Skip to content
  1. May 20, 2010
    • Sonic Zhang's avatar
      i2c-bfin-twi: integrate timeout timer with completion interface · dd7319a5
      Sonic Zhang authored
      
      
      There isn't much point in managing our own custom timeout timer when the
      completion interface already includes support for it.  This makes the
      resulting code much simpler and robust.
      
      Signed-off-by: default avatarSonic Zhang <sonic.zhang@analog.com>
      Signed-off-by: default avatarMike Frysinger <vapier@gentoo.org>
      Signed-off-by: default avatarBen Dooks <ben-linux@fluff.org>
      dd7319a5
    • Mark Brown's avatar
      i2c-s3c2410: Remove unconditional 1ms delay on each transfer · 1bc2962e
      Mark Brown authored
      
      
      The S3C I2C controller indicates completion of I2C transfers before
      the bus has a stop condition on it. In order to ensure that we do not
      attempt to start a new transfer before the bus is idle the driver
      currently inserts a 1ms delay. This is vastly larger than is generally
      required and has a visible effect on performance under load, such as
      when bringing up audio CODECs or reading back status information with
      non-bulk I2C reads.
      
      Replace the sleep with a spin on the IIC status register for up to 1ms.
      This will busy wait but testing on my SMDK6410 system indicates that
      the overwhelming majority of transactions complete on the first spin,
      with maximum latencies of less than 10 spins so the absolute overhead
      of busy waiting should be at worst comprable to msleep(), and the
      overall system performance is dramatically improved.
      
      The main risk is poor interaction with multimaster systems where
      we may miss the bus going idle before the next transaction. Defend
      against this by falling back to the original 1ms delay after 20 spins.
      
      The overall effect in my testing is an approximately 20% improvement
      in kernel startup time.
      
      Signed-off-by: default avatarMark Brown <broonie@opensource.wolfsonmicro.com>
      Signed-off-by: default avatarBen Dooks <ben-linux@fluff.org>
      1bc2962e
    • Linus Torvalds's avatar
      Merge branch 'davinci-for-linus' of... · 024a6b95
      Linus Torvalds authored
      Merge branch 'davinci-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/khilman/linux-davinci
      
      * 'davinci-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/khilman/linux-davinci: (48 commits)
        Revert "rtc: omap: let device wakeup capability be configured from chip init logic"
        DM365: Added more PINMUX configurations for AEMIF
        DM365: Make CLKOUTx available
        DM365: Added PINMUX definitions for GPIO30..32
        Davinci: iotable based ioremap() interception
        Davinci: pinmux - use ioremap()
        Davinci: aintc/cpintc - use ioremap()
        Davinci: psc - use ioremap()
        Davinci: timer - use ioremap()
        Davinci: jtag_id - use ioremap()
        Davinci: da8xx: rtc - use ioremap
        Davinci: gpio - use ioremap()
        davinci: edma: fix coding style issue related to breaking lines
        davinci: edma: use BIT() wherever possible
        davinci: edma: fix coding style issue related to usage of braces
        davinci: edma: use a more intuitive name for edma_info
        Davinci: serial - conditional reset via pwremu
        Davinci: serial - use ioremap()
        Davinci: serial - remove unnecessary define
        Davinci: watchdog reset separation across socs
        ...
      
