Граф коммитов

137 Коммитов

Автор SHA1 Сообщение Дата
Mauro Carvalho Chehab accf74fff3 i7core_edac: don't use a freed mci struct
This is a nasty bug. Since kobject count will be reduced by zero by
edac_mc_del_mc(), and this triggers the kobj release method, the
mci memory will be freed automatically. So, all we have left is ctl_name,
as shown by enabling debug:

[   80.822186] EDAC DEBUG: in drivers/edac/edac_mc_sysfs.c, line at 1020: edac_remove_sysfs_mci_device()  remove_link
[   80.832590] EDAC DEBUG: in drivers/edac/edac_mc_sysfs.c, line at 1024: edac_remove_sysfs_mci_device()  remove_mci_instance
[   80.843776] EDAC DEBUG: in drivers/edac/edac_mc_sysfs.c, line at 640: edac_mci_control_release() mci instance idx=0 releasing
[   80.855163] EDAC MC: Removed device 0 for i7core_edac.c i7 core #0: DEV 0000:3f:03.0
[   80.862936] EDAC DEBUG: in drivers/edac/i7core_edac.c, line at 2089: (null): free structs
[   80.871134] EDAC DEBUG: in drivers/edac/edac_mc.c, line at 238: edac_mc_free()
[   80.878379] EDAC DEBUG: in drivers/edac/edac_mc_sysfs.c, line at 726: edac_mc_unregister_sysfs_main_kobj()
[   80.888043] EDAC DEBUG: in drivers/edac/i7core_edac.c, line at 1232: drivers/edac/i7core_edac.c: i7core_put_devices()

Also, kfree(mci) shouldn't happen at the kobj.release, as it happens
when edac_remove_sysfs_mci_device() is called, but the logic is:
	edac_remove_sysfs_mci_device(mci);
	edac_printk(KERN_INFO, EDAC_MC,
		"Removed device %d for %s %s: DEV %s\n", mci->mc_idx,
		mci->mod_name, mci->ctl_name, edac_dev_name(mci));
So, as the edac_printk() needs the mci struct, this generates an OOPS.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-10-24 11:20:38 -02:00
Mauro Carvalho Chehab bbc560ae67 edac_core: Print debug messages at release calls
This is important to track a nasty bug at the free logic.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-10-24 11:20:38 -02:00
Mauro Carvalho Chehab 39300e7143 i7core_edac: explicitly remove PCI devices from the devices list
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-10-24 11:20:37 -02:00
Mauro Carvalho Chehab 41ba6c1058 i7core_edac: MCE NMI handling should stop first
Otherwise, a NMI may happen causing a race condition and a panic.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-10-24 11:20:37 -02:00
Mauro Carvalho Chehab 6ee7dd5044 i7core_edac: Initialize all priv vars before start polling
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-10-24 11:20:37 -02:00
Mauro Carvalho Chehab 3cfd01468b i7core_edac: Improve debug to seek for register/remove errors
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-10-24 11:20:37 -02:00
Mauro Carvalho Chehab e9144601d3 i7core_edac: move #if PAGE_SHIFT to edac_core.h
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-10-24 11:20:36 -02:00
Mauro Carvalho Chehab 1288c18f48 i7core_edac: Properly mark const static vars as such
There are two groups of sysfs attributes: one for rdimm and another
for udimm. Instead of changing dynamically the unique static struct
for handling udimm's, declare two vars and make them constant.

This avoids the risk of having two or more memory controllers, each
needing a different set of attributes.

While here, use const on all places where it is applicable.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

