdata = {
+static const struct ide_port_info ali15x3_chipset __devinitconst = {
 	.name		= DRV_NAME,
 	.init_chipset	= init_chipset_ali15x3,
 	.init_hwif	= init_hwif_ali15x3,

drivers/ide/amd74xx.c

@@ -223,7 +223,7 @@ static const struct ide_port_ops amd_port_ops = {
 		.udma_mask	= udma,					\
 	}
 
-static const struct ide_port_info amd74xx_chipsets[] __devinitdata = {
+static const struct ide_port_info amd74xx_chipsets[] __devinitconst = {
 	/* 0: AMD7401 */	DECLARE_AMD_DEV(0x00, ATA_UDMA2),
 	/* 1: AMD7409 */	DECLARE_AMD_DEV(ATA_SWDMA2, ATA_UDMA4),
 	/* 2: AMD7411/7441 */	DECLARE_AMD_DEV(ATA_SWDMA2, ATA_UDMA5),

drivers/ide/atiixp.c

@@ -139,7 +139,7 @@ static const struct ide_port_ops atiixp_port_ops = {
 	.cable_detect		= atiixp_cable_detect,
 };
 
-static const struct ide_port_info atiixp_pci_info[] __devinitdata = {
+static const struct ide_port_info atiixp_pci_info[] __devinitconst = {
 	{	/* 0: IXP200/300/400/700 */
 		.name		= DRV_NAME,
 		.enablebits	= {{0x48,0x01,0x00}, {0x48,0x08,0x00}},