AMD Zen-based systems report memory error addresses through machine
check banks representing Unified Memory Controllers (UMCs) in the form
of UMC relative "normalized" addresses. A normalized address must be
converted to a system physical address to be usable by the OS.
Future AMD platforms will provide a UEFI PRM module that implements a
number of address translation PRM handlers. This will provide an
interface for the OS to call platform specific code without requiring
the use of SMM or other heavy firmware operations.
Add support for the normalized to system physical address translation
PRM handler in the AMD Address Translation Library and prefer it over
native code if available. The GUID and parameter buffer structure are
specific to the normalized to system physical address handler provided
by the address translation PRM module included in future AMD systems.
The address translation PRM module is documented in chapter 22 of the
publicly available "AMD Family 1Ah Models 00h–0Fh and Models 10h–1Fh
ACPI v6.5 Porting Guide".
[ bp: Massage commit message. ]
Signed-off-by: John Allen <john.allen@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/r/20240730151731.15363-3-john.allen@amd.com
non-power-of-2 denormalization in the sense that certain bits of the
system physical address cannot be reconstructed from the normalized
address reported by the RAS hardware. Add support for handling such
addresses
- Switch the EDAC drivers to the new Intel CPU model defines
- The usual fixes and cleanups all over the place
-----BEGIN PGP SIGNATURE-----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=kLNX
-----END PGP SIGNATURE-----
Merge tag 'edac_updates_for_v6.11' of git://git.kernel.org/pub/scm/linux/kernel/git/ras/ras
Pull EDAC updates from Borislav Petkov:
- The AMD memory controllers data fabric version 4.5 supports
non-power-of-2 denormalization in the sense that certain bits of the
system physical address cannot be reconstructed from the normalized
address reported by the RAS hardware. Add support for handling such
addresses
- Switch the EDAC drivers to the new Intel CPU model defines
- The usual fixes and cleanups all over the place
* tag 'edac_updates_for_v6.11' of git://git.kernel.org/pub/scm/linux/kernel/git/ras/ras:
EDAC: Add missing MODULE_DESCRIPTION() macros
EDAC/dmc520: Use devm_platform_ioremap_resource()
EDAC/igen6: Add Intel Arrow Lake-U/H SoCs support
RAS/AMD/FMPM: Use atl internal.h for INVALID_SPA
RAS/AMD/ATL: Implement DF 4.5 NP2 denormalization
RAS/AMD/ATL: Validate address map when information is gathered
RAS/AMD/ATL: Expand helpers for adding and removing base and hole
RAS/AMD/ATL: Read DRAM hole base early
RAS/AMD/ATL: Add amd_atl pr_fmt() prefix
RAS/AMD/ATL: Add a missing module description
EDAC, i10nm: make skx_common.o a separate module
EDAC/skx: Switch to new Intel CPU model defines
EDAC/sb_edac: Switch to new Intel CPU model defines
EDAC, pnd2: Switch to new Intel CPU model defines
EDAC/i10nm: Switch to new Intel CPU model defines
EDAC/ghes: Add missing newline to pr_info() statement
RAS/AMD/ATL: Add missing newline to pr_info() statement
EDAC/thunderx: Remove unused struct error_syndrome
* ras/edac-amd-atl:
RAS/AMD/FMPM: Use atl internal.h for INVALID_SPA
RAS/AMD/ATL: Implement DF 4.5 NP2 denormalization
RAS/AMD/ATL: Validate address map when information is gathered
RAS/AMD/ATL: Expand helpers for adding and removing base and hole
RAS/AMD/ATL: Read DRAM hole base early
RAS/AMD/ATL: Add amd_atl pr_fmt() prefix
RAS/AMD/ATL: Add a missing module description
* ras/edac-misc:
EDAC: Add missing MODULE_DESCRIPTION() macros
EDAC/dmc520: Use devm_platform_ioremap_resource()
EDAC/igen6: Add Intel Arrow Lake-U/H SoCs support
EDAC, i10nm: make skx_common.o a separate module
EDAC/skx: Switch to new Intel CPU model defines
EDAC/sb_edac: Switch to new Intel CPU model defines
EDAC, pnd2: Switch to new Intel CPU model defines
EDAC/i10nm: Switch to new Intel CPU model defines
EDAC/ghes: Add missing newline to pr_info() statement
RAS/AMD/ATL: Add missing newline to pr_info() statement
EDAC/thunderx: Remove unused struct error_syndrome
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
The currently used normalized address format is not applicable to all
MI300 systems. This leads to incorrect results during address
translation.