edac_core: use const for constant sysfs arguments

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-10-24 11:20:14 -02:00
Mauro Carvalho Chehab 18c29002f9 i7core_edac: move static vars to the beginning of the file
While here, don't initialize probed with 0.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-10-24 11:20:12 -02:00
Mauro Carvalho Chehab 939747bd68 i7core_edac: Be sure that the edac pci handler will be properly released
With multi-sockets, more than one edac pci handler is enabled. Be sure to
un-register all instances.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-10-24 11:20:12 -02:00
Marcin Slusarz 64aab720bd i7core_edac: fix panic in udimm sysfs attributes registration
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: Marcin Slusarz <marcin.slusarz@gmail.com>
Cc: Mauro Carvalho Chehab <mchehab@redhat.com>
Acked-by: Doug Thompson <dougthompson@xmission.com>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2010-10-01 10:50:58 -07:00
Daniel J Blueman ab08937400 quiesce EDAC initialisation on desktop/mobile i7
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: Daniel J Blueman <daniel.blueman@gmail.com>
Acked-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2010-07-26 08:17:44 -07:00
Mauro Carvalho Chehab 2d95d8158b i7core_edac: Avoid doing multiple probes for the same card
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: Doug Thompson <dougthompson@xmission.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-07-02 18:04:29 -03:00
Mauro Carvalho Chehab bda142890e i7core_edac: Properly discover the first QPI device
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: Doug Thompson <dougthompson@xmission.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-07-02 18:04:05 -03:00
Mauro Carvalho Chehab 52707f918c i7core_edac: Better describe the supported devices
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-18 20:43:52 -03:00
Vernon Mauery bd9e19ca46 Add support for Westmere to i7core_edac driver
This adds new PCI IDs for the Westmere's memory controller
devices and modifies the i7core_edac driver to be able to
probe both Nehalem and Westmere processors.

Signed-off-by: Vernon Mauery <vernux@us.ibm.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-18 20:23:56 -03:00
Tony Luck d4d1ef4515 i7core_edac: don't free on success
Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-18 14:47:31 -03:00
Mauro Carvalho Chehab ac1ececea9 i7core_edac: Add support for X5670
As reported by Vernon Mauery <vernux@us.ibm.com>, X5670 (Westmere-EP) uses a
different register for one of the uncore PCI devices. Add support for
it.

Those are the PCI ID's on this new chipset:

fe:00.0 0600: 8086:2c70 (rev 02)
fe:00.1 0600: 8086:2d81 (rev 02)
fe:02.0 0600: 8086:2d90 (rev 02)
fe:02.1 0600: 8086:2d91 (rev 02)
fe:02.2 0600: 8086:2d92 (rev 02)
fe:02.3 0600: 8086:2d93 (rev 02)
fe:02.4 0600: 8086:2d94 (rev 02)
fe:02.5 0600: 8086:2d95 (rev 02)
fe:03.0 0600: 8086:2d98 (rev 02)
fe:03.1 0600: 8086:2d99 (rev 02)
fe:03.2 0600: 8086:2d9a (rev 02)
fe:03.4 0600: 8086:2d9c (rev 02)
fe:04.0 0600: 8086:2da0 (rev 02)
fe:04.1 0600: 8086:2da1 (rev 02)
fe:04.2 0600: 8086:2da2 (rev 02)
fe:04.3 0600: 8086:2da3 (rev 02)
fe:05.0 0600: 8086:2da8 (rev 02)
fe:05.1 0600: 8086:2da9 (rev 02)
fe:05.2 0600: 8086:2daa (rev 02)
fe:05.3 0600: 8086:2dab (rev 02)
fe:06.0 0600: 8086:2db0 (rev 02)
fe:06.1 0600: 8086:2db1 (rev 02)
fe:06.2 0600: 8086:2db2 (rev 02)
fe:06.3 0600: 8086:2db3 (rev 02)
(as usual, the same PCI devices repeat at ff: bus)

The PCI device 8086:2c70 is shown as:

fe:00.0 Host bridge: Intel Corporation QuickPath Architecture Generic
Non-core Registers (rev 02)

So, for this device to be recognized, it is only a matter of adding this
new PCI ID to the driver.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-18 13:15:42 -03:00
Vernon Mauery 8a311e179e Always call i7core_[ur]dimm_check_mc_ecc_err
This fixes an error in function i7core_check_error

In commit ca9c90ba09 which converts the
driver to use double buffering, there is a change in the logic.  Before,
if mce_count was zero, it skipped over a couple of statements and
finished out with a call to the *check_mc_ecc_err function.  The current
code checks to see if mce_count is 0 and then exits.

This change reverts the behavior back to the original where if there are
no errors to report, we skip to the end and call the *check_mc_ecc_err
function.

