Several functions can't fail anymore: ich9_pm_add_properties(),
device_add_bootindex_property(), ppc_compat_add_property(),
spapr_caps_add_properties(), PropertyInfo.create().  Drop their @errp
parameter.
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NEric Blake <eblake@redhat.com>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200505152926.18877-16-armbru@redhat.com>

The only way object_property_add() can fail is when a property with
the same name already exists.  Since our property names are all
hardcoded, failure is a programming error, and the appropriate way to
handle it is passing &error_abort.

Same for its variants, except for object_property_add_child(), which
additionally fails when the child already has a parent.  Parentage is
also under program control, so this is a programming error, too.

We have a bit over 500 callers.  Almost half of them pass
&error_abort, slightly fewer ignore errors, one test case handles
errors, and the remaining few callers pass them to their own callers.

The previous few commits demonstrated once again that ignoring
programming errors is a bad idea.

Of the few ones that pass on errors, several violate the Error API.
The Error ** argument must be NULL, &error_abort, &error_fatal, or a
pointer to a variable containing NULL.  Passing an argument of the
latter kind twice without clearing it in between is wrong: if the
first call sets an error, it no longer points to NULL for the second
call.  ich9_pm_add_properties(), sparc32_ledma_realize(),
sparc32_dma_realize(), xilinx_axidma_realize(), xilinx_enet_realize()
are wrong that way.

When the one appropriate choice of argument is &error_abort, letting
users pick the argument is a bad idea.

Drop parameter @errp and assert the preconditions instead.

There's one exception to "duplicate property name is a programming
error": the way object_property_add() implements the magic (and
undocumented) "automatic arrayification".  Don't drop @errp there.
Instead, rename object_property_add() to object_property_try_add(),
and add the obvious wrapper object_property_add().
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NEric Blake <eblake@redhat.com>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200505152926.18877-15-armbru@redhat.com>
[Two semantic rebase conflicts resolved]

Both qdev_connect_gpio_out_named() and device_set_realized() put
objects without a parent into the "/machine/unattached/" orphanage.

qdev_connect_gpio_out_named() needs a lengthy comment to explain how
it works.  It exploits that object_property_add_child() can fail only
when we got a parent already, and ignoring that error does what we
want.  True.  If it failed due to "duplicate property", we'd be in
trouble, but that would be a programming error.

device_set_realized() is cleaner: it checks whether we need a parent,
then calls object_property_add_child(), aborting on failure.  No need
for a comment, and programming errors get caught.

Change qdev_connect_gpio_out_named() to match.

Cc: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
Reviewed-by: NPhilippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200505152926.18877-14-armbru@redhat.com>

The "bcm2835-peripherals" device's .instance_init() method
bcm2835_peripherals_init() attempts to make two memory regions QOM
children of the device.  This is futile, because memory_region_init()
already did.  The errors are ignored (a later commit will change
that).  Drop the useless calls.

Cc: Peter Maydell <peter.maydell@linaro.org>
Cc: Andrew Baumann <Andrew.Baumann@microsoft.com>
Cc: "Philippe Mathieu-Daudé" <philmd@redhat.com>
Cc: qemu-arm@nongnu.org
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NPhilippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: NEric Blake <eblake@redhat.com>
Message-Id: <20200505152926.18877-13-armbru@redhat.com>

QOM object initialization runs .instance_init() for the type and all
its supertypes; see object_init_with_type().

Both TYPE_E1000_BASE and its concrete subtypes set .instance_init() to
e1000_instance_init().  For the concrete subtypes, it duly gets run
twice.  The second run fails, but the error gets ignored (a later
commit will change that).

Remove it from the subtypes.

Cc: Jason Wang <jasowang@redhat.com>
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Acked-by: NJason Wang <jasowang@redhat.com>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200505152926.18877-12-armbru@redhat.com>

isa_superio_realize() attempts to make isa-parallel and isa-serial QOM
children, but this does not work, because it calls
object_property_add_child() after realizing with qdev_init_nofail().
Realizing a device without a parent gives it one: it gets put into the
"/machine/unattached/" orphanage.  The extra
object_property_add_child() fails, and isa_superio_realize() ignores
the error.

Move the object_property_add_child() before qdev_init_nofail(), and
pass &error_abort.

