summaryrefslogtreecommitdiffstats
path: root/include
Commit message (Collapse)AuthorAgeFilesLines
...
| * | [ARM] 3239/1: Add ARM optimised swab32Andre McCurdy2006-01-071-1/+14
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Patch from Andre McCurdy Replaces generic swab32 routine with a more ARM friendly version. Reduces kernel text size by approx 1200 bytes when compiled with 3.4.4 and approx 2400 bytes with 4.0.2 Probably some performance benefit as well. Signed-off-by: Andre McCurdy <armccurdy@yahoo.co.uk> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
| * | [ARM] 3206/1: Modifications to the bus arbiter controller for the Intel PXA27xJared Hulbert2006-01-051-0/+12
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Patch from Jared Hulbert The following patch changes the bus arbiter controller settings for the Intel PXA27x Application Processor Family. Up to 5% better video performance. It parks the bus on the core while not in use and sets the arbitration for other bus items. The patch only applies changes to the Intel Mainstone development platform. This patch is not compatible with preproduction Intel PXA27x silicon. This patch is based on the Intel Linux Preview Kit released to the public on 25 Feb. 2005 found at ftp://ftp.arm.linux.org.uk/pub/linux/arm/people/xscale/mainstone/02-25-2005/. Signed-off-by: Justin A Treon <justin_treon@yahoo.com> Signed-off-by: Jared Hulbert <jaredeh@gmail.com> Signed-off-by: Nicolas Pitre <nico@cam.org> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
| * | [ARM] 3226/1: IXP4xx runtime expansion bus window size configurationDeepak Saxena2006-01-056-27/+12
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Patch from Deepak Saxena The expansion bus on the IXP46x NPU can be configured for either 32MiB or 16MiB windows and changing the configuration causes the base address for each chip select for each region to change. Because of this, we cannot hardcode the physical base as we currently do. This patch checks the expansion bus configuration registers at runtime to determine the appropriate window size. Note that this requires that the bootloader already configured the device sizes appropriately, but I feel that is valid assumption to make as the bootloader must configure and access the flash window, the output display (LCD, LEDs, etc) window, and other expansion bus devices. Signed-off-by: Deepak Saxena <dsaxena@plexity.net> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
| * | [ARM] 3228/1: SharpSL: Move PM code to arch/arm/commonRichard Purdie2006-01-051-0/+94
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Patch from Richard Purdie This patch moves a large chunk of the sharpsl_pm driver to arch/arm/common so that it can be reused on other devices such as the SL-5500 (collie). It also abstracts some functions from the core into the machine and platform specific parts of the driver to aid reuse. Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
| * | [ARM] 3225/1: add symbolic names for enp2611 gpio interruptsLennert Buytenhek2006-01-041-0/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Patch from Lennert Buytenhek Add symbolic names for the five ixp2400 GPIO lines on the enp2611 that are used as interrupts. Signed-off-by: Lennert Buytenhek <buytenh@wantstofly.org> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
| * | [ARM] 3224/1: add masked thread interrupt status registers for ixp2000Lennert Buytenhek2006-01-041-0/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Patch from Lennert Buytenhek In its interrupt handler, the (NAPI) ixp2000 netdev driver needs to use the masked thread interrupt status register (instead of the raw one) to prevent scheduling polling when polling is already running when a TXdone interrupt comes in. The definitions for the masked status registers were not in yet, so this patch adds them. Signed-off-by: Lennert Buytenhek <buytenh@wantstofly.org> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
| * | [ARM] 3223/1: remove ixdp2x01 cs89x0 hackLennert Buytenhek2006-01-041-98/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Patch from Lennert Buytenhek Remove the ixdp2x01 cs89x0 hack from ixp2000's io implementation. Since the cs89x0 driver has been made properly aware of the odd way the cs89x0 is hooked up on the ixdp2x01, we don't need this hack anymore. Signed-off-by: Lennert Buytenhek <buytenh@wantstofly.org> Signed-off-by: Deepak Saxena <dsaxena@plexity.net> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