This fix allows the driver to work again on my Nehalem based blades
again.

Signed-off-by: Vernon Mauery <vernux@us.ibm.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-18 12:43:23 -03:00
Alexander Beregalov 2a6fae3267 i7core_edac: fix memory leak of i7core_dev
Free already allocated i7core_dev.

Signed-off-by: Alexander Beregalov <a.beregalov@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-18 11:45:20 -03:00
Jiri Slaby 71753e0141 EDAC: add __init to i7core_xeon_pci_fixup
It's called only from an __init function and is the only user
of pcibios_scan_specific_bus which will be marked as __devinit in
the next patch.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-18 11:45:19 -03:00
Mauro Carvalho Chehab 508fa179f8 i7core_edac: Fix wrong device id for channel 1 devices
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 12:18:31 -03:00
Mauro Carvalho Chehab f05da2f785 i7core: add support for Lynnfield alternate address
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 12:18:29 -03:00
Mauro Carvalho Chehab 52a2e4fc37 i7core_edac: Add initial support for Lynnfield
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 12:18:28 -03:00
Randy Dunlap 3b918c12df edac: fix i7core build
Fix build warning (missing header file) and
build error when CONFIG_SMP=n.

drivers/edac/i7core_edac.c:860: error: implicit declaration of function 'msleep'
drivers/edac/i7core_edac.c:1700: error: 'struct cpuinfo_x86' has no member named 'phys_proc_id'

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:49:32 -03:00
Alan Cox 486dd09f12 edac: i7core_edac produces undefined behaviour on 32bit
Fix the shifts up

Signed-off-by: Alan Cox <alan@linux.intel.com>
Acked-by: Doug Thompson <dougthompson@xmission.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:49:32 -03:00
Mauro Carvalho Chehab de06eeef58 i7core_edac: Use a more generic approach for probing PCI devices
Currently, only one PCI set of tables is allowed. This prevents using
the driver for other devices like Lynnfield, with have a different
set of PCI ID's.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:49:31 -03:00
Mauro Carvalho Chehab fd3826549d i7core_edac: PCI device is called NONCORE, instead of NOCORE
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:49:31 -03:00
Mauro Carvalho Chehab 321ece4dda i7core_edac: Fix ringbuffer maxsize
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:49:31 -03:00
Mauro Carvalho Chehab 6e103be1c7 i7core_edac: First store, then increment
Fix ringbuffer store logic.

While here, add a few comments to the code and remove the undesired
printk that could otherwise be called during NMI time.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:49:31 -03:00
Mauro Carvalho Chehab 4f87fad1d3 i7core_edac: Better parse "any" addrmask
Instead of accepting just "any", accept also "any\n"

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:49:30 -03:00
Mauro Carvalho Chehab ca9c90ba09 i7core_edac: Use a lockless ringbuffer
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:49:30 -03:00
Mauro Carvalho Chehab f338d73691 i7core_edac: Convert UDIMM error counters into a proper sysfs group
Instead of displaying 3 values at the same var, break it into 3
different sysfs nodes:

/sys/devices/system/edac/mc/mc0/all_channel_counts/udimm0
/sys/devices/system/edac/mc/mc0/all_channel_counts/udimm1
/sys/devices/system/edac/mc/mc0/all_channel_counts/udimm2

For registered dimms, however, the error counters are already being
displayed at:
	/sys/devices/system/edac/mc/mc0/csrow*/ce_count

So, there's no need to add any extra sysfs nodes.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:45:02 -03:00
Mauro Carvalho Chehab cc301b3ae3 edac: store/show methods for device groups weren't working
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:45:02 -03:00
Mauro Carvalho Chehab a5538e531f i7core_edac: Add support for sysfs addrmatch group
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:45:01 -03:00
Mauro Carvalho Chehab 4af91889e0 i7core_edac: Avoid printing a warning when debug is disabled
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:45:00 -03:00
Mauro Carvalho Chehab 4253868034 i7core_edac: We need to use list_for_each_entry_safe to avoid errors
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:45:00 -03:00
Mauro Carvalho Chehab 22e6bcbdcf i7core_edac: change remove module strategy
The old remove module stragegy didn't work on devices with multiple
cores, since only one PCI device is used to open all mc's, due to
Nehalem nature.

