- Add support for parsing the Coherent Device Attribute Table (CDAT)
- Add support for calculating a platform CXL QoS class from CDAT data
- Unify the tracing of EFI CXL Events with native CXL Events.
- Add Get Timestamp support
- Miscellaneous cleanups and fixups
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQSbo+XnGs+rwLz9XGXfioYZHlFsZwUCZaHVvAAKCRDfioYZHlFs
Z3sCAQDPHSsHmj845k4lvKbWjys3eh78MKKEFyTXLQgYhOlsGAEAigQY2ZiSum52
nwdIgpOOADNt0Iq6yXuLsmn9xvY9bAU=
=HjCl
-----END PGP SIGNATURE-----
Merge tag 'cxl-for-6.8' of git://git.kernel.org/pub/scm/linux/kernel/git/cxl/cxl
Pull CXL (Compute Express Link) updates from Dan Williams:
"The bulk of this update is support for enumerating the performance
capabilities of CXL memory targets and connecting that to a platform
CXL memory QoS class. Some follow-on work remains to hook up this data
into core-mm policy, but that is saved for v6.9.
The next significant update is unifying how CXL event records (things
like background scrub errors) are processed between so called
"firmware first" and native error record retrieval. The CXL driver
handler that processes the record retrieved from the device mailbox is
now the handler for that same record format coming from an EFI/ACPI
notification source.
This also contains miscellaneous feature updates, like Get Timestamp,
and other fixups.
Summary:
- Add support for parsing the Coherent Device Attribute Table (CDAT)
- Add support for calculating a platform CXL QoS class from CDAT data
- Unify the tracing of EFI CXL Events with native CXL Events.
- Add Get Timestamp support
- Miscellaneous cleanups and fixups"
* tag 'cxl-for-6.8' of git://git.kernel.org/pub/scm/linux/kernel/git/cxl/cxl: (41 commits)
cxl/core: use sysfs_emit() for attr's _show()
cxl/pci: Register for and process CPER events
PCI: Introduce cleanup helpers for device reference counts and locks
acpi/ghes: Process CXL Component Events
cxl/events: Create a CXL event union
cxl/events: Separate UUID from event structures
cxl/events: Remove passing a UUID to known event traces
cxl/events: Create common event UUID defines
cxl/events: Promote CXL event structures to a core header
cxl: Refactor to use __free() for cxl_root allocation in cxl_endpoint_port_probe()
cxl: Refactor to use __free() for cxl_root allocation in cxl_find_nvdimm_bridge()
cxl: Fix device reference leak in cxl_port_perf_data_calculate()
cxl: Convert find_cxl_root() to return a 'struct cxl_root *'
cxl: Introduce put_cxl_root() helper
cxl/port: Fix missing target list lock
cxl/port: Fix decoder initialization when nr_targets > interleave_ways
cxl/region: fix x9 interleave typo
cxl/trace: Pass UUID explicitly to event traces
cxl/region: use %pap format to print resource_size_t
cxl/region: Add dev_dbg() detail on failure to allocate HPA space
...
BIOS can configure memory devices as firmware first. This will send CXL
events to the firmware instead of the OS. The firmware can then send
these events to the OS via UEFI.
UEFI v2.10 section N.2.14 defines a Common Platform Error Record (CPER)
format for CXL Component Events. The format is mostly the same as the
CXL Common Event Record Format. The difference is the use of a GUID in
the Section Type rather than a UUID as part of the event itself.
Add GHES support to detect CXL CPER records and call a registered
callback with the event.
A notifier chain was considered for the callback but the complexity did
not justify the use case as only the CXL subsystem requires this event.
Enforce that only one callback can be registered at any time.
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Rafael J. Wysocki <rafael@kernel.org>
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
Link: https://lore.kernel.org/r/20231220-cxl-cper-v5-7-1bb8a4ca2c7a@intel.com
[djbw: fixup checkpatch errors]
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
There are two major types of uncorrected recoverable (UCR) errors :
- Synchronous error: The error is detected and raised at the point of
the consumption in the execution flow, e.g. when a CPU tries to
access a poisoned cache line. The CPU will take a synchronous error
exception such as Synchronous External Abort (SEA) on Arm64 and
Machine Check Exception (MCE) on X86. OS requires to take action (for
example, offline failure page/kill failure thread) to recover this
uncorrectable error.
- Asynchronous error: The error is detected out of processor execution
context, e.g. when an error is detected by a background scrubber.
Some data in the memory are corrupted. But the data have not been
consumed. OS is optional to take action to recover this uncorrectable
error.
When APEI firmware first is enabled, a platform may describe one error
source for the handling of synchronous errors (e.g. MCE or SEA notification
), or for handling asynchronous errors (e.g. SCI or External Interrupt
notification). In other words, we can distinguish synchronous errors by
APEI notification. For synchronous errors, kernel will kill the current
process which accessing the poisoned page by sending SIGBUS with
BUS_MCEERR_AR. In addition, for asynchronous errors, kernel will notify the
process who owns the poisoned page by sending SIGBUS with BUS_MCEERR_AO in
early kill mode. However, the GHES driver always sets mf_flags to 0 so that
all synchronous errors are handled as asynchronous errors in memory failure.
To this end, set memory failure flags as MF_ACTION_REQUIRED on synchronous
events.
Signed-off-by: Shuai Xue <xueshuai@linux.alibaba.com>
Tested-by: Ma Wupeng <mawupeng1@huawei.com>
Reviewed-by: Kefeng Wang <wangkefeng.wang@huawei.com>
Reviewed-by: Xiaofei Tan <tanxiaofei@huawei.com>
Reviewed-by: Baolin Wang <baolin.wang@linux.alibaba.com>
Reviewed-by: James Morse <james.morse@arm.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Vendor-Defined Error types are supported by the platform apart from
standard error types if bit 31 is set in the output of GET_ERROR_TYPE
Error Injection Action.[1] While the errors themselves and the length
of their associated "OEM Defined data structure" might vary between
vendors, the physical address of this structure can be computed through
vendor_extension and length fields of "SET_ERROR_TYPE_WITH_ADDRESS" and
"Vendor Error Type Extension" Structures respectively.[2][3]
Currently, however, the einj module only computes the physical address of
Vendor Error Type Extension Structure. Neither does it compute the physical
address of OEM Defined structure nor does it establish the memory mapping
required for injecting Vendor-defined errors. Consequently, userspace
tools have to establish the very mapping through /dev/mem, nopat kernel
parameter and system calls like mmap/munmap initially before injecting
Vendor-defined errors.
Circumvent the issue by computing the physical address of OEM Defined data
structure and establishing the required mapping with the structure. Create
a new file "oem_error", if the system supports Vendor-defined errors, to
export this mapping, through debugfs_create_blob(). Userspace tools can
then populate their respective OEM Defined structure instances and just
write to the file as part of injecting Vendor-defined Errors. Similarly,
the tools can also read from the file if the system firmware provides some
information through the OEM defined structure after error injection.
[1] ACPI specification 6.5, section 18.6.4
[2] ACPI specification 6.5, Table 18.31
[3] ACPI specification 6.5, Table 18.32
Suggested-by: Yazen Ghannam <yazen.ghannam@amd.com>
Signed-off-by: Avadhut Naik <Avadhut.Naik@amd.com>
Reviewed-by: Tony Luck <tony.luck@intel.com>
Reviewed-by: Borislav Petkov (AMD) <bp@alien8.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
OSPM can discover the error injection capabilities of the platform by
executing GET_ERROR_TYPE error injection action.[1] The action returns
a DWORD representing a bitmap of platform supported error injections.[2]
The available_error_type_show() function determines the bits set within
this DWORD and provides a verbose output, from einj_error_type_string
array, through /sys/kernel/debug/apei/einj/available_error_type file.
The function however, assumes one to one correspondence between an error's
position in the bitmap and its array entry offset. Consequently, some
errors like Vendor Defined Error Type fail this assumption and will
incorrectly be shown as not supported, even if their corresponding bit is
set in the bitmap and they have an entry in the array.
Navigate around the issue by converting einj_error_type_string into an
array of structures with a predetermined mask for all error types
corresponding to their bit position in the DWORD returned by GET_ERROR_TYPE
action. The same breaks the aforementioned assumption resulting in all
supported error types by a platform being outputted through the above
available_error_type file.
[1] ACPI specification 6.5, Table 18.25
[2] ACPI specification 6.5, Table 18.30
Suggested-by: Alexey Kardashevskiy <alexey.kardashevskiy@amd.com>
Signed-off-by: Avadhut Naik <Avadhut.Naik@amd.com>
Reviewed-by: Borislav Petkov (AMD) <bp@alien8.de>
Reviewed-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Slow devices such as flash may not meet the default 1ms timeout value,
so use the ERST max execution time value that they provide as the
timeout if it is larger.