For the other components, isa_superio_realize() doesn't even try.  Add
object_property_add_child() there.

This affects machines 40p, clipper and fulong2e.

For instance, fulong2e has its vt82c686b-superio (which is an
isa-superio) at /machine/unattached/device[9].  Before the patch, its
components are at /machine/unattached/device[10] .. [14].  Afterwards,
they are at
/machine/unattached/device[9]/{parallel0,serial0,serial1,isa-fdc,i8042}.

Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NPhilippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: NPhilippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200505152926.18877-11-armbru@redhat.com>

object_property_set_description() and
object_class_property_set_description() fail only when property @name
is not found.

There are 85 calls of object_property_set_description() and
object_class_property_set_description().  None of them can fail:

* 84 immediately follow the creation of the property.

* The one in spapr_rng_instance_init() refers to a property created in
  spapr_rng_class_init(), from spapr_rng_properties[].

Every one of them still gets to decide what to pass for @errp.

51 calls pass &error_abort, 32 calls pass NULL, one receives the error
and propagates it to &error_abort, and one propagates it to
&error_fatal.  I'm actually surprised none of them violates the Error
API.

What are we gaining by letting callers handle the "property not found"
error?  Use when the property is not known to exist is simpler: you
don't have to guard the call with a check.  We haven't found such a
use in 5+ years.  Until we do, let's make life a bit simpler and drop
the @errp parameter.
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NEric Blake <eblake@redhat.com>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200505152926.18877-8-armbru@redhat.com>
[One semantic rebase conflict resolved]

qom/object.c provides object_property_get_TYPE() and
object_property_set_TYPE() for a number of common types.  These are
all convenience wrappers around object_property_get_qobject() and
object_property_set_qobject().

Except for object_property_get_uint16List(), which is unusual in two ways:

* It bypasses object_property_get_qobject().  Fixable; the previous
  commit did it for object_property_get_enum())

* It stores the value through a parameter.  Its contract claims it
  returns the value, like the other functions do.  Also fixable.

Fixing is not worthwhile, though: object_property_get_uint16List() has
seen exactly one user in six years.

Convert the lone user to do its job with the generic
object_property_get_qobject(), and drop object_property_get_uint16List().
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Message-Id: <20200505152926.18877-6-armbru@redhat.com>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
[Commit message typo fixed]

Uses of gchar * in qom/object.h:

* ObjectProperty member @name

  Functions that take a property name argument all use char *.  Change
  the member to match.

* ObjectProperty member @type

  Functions that take a property type argument or return it all use
  char *.  Change the member to match.

* ObjectProperty member @description

  Functions that take a property description argument all use char *.
  Change the member to match.

* object_resolve_path_component() parameter @part

  Path components are property names.  Most callers pass char *
  arguments.  Change the parameter to match.  Adjust the few callers
  that pass gchar * to pass char *.

* Return value of object_get_canonical_path_component(),
  object_get_canonical_path()

  Most callers convert their return values right back to char *.
  Change the return value to match.  Adjust the few callers where that
  would add a conversion to gchar * to use char * instead.
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NEric Blake <eblake@redhat.com>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200505152926.18877-3-armbru@redhat.com>

kvm_arch_on_sigbus_vcpu() error injection uses source_id as
index in etc/hardware_errors to find out Error Status Data
Block entry corresponding to error source. So supported source_id
values should be assigned here and not be changed afterwards to
make sure that guest will write error into expected Error Status
Data Block.

Before QEMU writes a new error to ACPI table, it will check whether
previous error has been acknowledged. If not acknowledged, the new
errors will be ignored and not be recorded. For the errors section
type, QEMU simulate it to memory section error.
Signed-off-by: NDongjiu Geng <gengdongjiu@huawei.com>
Signed-off-by: NXiang Zheng <zhengxiang9@huawei.com>
Reviewed-by: NIgor Mammedov <imammedo@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Message-id: 20200512030609.19593-9-gengdongjiu@huawei.com
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

Record the GHEB address via fw_cfg file, when recording
a error to CPER, it will use this address to find out
Generic Error Data Entries and write the error.

