Eric Lee / smarc-fsl-linux-kernel

09 Sep, 2015

3 commits

25 Jun, 2015

1 commit

  • fc6daaf93   mm/memblock: add extra "flags" to memblock to allow selection of memory based on attribute ... Browse Code »

    Some high end Intel Xeon systems report uncorrectable memory errors as a
    recoverable machine check. Linux has included code for some time to
    process these and just signal the affected processes (or even recover
    completely if the error was in a read only page that can be replaced by
    reading from disk).

    But we have no recovery path for errors encountered during kernel code
    execution. Except for some very specific cases were are unlikely to ever
    be able to recover.

    Enter memory mirroring. Actually 3rd generation of memory mirroing.

    Gen1: All memory is mirrored
    Pro: No s/w enabling - h/w just gets good data from other side of the
    mirror
    Con: Halves effective memory capacity available to OS/applications

    Gen2: Partial memory mirror - just mirror memory begind some memory controllers
    Pro: Keep more of the capacity
    Con: Nightmare to enable. Have to choose between allocating from
    mirrored memory for safety vs. NUMA local memory for performance

    Gen3: Address range partial memory mirror - some mirror on each memory
    controller
    Pro: Can tune the amount of mirror and keep NUMA performance
    Con: I have to write memory management code to implement

    The current plan is just to use mirrored memory for kernel allocations.
    This has been broken into two phases:

    1) This patch series - find the mirrored memory, use it for boot time
    allocations

    2) Wade into mm/page_alloc.c and define a ZONE_MIRROR to pick up the
    unused mirrored memory from mm/memblock.c and only give it out to
    select kernel allocations (this is still being scoped because
    page_alloc.c is scary).

    This patch (of 3):

    Add extra "flags" to memblock to allow selection of memory based on
    attribute. No functional changes

    Signed-off-by: Tony Luck
    Cc: Xishi Qiu
    Cc: Hanjun Guo
    Cc: Xiexiuqi
    Cc: Ingo Molnar
    Cc: Thomas Gleixner
    Cc: "H. Peter Anvin"
    Cc: Yinghai Lu
    Cc: Naoya Horiguchi
    Signed-off-by: Andrew Morton
    Signed-off-by: Linus Torvalds

    Tony Luck
     

15 Apr, 2015

2 commits

  • 7f70baeeb   memtest: use phys_addr_t for physical addresses ... Browse Code »

    Since memtest might be used by other architectures pass input parameters
    as phys_addr_t instead of long to prevent overflow.

    Signed-off-by: Vladimir Murzin
    Acked-by: Will Deacon
    Tested-by: Mark Rutland
    Cc: "H. Peter Anvin"
    Cc: Catalin Marinas
    Cc: Ingo Molnar
    Cc: Russell King
    Cc: Thomas Gleixner
    Signed-off-by: Andrew Morton
    Signed-off-by: Linus Torvalds

    Vladimir Murzin
     
  • 4a20799d1   mm: move memtest under mm ... Browse Code »

    Memtest is a simple feature which fills the memory with a given set of
    patterns and validates memory contents, if bad memory regions is detected
    it reserves them via memblock API. Since memblock API is widely used by
    other architectures this feature can be enabled outside of x86 world.

    This patch set promotes memtest to live under generic mm umbrella and
    enables memtest feature for arm/arm64.

    It was reported that this patch set was useful for tracking down an issue
    with some errant DMA on an arm64 platform.

    This patch (of 6):

    There is nothing platform dependent in the core memtest code, so other
    platforms might benefit from this feature too.

    [linux@roeck-us.net: MEMTEST depends on MEMBLOCK]
    Signed-off-by: Vladimir Murzin
    Acked-by: Will Deacon
    Tested-by: Mark Rutland
    Cc: Ingo Molnar
    Cc: Thomas Gleixner
    Cc: "H. Peter Anvin"
    Cc: Catalin Marinas
    Cc: Russell King
    Cc: Paul Bolle
    Signed-off-by: Andrew Morton
    Signed-off-by: Linus Torvalds

    Vladimir Murzin