Commit 18996f2d ("ACPICA: Events: Stop unconditionally clearing
ACPI IRQs during suspend/resume") was added to stop clearing event
status bits unconditionally in the system-wide suspend and resume
paths. This was done because of an issue with a laptop lid appaering
to be closed even when it was used to wake up the system from suspend
(see https://bugzilla.kernel.org/show_bug.cgi?id=196249), which
happened because event status bits were cleared unconditionally on
system resume. Though this change fixed the issue in the resume path,
it introduced regressions in a few suspend paths.

First regression was reported and fixed in the S5 entry path by commit
fa85015c ("ACPICA: Clear status of all events when entering S5").
Next regression was reported and fixed for all legacy sleep paths by
commit f317c7dc ("ACPICA: Clear status of all events when entering
sleep states").  However, there still is a suspend-to-idle regression,
since suspend-to-idle does not follow the legacy sleep paths.

In the suspend-to-idle case, wakeup is enabled as part of device
suspend.  If the status bits of wakeup GPEs are set when they are
enabled, it causes a premature system wakeup to occur.

To address that problem, partially revert commit 18996f2d to
restore GPE status bits clearing before the GPE is enabled in
acpi_ev_enable_gpe().

Fixes: 18996f2d ("ACPICA: Events: Stop unconditionally clearing ACPI IRQs during suspend/resume")
Signed-off-by: NFurquan Shaikh <furquan@google.com>
Cc: 4.17+ <stable@vger.kernel.org> # 4.17+
[ rjw: Subject & changelog ]
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

When commit 8661423e ("ACPI / utils: Add new acpi_dev_present
helper") introduced acpi_dev_present(), it missed the fact that
bus_find_device() took a reference on the device found by it and
the callers of acpi_dev_present() don't drop that reference.

Drop the reference on the device in acpi_dev_present().

Fixes: 8661423e ("ACPI / utils: Add new acpi_dev_present helper")
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: NHans de Goede <hdegoede@redhat.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

The return value of get_status() is passed to user space on errors,
so it should not return acpi_status values then.  Make it return
error values that are meaningful for user space instead.

This also makes a Clang warning regarding the initialization of a
local variable in get_status() go away.
Reported-by: NNathan Chancellor <natechancellor@gmail.com>
Reviewed-by: NNathan Chancellor <natechancellor@gmail.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

If SSDT overlay is loaded via ConfigFS and then unloaded the device,
we would like to have OF modalias for, already gone. Thus, acpi_get_name()
returns no allocated buffer for such case and kernel crashes afterwards:

 ACPI: Host-directed Dynamic ACPI Table Unload
 ads7950 spi-PRP0001:00: Dropping the link to regulator.0
 BUG: unable to handle kernel NULL pointer dereference at 0000000000000000
 #PF error: [normal kernel read fault]
 PGD 80000000070d6067 P4D 80000000070d6067 PUD 70d0067 PMD 0
 Oops: 0000 [#1] SMP PTI
 CPU: 0 PID: 40 Comm: kworker/u4:2 Not tainted 5.0.0+ #96
 Hardware name: Intel Corporation Merrifield/BODEGA BAY, BIOS 542 2015.01.21:18.19.48
 Workqueue: kacpi_hotplug acpi_device_del_work_fn
 RIP: 0010:create_of_modalias.isra.1+0x4c/0x150
 Code: 00 00 48 89 44 24 18 31 c0 48 8d 54 24 08 48 c7 44 24 10 00 00 00 00 48 c7 44 24 08 ff ff ff ff e8 7a b0 03 00 48 8b 4c 24 10 <0f> b6 01 84 c0 74 27 48 c7 c7 00 09 f4 a5 0f b6 f0 8d 50 20 f6 04
 RSP: 0000:ffffa51040297c10 EFLAGS: 00010246
 RAX: 0000000000001001 RBX: 0000000000000785 RCX: 0000000000000000
 RDX: 0000000000001001 RSI: 0000000000000286 RDI: ffffa2163dc042e0
 RBP: ffffa216062b1196 R08: 0000000000001001 R09: ffffa21639873000
 R10: ffffffffa606761d R11: 0000000000000001 R12: ffffa21639873218
 R13: ffffa2163deb5060 R14: ffffa216063d1010 R15: 0000000000000000
 FS:  0000000000000000(0000) GS:ffffa2163e000000(0000) knlGS:0000000000000000
 CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
 CR2: 0000000000000000 CR3: 0000000007114000 CR4: 00000000001006f0
 Call Trace:
  __acpi_device_uevent_modalias+0xb0/0x100
  spi_uevent+0xd/0x40

 ...