In order to avoid migration failure, make hardware
error table address to a part of GED device instead
of global variable, then this address will be migrated
to target QEMU.
Acked-by: NXiang Zheng <zhengxiang9@huawei.com>
Signed-off-by: NDongjiu Geng <gengdongjiu@huawei.com>
Reviewed-by: NIgor Mammedov <imammedo@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Message-id: 20200512030609.19593-7-gengdongjiu@huawei.com
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

This patch builds Hardware Error Source Table(HEST) via fw_cfg blobs.
Now it only supports ARMv8 SEA, a type of Generic Hardware Error
Source version 2(GHESv2) error source. Afterwards, we can extend
the supported types if needed. For the CPER section, currently it
is memory section because kernel mainly wants userspace to handle
the memory errors.

This patch follows the spec ACPI 6.2 to build the Hardware Error
Source table. For more detailed information, please refer to
document: docs/specs/acpi_hest_ghes.rst

build_ghes_hw_error_notification() helper will help to add Hardware
Error Notification to ACPI tables without using packed C structures
and avoid endianness issues as API doesn't need explicit conversion.
Signed-off-by: NXiang Zheng <zhengxiang9@huawei.com>
Signed-off-by: NDongjiu Geng <gengdongjiu@huawei.com>
Reviewed-by: NIgor Mammedov <imammedo@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Message-id: 20200512030609.19593-6-gengdongjiu@huawei.com
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

This patch builds error_block_address and read_ack_register fields
in hardware errors table , the error_block_address points to Generic
Error Status Block(GESB) via bios_linker. The max size for one GESB
is 1kb, For more detailed information, please refer to
document: docs/specs/acpi_hest_ghes.rst

Now we only support one Error source, if necessary, we can extend to
support more.
Suggested-by: NLaszlo Ersek <lersek@redhat.com>
Signed-off-by: NXiang Zheng <zhengxiang9@huawei.com>
Reviewed-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: NIgor Mammedov <imammedo@redhat.com>
Signed-off-by: NDongjiu Geng <gengdongjiu@huawei.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Message-id: 20200512030609.19593-5-gengdongjiu@huawei.com
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

RAS Virtualization feature is not supported now, so
add a RAS machine option and disable it by default.
Reviewed-by: NPeter Maydell <peter.maydell@linaro.org>
Signed-off-by: NDongjiu Geng <gengdongjiu@huawei.com>
Signed-off-by: NXiang Zheng <zhengxiang9@huawei.com>
Reviewed-by: NJonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: NIgor Mammedov <imammedo@redhat.com>
Message-id: 20200512030609.19593-3-gengdongjiu@huawei.com
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

The little end UUID is used in many places, so make
NVDIMM_UUID_LE to a common macro to convert the UUID
to a little end array.
Reviewed-by: NXiang Zheng <zhengxiang9@huawei.com>
Signed-off-by: NDongjiu Geng <gengdongjiu@huawei.com>
Message-id: 20200512030609.19593-2-gengdongjiu@huawei.com
Reviewed-by: NPeter Maydell <peter.maydell@linaro.org>
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

Sonora Pass is a 2 socket x86 motherboard designed by Facebook
and supported by OpenBMC.  Strapping configuration was obtained
from hardware and i2c configuration is based on dts found at:

https://github.com/facebook/openbmc-linux/blob/1633c87b8ba7c162095787c988979b748ba65dc8/arch/arm/boot/dts/aspeed-bmc-facebook-sonorapass.dts

Booted a test image of http://github.com/facebook/openbmc to login
prompt.
Signed-off-by: NPatrick Williams <patrick@stwcx.xyz>
Reviewed-by: NAmithash Prasad <amithash@fb.com>
Reviewed-by: NCédric Le Goater <clg@kaod.org>
[PMM: fixed block comment style nit]
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

Add support for stream fragments.
Reviewed-by: NAlistair Francis <alistair.francis@wdc.com>
Signed-off-by: NEdgar E. Iglesias <edgar.iglesias@xilinx.com>
Message-Id: <20200506082513.18751-9-edgar.iglesias@gmail.com>