Signed-off-by: Jeshua Smith <jeshuas@nvidia.com>
Reviewed-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
ghes_handle_aer() passes AER data to the PCI core for logging and
recovery by calling aer_recover_queue() with a pointer to struct
aer_capability_regs.
The problem was that aer_recover_queue() queues the pointer directly
without copying the aer_capability_regs data. The pointer was to
the ghes->estatus buffer, which could be reused before
aer_recover_work_func() reads the data.
To avoid this problem, allocate a new aer_capability_regs structure
from the ghes_estatus_pool, copy the AER data from the ghes->estatus
buffer into it, pass a pointer to the new struct to
aer_recover_queue(), and free it after aer_recover_work_func() has
processed it.
Reported-by: Bjorn Helgaas <helgaas@kernel.org>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Shiju Jose <shiju.jose@huawei.com>
[ rjw: Subject edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Since 315bada690 ("EDAC: Check for GHES preference in the
chipset-specific EDAC drivers"), vendor specific EDAC driver will not
probe correctly when CONFIG_ACPI_APEI_GHES is enabled but no GHES device
is present. Make ghes_get_devices() return NULL when the GHES device
list is empty to fix the problem.
Fixes: 9057a3f7ac ("EDAC/ghes: Prepare to make ghes_edac a proper module")
Signed-off-by: Li Yang <leoyang.li@nxp.com>
Reviewed-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
It's only used inside the __init section. Mark it __initdata.
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
ghes_estatus_pool_size_request() is unused now, so remove it.
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
[ rjw: Subject and changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The cper.c file needs to include an extra header, and efi_zboot_entry
needs an extern declaration to avoid these 'make W=1' warnings:
drivers/firmware/efi/libstub/zboot.c:65:1: error: no previous prototype for 'efi_zboot_entry' [-Werror=missing-prototypes]
drivers/firmware/efi/efi.c:176:16: error: no previous prototype for 'efi_attr_is_visible' [-Werror=missing-prototypes]
drivers/firmware/efi/cper.c:626:6: error: no previous prototype for 'cper_estatus_print' [-Werror=missing-prototypes]
drivers/firmware/efi/cper.c:649:5: error: no previous prototype for 'cper_estatus_check_header' [-Werror=missing-prototypes]
drivers/firmware/efi/cper.c:662:5: error: no previous prototype for 'cper_estatus_check' [-Werror=missing-prototypes]
To make this easier, move the cper specific declarations to
include/linux/cper.h.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
OSPM executes an EXECUTE_OPERATION action to instruct the platform to begin
the injection operation, then executes a GET_COMMAND_STATUS action to
determine the status of the completed operation. The ACPI Specification
documented error codes[1] are:
0 = Success (Linux #define EINJ_STATUS_SUCCESS)
1 = Unknown failure (Linux #define EINJ_STATUS_FAIL)
2 = Invalid Access (Linux #define EINJ_STATUS_INVAL)
The original code report -EBUSY for both "Unknown Failure" and "Invalid
Access" cases. Actually, firmware could do some platform dependent sanity
checks and returns different error codes, e.g. "Invalid Access" to indicate
to the user that the parameters they supplied cannot be used for injection.
To this end, fix to return -EINVAL in the __einj_error_inject() error
handling case instead of always -EBUSY, when explicitly indicated by the
platform in the status of the completed operation.
[1] ACPI Specification 6.5 18.6.1. Error Injection Table
Signed-off-by: Shuai Xue <xueshuai@linux.alibaba.com>
Reviewed-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
ACPI 6.5 added six new error types for CXL. See chapter 18
table 18.30.
Add strings for the new types so that Linux will list them in the
/sys/kernel/debug/apei/einj/available_error_types file.
It seems no other changes are needed. Linux already accepts
the CXL codes (on a BIOS that advertises them).
Signed-off-by: Tony Luck <tony.luck@intel.com>
Reviewed-by: Davidlohr Bueso <dave@stgolabs.net>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The bit map of error types to inject is 32-bit width [1].
Add parameter check to reflect the fact.
[1] ACPI Specification 6.4, Section 18.6.4. Error Types
Signed-off-by: Shuai Xue <xueshuai@linux.alibaba.com>
Reviewed-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Due to several bugs caused by timers being re-armed after they are
shutdown and just before they are freed, a new state of timers was added
called "shutdown". After a timer is set to this state, then it can no
longer be re-armed.
The following script was run to find all the trivial locations where
del_timer() or del_timer_sync() is called in the same function that the
object holding the timer is freed. It also ignores any locations where
the timer->function is modified between the del_timer*() and the free(),
as that is not considered a "trivial" case.
This was created by using a coccinelle script and the following
commands:
$ cat timer.cocci
@@
expression ptr, slab;
identifier timer, rfield;
@@
(
- del_timer(&ptr->timer);
+ timer_shutdown(&ptr->timer);
|
- del_timer_sync(&ptr->timer);
+ timer_shutdown_sync(&ptr->timer);
)
... when strict
when != ptr->timer
(
kfree_rcu(ptr, rfield);
|
kmem_cache_free(slab, ptr);
|
kfree(ptr);
)
$ spatch timer.cocci . > /tmp/t.patch
$ patch -p1 < /tmp/t.patch
Link: https://lore.kernel.org/lkml/20221123201306.823305113@linutronix.de/
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Acked-by: Pavel Machek <pavel@ucw.cz> [ LED ]
Acked-by: Kalle Valo <kvalo@kernel.org> [ wireless ]
Acked-by: Paolo Abeni <pabeni@redhat.com> [ networking ]
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
drop this forced built-in only configuration by disentangling it from
GHES. Work by Jia He.
- The usual small cleanups and improvements all over EDAC land
-----BEGIN PGP SIGNATURE-----
iQIzBAABCgAdFiEEzv7L6UO9uDPlPSfHEsHwGGHeVUoFAmOXPvoACgkQEsHwGGHe
VUq98xAAmhz4u9e9pXG0Ixkx25ZtnZ+YxANeQ53Hsa2gWicbcoFgL2E30gi97c1y
X9W361B2Q5dYq+J/YRUnEOXlI/KMWLxzNykSvVipUFNfxXZH+PijEAArz2V35/uE
6ZISRLUYVYEtHEoUXbTogeyBmBUnIaJfYheZCluDQlWPggsDESP1qmE+FTg25OBs
rDl5y+zUZYPxrWustNodVThPyhdMwGyYAUS6qYKCoNs9SNkAjGnrXoPc9j/U+cV+
qMY2dNS3uKnCujKEssQhcHucyWgCEDvmEKWMH4ItryV2UBBjpNRoM6HDe7XFKwVJ
riOKX8VDrpdSdlV1jbCx9KB47BUwFygOYsFdW7gIDJ1hb8usN4nSYQNDIlZKEIQG
cHNpv2XGT+pCSvyc4Iv2Fgyvnp25XensSQwQAtk5Y4/lJL1yrgcPjMOkPmRS+mmH
BclDWNbL+gwqkyWxgfoivDBOetLgwJYTr2ewBr6QbBtwLB8rL4BxXIdomcoFPuxi
jAxixZnTbS+Xq5S7uYK4r6KbaHGcJtwolXMGjx13IHmPfvYtTTQzfRcrBlAtQ/pV
BDLoygmDVlkhSVx6bi5V5QZ06rcWYR4cRsBQ54FnBGMr730ZljgFONOHFtUab28T
C+YUOaeLEYEYI0cIkkyoSuiz6avB6YvQAiyEPM0EdHZrQFwhBBw=
=DFp8
-----END PGP SIGNATURE-----
Merge tag 'edac_updates_for_6.2' of git://git.kernel.org/pub/scm/linux/kernel/git/ras/ras
Pull EDAC updates from Borislav Petkov:
- Make ghes_edac a simple module like the rest of the EDAC drivers and
drop the forced built-in only configuration by disentangling it from
GHES (Jia He)
- The usual small cleanups and improvements all over EDAC land
* tag 'edac_updates_for_6.2' of git://git.kernel.org/pub/scm/linux/kernel/git/ras/ras:
EDAC/i10nm: fix refcount leak in pci_get_dev_wrapper()
EDAC/i5400: Fix typo in comment: vaious -> various
EDAC/mc_sysfs: Increase legacy channel support to 12
MAINTAINERS: Make Mauro EDAC reviewer
MAINTAINERS: Make Manivannan Sadhasivam the maintainer of qcom_edac
EDAC/igen6: Return the correct error type when not the MC owner
apei/ghes: Use xchg_release() for updating new cache slot instead of cmpxchg()
EDAC: Check for GHES preference in the chipset-specific EDAC drivers
EDAC/ghes: Make ghes_edac a proper module
EDAC/ghes: Prepare to make ghes_edac a proper module
EDAC/ghes: Add a notifier for reporting memory errors
efi/cper: Export several helpers for ghes_edac to use
EDAC/i5000: Mark as BROKEN
Move error type descriptions into an array and loop over error types
to improve readability and maintainability.