In order to fix above let create_of_modalias() check the status returned
by acpi_get_name() and bail out in case of failure.

Fixes: 8765c5ba ("ACPI / scan: Rework modalias creation when "compatible" is present")
Link: https://bugzilla.kernel.org/show_bug.cgi?id=201381Reported-by: NFerry Toth <fntoth@gmail.com>
Tested-by: Ferry Toth<fntoth@gmail.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Cc: 4.1+ <stable@vger.kernel.org> # 4.1+
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

There is no need to have non-static local data structures. otherwise
sparse is not happy:

  CHECK   drivers/acpi/acpi_configfs.c
drivers/acpi/acpi_configfs.c:100:31: warning: symbol 'acpi_table_bin_attrs' was not declared. Should it be static?
drivers/acpi/acpi_configfs.c:196:27: warning: symbol 'acpi_table_attrs' was not declared. Should it be static?
drivers/acpi/acpi_configfs.c:236:34: warning: symbol 'acpi_table_group_ops' was not declared. Should it be static?
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

There is no need to have non-static local functions. otherwise
compiler is not happy:

  CC [M]  drivers/acpi/acpi_configfs.o
drivers/acpi/acpi_configfs.c:105:9: warning: no previous prototype for ‘acpi_table_signature_show’ [-Wmissing-prototypes]
 ssize_t acpi_table_signature_show(struct config_item *cfg, char *str)
         ^~~~~~~~~~~~~~~~~~~~~~~~~
drivers/acpi/acpi_configfs.c:115:9: warning: no previous prototype for ‘acpi_table_length_show’ [-Wmissing-prototypes]
 ssize_t acpi_table_length_show(struct config_item *cfg, char *str)
         ^~~~~~~~~~~~~~~~~~~~~~
...
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

ACPI 6.3 bumps the PPTT table revision and adds a LEAF_NODE flag.

This allows us to avoid a second pass through the table to assure
that the node in question is a leaf.
Signed-off-by: NJeremy Linton <jeremy.linton@arm.com>
Reviewed-by: NSudeep Holla <sudeep.holla@arm.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

ACPI NFIT flags field reports major errors on NVDIMM, which need
user's attention.

Update the current log to a proper error message with dev_err().
The current message string is kept for grep-compatibility.
Signed-off-by: NToshi Kani <toshi.kani@hpe.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: "Rafael J. Wysocki" <rjw@rjwysocki.net>
Cc: Robert Elliott <elliott@hpe.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

In the PRP0001 case, the compatible string may have additional data
affiliated with the device.  When we call device_get_match_data() on
such device, we will get nothing since currently
acpi_device_get_match_data() doesn't respect PRP0001.

To fix the above, try acpi_of_match_device() if there is no ACPI
table in the driver.

Anyway, note that the device is expected to get its own proper
ACPI ID.
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Acked-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Signed-off-by: NErik Schmauss <erik.schmauss@intel.com>
Signed-off-by: NBob Moore <robert.moore@intel.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

ACPICA commit 205ac8fc721073f1e609df963b14ef2237aeba73

Link: https://github.com/acpica/acpica/commit/205ac8fcReviewed-by: NColin Ian King <colin.king@canonical.com>
Signed-off-by: NErik Schmauss <erik.schmauss@intel.com>
Signed-off-by: NBob Moore <robert.moore@intel.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

ACPICA commit a4849944e80f97970e99843f4975850753584a4e

This change adds PCC operation region support in the AML interpreter
and a default handler for acpiexec. According to the specification,
the PCC operation region performs a transaction when the COMD field
is written. This allows ASL to write data to other fields before
sending the data.

In order to accommodate this protocol, a temorary buffer is added
to the regionfield object to accumulate writes. If any offset that
spans COMD is written, the temporary buffer is sent to the PCC
operation region handler to be processed.

This change also renames the PCC keyword to platform_comm_channel.

