Merge tag 'amd-thresholding-fixes-for-3.6' of git://git.kernel.org/pub/scm/linux...
authorIngo Molnar <mingo@kernel.org>
Fri, 8 Jun 2012 10:29:47 +0000 (12:29 +0200)
committerIngo Molnar <mingo@kernel.org>
Fri, 8 Jun 2012 10:29:47 +0000 (12:29 +0200)
commit707ecec1dc80d54163356a25d145c62083e7be60
treea78ba59e285b6148d012bc69816fd2429aae07ee
parent48d212a2eecaca2e1875925837ad27b2f43f48a3
parent11122570193a429f72014703ce6b849640f8d7e5
Merge tag 'amd-thresholding-fixes-for-3.6' of git://git./linux/kernel/git/ras/ras into x86/mce

Pull in AMD MCE thresholding fixes for v3.6, from Borislav Petkov:

" Those are a bunch of patches which give the MCE thresholding code a
  hard look and a scrubbing to remove a couple of annoyances like sysfs
  warnings when running CPU off-/online tests and the threshold_bank4 node
  under /sys/devices/system/machinecheck/ is a symlink.

  It also gives proper names to the thresholding banks instead of simply
  enumerating them, like this:

     /sys/devices/system/machinecheck/machinecheck0/
     |-- bank0
     |-- bank1
     |-- bank2
     |-- bank3
     |-- bank4
     |-- bank5
     |-- bank6
     |-- check_interval
     |-- cmci_disabled
     |-- combined_unit
     |   |-- combined_unit
     |       |-- error_count
     |       |-- threshold_limit
     |-- dont_log_ce
     |-- execution_unit
     |   |-- execution_unit
     |       |-- error_count
     |       |-- threshold_limit
     |-- ignore_ce
     |-- insn_fetch
     |   |-- insn_fetch
     |       |-- error_count
     |       |-- threshold_limit
     |-- load_store
     |   |-- load_store
     |       |-- error_count
     |       |-- threshold_limit
     |-- monarch_timeout
     |-- northbridge
     |   |-- dram
     |   |   |-- error_count
     |   |   |-- interrupt_enable
     |   |   |-- threshold_limit
     |   |-- ht_links
     |   |   |-- error_count
     |   |   |-- interrupt_enable
     |   |   |-- threshold_limit
     |   |-- l3_cache
     |       |-- error_count
     |       |-- interrupt_enable
     |       |-- threshold_limit
    ...

  It is tested on all our families >= K8."

Signed-off-by: Ingo Molnar <mingo@kernel.org>