Also, it were based at pdev value. However, this doesn't point to the
pci device used at mci->dev.

So, instead, it unregisters all devices at once, deleting them from the
device list.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:45:00 -03:00
Mauro Carvalho Chehab 0f062792b4 i7core_edac: remove static counter for max sockets
The number of sockets is now fully dynamic. Get rid of this obsolete
var.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:45:00 -03:00
Mauro Carvalho Chehab 13d6e9b653 i7core_edac: at remove, don't remove all pci devices at once
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:59 -03:00
Mauro Carvalho Chehab d88b85072f i7core_edac: Fix a bug when printing error counts with RDIMMs
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:59 -03:00
Mauro Carvalho Chehab d4c277957f i7core_edac: a few fixes for multiple mc's
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:59 -03:00
Mauro Carvalho Chehab 6c6aa3afdb i7core_edac: sanity check: print a warning if a mcelog is ignored
In thesis, the other mc controller should handle it.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:58 -03:00
Mauro Carvalho Chehab f47429494f i7core_edac: create one mc per socket/QPI
Instead of creating just one memory controller, create one per socket
(e. g. per Quick Link Path Interconnect).

This better reflects the Nehalem architecture.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:58 -03:00
Mauro Carvalho Chehab 66607706ce Dynamically allocate memory for PCI devices
Instead of using a static table assuming always 2 CPU sockets, allocate
space dynamically for Nehalem PCI devs.

This patch is part of a series of patches that changes i7core_edac to
allow more than 2 sockets and to properly report one memory controller
per socket.
2010-05-10 11:44:58 -03:00
Mauro Carvalho Chehab a55456f344 i7core: temporary workaround to allow it to compile against 2.6.30
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:58 -03:00
Mauro Carvalho Chehab 3a3bb4a647 i7core_edac: Improve corrected_error_counts output for RDIMM
Just cosmetics. instead of showing something like:

socket 0, channel 2dimm0: 1
dimm1: 0
dimm2: 0
socket 1, channel 2dimm0: 0
dimm1: 0
dimm2: 0

Show:

socket 0, channel 2 RDIMM0: 1 RDIMM1: 0 RDIMM2: 0
socket 0, channel 2 RDIMM0: 0 RDIMM1: 0 RDIMM2: 0

This is more synthetic and easier to parse.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:58 -03:00
Keith Mannthey bc2d7245ff i7core_edac: Probe on Xeons eariler
On the Xeon 55XX series cpus the pci deives are not exposed via acpi so
we much explicitly probe them to make the usable as a Linux PCI device.

This moves the detection of this state to before pci_register_driver is
called.  Its present position was not working on my systems, the driver
would complain about not finding a specific device.

This patch allows the driver to load on my systems.

Signed-off-by: Keith Mannthey <kmannth@us.ibm.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:57 -03:00
Mauro Carvalho Chehab 14d2c08343 i7core: Use registered memories per processor
Instead of assuming that the entire machine has either registered or
unregistered memories, do it at CPU socket based.

While here, fix a bug at i7core_mce_output_error(), where the we're
using m->cpu directly as if it would represent a socket. Instead, the
proper socket_id is given by cpu_data[m->cpu].phys_proc_id.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
---
2010-05-10 11:44:57 -03:00
Mauro Carvalho Chehab b4e8f0b6ea i7core_edac: Use Device 3 function 2 to report errors with RDIMM's
Nehalem and upper chipsets provide an special device that has corrected memory
error counters detected with registered dimms. This device is only seen if
there are registered memories plugged.

After this patch, on a machine fully equiped with RDIMM's, it will use the
Device 3 function 2 to count corrected errors instead on relying at mcelog.

For unregistered DIMMs, it will keep the old behavior, counting errors
via mcelog.