Link: https://github.com/acpica/acpica/commit/a4849944Reviewed-by: NKyle Pelton <kyle.d.pelton@intel.com>
Signed-off-by: NErik Schmauss <erik.schmauss@intel.com>
Signed-off-by: NBob Moore <robert.moore@intel.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

clk-lpss.h is solely x86 related header. Move it to correct folder.
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: NStephen Boyd <sboyd@kernel.org>

Gate ARS result consumption on whether the OS issued start-ARS since the
previous consumption. The BIOS may only clear its result buffers after a
successful start-ARS.

Fixes: 0caeef63 ("libnvdimm: Add a poison list and export badblocks")
Cc: <stable@vger.kernel.org>
Reported-by: NKrzysztof Rusocki <krzysztof.rusocki@intel.com>
Reported-by: NVishal Verma <vishal.l.verma@intel.com>
Reviewed-by: NToshi Kani <toshi.kani@hpe.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

The ARS implementation implements exponential back-off on the poll
interval to prevent high-frequency access to the DIMM / platform
interface. Depending on when the ARS completes the poll interval may
exceed the completion event by minutes. Allow root to reset the timeout
each time it probes the status. A one-second timeout is still enforced,
but root can otherwise can control the poll interval.

Fixes: bc6ba808 ("nfit, address-range-scrub: rework and simplify ARS...")
Cc: <stable@vger.kernel.org>
Reported-by: NErwin Tsaur <erwin.tsaur@oracle.com>
Reviewed-by: NToshi Kani <toshi.kani@hpe.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

In preparation for introducing new flags to gate whether ARS results are
stale, or poll the completion state, convert the existing flags to an
unsigned long with enumerated values. This conversion allows the flags
to be atomically updated outside of ->init_mutex.
Reviewed-by: NToshi Kani <toshi.kani@hpe.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

The ars_start_flags property of 'struct acpi_nfit_desc' is no longer
used since ARS_REQ_SHORT and ARS_REQ_LONG were added.
Reviewed-by: NToshi Kani <toshi.kani@hpe.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

The no_init_ars option is meant to prevent long-ARS, but short-ARS
should be allowed to grab any immediate results.

Fixes: bc6ba808 ("nfit, address-range-scrub: rework and simplify ARS...")
Cc: <stable@vger.kernel.org>
Reported-by: NErwin Tsaur <erwin.tsaur@oracle.com>
Reviewed-by: NToshi Kani <toshi.kani@hpe.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

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: NRoss Lagerwall <ross.lagerwall@citrix.com>
Acked-by: NBorislav Petkov <bp@suse.de>
Tested-by: NTyler Baicar <baicar.tyler@gmail.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

The ACPI idle driver will fallback to using the legacy P_LVL* SystemIO
method of entering C-states if the _CST method is disabled and P_BLK is
defined. However, in this case the C2 and C3 states won't have a
description set, so the user will see "<null>" when reading the
description from sysfs.

Give each of these states a description.
Signed-off-by: NYazen Ghannam <yazen.ghannam@amd.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

This patch add a helper to get the value of desired performance
register.
Signed-off-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
[ rjw: More white space ]
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

ACPICA commit 0015e2491bda996ddb9d56bfa4ee39644acbb22b

Link: https://github.com/acpica/acpica/commit/0015e249Signed-off-by: NErik Schmauss <erik.schmauss@intel.com>
Signed-off-by: NBob Moore <robert.moore@intel.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

ACPICA commit 2efd616e5b1c960f407763e6782f7dc259ea55df

Attempting to improve error messages to clarify that errors
are bubbled up from the original error, possibly across nested
methods.

Link: https://github.com/acpica/acpica/commit/2efd616eSigned-off-by: NBob Moore <robert.moore@intel.com>
Signed-off-by: NErik Schmauss <erik.schmauss@intel.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

ACPICA commit 349dd29335d6928f883bc95c614a0edd033141bb

 - Fault on Field Units
 - Some restructuring
 - General cleanup of dbtest module

Link: https://github.com/acpica/acpica/commit/349dd293Signed-off-by: NBob Moore <robert.moore@intel.com>
Signed-off-by: NErik Schmauss <erik.schmauss@intel.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

ACPICA commit 387c850c5d49d09d7c2e70b2711e584ad83956a1

Nothing can be done with such a region. Just emit a warning so as
not to abort a table load or running method.