| * | [ARM] 3218/1: PAGE_SHIFT undeclared in arch-ixp4xx/memory.h (adjust_zones ↵Rod Whitby2006-01-041-23/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | moved out of line) Patch from Rod Whitby PAGE_SHIFT is undeclared in include/asm-arm/arch-ixp4xx/memory.h, identified by the following kernel compilation error: CC [M] sound/core/memory.o In file included from include/asm/memory.h:27, from include/asm/io.h:28, from sound/core/memory.c:24: include/asm/arch/memory.h: In function `__arch_adjust_zones': include/asm/arch/memory.h:28: error: `PAGE_SHIFT' undeclared (first use in this function) This patch replaces my previous attempt at fixing this problem (Patch 3214/1) and is based on the following feedback: Russell King wrote: > The error you see came up on SA1100. The best solution was to move > the __arch_adjust_zones() function out of line. I suggest ixp4xx > does the same. I have moved the function out of line into arch/arm/mach-ixp4xx/common-pci.c as suggested. Signed-off-by: Rod Whitby <rod@whitby.id.au> Signed-off-by: Deepak Saxena <dsaxena@plexity.net> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
| * | [ARM] 3215/1: Iomega NAS 100d (MACH_NAS100D) machine supportRod Whitby2006-01-043-0/+85
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Patch from Rod Whitby This patch adds support for a new arm/ixp4xx machine - the Iomega NAS 100d network attached storage product. The NAS100D is a consumer device containing a 266MHz Intel IXP420 processor, 16MB of flash, 64MB of RAM, a 160Gb internal IDE hard disk, and 802.11b/g wireless on an Atheros mini-PCI card. Work on porting the latest 2.6.x kernel to this device is being done by the NSLU2-Linux project (the same team who maintains the port to the Linksys NSLU2 device). In particular, the majority of this patch was authored by Alessandro Zummo, based on the work done for MACH_NSLU2 support by the NSLU2-Linux core team of developers. MACH_NAS100D (as implemented by this patch) can be enabled in jumbo ixp4xx kernels without any affect on the other machines supported by that kernel. This patch applies cleanly against 2.6.15-rc7 and should be trivial to apply to later kernel versions. It does not depend upon any other patches. Modified files (and number of lines inserted): arch/arm/mach-ixp4xx/Kconfig | 8 arch/arm/mach-ixp4xx/Makefile | 1 include/asm-arm/arch-ixp4xx/hardware.h | 1 include/asm-arm/arch-ixp4xx/irqs.h | 9 include/asm-arm/arch-ixp4xx/nas100d.h | 75 arch/arm/mach-ixp4xx/nas100d-pci.c | 77 arch/arm/mach-ixp4xx/nas100d-power.c | 69 arch/arm/mach-ixp4xx/nas100d-setup.c | 133 -- Rod Whitby (NSLU2-Linux project lead) Signed-off-by: Rod Whitby <rod@whitby.id.au> Signed-off-by: Alessandro Zummo <a.zummo@towertech.it> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
| * | [ARM] 3192/1: Remove gpio_isr_line_clear() API from IXP4xxDeepak Saxena2006-01-041-5/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Patch from Deepak Saxena Other than interrupt masking purposes, this API is only used when configuring interrupt lines and this patch moves that functionality directly into the ixp4xx_set_irq_type() implementation as board level PCI code should not need to worry about those details. Signed-off-by: Deepak Saxena <dsaxena@plexity.net> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
| * | [ARM] Remove definition of MAX_DMA_CHANNELS to zeroRussell King2006-01-0418-94/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | Since we now only build arch/arm/kernel/dma.c on machine types which set ISA_DMA_API, we don't need to define MAX_DMA_CHANNELS to 0 to indicate this - this definition becomes superfluous. Remove it. Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
| * | [ARM] Move common definition of MAX_DMA_ADDRESS to asm/dma.hRussell King2006-01-0416-31/+7
| | | | | | | | | | | | Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
| * | [ARM] Remove '__address' from scatterlist and convert to DMA APIRussell King2006-01-042-2/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The old __address element in struct scatterlist remained from older kernels because the ARM DMA emulation code made use of it. Move this field into struct dma_struct, and convert DMA emulation code to setup a SG entry as required. Also, convert DMA emulation code to use the new DMA API rather than the PCI DMA API. Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
| * | [ARM] Move ISA DMA bus_to_virt() out of set_dma_addr()Russell King2006-01-041-1/+3