Replace seq_printf() with seq_puts() as recommended by checkpatch.pl.
Signed-off-by: Jay Lu <jaylu102@amd.com>
Co-developed-by: Ben Cheatham <benjamin.cheatham@amd.com>
Signed-off-by: Ben Cheatham <benjamin.cheatham@amd.com>
Reviewed-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Checkpatch reveals warnings and an error due to missing lines and
incorrect indentations. Add the missing lines after declarations and
fix the suspect indentations.
Signed-off-by: Jay Lu <jaylu102@amd.com>
Signed-off-by: Ben Cheatham <benjamin.cheatham@amd.com>
Reviewed-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
This file does not use rcu, so there is no point in including
<linux/rculist.h>.
So just remove it.
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Silence the following warnings when make W=1:
| CC drivers/acpi/apei/apei-base.c
| warning: no previous prototype for 'arch_apei_enable_cmcff' [-Wmissing-prototypes]
| int __weak arch_apei_enable_cmcff(struct acpi_hest_header *hest_hdr,
| ^
| CC drivers/acpi/apei/apei-base.c
| warning: no previous prototype for 'arch_apei_report_mem_error' [-Wmissing-prototypes]
| void __weak arch_apei_report_mem_error(int sev,
| ^
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Some documentation first, about how this machinery works:
It seems, the intent of the GHES error records cache is to collect
already reported errors - see the ghes_estatus_cached() checks. There's
even a sentence trying to say what this does:
/*
* GHES error status reporting throttle, to report more kinds of
* errors, instead of just most frequently occurred errors.
*/
New elements are added to the cache this way:
if (!ghes_estatus_cached(estatus)) {
if (ghes_print_estatus(NULL, ghes->generic, estatus))
ghes_estatus_cache_add(ghes->generic, estatus);
The intent being, once this new error record is reported, it gets cached
so that it doesn't get reported for a while due to too many, same-type
error records getting reported in burst-like scenarios. I.e., new,
unreported error types can have a higher chance of getting reported.
Now, the loop in ghes_estatus_cache_add() is trying to pick out the
oldest element in there. Meaning, something which got reported already
but a long while ago, i.e., a LRU-type scheme.
And the cmpxchg() is there presumably to make sure when that selected
element slot_cache is removed, it really *is* that element that gets
removed and not one which replaced it in the meantime.
Now, ghes_estatus_cache_add() selects a slot, and either succeeds in
replacing its contents with a pointer to a newly cached item, or it just
gives up and frees the new item again, without attempting to select
another slot even if one might be available.
Since only inserting new items is being done here, the race can only
cause a failure if the selected slot was updated with another new item
concurrently, which means that it is arbitrary which of those two items
gets dropped.
And "dropped" here means, the item doesn't get added to the cache so
the next time it is seen, it'll get reported again and an insertion
attempt will be done again. Eventually, it'll get inserted and all those
times when the insertion fails, the item will get reported although the
cache is supposed to prevent that and "ratelimit" those repeated error
records. Not a big deal in any case.
This means the cmpxchg() and the special case are not necessary.
Therefore, just drop the existing item unconditionally.
Move the xchg_release() and call_rcu() out of rcu_read_lock/unlock
section since there is no actually dereferencing the pointer at all.
[ bp:
- Flesh out and summarize what was discussed on the thread now
that that cache contraption is understood;
- Touch up code style. ]
Co-developed-by: Jia He <justin.he@arm.com>
Signed-off-by: Jia He <justin.he@arm.com>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lore.kernel.org/r/20221010023559.69655-7-justin.he@arm.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Since commit 0998d06310 ("device-core: Ensure drvdata = NULL when no
driver is bound") the driver core cares for cleaning driver data, so
don't do it in the driver, too.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Some documentation first, about how this machinery works:
It seems, the intent of the GHES error records cache is to collect
already reported errors - see the ghes_estatus_cached() checks. There's
even a sentence trying to say what this does:
/*
* GHES error status reporting throttle, to report more kinds of
* errors, instead of just most frequently occurred errors.
*/
New elements are added to the cache this way:
if (!ghes_estatus_cached(estatus)) {
if (ghes_print_estatus(NULL, ghes->generic, estatus))
ghes_estatus_cache_add(ghes->generic, estatus);
The intent being, once this new error record is reported, it gets cached
so that it doesn't get reported for a while due to too many, same-type
error records getting reported in burst-like scenarios. I.e., new,
unreported error types can have a higher chance of getting reported.
Now, the loop in ghes_estatus_cache_add() is trying to pick out the
oldest element in there. Meaning, something which got reported already
but a long while ago, i.e., a LRU-type scheme.
And the cmpxchg() is there presumably to make sure when that selected
element slot_cache is removed, it really *is* that element that gets
removed and not one which replaced it in the meantime.
Now, ghes_estatus_cache_add() selects a slot, and either succeeds in
replacing its contents with a pointer to a newly cached item, or it just
gives up and frees the new item again, without attempting to select
another slot even if one might be available.
Since only inserting new items is being done here, the race can only
cause a failure if the selected slot was updated with another new item
concurrently, which means that it is arbitrary which of those two items
gets dropped.
And "dropped" here means, the item doesn't get added to the cache so
the next time it is seen, it'll get reported again and an insertion
attempt will be done again. Eventually, it'll get inserted and all those
times when the insertion fails, the item will get reported although the
cache is supposed to prevent that and "ratelimit" those repeated error
records. Not a big deal in any case.
This means the cmpxchg() and the special case are not necessary.
Therefore, just drop the existing item unconditionally.
Move the xchg_release() and call_rcu() out of rcu_read_lock/unlock
section since there is no actually dereferencing the pointer at all.
[ bp:
- Flesh out and summarize what was discussed on the thread now
that that cache contraption is understood;
- Touch up code style. ]
Co-developed-by: Jia He <justin.he@arm.com>
Signed-off-by: Jia He <justin.he@arm.com>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lore.kernel.org/r/20221010023559.69655-7-justin.he@arm.com
Commit
dc4e8c07e9 ("ACPI: APEI: explicit init of HEST and GHES in apci_init()")
introduced a bug leading to ghes_edac_register() to be invoked before
edac_init(). Because at that time the bus "edac" hadn't been even
registered, this created sysfs nodes as /devices/mc0 instead of
/sys/devices/system/edac/mc/mc0 on an Ampere eMag server.
Fix this by turning ghes_edac into a proper module.
The list of GHES devices returned is not protected from being modified
concurrently but it is pretty static as it gets created only during GHES
init and latter is not a module so...
[ bp: Massage. ]
Fixes: dc4e8c07e9 ("ACPI: APEI: explicit init of HEST and GHES in apci_init()")
Co-developed-by: Borislav Petkov <bp@alien8.de>
Signed-off-by: Borislav Petkov <bp@alien8.de>
Signed-off-by: Jia He <justin.he@arm.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lore.kernel.org/r/20221010023559.69655-5-justin.he@arm.com
To make ghes_edac a proper module, prepare to decouple its dependencies
from GHES.
Move the ghes_edac.force_load parameter to ghes.c in order to
properly control whether ghes_edac should be force-loaded: In
ghes_edac_register() it is too late to set the module flag.
Introduce a helper ghes_get_devices(), which returns the list of GHES
devices which got probed when the platform-check passes on the system.
The previous force_load check is not needed in ghes_edac_unregister()
since it will be checked in the module's init function of ghes_edac
later.
[ bp: Massage. ]
Suggested-by: Toshi Kani <toshi.kani@hpe.com>
Suggested-by: Borislav Petkov <bp@alien8.de>
Signed-off-by: Jia He <justin.he@arm.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lore.kernel.org/r/20221010023559.69655-4-justin.he@arm.com
In order to make it a proper module and disentangle it from facilities,
add a notifier for reporting memory errors. Use an atomic notifier
because calls sites like ghes_proc_in_irq() run in interrupt context.
[ bp: Massage commit message. ]
Suggested-by: Borislav Petkov <bp@alien8.de>
Signed-off-by: Jia He <justin.he@arm.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lore.kernel.org/r/20221010023559.69655-3-justin.he@arm.com
If an error is detected as a result of user-space process accessing a
corrupt memory location, the CPU may take an abort. Then the platform
firmware reports kernel via NMI like notifications, e.g. NOTIFY_SEA,
NOTIFY_SOFTWARE_DELEGATED, etc.
For NMI like notifications, commit 7f17b4a121 ("ACPI: APEI: Kick the
memory_failure() queue for synchronous errors") keep track of whether
memory_failure() work was queued, and make task_work pending to flush out
the queue so that the work is processed before return to user-space.
The code use init_mm to check whether the error occurs in user space:
if (current->mm != &init_mm)
The condition is always true, becase _nobody_ ever has "init_mm" as a real
VM any more.