This patch were developed together with Keith Mannthey <kmannth@us.ibm.com>

Signed-off-by: Keith Mannthey <kmannth@us.ibm.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:56 -03:00
Keith Mannthey 61053fdedb i7core_edac: Fix ecc enable shift
From: Keith Mannthey <kmannth@us.ibm.com>

Simple correction to a shift value.
ECC_ENABLED is bit 4 of MC_STATUS, Dev 3 Fun 0 Offset 0x4c

This correctly identifies the state of the ECC at the machine.

Signed-off-by: Keith Mannthey <kmannth@us.ibm.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:56 -03:00
Mauro Carvalho Chehab 3ef288a983 i7core_edac: Print an error message if pci register fails
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:56 -03:00
Mauro Carvalho Chehab b990538a78 i7core_edac: CodingSyle fixes/cleanups
No functional changes.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:56 -03:00
Mauro Carvalho Chehab 4157d9f554 i7core_edac: fix error injection
There were two stupid error injection bugs introduced by wrong
cut-and-paste: one at socket store, and another at the error inject
register. The last one were causing the code to not work at all.

While here, adds debug messages to allow seeing what registers are being
set while sending error injection.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:55 -03:00
Mauro Carvalho Chehab 2068def56c i7core_edac: fix error codes for sysfs error injection interface
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:55 -03:00
Mauro Carvalho Chehab 276b824c30 i7core_edac: some fixes at error injection code
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:54 -03:00
Mauro Carvalho Chehab 17cb7b0cf7 i7core_edac: Some cleanups at displayed info
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:54 -03:00
Mauro Carvalho Chehab 086271a037 i7core: remove some uneeded noisy debug messages
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:54 -03:00
Mauro Carvalho Chehab 3a7dde7fcd i7core: add socket info at the debug msg
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:53 -03:00
Mauro Carvalho Chehab ec6df24c15 i7core: better document i7core_get_active_channels()
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:53 -03:00
Mauro Carvalho Chehab c77720b954 i7core: fix get_devices routine for Xeon55xx
i7core_get_devices() were preparet to get just the first found device of each type.
Due to that, on Xeon 55xx, only socket 1 were retrived.

Rework i7core_get_devices() to clean it and to properly support Xeon 55xx.

While here, fix a small typo.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:53 -03:00
Mauro Carvalho Chehab a639539fa2 i7core: enrich error information based on memory transaction type
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:53 -03:00
Mauro Carvalho Chehab c5d3452869 i7core: check if the memory error is fatal or non-fatal
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:53 -03:00
Mauro Carvalho Chehab 310cbb7284 i7core: fix probing on Xeon55xx
Xeon55xx fails to probe with this error message:

EDAC DEBUG: in drivers/edac/i7core_edac.c, line at 1660: MC: drivers/edac/i7core_edac.c: i7core_init()
EDAC i7core: Device not found: dev 00:00.0 PCI ID 8086:2c41
i7core_edac: probe of 0000:00:14.0 failed with error -22

This is due to the fact that, on Xeon35xx (and i7core), device 00.0 has
PCI ID 8086:2c40.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:52 -03:00
Mauro Carvalho Chehab f237fcf2b7 i7core_edac: some fixes at memory error parser
m->bank is not related to the memory bank but, instead, to the MCA Error
register bank. Fix it accordingly. While here, improves the comments for
Nehalem bank.

A later fix is needed, in order to get bank/rank information from MCA
error log.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:52 -03:00
Mauro Carvalho Chehab 8a2f118e3a i7core_edac: decode mcelog error and send it via edac interface
Enriches mcelog error by using the encoded information at MCE status and
misc registers (IA32_MCx_STATUS, IA32_MCx_MISC).

Some fixes are still needed here, in order to properly fill the EDAC
fields.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:52 -03:00
Mauro Carvalho Chehab ba6c5c62ee i7core_edac: maps all sockets as if ther are one MC controller
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:52 -03:00
Mauro Carvalho Chehab 67166af4ab i7core_edac: add support for more than one MC socket
Some Nehalem architectures have more than one MC socket. Socket 0 is
located at bus 255.

Currently, it is using up to 2 sockets, but increasing it to a larger
number is just a matter of increasing MAX_SOCKETS definition.

This seems to be required for properly support of Xeon 55xx.