| | | | | | | | | | | | | | | | | | | | | Allow the compiler to optimise the bus_to_virt(virt_to_bus()) transformation in the ARM ISA DMA interface. Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
| * | [ARM] Remove clk_use()/clk_unuse()Russell King2006-01-031-14/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | It seems that clk_use() and clk_unuse() are additional complexity which isn't required anymore. Remove them from the clock framework to avoid the additional confusion which they cause, and update all ARM machine types except for OMAP. Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
| * | [ARM] Add additional documentation to the clock source frameworkRussell King2006-01-031-0/+14
| | | | | | | | | | | | | | | | | | | | | | | | It seems that there's some confusion over how the clock source framework should be used. Add some additional comments to explain the ambiguous areas. Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
| * | [ARM] Cleanup ARM includesRussell King2006-01-0318-5/+28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | arch/arm/kernel/entry-armv.S has contained a comment suggesting that asm/hardware.h and asm/arch/irqs.h should be moved into the asm/arch/entry-macro.S include. So move the includes to these two files as required. Add missing includes (asm/hardware.h, asm/io.h) to asm/arch/system.h includes which use those facilities, and remove asm/io.h from kernel/process.c. Remove other unnecessary includes from arch/arm/kernel, arch/arm/mm and arch/arm/mach-footbridge. Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
* | | Merge master.kernel.org:/home/rmk/linux-2.6-serialLinus Torvalds2006-01-071-0/+17
|\ \ \
| * | | [SERIAL] Add support for more Connect Tech PCI serial boardsGareth Howlett2006-01-041-0/+17
| |/ / | | | | | | | | | | | | | | | | | | I've also fixed the sort-ordering comments on this naming convention. Signed-off-by: Stuart MacDonald <stuartm@connecttech.com> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
* | | Merge master.kernel.org:/home/rmk/linux-2.6-mmcLinus Torvalds2006-01-071-0/+5
|\ \ \ | |_|/ |/| |
| * | [MMC] Fix missing ','Russell King2006-01-041-1/+1
| | | | | | | | | | | | Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
| * | [MMC] Improve MMC card block size selectionRussell King2006-01-031-0/+5
| |/ | | | | | | | | | | | | | | | | | | | | Select a block size for IO based on the read and write block size combinations, and whether the card supports partial block reads and/or partial block writes. If we are able to satisfy block reads but not block writes, mark the device read only. Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
* | Merge master.kernel.org:/pub/scm/linux/kernel/git/gregkh/i2c-2.6Linus Torvalds2006-01-065-55/+20
|\ \
| * \ Merge ../torvalds-2.6/Greg Kroah-Hartman2006-01-06124-1969/+1843
| |\ \
| * | | [PATCH] i2c: i2c-nforce2 add nforce4 MCP-04 device IDJean Delvare2006-01-051-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | One more supported PCI ID for the i2c-nforce2 driver. Signed-off-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
| * | | [PATCH] i2c: driver ID list cleanupsJean Delvare2006-01-051-20/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Cleanups to i2c driver ID list: * Remove mostly bogus comments about driver ID ranges. * Drop experimental driver IDs, as the concept is pretty broken. * Drop now unused IDs of non-I2C (ISA) drivers. * Drop a few more IDs which are no more used. Signed-off-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
| * | | [PATCH] hwmon: add VRM/VID support for some VIA CPUsRudolf Marek2006-01-051-3/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch adds the VIA CENTAUR CPUs to detection table. Table was updated to treat future Intel x86 CPUs as VRD10. Stepping field was added, because some VIA CPUs have different VRM specs across stepping. I changed the vrm type to u8 because all drivers use u8 anyway. Signed-off-by: Rudolf Marek <r.marek@sh.cvut.cz> Signed-off-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
| * | | [PATCH] I2C: Make i2c_add_driver automatically set the proper module ownerGreg Kroah-Hartman2006-01-051-1/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This prevents i2c drivers from messing up and forgetting to set the module owner of their driver. It also reduces the size of their drivers by one line :) Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Cc: Jean Delvare <khali@linux-fr.org>