In addition to abort, errors can also be signaled as asynchronous
exceptions, such as interrupt and SError. In such case, the interrupted
current process could be any kind of thread. When a kernel thread is
interrupted, the work ghes_kick_task_work deferred to task_work will never
be processed because entry_handler returns to call ret_to_kernel() instead
of ret_to_user(). Consequently, the estatus_node alloced from
ghes_estatus_pool in ghes_in_nmi_queue_one_entry() will not be freed.
After around 200 allocations in our platform, the ghes_estatus_pool will
run of memory and ghes_in_nmi_queue_one_entry() returns ENOMEM. As a
result, the event failed to be processed.
sdei: event 805 on CPU 113 failed with error: -2
Finally, a lot of unhandled events may cause platform firmware to exceed
some threshold and reboot.
The condition should generally just do
if (current->mm)
as described in active_mm.rst documentation.
Then if an asynchronous error is detected when a kernel thread is running,
(e.g. when detected by a background scrubber), do not add task_work to it
as the original patch intends to do.
Fixes: 7f17b4a121 ("ACPI: APEI: Kick the memory_failure() queue for synchronous errors")
Signed-off-by: Shuai Xue <xueshuai@linux.alibaba.com>
Reviewed-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Return the erst_get_record_id_begin() and apei_exec_write_register()
return values directly instead of storing them in redundant local
variables.
Reported-by: Zeal Robot <zealci@zte.com.cn>
Signed-off-by: ye xingchen <ye.xingchen@zte.com.cn>
[ rjw: Changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Print total number of records found during BERT log parsing.
This also simplify dmesg parser implementation for BERT events.
Signed-off-by: Dmitry Monakhov <dmtrmonakhov@yandex-team.ru>
Acked-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
When a platform marks a memory range as "special purpose" it is not
onlined as System RAM by default. However, it is still suitable for
error injection. Add IORES_DESC_SOFT_RESERVED to einj_error_inject() as
a permissible memory type in the sanity checking of the arguments to
_EINJ.
Fixes: 262b45ae3a ("x86/efi: EFI soft reservation to E820 enumeration")
Reviewed-by: Tony Luck <tony.luck@intel.com>
Reported-by: Omar Avelar <omar.avelar@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The fix in commit 3f8dec1162 ("ACPI/APEI: Limit printable size of BERT
table data") does not work as intended on systems where the BIOS has a
fixed size block of memory for the BERT table, relying on s/w to quit
when it finds a record with estatus->block_status == 0. On these systems
all errors are suppressed because the check:
if (region_len < ACPI_BERT_PRINT_MAX_LEN)
always fails.
New scheme skips individual CPER records that are too large, and also
limits the total number of records that will be printed to 5.
Fixes: 3f8dec1162 ("ACPI/APEI: Limit printable size of BERT table data")
Cc: All applicable <stable@vger.kernel.org>
Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Delete the redundant word 'the'.
Signed-off-by: Xiang wangx <wangxiang@cdjrlc.com>
[ rjw: New subject ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Some validation tests dynamically inject errors into memory used by
applications to check that the system can recover from a variety of
poison consumption sceenarios.
But sometimes the virtual address picked by these tests is mapped to
the zero page.
This causes additional unexpected machine checks as other processes that
map the zero page also consume the poison.
Disallow injection to the zero page.
Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Read a record is cleared by others, but the deleted record cache entry is
still created by erst_get_record_id_next. When next enumerate the records,
get the cached deleted record, then erst_read() return -ENOENT and try to
get next record, loop back to first ID will return 0 in function
__erst_record_id_cache_add_one and then set record_id as
APEI_ERST_INVALID_RECORD_ID, finished this time read operation.
It will result in read the records just in the cache hereafter.
This patch cleared the deleted record cache, fix the issue that
"./erst-inject -p" shows record counts not equal to "./erst-inject -n".
A reproducer of the problem(retry many times):
[root@localhost erst-inject]# ./erst-inject -c 0xaaaaa00011
[root@localhost erst-inject]# ./erst-inject -p
rc: 273
rcd sig: CPER
rcd id: 0xaaaaa00012
rc: 273
rcd sig: CPER
rcd id: 0xaaaaa00013
rc: 273
rcd sig: CPER
rcd id: 0xaaaaa00014
[root@localhost erst-inject]# ./erst-inject -i 0xaaaaa000006
[root@localhost erst-inject]# ./erst-inject -i 0xaaaaa000007
[root@localhost erst-inject]# ./erst-inject -i 0xaaaaa000008
[root@localhost erst-inject]# ./erst-inject -p
rc: 273
rcd sig: CPER
rcd id: 0xaaaaa00012
rc: 273
rcd sig: CPER
rcd id: 0xaaaaa00013
rc: 273
rcd sig: CPER
rcd id: 0xaaaaa00014
[root@localhost erst-inject]# ./erst-inject -n
total error record count: 6
Signed-off-by: Liu Xinpeng <liuxp11@chinatelecom.cn>
Reviewed-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
While the original code is valid, it is not the obvious choice for the
sizeof() call and in preparation to limit the scope of the list iterator
variable the sizeof should be changed to the size of the variable
being allocated.
Signed-off-by: Jakob Koschel <jakobkoschel@gmail.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Platforms with large BERT table data can trigger soft lockup errors
while attempting to print the entire BERT table data to the console at
boot:
watchdog: BUG: soft lockup - CPU#160 stuck for 23s! [swapper/0:1]
Observed on Ampere Altra systems with a single BERT record of ~250KB.
The original bert driver appears to have assumed relatively small table
data. Since it is impractical to reassemble large table data from
interwoven console messages, and the table data is available in
/sys/firmware/acpi/tables/data/BERT
limit the size for tables printed to the console to 1024 (for no reason
other than it seemed like a good place to kick off the discussion, would
appreciate feedback from existing users in terms of what size would
maintain their current usage model).
Alternatively, we could make printing a CONFIG option, use the
bert_disable boot arg (or something similar), or use a debug log level.
However, all those solutions require extra steps or change the existing
behavior for small table data. Limiting the size preserves existing
behavior on existing platforms with small table data, and eliminates the
soft lockups for platforms with large table data, while still making it
available.
Signed-off-by: Darren Hart <darren@os.amperecomputing.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
__setup() handlers should return 1 to indicate that the boot option
has been handled. Returning 0 causes a boot option to be listed in
the Unknown kernel command line parameters and also added to init's
arg list (if no '=' sign) or environment list (if of the form 'a=b').
Unknown kernel command line parameters "erst_disable
bert_disable hest_disable BOOT_IMAGE=/boot/bzImage-517rc6", will be
passed to user space.
Run /sbin/init as init process
with arguments:
/sbin/init
erst_disable
bert_disable
hest_disable
with environment:
HOME=/
TERM=linux
BOOT_IMAGE=/boot/bzImage-517rc6
Fixes: a3e2acc5e3 ("ACPI / APEI: Add Boot Error Record Table (BERT) support")
Fixes: a08f82d080 ("ACPI, APEI, Error Record Serialization Table (ERST) support")
Fixes: 9dc9666416 ("ACPI, APEI, HEST table parsing")
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Reported-by: Igor Zhbanov <i.zhbanov@omprussia.ru>
Link: lore.kernel.org/r/64644a2f-4a20-bab3-1e15-3b2cdd0defe3@omprussia.ru
Reviewed-by: "Huang, Ying" <ying.huang@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
ghes_init() sticks out in acpi_init() because it is the only functions
without an "acpi_" prefix.
Rename ghes_init with an "acpi_" prefix, then all looks fine.
Signed-off-by: Shuai Xue <xueshuai@linux.alibaba.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
From commit e147133a42 ("ACPI / APEI: Make hest.c manage the estatus
memory pool") was merged, ghes_init() relies on acpi_hest_init() to manage
the estatus memory pool. On the other hand, ghes_init() relies on
sdei_init() to detect the SDEI version and (un)register events. The
dependencies are as follows:
ghes_init() => acpi_hest_init() => acpi_bus_init() => acpi_init()
ghes_init() => sdei_init()
HEST is not PCI-specific and initcall ordering is implicit and not
well-defined within a level.
Based on above, remove acpi_hest_init() from acpi_pci_root_init() and
convert ghes_init() and sdei_init() from initcalls to explicit calls in the
following order:
acpi_hest_init()
ghes_init()
sdei_init()
Signed-off-by: Shuai Xue <xueshuai@linux.alibaba.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
SGX EPC pages do not have a "struct page" associated with them so the
pfn_valid() sanity check fails and results in a warning message to
the console.
Add an additional check to skip the warning if the address of the error
is in an SGX EPC page.
Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Jarkko Sakkinen <jarkko@kernel.org>
Tested-by: Reinette Chatre <reinette.chatre@intel.com>
Link: https://lkml.kernel.org/r/20211026220050.697075-8-tony.luck@intel.com
SGX reserved memory does not appear in the standard address maps.
Add hook to call into the SGX code to check if an address is located
in SGX memory.
