F11h has almost the same MCE signatures as K8 except DRAM ECC and MC5
bank errors. Reuse functionality from the other families.
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>

Now that all decoders have been taught about F14h, models < 0x10
MCEs, enable decoding on this family of CPUs. Also, issue a short
informational message upon boot that MCE decoding gets enabled.
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>

Those are N/A on K8, so don't decode them there.
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>

Add support for decoding F14h BU MCEs and improve decoding of the
remaining families.
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>

F14h CPUs do not generate LS MCEs so exit early and warn the user in
case this path is ever hit that something else might be going haywire.
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>

Add support for IC MCEs for F14h CPUs. K8 and F10h are almost identical
so use one function for both.
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>

Add a per-family data cache decoders. Since there is a certain overlap
between the different DC MCE signatures, reuse functionality between the
families as far as possible.
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>

Drop "edac_" string from the filenames since they're prefixed with edac/
in their pathname anyway.
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>

Add sysfs injection facilities for testing of the MCE decoding code.
Remove large parts of amd64_edac_dbg.c, as a result, which did only
NB MCE injection anyway and the new injection code supports that
functionality already.

Add an injection module so that MCE decoding code in production kernels
like those in RHEL and SLES can be tested.
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>

Move toplevel sysfs class to the stub and make it available to
non-modularized code too. Add proper refcounting of its users and move
the registration functionality into the reference counting routines.
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>

... instead of the MCi_STATUS info only for improved handling of certain
types of errors later.
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>

Clean up error codes names, shorten to mnemonics, add RRRR boundary
checking.
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>

Remove remains from previous functionality.
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>

.. so that the user knows what she's looking at there in dmesg. Also,
fix a minor cosmetic output inconsistency.
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>

We should return a negative value when we cannot get the toplevel edac
sysfs class.
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>

Array of udimm sysfs attributes was not ended with NULL marker, leading to
dereference of random memory.

  EDAC DEBUG: edac_create_mci_instance_attributes: edac_create_mci_instance_attributes() file udimm0
  EDAC DEBUG: edac_create_mci_instance_attributes: edac_create_mci_instance_attributes() file udimm1
  EDAC DEBUG: edac_create_mci_instance_attributes: edac_create_mci_instance_attributes() file udimm2
  BUG: unable to handle kernel NULL pointer dereference at 00000000000001a4
  IP: [<ffffffff81330b36>] edac_create_mci_instance_attributes+0x148/0x1f1
  Pid: 1, comm: swapper Not tainted 2.6.36-rc3-nv+ #483 P6T SE/System Product Name
  RIP: 0010:[<ffffffff81330b36>]  [<ffffffff81330b36>] edac_create_mci_instance_attributes+0x148/0x1f1
  (...)
  Call Trace:
   [<ffffffff81330b86>] edac_create_mci_instance_attributes+0x198/0x1f1
   [<ffffffff81330c9a>] edac_create_sysfs_mci_device+0xbb/0x2b2
   [<ffffffff8132f533>] edac_mc_add_mc+0x46b/0x557
   [<ffffffff81428901>] i7core_probe+0xccf/0xec0
  RIP  [<ffffffff81330b36>] edac_create_mci_instance_attributes+0x148/0x1f1
  ---[ end trace 20de320855b81d78 ]---
  Kernel panic - not syncing: Attempted to kill init!
Signed-off-by: NMarcin Slusarz <marcin.slusarz@gmail.com>
Cc: Mauro Carvalho Chehab <mchehab@redhat.com>
Acked-by: NDoug Thompson <dougthompson@xmission.com>
Cc: <stable@kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

f4347553 removed the edac polling
mechanism in favor of using a notifier chain for conveying MCE
information to edac. However, the module removal path didn't test
whether the driver had setup the polling function workqueue at all and
the rmmod process was hanging in the kernel at try_to_del_timer_sync()
in the cancel_delayed_work() path, trying to cancel an uninitialized
work struct.