Drop the fixed layout and construct the normalized address from system
settings.
Fixes: 87a6123753 ("RAS/AMD/ATL: Add MI300 DRAM to normalized address translation support")
Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Cc: <stable@kernel.org>
Link: https://lore.kernel.org/r/20240607-mi300-dram-xl-fix-v1-2-2f11547a178c@amd.com
Apply the SID bits to the correct offset in the Bank value. Do this in
the temporary value so they don't need to be masked off later.
Fixes: 87a6123753 ("RAS/AMD/ATL: Add MI300 DRAM to normalized address translation support")
Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Cc: <stable@kernel.org>
Link: https://lore.kernel.org/r/20240607-mi300-dram-xl-fix-v1-1-2f11547a178c@amd.com
Both the AMD ATL and the FMPM driver define INVALID_SPA. Include the
definition from the ATL internal.h header in the FMPM driver.
Signed-off-by: John Allen <john.allen@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/r/20240606203313.51197-7-john.allen@amd.com
Unlike with previous Data Fabric versions, with Data Fabric 4.5
non-power-of-2 denormalization, there are bits of the system physical
address that can't be fully reconstructed from the normalized address.
To determine the proper combination of missing system physical address
bits, iterate through each possible combination of these bits, normalize
the resulting system physical address, and compare to the original
address that is being translated. If the addresses match, then the
correct permutation of bits has been found.
Signed-off-by: John Allen <john.allen@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Reviewed-by: Yazen Ghannam <yazen.ghannam@amd.com>
Link: https://lore.kernel.org/r/20240606203313.51197-6-john.allen@amd.com
Validate address maps at the time the information is gathered as the
address map will not change during translation.
Signed-off-by: John Allen <john.allen@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Reviewed-by: Yazen Ghannam <yazen.ghannam@amd.com>
Link: https://lore.kernel.org/r/20240606203313.51197-5-john.allen@amd.com
The ret_addr field in struct addr_ctx contains the intermediate value of
the returned address as it passes through multiple steps in the
translation process. Currently, adding the DRAM base and legacy hole
is only done once, so it operates directly on the intermediate value.
However, for DF 4.5 non-power-of-2 denormalization, adding and removing
the DRAM base and legacy hole needs to be done for multiple temporary
address values. During this process, the intermediate value should not be
lost so the ret_addr value can't be reused.
Update the existing 'add' helper to operate on an arbitrary address
and introduce a new 'remove' helper to do the inverse operations.
Signed-off-by: John Allen <john.allen@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Reviewed-by: Yazen Ghannam <yazen.ghannam@amd.com>
Link: https://lore.kernel.org/r/20240606203313.51197-4-john.allen@amd.com
Read DRAM hole base when constructing the address map as the value will
not change during run time.
Signed-off-by: John Allen <john.allen@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Reviewed-by: Yazen Ghannam <yazen.ghannam@amd.com>
Link: https://lore.kernel.org/r/20240606203313.51197-3-john.allen@amd.com
Add a missing newline character even if printk() adds newlines to
non-\n-terminated strings because in the unlikely case a KERN_CONT print
statement is added after the unterminated statement, the two will get
glued together which is not the expected behavior.