| * | | [PATCH] i2c: Drop i2c_driver.{owner,name}, 5 of 11Laurent Riffard2006-01-051-3/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We should use the i2c_driver.driver's .name and .owner fields instead of the i2c_driver's ones. This patch updates the drivers/media/video and usb/media drivers. Signed-off-by: Laurent Riffard <laurent.riffard@free.fr> Signed-off-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
| * | | [PATCH] i2c: Drop i2c_driver.{owner,name}, 1 of 11Laurent Riffard2006-01-051-2/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We should use the i2c_driver.driver's .name and .owner fields instead of the i2c_driver's ones. This patch updates the core of the i2c drivers: it removes .name and .owner fields from the struct i2c_device and modify various functions to use struct device fields instead. Signed-off-by: Laurent Riffard <laurent.riffard@free.fr> Signed-off-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
| * | | [PATCH] i2c: i2c_get_client is goneJean Delvare2006-01-051-11/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The i2c_get_client function doesn't exist anymore, so we shouldn't have a definition for it in i2c.h. Signed-off-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
| * | | [PATCH] i2c: Rework client usage count, 3 of 3Jean Delvare2006-01-051-3/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Do not limit the usage count of i2c clients to 1. In other words, change the client usage count behavior from the old I2C_CLIENT_ALLOW_USE to the old I2C_CLIENT_ALLOW_MULTIPLE_USE. The rationale is that no driver actually needs the limiting behavior, and the unlimiting behavior is slightly easier to implement. Update the documentation to reflect this change. Signed-off-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
| * | | [PATCH] i2c: Rework client usage count, 2 of 3Jean Delvare2006-01-051-1/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Make I2C_CLIENT_ALLOW_USE the default for all i2c clients. It doesn't hurt if the usage count is actually never used for any given driver, and allows for nice code simplifications in i2c-core. Signed-off-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
| * | | [PATCH] i2c: Rework client usage count, 1 of 3Jean Delvare2006-01-051-2/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | No i2c client uses the I2C_CLIENT_ALLOW_MULTIPLE_USE flag, drop it. Signed-off-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
| * | | [PATCH] i2c: Drop i2c_driver.flags, 3 of 3Jean Delvare2006-01-051-3/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | The flags member of the i2c_driver structure is no more used. Drop it. Signed-off-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
| * | | [PATCH] i2c: Drop i2c_driver.flags, 2 of 3Jean Delvare2006-01-051-1/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Just about every i2c chip driver sets the I2C_DF_NOTIFY flag, so we can simply make it the default and drop the flag. If any driver really doesn't want to be notified when i2c adapters are added, that driver can simply omit to set .attach_adapter. This approach is also more robust as it prevents accidental NULL pointer dereferences. Signed-off-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
| * | | [PATCH] i2c: Drop i2c_driver.flags, 1 of 3Jean Delvare2006-01-051-5/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The I2C_DF_DUMMY flag is gone since 2.5.70, it's about time to drop all ifdef'd out references thereto. Signed-off-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
* | | | Merge master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6Linus Torvalds2006-01-064-9/+9
|\ \ \ \
| * | | | [NET]: Endian-annotate in_aton()Alexey Dobriyan2006-01-061-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
| * | | | [NET]: Endian-annotate struct iphdrAlexey Dobriyan2006-01-061-5/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | And fix trivial warnings that emerged. Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
| * | | | [NET]: Change sk_run_filter()'s return type in net/core/filter.cKris Katterjohn2006-01-062-3/+3
| | |/ / | |/| | | | | | | | | | | | | | | | | | | | | | It should return an unsigned value, and fix sk_filter() as well. Signed-off-by: Kris Katterjohn <kjak@ispwest.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | Merge git://git.kernel.org/pub/scm/linux/kernel/git/sam/kbuildLinus Torvalds2006-01-062-4/+4
|\ \ \ \