Fix that by adding a balancing check to the workqueue removal path.
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>

When the Overflow MCi_STATUS bit is set, EDAC reports the lost error
with a "no information available" message which often puzzles users
parsing the dmesg. This doesn't make much sense since this error has
been lost anyway so no need for reporting it separately. Thus, report
the overflow bit setting in the MCE dump instead. While at it, remove
reporting of MiscV and ErrorEnable (en) which are superfluous.

Now it looks like this:

[ 1501.650024] MC4_STATUS: Corrected error, other errors lost: yes, CPU context corrupt: no, CECC Error
[ 1501.666887] Northbridge Error, node 2
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>

Limit MCE error decoding to current and older families only (K8-F11h).
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>

Simply add proper IDs into the device table.
Signed-off-by: NAnton Vorontsov <avorontsov@mvista.com>
Cc: Scott Wood <scottwood@freescale.com>
Cc: Peter Tyser <ptyser@xes-inc.com>
Cc: Dave Jiang <djiang@mvista.com>
Cc: Doug Thompson <dougthompson@xmission.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

-EIO is not the only error code that pci_enable_device() may return, also
the set of errors can be enhanced in future.  We should compare return
code with zero, not with concrete error value.
Signed-off-by: NKulikov Vasiliy <segooon@gmail.com>
Acked-by: NMauro Carvalho Chehab <mchehab@redhat.com>
Cc: Jeff Roberson <jroberson@jroberson.net>
Cc: Doug Thompson <dougthompson@xmission.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

-EIO is not the only error code that pci_enable_device() may return, also
the set of errors can be enhanced in future.  We should compare return
code with zero, not with concrete error value.
Signed-off-by: NKulikov Vasiliy <segooon@gmail.com>
Acked-by: NMauro Carvalho Chehab <mchehab@redhat.com>
Cc: Jeff Roberson <jroberson@jroberson.net>
Cc: Doug Thompson <dougthompson@xmission.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

In 5753c082 ("powerpc/85xx: Kconfig
cleanup") menuconfig MPC85xx was replaced by FSL_SOC_BOOKE but some
references insider the code were not adjusted accordingly.  This patch
adresses these missing pieces.
Signed-off-by: NChristoph Egger <siccegge@cs.fau.de>
Cc: Doug Thompson <dougthompson@xmission.com>
Cc: Peter Tyser <ptyser@xes-inc.com>
Cc: Kumar Gala <galak@kernel.crashing.org>
Cc: Scott Wood <scottwood@freescale.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

of_device is just an alias for platform_device, so remove it entirely.  Also
replace to_of_device() with to_platform_device() and update comment blocks.

This patch was initially generated from the following semantic patch, and then
edited by hand to pick up the bits that coccinelle didn't catch.

@@
@@
-struct of_device
+struct platform_device
Signed-off-by: NGrant Likely <grant.likely@secretlab.ca>
Reviewed-by: NDavid S. Miller <davem@davemloft.net>

EDAC MC3: CE page 0xc32281, offset 0x8a0, grain 0, syndrome 0x1, row 2, channel 1, label "": amd64_edac
EDAC MC3: CE - no information available: amd64_edacError Overflow

Add the missing space before "Error Overflow" on the second line.
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>

The bitwise AND is of higher precedence, make that explicit.

Cc: <stable@kernel.org> # 34.x
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>

Fortify the interface to not accept negative values, remove
memctrl_int_store() as a result. Also, sanitize bandwidth setting by
making the argument a simple u32 instead of strange u32 pointer being
passed around for no obvious reason. Then, fix error handling and teach
it to return proper error values. Finally, make code more readable,
simplify debug messages.

Cc: Mauro Carvalho Chehab <mchehab@redhat.com>
Cc: Arthur Jones <ajones@riverbed.com>
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>
Acked-by: NDoug Thompson <dougthompson@xmission.com>

Exit early when setting scrub rate on unknown/unsupported families.

Cc: <stable@kernel.org> # 32.x 33.x 34.x
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>
Acked-by: NDoug Thompson <dougthompson@xmission.com>

The correct check is to verify whether in high range we're below 4GB
and not to extract the DctSelBaseAddr again. See "2.8.5 Routing DRAM
Requests" in the F10h BKDG.