Stream descriptor by descriptor from memory instead of
buffering entire packets before pushing. This enables
non-packet streaming clients to work and also lifts the
limitation that our internal DMA buffer needs to be able
to hold entire packets.
Reviewed-by: NAlistair Francis <alistair.francis@wdc.com>
Signed-off-by: NEdgar E. Iglesias <edgar.iglesias@xilinx.com>
Message-Id: <20200506082513.18751-8-edgar.iglesias@gmail.com>

Add support for fragmented packets from the DMA.
Reviewed-by: NFrancisco Iglesias <frasse.iglesias@gmail.com>
Signed-off-by: NEdgar E. Iglesias <edgar.iglesias@xilinx.com>
Message-Id: <20200506082513.18751-7-edgar.iglesias@gmail.com>

Some stream clients stream an endless stream of data while
other clients stream data in packets. Stream interfaces
usually have a way to signal the end of a packet or the
last beat of a transfer.

This adds an end-of-packet flag to the push interface.
Reviewed-by: NAlistair Francis <alistair.francis@wdc.com>
Reviewed-by: NFrancisco Iglesias <frasse.iglesias@gmail.com>
Signed-off-by: NEdgar E. Iglesias <edgar.iglesias@xilinx.com>
Message-Id: <20200506082513.18751-6-edgar.iglesias@gmail.com>

Add DMA memory-region property to externally control what
address-space this DMA operates on.
Reviewed-by: NAlistair Francis <alistair.francis@wdc.com>
Reviewed-by: NFrancisco Iglesias <frasse.iglesias@gmail.com>
Signed-off-by: NEdgar E. Iglesias <edgar.iglesias@xilinx.com>
Reviewed-by: NPhilippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20200506082513.18751-5-edgar.iglesias@gmail.com>

Remove unncessary cast, buf is already uint8_t *.
No functional change.
Reviewed-by: NAlistair Francis <alistair.francis@wdc.com>
Reviewed-by: NFrancisco Iglesias <frasse.iglesias@gmail.com>
Signed-off-by: NEdgar E. Iglesias <edgar.iglesias@xilinx.com>
Reviewed-by: NPhilippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20200506082513.18751-4-edgar.iglesias@gmail.com>

Split the shared stream_class_init function to assign
stream->push with better type-safety.
Reviewed-by: NAlistair Francis <alistair.francis@wdc.com>
Reviewed-by: NFrancisco Iglesias <frasse.iglesias@gmail.com>
Signed-off-by: NEdgar E. Iglesias <edgar.iglesias@xilinx.com>
Reviewed-by: NPhilippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20200506082513.18751-3-edgar.iglesias@gmail.com>

Auto-clear PHY CR Autoneg bits. This makes this model
work with recent Linux kernels.
Reviewed-by: NFrancisco Iglesias <frasse.iglesias@gmail.com>
Signed-off-by: NEdgar E. Iglesias <edgar.iglesias@xilinx.com>
Message-Id: <20200506082513.18751-2-edgar.iglesias@gmail.com>

If delivery of some 9pfs response fails for some reason, log the
error message by mentioning the 9P protocol reply type, not by
client's request type. The latter could be misleading that the
error occurred already when handling the request input.
Signed-off-by: NChristian Schoenebeck <qemu_oss@crudebyte.com>
Acked-by: NStefano Stabellini <sstabellini@kernel.org>
Message-Id: <ad0e5a9b6abde52502aa40b30661d29aebe1590a.1589132512.git.qemu_oss@crudebyte.com>
Signed-off-by: NGreg Kurz <groug@kaod.org>

QEMU's local 9pfs server passes through O_NOATIME from the client. If
the QEMU process doesn't have permissions to use O_NOATIME (namely, it
does not own the file nor have the CAP_FOWNER capability), the open will
fail. This causes issues when from the client's point of view, it
believes it has permissions to use O_NOATIME (e.g., a process running as
root in the virtual machine). Additionally, overlayfs on Linux opens
files on the lower layer using O_NOATIME, so in this case a 9pfs mount
can't be used as a lower layer for overlayfs (cf.
https://github.com/osandov/drgn/blob/dabfe1971951701da13863dbe6d8a1d172ad9650/vmtest/onoatimehack.c
and https://github.com/NixOS/nixpkgs/issues/54509).