[ bp: Rewrite commit message. ]
Signed-off-by: Vasyl Gomonovych <gomonovych@gmail.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/r/20240517215452.2020680-1-gomonovych@gmail.com
A new helper was introduced for RAS modules to be able to get the RAS
subsystem debugfs root directory. The helper is defined in debugfs.c
which is only built when CONFIG_DEBUG_FS=y.
However, it's possible that the modules would include debugfs support
for optional functionality. One current example is the fmpm module. In
this case, a build error will occur when CONFIG_RAS_FMPM is selected and
CONFIG_DEBUG_FS=n.
Add an inline helper function stub for the CONFIG_DEBUG_FS=n case as the
fmpm module can function without the debugfs functionality too.
Fixes: 9d2b6fa09d ("RAS: Export helper to get ras_debugfs_dir")
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=218640
Reported-by: anthony s. knowles <akira.2020@protonmail.com>
Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Tested-by: anthony s. knowles <akira.2020@protonmail.com>
Link: https://lore.kernel.org/r/20240325183755.776-1-bp@alien8.de
Currently, the size of the locally cached FRU record structures is
based on the module parameter "max_nr_entries".
This creates issues when restoring records if a user changes the
parameter.
If the number of entries is reduced, then old, larger records will not
be restored. The opportunity to take action on the saved data is missed.
Also, new records will be created and written to storage, even as the old
records remain in storage, resulting in wasted space.
If the number of entries is increased, then the length of the old,
smaller records will not be adjusted. This causes a checksum failure
which leads to the old record being cleared from storage. Again this
results in another missed opportunity for action on the saved data.
Allocate the temporary record with the maximum possible size based on
the current maximum number of supported entries (255). This allows the
ERST read operation to succeed if max_nr_entries has been increased.
Warn the user if a saved record exceeds the expected size and fail to
load the module. This allows the user to adjust the module parameter
without losing data or the opportunity to restore larger records.
Increase the size of a saved record up to the current max_rec_len. The
checksum will be recalculated, and the updated record will be written to
storage.
Fixes: 6f15e617cc ("RAS: Introduce a FRU memory poison manager")
Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Tested-by: Muralidhara M K <muralidhara.mk@amd.com>
Link: https://lore.kernel.org/r/20240319113322.280096-3-yazen.ghannam@amd.com
An old, invalid record should be cleared and skipped.
Currently, the record is cleared in ERST, but it is not skipped. This
leads to a NULL pointer dereference when attempting to copy the old
record to the new record.
Continue the loop after clearing an old, invalid record to skip it.
Fixes: 6f15e617cc ("RAS: Introduce a FRU memory poison manager")
Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Tested-by: Muralidhara M K <muralidhara.mk@amd.com>
Link: https://lore.kernel.org/r/20240319113322.280096-2-yazen.ghannam@amd.com
collects and manages previously encountered hw errors in order to
save them to persistent storage across reboots. Previously recorded
errors are "replayed" upon reboot in order to poison memory which has
caused said errors in the past.
The main use case is stacked, on-chip memory which cannot simply be
replaced so poisoning faulty areas of it and thus making them
inaccessible is the only strategy to prolong its lifetime.
- Add an AMD address translation library glue which converts the
reported addresses of hw errors into system physical addresses in
order to be used by other subsystems like memory failure, for
example. Add support for MI300 accelerators to that library.
- igen6: Add support for Alder Lake-N SoC
- i10nm: Add Grand Ridge support
- The usual fixlets and cleanups
-----BEGIN PGP SIGNATURE-----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=FR2l
-----END PGP SIGNATURE-----
Merge tag 'edac_updates_for_v6.9' of git://git.kernel.org/pub/scm/linux/kernel/git/ras/ras
Pull EDAC updates from Borislav Petkov:
- Add a FRU (Field Replaceable Unit) memory poison manager which
collects and manages previously encountered hw errors in order to
save them to persistent storage across reboots. Previously recorded
errors are "replayed" upon reboot in order to poison memory which has
caused said errors in the past.