There are other challenges in injecting errors into SGX. Update the
documentation with a sequence of operations to inject.
Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Jarkko Sakkinen <jarkko@kernel.org>
Tested-by: Reinette Chatre <reinette.chatre@intel.com>
Link: https://lkml.kernel.org/r/20211026220050.697075-7-tony.luck@intel.com
apei_hest_parse() is only used in hest.c, so mark it static.
Signed-off-by: Christoph Hellwig <hch@lst.de>
[ rjw: Minor subject and changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
When injecting an error into the platform, the OSPM executes an
EXECUTE_OPERATION action to instruct the platform to begin the injection
operation. And then, the OSPM busy waits for a while by continually
executing CHECK_BUSY_STATUS action until the platform indicates that the
operation is complete. More specifically, the platform is limited to
respond within 1 millisecond right now. This is too strict for some
platforms.
For example, in Arm platform, when injecting a Processor Correctable error,
the OSPM will warn:
Firmware does not respond in time.
And a message is printed on the console:
echo: write error: Input/output error
We observe that the waiting time for DDR error injection is about 10 ms and
that for PCIe error injection is about 500 ms in Arm platform.
In this patch, we relax the response timeout to 1 second.
Signed-off-by: Shuai Xue <xueshuai@linux.alibaba.com>
Reviewed-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Before commit 8fcc4ae6fa ("arm64: acpi: Make apei_claim_sea()
synchronise with APEI's irq work"), do_sea() would unconditionally
signal the affected task from the arch code. Since that change,
the GHES driver sends the signals.
This exposes a problem as errors the GHES driver doesn't understand
or doesn't handle effectively are silently ignored. It will cause
the errors get taken again, and circulate endlessly. User-space task
get stuck in this loop.
Existing firmware on Kunpeng9xx systems reports cache errors with the
'ARM Processor Error' CPER records.
Do memory failure handling for ARM Processor Error Section just like
for Memory Error Section.
Fixes: 8fcc4ae6fa ("arm64: acpi: Make apei_claim_sea() synchronise with APEI's irq work")
Signed-off-by: Xiaofei Tan <tanxiaofei@huawei.com>
Reviewed-by: James Morse <james.morse@arm.com>
[ rjw: Subject edit ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
If ACPI is not enabled but support for ACPI and APEI is enabled in the
kernel, then the following warning is seen on boot ...
WARNING KERN EINJ: ACPI disabled.
For ARM64 platforms, the 'acpi_disabled' variable is true by default
and hence, the above is often seen on ARM64. Given that it can be
normal for ACPI to be disabled, make this an informational print rather
that a warning.
Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
* acpi-misc:
ACPI: dock: fix some coding style issues
ACPI: sysfs: fix some coding style issues
ACPI: PM: add a missed blank line after declarations
ACPI: custom_method: fix a coding style issue
ACPI: CPPC: fix some coding style issues
ACPI: button: fix some coding style issues
ACPI: battery: fix some coding style issues
ACPI: acpi_pad: add a missed blank line after declarations
ACPI: LPSS: add a missed blank line after declarations
ACPI: ipmi: remove useless return statement for void function
ACPI: processor: fix some coding style issues
ACPI: APD: fix a block comment align issue
ACPI: AC: fix some coding style issues
ACPI: fix various typos in comments
The variable rc is being assigned a value that is never read,
the assignment is redundant and can be removed.
Addresses-Coverity: ("Unused value")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Eliminate the following coccicheck warning:
./drivers/acpi/apei/erst.c:691:2-3: Unneeded semicolon
Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Signed-off-by: Yang Li <yang.lee@linux.alibaba.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
From commit 6915564dc5 ("ACPI: OSL: Change the type of
acpi_os_map_generic_address() return value"),
acpi_os_map_generic_address() will return logical address or NULL
for error, but for ACPI_ADR_SPACE_SYSTEM_IO case, it should be also
return 0 as it's a normal case, but now it will return -ENXIO.
So check it out for such case to avoid einj module initialization
fail.
Fixes: 6915564dc5 ("ACPI: OSL: Change the type of acpi_os_map_generic_address() return value")
Cc: <stable@vger.kernel.org>
Reviewed-by: James Morse <james.morse@arm.com>
Tested-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Aili Yao <yaoaili@kingsoft.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
-----BEGIN PGP SIGNATURE-----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=aVWo
-----END PGP SIGNATURE-----
Merge tag 'arch-cleanup-2020-10-22' of git://git.kernel.dk/linux-block
Pull arch task_work cleanups from Jens Axboe:
"Two cleanups that don't fit other categories:
- Finally get the task_work_add() cleanup done properly, so we don't
have random 0/1/false/true/TWA_SIGNAL confusing use cases. Updates
all callers, and also fixes up the documentation for
task_work_add().
- While working on some TIF related changes for 5.11, this
TIF_NOTIFY_RESUME cleanup fell out of that. Remove some arch
duplication for how that is handled"
* tag 'arch-cleanup-2020-10-22' of git://git.kernel.dk/linux-block:
task_work: cleanup notification modes
tracehook: clear TIF_NOTIFY_RESUME in tracehook_notify_resume()
A previous commit changed the notification mode from true/false to an
int, allowing notify-no, notify-yes, or signal-notify. This was
backwards compatible in the sense that any existing true/false user
would translate to either 0 (on notification sent) or 1, the latter
which mapped to TWA_RESUME. TWA_SIGNAL was assigned a value of 2.
Clean this up properly, and define a proper enum for the notification
mode. Now we have:
- TWA_NONE. This is 0, same as before the original change, meaning no
notification requested.
- TWA_RESUME. This is 1, same as before the original change, meaning
that we use TIF_NOTIFY_RESUME.
- TWA_SIGNAL. This uses TIF_SIGPENDING/JOBCTL_TASK_WORK for the
notification.
Clean up all the callers, switching their 0/1/false/true to using the
appropriate TWA_* mode for notifications.
Fixes: e91b481623 ("task_work: teach task_work_add() to do signal_wake_up()")
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
This eliminates the following sparse warning:
drivers/acpi/apei/apei-base.c:290:23: warning: symbol
'apei_resources_all' was not declared. Should it be static?
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Jason Yan <yanaijie@huawei.com>
Reviewed-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
CPER records describing a firmware-first error are identified by GUID.
The ghes driver currently logs, but ignores any unknown CPER records.
This prevents describing errors that can't be represented by a standard
entry, that would otherwise allow a driver to recover from an error.
The UEFI spec calls these 'Non-standard Section Body' (N.2.3 of
version 2.8).
Add a notifier chain for these non-standard/vendor-records. Callers
must identify their type of records by GUID.
Record data is copied to memory from the ghes_estatus_pool to allow
us to keep it until after the notifier has run.
Co-developed-by: James Morse <james.morse@arm.com>
Link: https://lore.kernel.org/r/20200903123456.1823-2-shiju.jose@huawei.com
Signed-off-by: James Morse <james.morse@arm.com>
Signed-off-by: Shiju Jose <shiju.jose@huawei.com>
Signed-off-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Acked-by: "Rafael J. Wysocki" <rjw@rjwysocki.net>
Modify acpi_os_map_generic_address() to return the pointer returned
by acpi_os_map_iomem() which represents the logical address
corresponding to the struct acpi_generic_address argument passed to
it or NULL if that address cannot be obtained (for example, the
argument does not represent an address in system memory or it could
not be mapped by the OS).
Among other things, that will allow the ACPI OS layer to pass the
logical addresses of the FADT GPE blocks 0 and 1 to ACPICA going
forward.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The variable rc is being initialized with a value that is
never read and it is being updated later with a new value. The
initialization is redundant and can be removed.
Addresses-Coverity: ("Unused value")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
- Update the ACPICA code in the kernel to upstream revision
20200430:
* Move acpi_gbl_next_cmd_num definition (Erik Kaneda).
* Ignore AE_ALREADY_EXISTS status in the disassembler when parsing
create operators (Erik Kaneda).
* Add status checks to the dispatcher (Erik Kaneda).
* Fix required parameters for _NIG and _NIH (Erik Kaneda).
* Make acpi_protocol_lengths static (Yue Haibing).
- Fix ACPI table reference counting errors in several places, mostly
in error code paths (Hanjun Guo).
- Extend the Generic Event Device (GED) driver to support _Exx and
_Lxx handler methods (Ard Biesheuvel).
- Add new acpi_evaluate_reg() helper and modify the ACPI PCI hotplug
code to use it (Hans de Goede).
- Add new DPTF battery participant driver and make the DPFT power
participant driver create more sysfs device attributes (Srinivas
Pandruvada).
- Improve the handling of memory failures in APEI (James Morse).
- Add new blacklist entry for Acer TravelMate 5735Z to the backlight
driver (Paul Menzel).
- Add i2c address for thermal control to the PMIC driver (Mauro
Carvalho Chehab).