| * | | | kbuild: un-stringnify KBUILD_MODNAMESam Ravnborg2006-01-062-4/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Now when kbuild passes KBUILD_MODNAME with "" do not __stringify it when used. Remove __stringnify for all users. This also fixes the output of: $ ls -l /sys/module/ drwxr-xr-x 4 root root 0 2006-01-05 14:24 pcmcia drwxr-xr-x 4 root root 0 2006-01-05 14:24 pcmcia_core drwxr-xr-x 3 root root 0 2006-01-05 14:24 "processor" drwxr-xr-x 3 root root 0 2006-01-05 14:24 "psmouse" The quoting of the module names will be gone again. Thanks to GregKH + Kay Sievers for reproting this. Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
* | | | | SUNRPC: Update the spkm3 code to use the make_checksum interfaceJ. Bruce Fields2006-01-061-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Also update the tokenlen calculations to accomodate g_token_size(). Signed-off-by: Andy Adamson <andros@citi.umich.edu> Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
* | | | | NFSv4: Allow entries in the idmap cache to expireTrond Myklebust2006-01-061-0/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | If someone changes the uid/gid mapping in userland, then we do eventually want those changes to be propagated to the kernel. Currently the kernel assumes that it may cache entries forever. Add an expiration time + garbage collector for idmap entries. Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
* | | | | SUNRPC: Ensure client closes the socket when server initiates a closeTrond Myklebust2006-01-061-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | If the server decides to close the RPC socket, we currently don't actually respond until either another RPC call is scheduled, or until xprt_autoclose() gets called by the socket expiry timer (which may be up to 5 minutes later). This patch ensures that xprt_autoclose() is called much sooner if the server closes the socket. Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
* | | | | SUNRPC: get rid of cl_chattyChuck Lever2006-01-061-1/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Clean up: Every ULP that uses the in-kernel RPC client, except the NLM client, sets cl_chatty. There's no reason why NLM shouldn't set it, so just get rid of cl_chatty and always be verbose. Test-plan: Compile with CONFIG_NFS enabled. Signed-off-by: Chuck Lever <cel@netapp.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
* | | | | SUNRPC: transport switch API for setting port numberChuck Lever2006-01-061-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | At some point, transport endpoint addresses will no longer be IPv4. To hide the structure of the rpc_xprt's address field from ULPs and port mappers, add an API for setting the port number during an RPC bind operation. Test-plan: Destructive testing (unplugging the network temporarily). Connectathon with UDP and TCP. NFSv2/3 and NFSv4 mounting should be carefully checked. Probably need to rig a server where certain services aren't running, or that returns an error for some typical operation. Signed-off-by: Chuck Lever <cel@netapp.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
* | | | | SUNRPC: new interface to force an RPC rebindChuck Lever2006-01-061-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We'd like to hide fields in rpc_xprt and rpc_clnt from upper layer protocols. Start by creating an API to force RPC rebind, replacing logic that simply sets cl_port to zero. Test-plan: Destructive testing (unplugging the network temporarily). Connectathon with UDP and TCP. NFSv2/3 and NFSv4 mounting should be carefully checked. Probably need to rig a server where certain services aren't running, or that returns an error for some typical operation. Signed-off-by: Chuck Lever <cel@netapp.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
* | | | | SUNRPC: switchable buffer allocationChuck Lever2006-01-062-7/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add RPC client transport switch support for replacing buffer management on a per-transport basis. In the current IPv4 socket transport implementation, RPC buffers are allocated as needed for each RPC message that is sent. Some transport implementations may choose to use pre-allocated buffers for encoding, sending, receiving, and unmarshalling RPC messages, however. For transports capable of direct data placement, the buffers can be carved out of a pre-registered area of memory rather than from a slab cache. Test-plan: Millions of fsx operations. Performance characterization with "sio" and "iozone". Use oprofile and other tools to look for significant regression in CPU utilization. Signed-off-by: Chuck Lever <cel@netapp.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
OpenPOWER on IntegriCloud