Luckily, O_NOATIME is effectively a hint, and is often ignored by, e.g.,
network filesystems. open(2) notes that O_NOATIME "may not be effective
on all filesystems. One example is NFS, where the server maintains the
access time." This means that we can honor it when possible but fall
back to ignoring it.
Acked-by: NChristian Schoenebeck <qemu_oss@crudebyte.com>
Signed-off-by: NOmar Sandoval <osandov@fb.com>
Message-Id: <e9bee604e8df528584693a4ec474ded6295ce8ad.1587149256.git.osandov@fb.com>
Signed-off-by: NGreg Kurz <groug@kaod.org>

Clean up wrong usage of FALSE and TRUE in places that use "bool" from stdbool.h.

FALSE and TRUE (with capital letters) are the constants defined by glib for
being used with the "gboolean" type of glib. But some parts of the code also use
TRUE and FALSE for variables that are declared as "bool" (the type from <stdbool.h>).
Signed-off-by: NJafar Abdi <cafer.abdi@gmail.com>
Reviewed-by: NThomas Huth <thuth@redhat.com>
Reviewed-by: NStefan Berger <stefanb@linux.ibm.com>
Signed-off-by: NStefan Berger <stefanb@linux.ibm.com>

I can't find proper documentation or datasheet, but it is likely
a MMIO mapped serial device mapped in the 0x80000000..0x8000ffff
range belongs to the SoC address space, thus is always mapped in
the memory bus.
Map the devices on the bus regardless a chardev is attached to it.
Signed-off-by: NPhilippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: NJan Kiszka <jan.kiszka@web.de>
Message-id: 20200505095945.23146-1-f4bug@amsat.org
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

Add trace event to display timer's counter value updates.
Signed-off-by: NPhilippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: NRichard Henderson <richard.henderson@linaro.org>
Message-id: 20200504072822.18799-5-f4bug@amsat.org
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

The NRF51 series SoC have 3 timer peripherals, each having
4 counters. To help differentiate which peripheral is accessed,
display the timer ID in the trace events.
Signed-off-by: NPhilippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: NRichard Henderson <richard.henderson@linaro.org>
Message-id: 20200504072822.18799-4-f4bug@amsat.org
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

On the NRF51 series, all peripherals have a fixed I/O size
of 4KiB. Define NRF51_PERIPHERAL_SIZE and use it.
Signed-off-by: NPhilippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: NRichard Henderson <richard.henderson@linaro.org>
Message-id: 20200504072822.18799-2-f4bug@amsat.org
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

The AST2600 handles this differently with the extra 'hardlock' state, so
move the testing to the soc specific class' write callback.
Signed-off-by: NJoel Stanley <joel@jms.id.au>
Reviewed-by: NCédric Le Goater <clg@kaod.org>
Message-id: 20200505090136.341426-1-joel@jms.id.au
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

There are minimal differences from Qemu's point of view between the A0
and A1 silicon revisions.

As the A1 exercises different code paths in u-boot it is desirable to
emulate that instead.
Signed-off-by: NJoel Stanley <joel@jms.id.au>
Reviewed-by: NAndrew Jeffery <andrew@aj.id.au>
Reviewed-by: NCédric Le Goater <clg@kaod.org>
Message-id: 20200504093703.261135-1-joel@jms.id.au
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

This is a boot stub that is similar to the code u-boot runs, allowing
the kernel to boot the secondary CPU.

u-boot works as follows:

 1. Initialises the SMP mailbox area in the SCU at 0x1e6e2180 with default values

 2. Copies a stub named 'mailbox_insn' from flash to the SCU, just above the
    mailbox area

 3. Sets AST_SMP_MBOX_FIELD_READY to a magic value to indicate the
    secondary can begin execution from the stub

 4. The stub waits until the AST_SMP_MBOX_FIELD_GOSIGN register is set to
    a magic value

 5. Jumps to the address in AST_SMP_MBOX_FIELD_ENTRY, starting Linux

Linux indicates it is ready by writing the address of its entrypoint
function to AST_SMP_MBOX_FIELD_ENTRY and the 'go' magic number to
AST_SMP_MBOX_FIELD_GOSIGN. The secondary CPU sees this at step 4 and
breaks out of it's loop.