- Allow the ACPI processor idle driver to work on platforms with
only one ACPI C-state present (Zhang Rui).
- Fix kobject reference count leaks in error code paths in two
places (Qiushi Wu).
- Delete unused proc filename macros and make some symbols static
(Pascal Terjan, Zheng Zengkai, Zou Wei).
-----BEGIN PGP SIGNATURE-----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=xH3i
-----END PGP SIGNATURE-----
Merge tag 'acpi-5.8-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull ACPI updates from Rafael Wysocki:
"These update the ACPICA code in the kernel to upstream revision
20200430, fix several reference counting errors related to ACPI
tables, add _Exx / _Lxx support to the GED driver, add a new
acpi_evaluate_reg() helper, add new DPTF battery participant driver
and extend the DPFT power participant driver, improve the handling of
memory failures in the APEI code, add a blacklist entry to the
backlight driver, update the PMIC driver and the processor idle
driver, fix two kobject reference count leaks, and make a few janitory
changes.
Specifics:
- Update the ACPICA code in the kernel to upstream revision 20200430:
- Move acpi_gbl_next_cmd_num definition (Erik Kaneda).
- Ignore AE_ALREADY_EXISTS status in the disassembler when parsing
create operators (Erik Kaneda).
- Add status checks to the dispatcher (Erik Kaneda).
- Fix required parameters for _NIG and _NIH (Erik Kaneda).
- Make acpi_protocol_lengths static (Yue Haibing).
- Fix ACPI table reference counting errors in several places, mostly
in error code paths (Hanjun Guo).
- Extend the Generic Event Device (GED) driver to support _Exx and
_Lxx handler methods (Ard Biesheuvel).
- Add new acpi_evaluate_reg() helper and modify the ACPI PCI hotplug
code to use it (Hans de Goede).
- Add new DPTF battery participant driver and make the DPFT power
participant driver create more sysfs device attributes (Srinivas
Pandruvada).
- Improve the handling of memory failures in APEI (James Morse).
- Add new blacklist entry for Acer TravelMate 5735Z to the backlight
driver (Paul Menzel).
- Add i2c address for thermal control to the PMIC driver (Mauro
Carvalho Chehab).
- Allow the ACPI processor idle driver to work on platforms with only
one ACPI C-state present (Zhang Rui).
- Fix kobject reference count leaks in error code paths in two places
(Qiushi Wu).
- Delete unused proc filename macros and make some symbols static
(Pascal Terjan, Zheng Zengkai, Zou Wei)"
* tag 'acpi-5.8-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: (32 commits)
ACPI: CPPC: Fix reference count leak in acpi_cppc_processor_probe()
ACPI: sysfs: Fix reference count leak in acpi_sysfs_add_hotplug_profile()
ACPI: GED: use correct trigger type field in _Exx / _Lxx handling
ACPI: DPTF: Add battery participant driver
ACPI: DPTF: Additional sysfs attributes for power participant driver
ACPI: video: Use native backlight on Acer TravelMate 5735Z
arm64: acpi: Make apei_claim_sea() synchronise with APEI's irq work
ACPI: APEI: Kick the memory_failure() queue for synchronous errors
mm/memory-failure: Add memory_failure_queue_kick()
ACPI / PMIC: Add i2c address for thermal control
ACPI: GED: add support for _Exx / _Lxx handler methods
ACPI: Delete unused proc filename macros
ACPI: hotplug: PCI: Use the new acpi_evaluate_reg() helper
ACPI: utils: Add acpi_evaluate_reg() helper
ACPI: debug: Make two functions static
ACPI: sleep: Put the FACS table after using it
ACPI: scan: Put SPCR and STAO table after using it
ACPI: EC: Put the ACPI table after using it
ACPI: APEI: Put the HEST table for error path
ACPI: APEI: Put the error record serialization table for error path
...
These functions are not needed anymore because the vmalloc and ioremap
mappings are now synchronized when they are created or torn down.
Remove all callers and function definitions.
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Tested-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Acked-by: Andy Lutomirski <luto@kernel.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: "Rafael J. Wysocki" <rjw@rjwysocki.net>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200515140023.25469-7-joro@8bytes.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
memory_failure() offlines or repairs pages of memory that have been
discovered to be corrupt. These may be detected by an external
component, (e.g. the memory controller), and notified via an IRQ.
In this case the work is queued as not all of memory_failure()s work
can happen in IRQ context.
If the error was detected as a result of user-space accessing a
corrupt memory location the CPU may take an abort instead. On arm64
this is a 'synchronous external abort', and on a firmware first
system it is replayed using NOTIFY_SEA.
This notification has NMI like properties, (it can interrupt
IRQ-masked code), so the memory_failure() work is queued. If we
return to user-space before the queued memory_failure() work is
processed, we will take the fault again. This loop may cause platform
firmware to exceed some threshold and reboot when Linux could have
recovered from this error.
For NMIlike notifications keep track of whether memory_failure() work
was queued, and make task_work pending to flush out the queue.
To save memory allocations, the task_work is allocated as part of
the ghes_estatus_node, and free()ing it back to the pool is deferred.
Signed-off-by: James Morse <james.morse@arm.com>
Tested-by: Tyler Baicar <baicar@os.amperecomputing.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
hest_tab will be used after hest_init(), but we need to release
it for error path.
Signed-off-by: Hanjun Guo <guohanjun@huawei.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The mapped error record serialization table needs to be
released for error path of erst_init().
Signed-off-by: Hanjun Guo <guohanjun@huawei.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The mapped error injection table will be used after einj_init()
for debugfs, but it should be released for module exit and error
path of einj_init().
Signed-off-by: Hanjun Guo <guohanjun@huawei.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The mapped boot error record table is not used after
bert_init(), release it.
Signed-off-by: Hanjun Guo <guohanjun@huawei.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Commit 3f8fd02b1b ("mm/vmalloc: Sync unmappings in
__purge_vmap_area_lazy()") introduced a call to vmalloc_sync_all() in
the vunmap() code-path. While this change was necessary to maintain
correctness on x86-32-pae kernels, it also adds additional cycles for
architectures that don't need it.
Specifically on x86-64 with CONFIG_VMAP_STACK=y some people reported
severe performance regressions in micro-benchmarks because it now also
calls the x86-64 implementation of vmalloc_sync_all() on vunmap(). But
the vmalloc_sync_all() implementation on x86-64 is only needed for newly
created mappings.
To avoid the unnecessary work on x86-64 and to gain the performance
back, split up vmalloc_sync_all() into two functions:
* vmalloc_sync_mappings(), and
* vmalloc_sync_unmappings()
Most call-sites to vmalloc_sync_all() only care about new mappings being
synchronized. The only exception is the new call-site added in the
above mentioned commit.
Shile Zhang directed us to a report of an 80% regression in reaim
throughput.
Fixes: 3f8fd02b1b ("mm/vmalloc: Sync unmappings in __purge_vmap_area_lazy()")
Reported-by: kernel test robot <oliver.sang@intel.com>
Reported-by: Shile Zhang <shile.zhang@linux.alibaba.com>
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Tested-by: Borislav Petkov <bp@suse.de>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> [GHES]
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: <stable@vger.kernel.org>
Link: http://lkml.kernel.org/r/20191009124418.8286-1-joro@8bytes.org
Link: https://lists.01.org/hyperkitty/list/lkp@lists.01.org/thread/4D3JPPHBNOSPFK2KEPC6KGKS6J25AIDB/
Link: http://lkml.kernel.org/r/20191113095530.228959-1-shile.zhang@linux.alibaba.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Currently, the ghes_poll_func() timer callback is registered with the
TIMER_DEFERRABLE flag. Thus, it is run when the CPU eventually wakes
up together with a subsequent non-deferrable timer and not at the precisely
configured polling interval.
For polling mode, the polling interval configured by firmware should not
be exceeded according to the ACPI spec 6.3, Table 18-394. The definition
of the polling interval is:
"Indicates the poll interval in milliseconds OSPM should use to
periodically check the error source for the presence of an error
condition."
If this interval is extended due to the timer callback deferring, error
records can get lost. Which we are observing on our ThunderX2 platforms.
Therefore, remove the TIMER_DEFERRABLE flag so that the timer callback
executes at the precise interval.
Signed-off-by: Bhaskar Upadhaya <bupadhaya@marvell.com>
[ bp: Subject & changelog ]
Acked-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
-----BEGIN PGP SIGNATURE-----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=QCEy
-----END PGP SIGNATURE-----
Merge tag 'printk-for-5.5' of git://git.kernel.org/pub/scm/linux/kernel/git/pmladek/printk
Pull printk updates from Petr Mladek:
- Allow to print symbolic error names via new %pe modifier.
- Use pr_warn() instead of the remaining pr_warning() calls. Fix
formatting of the related lines.
- Add VSPRINTF entry to MAINTAINERS.