Still needs testing with Xeon 55xx.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:51 -03:00
Mauro Carvalho Chehab d1fd4fb69e i7core_edac: Add a code to probe Xeon 55xx bus
This code changes the detection procedure of i7core_edac. Instead of
directly probing for MC registers, it probes for another register found
on Nehalem. If found, it tries to pick the first MC PCI BUS. This should
work fine with Xeon 35xx, but, on Xeon 55xx, this is at bus 254 and 255
that are not properly detected by the non-legacy PCI methods.

The new detection code scans specifically at buses 254 and 255 for the
Xeon 55xx devices.

This code has not tested yet. After working, a change at the code will
be needed, since the i7core is not yet ready for working with 2 sets of
MC.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:51 -03:00
Mauro Carvalho Chehab e9bd2e7379 i7core_edac: Adds write unlock to MC registers
The public Intel Xeon 5500 volume 2 datasheet describes, on page 53,
session 2.6.7 a register that can lock/unlock Memory Controller the
configuration register, called MC_CFG_CONTROL.

Adds support for it in the hope that software error injection would
work. With my tests with Xeon 35xx, there's still something missing.
With a program that does sequencial bit writes at dev 0.0, sometimes, it
produces error injection, after unblocking the MC_CFG_CONTROL (and,
sometimes, it just locks my testing machine).

I'll try later to discover by trial and error what's the register that
solves this issue on Xeon 35xx.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:50 -03:00
Mauro Carvalho Chehab d5381642ab i7core_edac: Add edac_mce glue
Adds a glue code to allow i7core to work with mcelog. With the glue,
i7core registers itself on edac_mce. At mce, when an error is detected,
it calls all registered drivers (in this case, i7core), for EDAC error
handling.

TODO: It currently just prints the MCE error log using about the same
      format as mce panic messages. The error message should be enhanced
      with mcelog userspace info and converted into the proper EDAC format,
      to feed the EDAC error counts.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:50 -03:00
Mauro Carvalho Chehab 41fcb7feed i7core_edac: CodingStyle fixes
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:48 -03:00
Mauro Carvalho Chehab eb94fc402f i7core_edac: fill csrows edac sysfs info
csrows is still fake, since we can't identify its representation with
Nehalem registers.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:48 -03:00
Mauro Carvalho Chehab 5566cb7c91 i7core_edac: Memory info fixes and preparation for properly filling cswrow data
Now, memory size is properly displayed:

    EDAC i7core: DOD Max limits: DIMMS: 2, 1-ranked, 8-banked
    EDAC i7core: DOD Max rows x colums = 0x4000 x 0x400
    EDAC i7core: Memory channel configuration:
    EDAC i7core: Ch0 phy rd0, wr0 (0x063f7c31): 2 ranks, UDIMMs
    EDAC i7core:    dimm 0 (0x00000288) 1024 Mb offset: 0, numbank: 8,
                    numrank: 1, numrow: 0x4000, numcol: 0x400
    EDAC i7core:    dimm 1 (0x00001288) 1024 Mb offset: 4, numbank: 8,
                    numrank: 1, numrow: 0x4000, numcol: 0x400
    EDAC i7core: Ch1 phy rd1, wr1 (0x063f7c31): 2 ranks, UDIMMs
    EDAC i7core:    dimm 0 (0x00000288) 1024 Mb offset: 0, numbank: 8,
                    numrank: 1, numrow: 0x4000, numcol: 0x400
    EDAC i7core: Ch2 phy rd3, wr3 (0x063f7c31): 2 ranks, UDIMMs
    EDAC i7core:    dimm 0 (0x00000288) 1024 Mb offset: 0, numbank: 8,
                    numrank: 1, numrow: 0x4000, numcol: 0x400

Still, as the way to retrieve csrows info is not known, it does a
mapping of what's available to csrows basic unit at edac core.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:48 -03:00
Mauro Carvalho Chehab 854d334997 i7core_edac: Get more info about the memory DIMMs
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:48 -03:00
Mauro Carvalho Chehab 7dd6953c5f i7core_edac: Add more information about each active dimm
Thanks-to: Aristeu Rozanski <aris@redhat.com> for part of the code