The main use case is stacked, on-chip memory which cannot simply be
replaced so poisoning faulty areas of it and thus making them
inaccessible is the only strategy to prolong its lifetime.
- Add an AMD address translation library glue which converts the
reported addresses of hw errors into system physical addresses in
order to be used by other subsystems like memory failure, for
example. Add support for MI300 accelerators to that library.
- igen6: Add support for Alder Lake-N SoC
- i10nm: Add Grand Ridge support
- The usual fixlets and cleanups
* tag 'edac_updates_for_v6.9' of git://git.kernel.org/pub/scm/linux/kernel/git/ras/ras:
EDAC/versal: Convert to platform remove callback returning void
RAS/AMD/FMPM: Fix off by one when unwinding on error
RAS/AMD/FMPM: Add debugfs interface to print record entries
RAS/AMD/FMPM: Save SPA values
RAS: Export helper to get ras_debugfs_dir
RAS/AMD/ATL: Fix bit overflow in denorm_addr_df4_np2()
RAS: Introduce a FRU memory poison manager
RAS/AMD/ATL: Add MI300 row retirement support
Documentation: Move RAS section to admin-guide
EDAC/versal: Make the bit position of injected errors configurable
EDAC/i10nm: Add Intel Grand Ridge micro-server support
EDAC/igen6: Add one more Intel Alder Lake-N SoC support
RAS/AMD/ATL: Add MI300 DRAM to normalized address translation support
RAS/AMD/ATL: Fix array overflow in get_logical_coh_st_fabric_id_mi300()
RAS/AMD/ATL: Add MI300 support
Documentation: RAS: Add index and address translation section
EDAC/amd64: Use new AMD Address Translation Library
RAS: Introduce AMD Address Translation Library
EDAC/synopsys: Convert to devm_platform_ioremap_resource()
Decrement the index variable i before the first iteration when freeing
the remaining elements on error. Depending on where this fails it could
free something from one element beyond the end of the fru_records[]
array.
[ bp: Massage commit message. ]
Fixes: 6f15e617cc ("RAS: Introduce a FRU memory poison manager")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/r/6fdec71a-846b-4cd0-af69-e5f6cd12f4f6@moroto.mountain
It is helpful to see the saved record entries during run time in
human-readable format. This is useful for testing during module
development. It can also be used by system admins to quickly and easily
see the state of the system.
Provide a sequential file in debugfs to print fields of interest from
the FRU records and their entries.
Don't fail to load the module if the debugfs interface is not available.
This is a convenience feature which does not affect other module
functionality.
The new interface reads the record entries and should hold the mutex.
Expand the mutex code comment to clarify when it should be held.
Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/r/20240301143748.854090-4-yazen.ghannam@amd.com
The system physical address (SPA) of an error is not a stable value. It
will change depending on the location of the memory: parts can be
swapped. And it will change depending on memory topology: NUMA nodes
and/or interleaving can be adjusted.
Therefore, the SPA value is not part of the "FRU Memory Poison" record
format. And it will not be saved to persistent storage.
However, the SPA values can be helpful during debug and for system
admins during run time.
Save the SPA values in a separate structure. This is updated when
records are restored and when new errors are saved.
[ bp: Make error messages more user friendly and add and correct
comments. ]
Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/r/20240301143748.854090-3-yazen.ghannam@amd.com
Export a getter instead of the debugfs node directly so that, other
in-tree-only RAS modules can use it.
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
Link: https://lore.kernel.org/r/20240301143748.854090-2-yazen.ghannam@amd.com
The hash_pa8 and hashed_bit values in denorm_addr_df4_np2() are
currently defined as u8 types. These variables represent single bits.
'hash_pa8' is set based on logical AND operations using masks with more
than 8 bits. So the calculated value will not fit in this variable. It
will always be '0'. The 'hash_pa8' check later in the function will fail
which produces incorrect results for some cases.