Cc: <stable@kernel.org> # .32.x .33.x .34.x
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>
Acked-by: NDoug Thompson <dougthompson@xmission.com>

Switch to reusing the mcheck core's machine check polling mechanism
instead of duplicating functionality by using the EDAC polling routine.

Correct formatting while at it.
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>
Acked-by: NDoug Thompson <dougthompson@xmission.com>

All F2x110-related bit defines are used at only one place so replace
them with simple BIT() macros.
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>
Acked-by: NDoug Thompson <dougthompson@xmission.com>

This option differs from EDAC_DEBUG only by printing the file and
line of where the debug statement is placed, which contains unneeded
information. So remove it.
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>
Acked-by: NDoug Thompson <dougthompson@xmission.com>

Remove the two syndrome extraction macros and add a single function
which does the same thing but with proper typechecking. While at it,
make sure to cache ECC syndrome size and dump it in debug output.
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>

The MPC85xx EDAC driver is missing module device aliases, so the driver
won't load automatically on boot.  This patch fixes the issue by adding
proper MODULE_DEVICE_TABLE() macros.
Signed-off-by: NAnton Vorontsov <avorontsov@mvista.com>
Cc: Doug Thompson <dougthompson@xmission.com>
Cc: Peter Tyser <ptyser@xes-inc.com>
Cc: Dave Jiang <djiang@mvista.com>
Cc: Kumar Gala <galak@kernel.crashing.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Don't print failure to detect Core i7 EDAC facilities to the console at
boot time, most often occurring on Core i7 desktops and laptops.
Signed-off-by: NDaniel J Blueman <daniel.blueman@gmail.com>
Acked-by: NMauro Carvalho Chehab <mchehab@redhat.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Simply add a proper ID into the device table.
Signed-off-by: NAnton Vorontsov <avorontsov@mvista.com>
Cc: Doug Thompson <dougthompson@xmission.com>
Cc: Peter Tyser <ptyser@xes-inc.com>
Cc: Dave Jiang <djiang@mvista.com>
Cc: Kumar Gala <galak@kernel.crashing.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Since commit 5753c082 ("powerpc/85xx:
Kconfig cleanup"), there is no MPC85xx Kconfig symbol anymore, so the
driver became non-selectable.

This patch fixes the issue by switching to PPC_85xx symbol.
Signed-off-by: NAnton Vorontsov <avorontsov@mvista.com>
Cc: Doug Thompson <dougthompson@xmission.com>
Cc: Peter Tyser <ptyser@xes-inc.com>
Cc: Dave Jiang <djiang@mvista.com>
Cc: Kumar Gala <galak@kernel.crashing.org>
Cc: <stable@kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

As Nehalem/Nehalem-EP/Westmere devices uses several devices for the same
functionality (memory controller), the default way of proping devices doesn't
work. So, instead of a per-device probe, all devices should be probed at once.

This means that we should block any new attempt of probe, otherwise, it will
try to register the same device several times.
Acked-by: NDoug Thompson <dougthompson@xmission.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

On Nehalem/Nehalem-EP/Westmere, the first QPI device is the last PCI bus.
The last bus is generally at 0x3f or 0xff, but there are also other systems
using different setups. For example, HP Z800 has 0x7f as the last bus.

This patch adds a logic to discover the last bus, dynamically detecting it
at runtime.
Acked-by: NDoug Thompson <dougthompson@xmission.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

When calculating the DCT channel from the syndrome we need to know the
syndrome type (x4 vs x8). On F10h, this is read out from extended PCI
cfg space register F3x180 while on K8 we only support x4 syndromes and
don't have extended PCI config space anyway.

Make the code accessing F3x180 F10h only and fall back to x4 syndromes
on everything else.

Cc: <stable@kernel.org> # .33.x .34.x
Reported-by: NJeffrey Merkey <jeffmerkey@gmail.com>
Signed-off-by: NBorislav Petkov <borislav.petkov@amd.com>