To be compatible, a fixed qemu stub is loaded into the mailbox area. As
qemu can ensure the stub is loaded before execution starts, we do not
need to emulate the AST_SMP_MBOX_FIELD_READY behaviour of u-boot. The
secondary CPU's program counter points to the beginning of the stub,
allowing qemu to start secondaries at step four.

Reboot behaviour is preserved by resetting AST_SMP_MBOX_FIELD_GOSIGN
when the secondaries are reset.

This is only configured when the system is booted with -kernel and qemu
does not execute u-boot first.
Reviewed-by: NCédric Le Goater <clg@kaod.org>
Tested-by: NCédric Le Goater <clg@kaod.org>
Signed-off-by: NJoel Stanley <joel@jms.id.au>
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

The restrictions here (which are checked at pre-plug time) are PAPR
specific, rather than being inherent to the NVDIMM devices.  Adjust the
error messages to be clearer about this.
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>

The pseries machine does not support NVDIMM modules without label.
Attempting to do so, even if the overall block size is aligned with
256MB, will seg fault the guest kernel during NVDIMM probe. This
can be avoided by forcing 'label-size' to always be present for
sPAPR NVDIMMs.

The verification was put before the alignment check because the
presence of label-size affects the alignment calculation, so
it's not optimal to warn the user about an alignment error,
then about the lack of label-size, then about a new alignment
error when the user sets a label-size.
Signed-off-by: NDaniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20200413203628.31636-1-danielhb413@gmail.com>
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>

Currently, we can't properly handle unplug of NVLink2 devices, because we
don't have code to tear down their special memory resources.  There's not
a lot of impetus to implement that: since hardware NVLink2 devices can't
be hot unplugged, the guest side drivers don't usually support unplug
anyway.

Therefore, simply prevent unplug of NVLink2 devices.
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>
Reviewed-by: NAlexey Kardashevskiy <aik@ozlabs.ru>

The CAS reboot flag is false by default and all the locations that
could set it to true have been dropped. This means that all code
blocks depending on the flag being set is dead code and the other
code blocks should be executed always.

Just do that and drop the now uneeded CAS reboot flag. Fix a
comment on the way to make checkpatch happy.
Signed-off-by: NGreg Kurz <groug@kaod.org>
Message-Id: <158514994893.478799.11772512888322840990.stgit@bahia.lan>
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>

At the moment "ibm,client-architecture-support" ("CAS") is implemented
in SLOF and QEMU assists via the custom H_CAS hypercall which copies
an updated flatten device tree (FDT) blob to the SLOF memory which
it then uses to update its internal tree.

When we enable the OpenFirmware client interface in QEMU, we won't need
to copy the FDT to the guest as the client is expected to fetch
the device tree using the client interface.

This moves FDT rebuild out to a separate helper which is going to be
called from the "ibm,client-architecture-support" handler and leaves
writing FDT to the guest in the H_CAS handler.

This should not cause any behavioral change.
Signed-off-by: NAlexey Kardashevskiy <aik@ozlabs.ru>
Message-Id: <20200310050733.29805-3-aik@ozlabs.ru>
Signed-off-by: NGreg Kurz <groug@kaod.org>
Message-Id: <158514994229.478799.2178881312094922324.stgit@bahia.lan>
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>

The guest can select the MMU mode by setting bits 0-1 of byte 24
in OV5 to to 0b00 for hash or 0b01 for radix. As required by the
architecture, we terminate the boot process if any other value
is found there.

The usual way to negotiate features in OV5 is basically ANDing
the bitfield provided by the guest and the bitfield of features
supported by QEMU, previously populated at machine init.

For some not documented reason, MMU is treated differently : bit 1
of byte 24 (the radix/hash bit) is cleared from the guest OV5 and
explicitely set in the final negotiated OV5 if radix was requested.

Since the only expected input from the guest is the radix/hash bit
being set or not, it seems more appropriate to handle this like we
do for XIVE.

Set the radix bit in spapr->ov5 at machine init if it has a chance
to work (ie. power9, either TCG or a radix capable KVM) and rely
exclusively on spapr_ovec_intersect() to set the radix bit in
spapr->ov5_cas.
Signed-off-by: NGreg Kurz <groug@kaod.org>
Message-Id: <158514993621.478799.4204740354545734293.stgit@bahia.lan>
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>