Change these variables to bool type. This clarifies that they are
single bit values. Also, this allows the compiler to ensure they hold
the proper results. Remove an unnecessary shift operation.
[ bp: Remove the unnecessary brackets in the else-branch of the
hash_pa8 assignment. ]
Fixes: 3f3174996b ("RAS: Introduce AMD Address Translation Library")
Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/r/20240222165449.23582-1-yazen.ghannam@amd.com
Memory errors are an expected occurrence on systems with high memory
density. Generally, errors within a small number of unique physical
locations are acceptable, based on manufacturer and/or admin policy.
During run time, memory with errors may be retired so it is no longer
used by the system. This is done in mm through page poisoning, and the
effect will remain until the system is restarted.
If a memory location is consistently faulty, then the same run time
error handling may occur in the next reboot cycle, leading to
terminating jobs due to that already known bad memory. This could be
prevented if information from the previous boot was not lost.
Some add-in cards with driver-managed memory have on-board persistent
storage. Their driver saves memory error information to the persistent
storage during run time. The information is then restored after reset,
and known bad memory will be retired before the hardware is used.
A running log of bad memory locations is kept across multiple resets.
A similar solution is desirable for CPUs. However, this solution should
leverage industry-standard components as much as possible, rather than
a bespoke platform driver.
Two components are needed: a record format and a persistent storage
interface.
Implement a new module to manage the record formats on persistent
storage. Use the requirements for an AMD MI300-based system to start.
Vendor- and platform-specific details can be abstracted later as needed.
[ bp: Massage commit message and code, squash 30-ish more fixes from
Yazen and me. ]
Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
Co-developed-by: <naveenkrishna.chatradhi@amd.com>
Signed-off-by: <naveenkrishna.chatradhi@amd.com>
Co-developed-by: <muralidhara.mk@amd.com>
Signed-off-by: <muralidhara.mk@amd.com>
Tested-by: <sathyapriya.k@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/r/20240214033516.1344948-3-yazen.ghannam@amd.com
DRAM row retirement depends on model-specific information that is best
done within the AMD Address Translation Library.
Export a generic wrapper function for other modules to use. Add any
model-specific helpers here.
Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/r/20240214033516.1344948-2-yazen.ghannam@amd.com
Zen-based AMD systems report DRAM ECC errors through Unified Memory
Controller (UMC) MCA banks. The value provided in MCA_ADDR is
a "normalized" address which represents the UMC's view of its managed
memory. The normalized address must be translated to a system physical
address for software to take action.
MI300 systems, uniquely, do not provide a normalized address in MCA_ADDR
for DRAM ECC errors. Rather, the "DRAM" address is reported. This value
includes identifiers for the bank, row, column, pseudochannel and stack
of the memory location.
The DRAM address must be converted to a normalized address in order to
be further translated to a system physical address.
Add helper functions to do the DRAM to normalized translation for MI300
systems. The method is based on the fixed hardware layout of the on-chip
memory.
[ bp: Massage commit message, decapitalize some, rename function. ]
Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
Co-developed-by: Muralidhara M K <muralidhara.mk@amd.com>
Signed-off-by: Muralidhara M K <muralidhara.mk@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Tested-by: Muralidhara M K <muralidhara.mk@amd.com>
Link: https://lore.kernel.org/r/20240131165732.88297-1-yazen.ghannam@amd.com
Check against ARRAY_SIZE() which is the number of elements instead of
sizeof() which is the number of bytes.
Fixes: 453f0ae797 ("RAS/AMD/ATL: Add MI300 support")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/r/279c8b5e-6c00-467a-9071-9c67926abea4@moroto.mountain
AMD MI300 systems include on-die HBM3 memory and a unique topology. And
they fall under Data Fabric version 4.5 in overall design.
Generally, topology information (IDs, etc.) is gathered from Data Fabric
registers. However, the unique topology for MI300 means that some
topology information is fixed in hardware and follows arbitrary
mappings. Furthermore, not all hardware instances are software-visible,
so register accesses must be adjusted.