* tag 'printk-for-5.5' of git://git.kernel.org/pub/scm/linux/kernel/git/pmladek/printk: (32 commits)
checkpatch: don't warn about new vsprintf pointer extension '%pe'
MAINTAINERS: Add VSPRINTF
tools lib api: Renaming pr_warning to pr_warn
ASoC: samsung: Use pr_warn instead of pr_warning
lib: cpu_rmap: Use pr_warn instead of pr_warning
trace: Use pr_warn instead of pr_warning
dma-debug: Use pr_warn instead of pr_warning
vgacon: Use pr_warn instead of pr_warning
fs: afs: Use pr_warn instead of pr_warning
sh/intc: Use pr_warn instead of pr_warning
scsi: Use pr_warn instead of pr_warning
platform/x86: intel_oaktrail: Use pr_warn instead of pr_warning
platform/x86: asus-laptop: Use pr_warn instead of pr_warning
platform/x86: eeepc-laptop: Use pr_warn instead of pr_warning
oprofile: Use pr_warn instead of pr_warning
of: Use pr_warn instead of pr_warning
macintosh: Use pr_warn instead of pr_warning
idsn: Use pr_warn instead of pr_warning
ide: Use pr_warn instead of pr_warning
crypto: n2: Use pr_warn instead of pr_warning
...
As said in commit f2c2cbcc35 ("powerpc: Use pr_warn instead of
pr_warning"), removing pr_warning so all logging messages use a
consistent <prefix>_warn style. Let's do it.
Link: http://lkml.kernel.org/r/20191018031850.48498-8-wangkefeng.wang@huawei.com
To: linux-kernel@vger.kernel.org
Cc: "Rafael J. Wysocki" <rjw@rjwysocki.net>
Cc: Len Brown <lenb@kernel.org>
Cc: James Morse <james.morse@arm.com>
Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
Reviewed-by: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
[pmladek@suse.com: two more indentation fixes]
Signed-off-by: Petr Mladek <pmladek@suse.com>
Destroy ghes_estatus_pool and release memory allocated via vmalloc() on
errors in ghes_estatus_pool_init() in order to avoid memory leaks.
[ bp: do the labels properly and with descriptive names and massage. ]
Signed-off-by: Liguang Zhang <zhangliguang@linux.alibaba.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/1563173924-47479-1-git-send-email-zhangliguang@linux.alibaba.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
This is a missed part of the commit 5b53696a30
("ACPI / APEI: Switch to use new generic UUID API"), i.e.
replacing old definition with a global constant variable.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: James Morse <james.morse@arm.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Function __ghes_check_estatus() is always called after
__ghes_peek_estatus(), but it is already called in __ghes_peek_estatus().
So we should remove some needless __ghes_check_estatus() calls.
Signed-off-by: Liguang Zhang <zhangliguang@linux.alibaba.com>
Reviewed-by: James Morse <james.morse@arm.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Based on 1 normalized pattern(s):
this file is licensed under gplv2
extracted by the scancode license scanner the SPDX license identifier
GPL-2.0-only
has been chosen to replace the boilerplate/reference in 22 file(s).
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Armijn Hemel <armijn@tjaldur.nl>
Reviewed-by: Allison Randal <allison@lohutok.net>
Cc: linux-spdx@vger.kernel.org
Link: https://lkml.kernel.org/r/20190531190115.129548190@linutronix.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Based on 1 normalized pattern(s):
this program is free software you can redistribute it and or modify
it under the terms of the gnu general public license version 2 as
published by the free software foundation this program is
distributed in the hope that it will be useful but without any
warranty without even the implied warranty of merchantability or
fitness for a particular purpose see the gnu general public license
for more details
extracted by the scancode license scanner the SPDX license identifier
GPL-2.0-only
has been chosen to replace the boilerplate/reference in 655 file(s).
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Allison Randal <allison@lohutok.net>
Reviewed-by: Kate Stewart <kstewart@linuxfoundation.org>
Reviewed-by: Richard Fontana <rfontana@redhat.com>
Cc: linux-spdx@vger.kernel.org
Link: https://lkml.kernel.org/r/20190527070034.575739538@linutronix.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
* acpi-apei: (29 commits)
efi: cper: Fix possible out-of-bounds access
ACPI: APEI: Fix possible out-of-bounds access to BERT region
MAINTAINERS: Add James Morse to the list of APEI reviewers
ACPI / APEI: Add support for the SDEI GHES Notification type
firmware: arm_sdei: Add ACPI GHES registration helper
ACPI / APEI: Use separate fixmap pages for arm64 NMI-like notifications
ACPI / APEI: Only use queued estatus entry during in_nmi_queue_one_entry()
ACPI / APEI: Split ghes_read_estatus() to allow a peek at the CPER length
ACPI / APEI: Make GHES estatus header validation more user friendly
ACPI / APEI: Pass ghes and estatus separately to avoid a later copy
ACPI / APEI: Let the notification helper specify the fixmap slot
ACPI / APEI: Move locking to the notification helper
arm64: KVM/mm: Move SEA handling behind a single 'claim' interface
KVM: arm/arm64: Add kvm_ras.h to collect kvm specific RAS plumbing
ACPI / APEI: Switch NOTIFY_SEA to use the estatus queue
ACPI / APEI: Move NOTIFY_SEA between the estatus-queue and NOTIFY_NMI
ACPI / APEI: Don't allow ghes_ack_error() to mask earlier errors
ACPI / APEI: Generalise the estatus queue's notify code
ACPI / APEI: Don't update struct ghes' flags in read/clear estatus
ACPI / APEI: Remove spurious GHES_TO_CLEAR check
...
Check that the length recorded in the generic error status block is
within the region before checking the contents of the region itself.
Otherwise it may result in an out-of-bounds access if the system
firmware has generated a status block with an invalid length (larger
than the mapped region). Also move the block_status check so that it
only happens after the block has been verified to be within the mapped
region.
Signed-off-by: Ross Lagerwall <ross.lagerwall@citrix.com>
Acked-by: Borislav Petkov <bp@suse.de>
Tested-by: Tyler Baicar <baicar.tyler@gmail.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
If the GHES notification type is SDEI, register the provided event
using the SDEI-GHES helper.
SDEI may be one of two types of event, normal and critical. Critical
events can interrupt normal events, so these must have separate
fixmap slots and locks in case both event types are in use.
Signed-off-by: James Morse <james.morse@arm.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Now that ghes notification helpers provide the fixmap slots and
take the lock themselves, multiple NMI-like notifications can
be used on arm64.
These should be named after their notification method as they can't
all be called 'NMI'. x86's NOTIFY_NMI already is, change the SEA
fixmap entry to be called FIX_APEI_GHES_SEA.
Future patches can add support for FIX_APEI_GHES_SEI and
FIX_APEI_GHES_SDEI_{NORMAL,CRITICAL}.
Because all of ghes.c builds on both architectures, provide a
constant for each fixmap entry that the architecture will never
use.
Signed-off-by: James Morse <james.morse@arm.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Each struct ghes has an worst-case sized buffer for storing the
estatus. If an error is being processed by ghes_proc() in process
context this buffer will be in use. If the error source then triggers
an NMI-like notification, the same buffer will be used by
in_nmi_queue_one_entry() to stage the estatus data, before
__process_error() copys it into a queued estatus entry.
Merge __process_error()s work into in_nmi_queue_one_entry() so that
the queued estatus entry is used from the beginning. Use the new
ghes_peek_estatus() to know how much memory to allocate from
the ghes_estatus_pool before reading the records.
Reported-by: Borislav Petkov <bp@suse.de>
Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: Borislav Petkov <bp@suse.de>
Change since v6:
* Added a comment explaining the 'ack-error, then goto no_work'.
* Added missing esatus-clearing, which is necessary after reading the GAS,
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
ghes_read_estatus() reads the record address, then the record's
header, then performs some sanity checks before reading the
records into the provided estatus buffer.
To provide this estatus buffer the caller must know the size of the
records in advance, or always provide a worst-case sized buffer as
happens today for the non-NMI notifications.
Add a function to peek at the record's header to find the size. This
will let the NMI path allocate the right amount of memory before reading
the records, instead of using the worst-case size, and having to copy
the records.
Split ghes_read_estatus() to create __ghes_peek_estatus() which
returns the address and size of the CPER records.
Signed-off-by: James Morse <james.morse@arm.com>
Changes since v7:
* Grammar
* concistent argument ordering
Changes since v6:
* Additional buf_addr = 0 error handling
* Moved checking out of peek-estatus
* Reworded an error message so we can tell them apart
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
ghes_read_estatus() checks various lengths in the top-level header to
ensure the CPER records to be read aren't obviously corrupt.
Take the opportunity to make this more user-friendly, printing a
(ratelimited) message about the nature of the header format error.
Suggested-by: Borislav Petkov <bp@alien8.de>
Signed-off-by: James Morse <james.morse@arm.com>
[ rjw: Add missing 'static' ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The NMI-like notifications scribble over ghes->estatus, before
copying it somewhere else. If this interrupts the ghes_probe() code
calling ghes_proc() on each struct ghes, the data is corrupted.