Link: https://github.com/acpica/acpica/commit/387c850cSigned-off-by: NBob Moore <robert.moore@intel.com>
Signed-off-by: NErik Schmauss <erik.schmauss@intel.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

ACPICA commit 47f5607c204719d9239a12b889df725225098c8f

Module-level code refers to executable ASL code that runs during
table load. This is typically used in ASL to declare named objects
based on a condition evaluated during table load like so:

definition_block(...)
{
  opreation_region (OPR1, system_memory, ...)
  Field (OPR1)
  {
    FLD1, 8 /* Assume that FLD1's value is 0x1 */
  }

  /* The if statement below is referred to as module-level code */

  If (FLD1)
  {
    /* Declare DEV1 conditionally */
    Device (DEV1) {...}
  }

  Device (DEV2)
  {
    ...
  }
}

In legacy module-level code, the execution of the If statement
was deferred after other modules were loaded. The order of
code execution for the table above is the following:

1.) Load OPR1 to the ACPI Namespace
2.) Load FLD1 to the ACPI Namespace (not intended for drivers)
3.) Load DEV2 to the ACPI Namespace
4.) Execute If (FLD1) and load DEV1 if the condition is true

This legacy approach can be problematic for tables that look like the
following:

definition_block(...)
{
  opreation_region (OPR1, system_memory, ...)
  Field (OPR1)
  {
    FLD1, 8 /* Assume that FLD1's value is 0x1 */
  }

  /* The if statement below is referred to as module-level code */

  If (FLD1)
  {
    /* Declare DEV1 conditionally */
    Device (DEV1) {...}
  }

  Scope (DEV1)
  {
    /* Add objects DEV1's scope */
    Name (OBJ1, 0x1234)
  }
}

When loading this in the legacy approach, Scope DEV1 gets evaluated
before the If statement. The following is the order of execution:

1.) Load OPR1 to the ACPI Namespace
2.) Load FLD1 to the ACPI Namespace (not intended for drivers)
3.) Add OBJ1 under DEV1's scope -- ERROR. DEV1 does not exist
4.) Execute If (FLD1) and load DEV1 if the condition is true

The legacy approach can never succeed for tables like this due to the
deferral of the module-level code. Due to this limitation, a new
module-level code was developed. This new approach exeutes if
statements in the order that they appear in the definition block.
With this approach, the order of execution for the above defintion
block is as follows:

1.) Load OPR1 to the ACPI Namespace
2.) Load FLD1 to the ACPI Namespace (not intended for drivers)
3.) Execute If (FLD1) and load DEV1 because the condition is true
4.) Add OBJ1 under DEV1's scope.

Since DEV1 is loaded in the namespace in step 3, step 4 executes
successfully.

This change removes support for the legacy module-level code
execution. From this point onward, the new module-level code
execution will be the official approach.

Link: https://github.com/acpica/acpica/commit/47f5607cSigned-off-by: NErik Schmauss <erik.schmauss@intel.com>
Signed-off-by: NBob Moore <robert.moore@intel.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Lenovo Yoga Book uses PWM2 for controlling keyboard backlight but this
device is hidden in the DSDT in Windows version of notebook (if OSID == 1).

Make this device always present for this notebook.
Signed-off-by: NYauhen Kharuzhy <jekhor@gmail.com>
Reviewed-by: NHans de Goede <hdegoede@redhat.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Commit 53fa1f6e ("ACPI / video: Only default only_lcd to true on
Win8-ready _desktops_") introduced chassis type detection, limiting the
lcd_only check for the backlight to devices where the chassis-type
indicates their is no builtin LCD panel.

The purpose of the lcd_only check is to avoid advertising a backlight
interface on desktops, since skylake and newer machines seem to always
have a backlight interface even if there is no LCD panel. The limiting
of this check to desktops only was done to avoid breaking backlight
support on some laptops which do not have the lcd flag set.

The Fujitsu ESPRIMO Q910 which is a compact (NUC like) desktop machine
has a chassis type of 0x10 aka "Lunch Box". Without the lcd_only check
we end up falsely advertising backlight/brightness control on this
device. This commit extend the dmi_is_desktop check to return true
for type 0x10 to fix this.