Recognize and add helper functions for the new MI300 interleave modes.
Add lookup tables for fixed values where appropriate. Adjust how Die and
Node IDs are found and used.
Also, fix some register bitmasks that were mislabeled.
Signed-off-by: Muralidhara M K <muralidhara.mk@amd.com>
Co-developed-by: Yazen Ghannam <yazen.ghannam@amd.com>
Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/r/20240128155950.1434067-1-yazen.ghannam@amd.com
AMD Zen-based systems report memory errors through Machine Check banks
representing Unified Memory Controllers (UMCs). The address value
reported for DRAM ECC errors is a "normalized address" that is relative
to the UMC. This normalized address must be converted to a system
physical address to be usable by the OS.
Support for this address translation was introduced to the MCA subsystem
with Zen1 systems. The code was later moved to the AMD64 EDAC module,
since this was the only user of the code at the time.
However, there are uses for this translation outside of EDAC. The system
physical address can be used in MCA for preemptive page offlining as done
in some MCA notifier functions. Also, this translation is needed as the
basis of similar functionality needed for some CXL configurations on AMD
systems.
Introduce a common address translation library that can be used for
multiple subsystems including MCA, EDAC, and CXL.
Include support for UMC normalized to system physical address
translation for current CPU systems.
The Data Fabric Indirect register access offsets and one of the register
fields were changed. Default to the current offsets and register field
definition. And fallback to the older values if running on a "legacy"
system.
Provide built-in code to facilitate the loading and unloading of the
library module without affecting other modules or built-in code.
Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/r/20240123041401.79812-2-yazen.ghannam@amd.com
Documentation/filesystems/seq_file.rst describes the possible return
values from a "show()" function used by single_open().
show_trace() returns the value of "trace_count". This could be
interpreted as "SEQ_SKIP", or just confuse the calling function.
Change to just return "0" to avoid confusing anyone reading this code
and possibly using as a template. Reading "daemon_active" was never
an intended use case.
Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lore.kernel.org/r/20221018165900.109029-1-tony.luck@intel.com
A large scale study of memory errors on Intel systems in data centers
showed that aggressively taking pages with corrected errors offline is
the best strategy of using corrected errors as a predictor of future
uncorrected errors.
Set the threshold to "2" on Intel systems. AMD guidance is that this is
not necessary for their systems.
Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Reviewed-by: Yazen Ghannam <yazen.ghannam@amd.com>
Link: https://lore.kernel.org/r/20220607212015.175591-1-tony.luck@intel.com
Link: https://lore.kernel.org/r/YulOZ/Eso0bwUcC4@agluck-desk3.sc.intel.com
ce_add_elem() uses different return values to signal a result from
adding an element to the collector. Commit in Fixes: broke the case
where the element being added is not found in the array. Correct that.
[ bp: Rewrite commit message, add kernel-doc comments. ]
Fixes: de0e0624d8 ("RAS/CEC: Check count_threshold unconditionally")
Signed-off-by: William Roche <william.roche@oracle.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: <stable@vger.kernel.org>
Link: https://lkml.kernel.org/r/1617722939-29670-1-git-send-email-william.roche@oracle.com
late_initcall() expects a function that returns an integer. Update the
function signature to match.
[ bp: Massage commit message into proper sentences. ]
Fixes: 9554bfe403 ("x86/mce: Convert the CEC to use the MCE notifier")
Signed-off-by: Luca Stefani <luca.stefani.ge1@gmail.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Reviewed-by: Sami Tolvanen <samitolvanen@google.com>
Tested-by: Sami Tolvanen <samitolvanen@google.com>
Link: https://lkml.kernel.org/r/20200805095708.83939-1-luca.stefani.ge1@gmail.com
If the handler took any action to log or deal with the error, set a bit
in mce->kflags so that the default handler on the end of the machine
check chain can see what has been done.