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:47 -03:00
Mauro Carvalho Chehab b7c761512c i7core_edac: Improve error handling
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:47 -03:00
Mauro Carvalho Chehab 1c6fed808f i7core_edac: Properly fill struct csrow_info
Thanks-to: Aristeu Rozanski <aris@redhat.com> for part of the code

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:47 -03:00
Mauro Carvalho Chehab ef708b53b9 i7core_edac: Add additional tests for error detection
Properly check the number of channels and improve probing error detection

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:47 -03:00
Mauro Carvalho Chehab 442305b152 i7core_edac: Add a memory check routine, based on device 3 function 4
This function appears only on Xeon 5500 datasheet. Yet, testing with a
Xeon 3503 showed that this is also implemented on other Nehalem
processors.

At the first read, MC_TEST_ERR_RCV1 and MC_TEST_ERR_RCV0 can contain any
value. Modify CE error logic to update the error count only after the
second read.

An alternative approach would be to do a write at rcv0 and rcv1
registers, but it seemed better to keep they untouched, since BIOS might
eventually assume that they are exclusive for their usage.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:46 -03:00
Mauro Carvalho Chehab 87d1d272ba i7core_edac: need mci->edac_check, otherwise module removal doesn't work
There are some locking troubles with edac_core: if you don't declare an
edac_check, module may suffer from soft lock.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:46 -03:00
Mauro Carvalho Chehab 7b029d03c3 i7core_edac: A few fixes at error injection code
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:46 -03:00
Mauro Carvalho Chehab f122a89222 i7core_edac: Show read/write virtual/physical channel association
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:46 -03:00
Mauro Carvalho Chehab 8f33190757 i7core_edac: Registers all supported MC functions
Now, it will try to register on all supported Memory Controller
functions.

It should be noticed that dev3, function 2 is present only on chips with
Registered DIMM's, according to the datasheet. So, the driver doesn't
return -ENODEV is all functions but this one were successfully
registered and enabled:

    EDAC i7core: Registered device 8086:2c18 fn=3 0
    EDAC i7core: Registered device 8086:2c19 fn=3 1
    EDAC i7core: Device not found: PCI ID 8086:2c1a (dev 3, func 2)
    EDAC i7core: Registered device 8086:2c1c fn=3 4
    EDAC i7core: Registered device 8086:2c20 fn=4 0
    EDAC i7core: Registered device 8086:2c21 fn=4 1
    EDAC i7core: Registered device 8086:2c22 fn=4 2
    EDAC i7core: Registered device 8086:2c23 fn=4 3
    EDAC i7core: Registered device 8086:2c28 fn=5 0
    EDAC i7core: Registered device 8086:2c29 fn=5 1
    EDAC i7core: Registered device 8086:2c2a fn=5 2
    EDAC i7core: Registered device 8086:2c2b fn=5 3
    EDAC i7core: Registered device 8086:2c30 fn=6 0
    EDAC i7core: Registered device 8086:2c31 fn=6 1
    EDAC i7core: Registered device 8086:2c32 fn=6 2
    EDAC i7core: Registered device 8086:2c33 fn=6 3
    EDAC i7core: Driver loaded.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:45 -03:00
Mauro Carvalho Chehab 0b2b7b7ec0 i7core_edac: Add more status functions to EDAC driver
This patch were co-authored with Aristeu Rozanski.

Signed-off-by: Aristeu Sergio <arozansk@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:45 -03:00
Mauro Carvalho Chehab 194a40feab i7core_edac: Add error insertion code for Nehalem
Implements set_inject_error() with the low-level code needed to inject
memory errors at Nehalem, and adds some sysfs nodes to allow error injection

The next patch will add an API for error injection.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:45 -03:00
Mauro Carvalho Chehab a0c36a1f0f i7core_edac: Add an EDAC memory controller driver for Nehalem chipsets
This driver is meant to support i7 core/i7core extreme desktop
processors and Xeon 35xx/55xx series with integrated memory controller.
It is likely that it can be expanded in the future to work with other
processor series based at the same Memory Controller design.

For now, it has just a few MCH status reads.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2010-05-10 11:44:45 -03:00