All the NMI-like notifications should use a queued estatus entry
from the beginning, instead of the ghes version, then copying it.
To do this, break up any use of "ghes->estatus" so that all
functions take the estatus as an argument.
This patch just moves these ghes->estatus dereferences into separate
arguments, no change in behaviour. struct ghes becomes unused in
ghes_clear_estatus() as it only wanted ghes->estatus, which we now
pass directly. This is removed.
Signed-off-by: James Morse <james.morse@arm.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
ghes_copy_tofrom_phys() uses a different fixmap slot depending on in_nmi().
This doesn't work when there are multiple NMI-like notifications, that
could interrupt each other.
As with the locking, move the chosen fixmap_idx to the notification helper.
This only matters for NMI-like notifications, anything calling
ghes_proc() can use the IRQ fixmap slot as its already holding an irqsave
spinlock.
This lets us collapse the ghes_ioremap_pfn_*() helpers.
Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
ghes_copy_tofrom_phys() takes different locks depending on in_nmi().
This doesn't work if there are multiple NMI-like notifications, that
can interrupt each other.
Now that NOTIFY_SEA is always called in the same context, move the
lock-taking to the notification helper. The helper will always know
which lock to take. This avoids ghes_copy_tofrom_phys() taking a guess
based on in_nmi().
This splits NOTIFY_NMI and NOTIFY_SEA to use different locks. All
the other notifications use ghes_proc(), and are called in process
or IRQ context. Move the spin_lock_irqsave() around their ghes_proc()
calls.
Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Now that the estatus queue can be used by more than one notification
method, we can move notifications that have NMI-like behaviour over.
Switch NOTIFY_SEA over to use the estatus queue. This makes it behave
in the same way as x86's NOTIFY_NMI.
Remove Kconfig's ability to turn ACPI_APEI_SEA off if ACPI_APEI_GHES
is selected. This roughly matches the x86 NOTIFY_NMI behaviour, and means
each architecture has at least one user of the estatus-queue, meaning it
doesn't need guarding with ifdef.
Signed-off-by: James Morse <james.morse@arm.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The estatus-queue code is currently hidden by the NOTIFY_NMI #ifdefs.
Once NOTIFY_SEA starts using the estatus-queue we can stop hiding
it as each architecture has a user that can't be turned off.
Split the existing CONFIG_HAVE_ACPI_APEI_NMI block in two, and move
the SEA code into the gap.
Move the code around ... and changes the stale comment describing
why the status queue is necessary: printk() is no longer the issue,
its the helpers like memory_failure_queue() that aren't nmi safe.
Signed-off-by: James Morse <james.morse@arm.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
During ghes_proc() we use ghes_ack_error() to tell an external agent
we are done with these records and it can re-use the memory.
rc may hold an error returned by ghes_read_estatus(), ENOENT causes
us to skip ghes_ack_error() (as there is nothing to ack), but rc may
also by EIO, which gets supressed.
ghes_clear_estatus() is where we mark the records as processed for
non GHESv2 error sources, and already spots the ENOENT case as
buf_paddr is set to 0 by ghes_read_estatus().
Move the ghes_ack_error() call in here to avoid extra logic with
the return code in ghes_proc().
This enables GHESv2 acking for NMI-like error sources. This is safe
as the buffer is pre-mapped by map_gen_v2() before the GHES is added
to any NMI handler lists.
This same pre-mapping step means we can't receive an error from
apei_read()/write() here as apei_check_gar() succeeded when it
was mapped, and the mapping was cached, so the address can't be
rejected at runtime. Remove the error-returns as this is now
called from a function with no return.
Signed-off-by: James Morse <james.morse@arm.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Refactor the estatus queue's pool notification routine from
NOTIFY_NMI's handlers. This will allow another notification
method to use the estatus queue without duplicating this code.
Add rcu_read_lock()/rcu_read_unlock() around the list
list_for_each_entry_rcu() walker. These aren't strictly necessary as
the whole nmi_enter/nmi_exit() window is a spooky RCU read-side
critical section.
in_nmi_queue_one_entry() is separate from the rcu-list walker for a
later caller that doesn't need to walk a list.
Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: Punit Agrawal <punit.agrawal@arm.com>
Tested-by: Tyler Baicar <tbaicar@codeaurora.org>
[ rjw: Drop unnecessary err variable in two places ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
ghes_read_estatus() sets a flag in struct ghes if the buffer of
CPER records needs to be cleared once the records have been
processed. This flag value is a problem if a struct ghes can be
processed concurrently, as happens at probe time if an NMI arrives
for the same error source. The NMI clears the flag, meaning the
interrupted handler may never do the ghes_estatus_clear() work.
The GHES_TO_CLEAR flags is only set at the same time as
buffer_paddr, which is now owned by the caller and passed to
ghes_clear_estatus(). Use this value as the flag.
A non-zero buf_paddr returned by ghes_read_estatus() means
ghes_clear_estatus() should clear this address. ghes_read_estatus()
already checks for a read of error_status_address being zero,
so CPER records cannot be written here.
Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
ghes_notify_nmi() checks ghes->flags for GHES_TO_CLEAR before going
on to __process_error(). This is pointless as ghes_read_estatus()
will always set this flag if it returns success, which was checked
earlier in the loop. Remove it.
Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
When CPER records are found the address of the records is stashed
in the struct ghes. Once the records have been processed, this
address is overwritten with zero so that it won't be processed
again without being re-populated by firmware.
This goes wrong if a struct ghes can be processed concurrently,
as can happen at probe time when an NMI occurs. If the NMI arrives
on another CPU, the probing CPU may call ghes_clear_estatus() on the
records before the handler had finished with them.
Even on the same CPU, once the interrupted handler is resumed, it
will call ghes_clear_estatus() on the NMIs records, this memory may
have already been re-used by firmware.
Avoid this stashing by letting the caller hold the address. A
later patch will do away with the use of ghes->flags in the
read/clear code too.
Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Adding new NMI-like notifications duplicates the calls that grow
and shrink the estatus pool. This is all pretty pointless, as the
size is capped to 64K. Allocate this for each ghes and drop
the code that grows and shrinks the pool.
Suggested-by: Borislav Petkov <bp@suse.de>
Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
ghes.c has a memory pool it uses for the estatus cache and the estatus
queue. The cache is initialised when registering the platform driver.
For the queue, an NMI-like notification has to grow/shrink the pool
as it is registered and unregistered.
This is all pretty noisy when adding new NMI-like notifications, it
would be better to replace this with a static pool size based on the
number of users.
As a precursor, move the call that creates the pool from ghes_init(),
into hest.c. Later this will take the number of ghes entries and
consolidate the queue allocations.
Remove ghes_estatus_pool_exit() as hest.c doesn't have anywhere to put
this.
The pool is now initialised as part of ACPI's subsys_initcall():
(acpi_init(), acpi_scan_init(), acpi_pci_root_init(), acpi_hest_init())
Before this patch it happened later as a GHES specific device_initcall().
Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The ghes code is careful to parse and round firmware's advertised
memory requirements for CPER records, up to a maximum of 64K.
However when ghes_estatus_pool_expand() does its work, it splits
the requested size into PAGE_SIZE granules.
This means if firmware generates 5K of CPER records, and correctly
describes this in the table, __process_error() will silently fail as it
is unable to allocate more than PAGE_SIZE.
Switch the estatus pool to vmalloc() memory. On x86 vmalloc() memory
may fault and be fixed up by vmalloc_fault(). To prevent this call
vmalloc_sync_all() before an NMI handler could discover the memory.
Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Subsequent patches will split up ghes_read_estatus(), at which
point passing around the 'silent' flag gets annoying. This is to
suppress prink() messages, which prior to commit 42a0bb3f71
("printk/nmi: generic solution for safe printk in NMI"), were
unsafe in NMI context.
This is no longer necessary, remove the flag. printk() messages
are batched in a per-cpu buffer and printed via irq-work, or a call
back from panic().
Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
oops_begin() exists to group printk() messages with the oops message
printed by die(). To reach this caller we know that platform firmware
took this error first, then notified the OS via NMI with a 'panic'
severity.
Don't wait for another CPU to release the die-lock before panic()ing,
our only goal is to print this fatal error and panic().
This code is always called in_nmi(), and since commit 42a0bb3f71
("printk/nmi: generic solution for safe printk in NMI"), it has been
safe to call printk() from this context. Messages are batched in a
per-cpu buffer and printed via irq-work, or a call back from panic().
Link: https://patchwork.kernel.org/patch/10313555/
Acked-by: Borislav Petkov <bp@suse.de>
Signed-off-by: James Morse <james.morse@arm.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
When calling debugfs functions, there is no need to ever check the
return value. The function can work or not, but the code logic should
never do something different based on this.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