Get rid of NOTIFY_STOP returns. Make the EDAC and dev-mcelog handlers
skip over errors already processed by CEC.
Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Tested-by: Tony Luck <tony.luck@intel.com>
Link: https://lkml.kernel.org/r/20200214222720.13168-5-tony.luck@intel.com
The CEC code has its claws in a couple of routines in mce/core.c.
Convert it to just register itself on the normal MCE notifier chain.
[ bp: Make cec_add_elem() and cec_init() static. ]
Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Tested-by: Tony Luck <tony.luck@intel.com>
Link: https://lkml.kernel.org/r/20200214222720.13168-3-tony.luck@intel.com
In addition, the 0day bot reported this build error:
>> drivers/ras/debugfs.c:10:5: error: redefinition of 'ras_userspace_consumers'
int ras_userspace_consumers(void)
^~~~~~~~~~~~~~~~~~~~~~~
In file included from drivers/ras/debugfs.c:3:0:
include/linux/ras.h:14:19: note: previous definition of 'ras_userspace_consumers' was here
static inline int ras_userspace_consumers(void) { return 0; }
^~~~~~~~~~~~~~~~~~~~~~~
for a riscv-specific .config where CONFIG_DEBUG_FS is not set. Fix all
that by making debugfs.o depend on that define.
[ bp: Rewrite commit message. ]
Reported-by: kbuild test robot <lkp@intel.com>
Signed-off-by: Valdis Kletnieks <valdis.kletnieks@vt.edu>
Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: Tony Luck <tony.luck@intel.com>
Cc: linux-edac@vger.kernel.org
Cc: x86@kernel.org
Link: http://lkml.kernel.org/r/7053.1565218556@turing-police
When building with C=2 and/or W=1, legitimate warnings are issued about
missing prototypes:
CHECK drivers/ras/debugfs.c
drivers/ras/debugfs.c:4:15: warning: symbol 'ras_debugfs_dir' was not declared. Should it be static?
drivers/ras/debugfs.c:8:5: warning: symbol 'ras_userspace_consumers' was not declared. Should it be static?
drivers/ras/debugfs.c:38:12: warning: symbol 'ras_add_daemon_trace' was not declared. Should it be static?
drivers/ras/debugfs.c:54:13: warning: symbol 'ras_debugfs_init' was not declared. Should it be static?
CC drivers/ras/debugfs.o
drivers/ras/debugfs.c:8:5: warning: no previous prototype for 'ras_userspace_consumers' [-Wmissing-prototypes]
8 | int ras_userspace_consumers(void)
| ^~~~~~~~~~~~~~~~~~~~~~~
drivers/ras/debugfs.c:38:12: warning: no previous prototype for 'ras_add_daemon_trace' [-Wmissing-prototypes]
38 | int __init ras_add_daemon_trace(void)
| ^~~~~~~~~~~~~~~~~~~~
drivers/ras/debugfs.c:54:13: warning: no previous prototype for 'ras_debugfs_init' [-Wmissing-prototypes]
54 | void __init ras_debugfs_init(void)
| ^~~~~~~~~~~~~~~~
Provide the proper includes.
[ bp: Take care of the same warnings for cec.c too. ]
Signed-off-by: Valdis Kletnieks <valdis.kletnieks@vt.edu>
Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: Tony Luck <tony.luck@intel.com>
Cc: linux-edac@vger.kernel.org
Cc: x86@kernel.org
Link: http://lkml.kernel.org/r/7168.1565218769@turing-police
The pfn and array files in (debugfs)/ras/cec are intended for debugging
the CEC code itself. They are not needed on production systems, so the
default setting for this CONFIG option is "n".
[ bp: Have it with less ifdeffery by using IS_ENABLED(). ]
Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
When dumping the array elements, print them in the following format:
[ PFN | generation in binary | count ]
to be perfectly clear what all those sections are.
Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: Tony Luck <tony.luck@intel.com>
Cc: linux-edac <linux-edac@vger.kernel.org>
... which is the better, more-fitting name anyway.
Tony:
- make action_threshold u64 due to debugfs accessors expecting u64.
- rename the remaining: s/count_threshold/action_threshold/g
Co-developed-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: linux-edac <linux-edac@vger.kernel.org>
Check the elements order in the array after every insertion.
Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: Tony Luck <tony.luck@intel.com>
Cc: linux-edac <linux-edac@vger.kernel.org>
Free the array page if a failure is encountered while creating the
debugfs nodes.
Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: Tony Luck <tony.luck@intel.com>
Cc: linux-edac <linux-edac@vger.kernel.org>
When the value requested doesn't match the allowed (min,max) range,
the @data buffer should not be modified with the invalid value because
reading "decay_interval" shows it otherwise as if the previous write
succeeded.
Move the data write after the check.
Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: Tony Luck <tony.luck@intel.com>
Cc: linux-edac <linux-edac@vger.kernel.org>
The count_threshold should be checked unconditionally, after insertion
too, so that a count_threshold value of 1 can cause an immediate
offlining. I.e., offline the page on the *first* error encountered.
Add comments to make it clear what cec_add_elem() does, while at it.
Reported-by: WANG Chao <chao.wang@ucloud.cn>
Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: Tony Luck <tony.luck@intel.com>
Cc: linux-edac@vger.kernel.org
Link: https://lkml.kernel.org/r/20190418034115.75954-3-chao.wang@ucloud.cn
When inserting random PFNs for debugging the CEC through
(debugfs)/ras/cec/pfn, depending on the return value of pfn_set(),
multiple values get inserted per a single write.
That is because simple_attr_write() interprets a retval of 0 as
success and claims the whole input. However, pfn_set() returns the
cec_add_elem() value, which, if > 0 and smaller than the whole input
length, makes glibc continue issuing the write syscall until there's
input left:
pfn_set
simple_attr_write
debugfs_attr_write
full_proxy_write
vfs_write
ksys_write
do_syscall_64
entry_SYSCALL_64_after_hwframe
leading to those repeated calls.
Return 0 to fix that.
Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: Tony Luck <tony.luck@intel.com>
Cc: linux-edac <linux-edac@vger.kernel.org>
cec_timer_fn() is a timer callback which reads ce_arr.array[] and
updates its decay values. However, it runs in interrupt context and the
mutex protection the CEC uses for that array, is inadequate. Convert the
used timer to a workqueue to keep the tasks the CEC performs preemptible
and thus low-prio.
[ bp: Rewrite commit message.
s/timer/decay/gi to make it agnostic as to what facility is used. ]
Fixes: 011d826111 ("RAS: Add a Corrected Errors Collector")
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Tony Luck <tony.luck@intel.com>
Cc: linux-edac <linux-edac@vger.kernel.org>
Cc: <stable@vger.kernel.org>
Link: https://lkml.kernel.org/r/20190416213351.28999-2-xiyou.wangcong@gmail.com
Switch to using Donald Knuth's binary search algorithm (The Art of
Computer Programming, vol. 3, section 6.2.1). This should've been done
from the very beginning but the author must've been smoking something
very potent at the time.
The problem with the current one was that it would return the wrong
element index in certain situations:
https://lkml.kernel.org/r/CAM_iQpVd02zkVJ846cj-Fg1yUNuz6tY5q1Vpj4LrXmE06dPYYg@mail.gmail.com
and the noodling code after the loop was fishy at best.
So switch to using Knuth's binary search. The final result is much
cleaner and straightforward.
Fixes: 011d826111 ("RAS: Add a Corrected Errors Collector")
Reported-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: Tony Luck <tony.luck@intel.com>
Cc: linux-edac <linux-edac@vger.kernel.org>
Cc: <stable@vger.kernel.org>