      Fix up trivial conflict in arch/arm/Kconfig due to removal of "select
      GENERIC_TIME"
      024a6b95
    • Linus Torvalds's avatar
      Merge branch 'for-linus/samsung4' of git://git.fluff.org/bjdooks/linux · 349e1fba
      Linus Torvalds authored
      * 'for-linus/samsung4' of git://git.fluff.org/bjdooks/linux: (98 commits)
        Input: s3c24xx_ts - depend on SAMSUNG_DEV_TS and update menu entry
        Input: s3c24xx_ts - Add FEAT for Samsung touchscreen support
        Input: s3c24xx_ts - Implement generic GPIO configuration callback
        ARM: SAMSUNG: Move s3c64xx dev-ts.c to plat-samsung and rename configuration
        ARM: SAMSUNG: Implements cfg_gpio function for Samsung touchscreen
        ARM: S3C64XX: Add touchscreen platform device definition
        ARM: SAMSUNG: Move mach/ts.h to plat/ts.h
        ARM: S5PC100: Move i2c helpers from plat-s5pc1xx to mach-s5pc100
        ARM: S5PC100: Move frame buffer helpers from plat-s5pc1xx to mach-s5pc100
        ARM: S5PC100: gpio.h cleanup
        ARM: S5PC100: Move gpio support from plat-s5pc1xx to mach-s5pc100
        ARM: S5PC100: Use common functions for gpiolib implementation
        drivers: serial: S5PC100 serial driver cleanup
        ARM: S5PC100: Pre-requisite clock patch for plat-s5pc1xx to plat-s5p move
        ARM: SAMSUNG: Copy common I2C0 device helpers to machine directories
        ARM: SAMSUNG: move driver strength gpio configuration helper to common dir
        ARM: S5PV210: Add GPIOlib support
        ARM: SAMSUNGy: fix broken timer irq base
        ARM: SMDK6440: Add audio devices on board
        ARM: S5P6440: Add audio platform devices
        ...
      349e1fba
    • KOSAKI Motohiro's avatar
      cpumask: fix compat getaffinity · fa9dc265
      KOSAKI Motohiro authored
      Commit a45185d2
      
       "cpumask: convert kernel/compat.c" broke libnuma, which
      abuses sched_getaffinity to find out NR_CPUS in order to parse
      /sys/devices/system/node/node*/cpumap.
      
      On NUMA systems with less than 32 possibly CPUs, the current
      compat_sys_sched_getaffinity now returns '4' instead of the actual
      NR_CPUS/8, which makes libnuma bail out when parsing the cpumap.
      
      The libnuma call sched_getaffinity(0, bitmap, 4096) at first.  It mean
      the libnuma expect the return value of sched_getaffinity() is either len
      argument or NR_CPUS.  But it doesn't expect to return nr_cpu_ids.
      
      Strictly speaking, userland requirement are
      
      1) Glibc assume the return value mean the lengh of initialized
         of mask argument. E.g. if sched_getaffinity(1024) return 128,
         glibc make zero fill rest 896 byte.
      2) Libnuma assume the return value can be used to guess NR_CPUS
         in kernel. It assume len-arg<NR_CPUS makes -EINVAL. But
         it try len=4096 at first and 4096 is always bigger than
         NR_CPUS. Then, if we remove strange min_length normalization,
         we never hit -EINVAL case.
      
      sched_getaffinity() already solved this issue.  This patch adapts
      compat_sys_sched_getaffinity() to match the non-compat case.
      