Fixes: 53fa1f6e ("ACPI / video: Only default only_lcd to true ...")
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

This commit refactors the chassis-type detection introduced by
commit 53fa1f6e ("ACPI / video: Only default only_lcd to true on
Win8-ready _desktops_") (where desktop means anything without a builtin
screen).

The DMI chassis_type is an unsigned integer, so rather then doing a
whole bunch of string-compares on it, convert it to an int and feed
the result to a switch case.

Note the switch case uses hex values, this is done because the spec
uses hex values too. This changes the check for "Main Server Chassis"
from checking for 11 decimal to 11 hexadecimal, this is a bug fix,
the original check for 11 decimal was wrong.

Fixes: 53fa1f6e ("ACPI / video: Only default only_lcd to true ...")
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
[ rjw: Drop redundant return statements ]
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

If query-ARS reports that ARS has stopped and requires continuation
attempt to retrieve short-ARS results before continuing the long
operation.

Fixes: bc6ba808 ("nfit, address-range-scrub: rework and simplify ARS...")
Cc: <stable@vger.kernel.org>
Reported-by: NKrzysztof Rusocki <krzysztof.rusocki@intel.com>
Reviewed-by: NToshi Kani <toshi.kani@hpe.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

Recent fixes to command handling enabled Linux to read label
configurations that it could not before. Unfortunately that means that
configurations that were operating in label-less mode will be broken as
the kernel ignores the existing namespace configuration and tries to
honor the new found labels.

Fortunately this seems limited to a case where Linux can quirk the
behavior and maintain the existing label-less semantics by default.
When the platform does not emit an _LSW method, disable all label access
methods. Provide a 'force_labels' module parameter to allow read-only
label operation.

Fixes: 11189c10 ("acpi/nfit: Fix command-supported detection")
Reported-by: NDexuan Cui <decui@microsoft.com>
Reviewed-by: NDexuan Cui <decui@microsoft.com>
Signed-off-by: NDan Williams <dan.j.williams@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: NJames Morse <james.morse@arm.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Commit 11189c10 "acpi/nfit: Fix command-supported detection" broke
ND_CMD_CALL for bus-level commands. The "func = cmd" assumption is only
valid for:

    ND_CMD_ARS_CAP
    ND_CMD_ARS_START
    ND_CMD_ARS_STATUS
    ND_CMD_CLEAR_ERROR

The function number otherwise needs to be pulled from the command
payload for:

    NFIT_CMD_TRANSLATE_SPA
    NFIT_CMD_ARS_INJECT_SET
    NFIT_CMD_ARS_INJECT_CLEAR
    NFIT_CMD_ARS_INJECT_GET

Update cmd_to_func() for the bus case and call it in the common path.

Fixes: 11189c10 ("acpi/nfit: Fix command-supported detection")
Cc: <stable@vger.kernel.org>
Reviewed-by: NVishal Verma <vishal.l.verma@intel.com>
Reported-by: NGrzegorz Burzynski <grzegorz.burzynski@intel.com>
Tested-by: NJeff Moyer <jmoyer@redhat.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

For a system using ACPI-based FW without a PPTT, we may get many warnings
about the lack of a PPTT, as shown:

root@(none)$ dmesg | grep -i pptt
[    0.010125] ACPI PPTT: No PPTT table found, cpu topology may be inaccurate
[    7.138339] ACPI PPTT: No PPTT table found, cache topology may be inaccurate
[    7.145368] ACPI PPTT: No PPTT table found, cache topology may be inaccurate

These logs are generated with pr_warn_once(), so the intention was for a
single log, but the logs overlap, so consolidate them.
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Reviewed-by: NJeremy Linton <jeremy.linton@arm.com>
Signed-off-by: NRafael 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: NJames Morse <james.morse@arm.com>
Signed-off-by: NRafael 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: NBorislav Petkov <bp@suse.de>
Signed-off-by: NJames Morse <james.morse@arm.com>
Reviewed-by: NBorislav 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: NRafael 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: NJames 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: NRafael 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: NBorislav Petkov <bp@alien8.de>
Signed-off-by: NJames Morse <james.morse@arm.com>
[ rjw: Add missing 'static' ]
Signed-off-by: NRafael 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: NJames Morse <james.morse@arm.com>
Signed-off-by: NRafael 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: NJames Morse <james.morse@arm.com>
Reviewed-by: NBorislav Petkov <bp@suse.de>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>