      Signed-off-by: default avatarKOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
      Acked-by: default avatarRusty Russell <rusty@rustcorp.com.au>
      Acked-by: default avatarArnd Bergmann <arnd@arndb.de>
      Reported-by: default avatarKen Werner <ken.werner@web.de>
      Cc: stable@kernel.org
      Cc: Andi Kleen <andi@firstfloor.org>
      Signed-off-by: default avatarLinus Torvalds <torvalds@linux-foundation.org>
      fa9dc265
    • Linus Torvalds's avatar
      Merge branch 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6 · cf77e988
      Linus Torvalds authored
      * 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6:
        [IA64] Drop duplicated "config IOMMU_HELPER"
        [IA64] invoke oom-killer from page fault
        [IA64] use __ratelimit
        [IA64] Use set_cpus_allowed_ptr
        [IA64] Use set_cpus_allowed_ptr
        [IA64] arch/ia64/hp/common/sba_iommu.c: Rename dev_info to adi
        [IA64] removing redundant ifdef
      cf77e988
    • Linus Torvalds's avatar
      Merge branch 'devel' of master.kernel.org:/home/rmk/linux-2.6-arm · 1d3c6ff4
      Linus Torvalds authored
      * 'devel' of master.kernel.org:/home/rmk/linux-2.6-arm: (224 commits)
        ARM: remove 'select GENERIC_TIME'
        ARM: 6136/1: ARCH_REQUIRE_GPIOLIB selects GENERIC_GPIO
        ARM: 6074/1: oprofile: convert from sysdev to platform device
        ARM: 6073/1: oprofile: remove old files and update KConfig
        ARM: 6072/1: oprofile: use perf-events framework as backend
        ARM: 6071/1: perf-events: allow modules to query the number of hardware counters
        ARM: 6070/1: perf-events: add support for xscale PMUs
        ARM: 6069/1: perf-events: use numeric ID to identify PMU
        ARM: 6064/1: pmu: register IRQs at runtime
        ARM: Optionally allow ARMv6 to use 'normal, bufferable' memory for DMA
        ARM: 6134/1: Handle instruction cache maintenance fault properly
        ARM: nwfpe: allow debugging output to be configured at runtime
        ARM: rename mach_cpu_disable() to platform_cpu_disable()
        ARM: 6132/1: PL330: Add common core driver
        ARM: 6094/1: Extend cache-l2x0 to support the 16-way PL310
        ARM: Move memory mapping into mmu.c
        ARM: Ensure meminfo is sorted prior to sanity_check_meminfo
        ARM: Remove useless linux/bootmem.h includes
        ARM: convert /proc/cpu/aligment to seq_file
        arm: use asm-generic/scatterlist.h
        ...
      1d3c6ff4
    • Kevin Hilman's avatar
      Revert "rtc: omap: let device wakeup capability be configured from chip init logic" · 5e682ef8
      Kevin Hilman authored
      This reverts commit 9c0a342c because
      it was included without proper signoffs from RTC maintainers.
      5e682ef8
    • Linus Torvalds's avatar
      Merge git://git.kernel.org/pub/scm/linux/kernel/git/lethal/sh-2.6 · 7c7cbaf5
      Linus Torvalds authored
      * git://git.kernel.org/pub/scm/linux/kernel/git/lethal/sh-2.6: (127 commits)
        sh: update defconfigs.
        sh: Fix up the NUMA build for recent LMB changes.
        sh64: provide a stub per_cpu_trap_init() definition.
        sh: fix up CONFIG_KEXEC=n build.
        sh: fixup the docbook paths for clock framework shuffling.
        driver core: Early dev_name() depends on slab_is_available().
        sh: simplify WARN usage in SH clock driver
        sh: Check return value of clk_get on ms7724
        sh: Check return value of clk_get on ecovec24
        sh: move sh clock-cpg.c contents to drivers/sh/clk-cpg.c
        sh: move sh clock.c contents to drivers/sh/clk.
        sh: move sh asm/clock.h contents to linux/sh_clk.h V2
        sh: remove unused clock lookup
        sh: switch boards to clkdev
        sh: switch sh4-202 to clkdev
        sh: switch shx3 to clkdev
        sh: switch sh7757 to clkdev
        sh: switch sh7763 to clkdev
        sh: switch sh7780 to clkdev
        sh: switch sh7786 to clkdev
        ...
      7c7cbaf5
    • Linus Torvalds's avatar
      Merge branch 'for-linus' of git://git390.marist.edu/pub/scm/linux-2.6 · ba0234ec
      Linus Torvalds authored
      * 'for-linus' of git://git390.marist.edu/pub/scm/linux-2.6: (24 commits)
        [S390] drivers/s390/char: Use kmemdup
        [S390] drivers/s390/char: Use kstrdup
        [S390] debug: enable exception-trace debug facility
        [S390] s390_hypfs: Add new attributes
        [S390] qdio: remove API wrappers
        [S390] qdio: set correct bit in dsci
        [S390] qdio: dont convert timestamps to microseconds
        [S390] qdio: remove memset hack
        [S390] qdio: prevent starvation on PCI devices
        [S390] qdio: count number of qdio interrupts
        [S390] user space fault: report fault before calling do_exit
        [S390] topology: expose core identifier
        [S390] dasd: remove uid from devmap
        [S390] dasd: add dynamic pav toleration
        [S390] vdso: add missing vdso_install target
        [S390] vdso: remove redundant check for CONFIG_64BIT
        [S390] avoid default_llseek in s390 drivers
        [S390] vmcp: disallow modular build
        [S390] add breaking event address for user space
        [S390] virtualization aware cpu measurement
        ...
      ba0234ec
  2. May 19, 2010