The FWNMI option must deliver system reset interrupts to their
registered address, and there are a few constraints on the handler
addresses specified in PAPR. Add the system reset address state and
checks.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Message-Id: <20200316142613.121089-4-npiggin@gmail.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Reviwed-by: Mahesh Salgaonkar <mahesh@linux.ibm.com>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
The option is called "FWNMI", and it involves more than just machine
checks, also machine checks can be delivered without the FWNMI option,
so re-name various things to reflect that.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Message-Id: <20200316142613.121089-3-npiggin@gmail.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
In the spapr code we've been gradually moving towards a convention that
functions which create pieces of the device tree are called spapr_dt_*().
This patch speeds that along by renaming most of the things that don't yet
match that so that they do.
For now we leave the *_dt_populate() functions which are actual methods
used in the DRCClass::dt_populate method.
While we're there we remove a few comments that don't really say anything
useful.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
This is currently called from spapr_dt_cas_updates() which is a hang
over from when we created this only as a diff to the DT at CAS time.
Now that we fully rebuild the DT at CAS time, just create it along
with the rest of the properties in /chosen.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Greg Kurz <groug@kaod.org>
Currently this node with information about hotpluggable memory is created
from spapr_dt_cas_updates(). But that's just a hangover from when we
created it only as a diff to the device tree at CAS time. Now that we
fully rebuild the DT as CAS time, it makes more sense to create this along
with the rest of the memory information in the device tree.
So, move it to spapr_populate_memory(). The patch is huge, but it's nearly
all just code motion.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Greg Kurz <groug@kaod.org>
At the moment SLOF reserves space for RTAS and instantiates the RTAS blob
which is 20 bytes binary blob calling an hypercall. The rest of the RTAS
area is a log which SLOF has no idea about but QEMU does.
This moves RTAS sizing to QEMU and this overrides the size from SLOF.
The only remaining problem is that SLOF copies the number of bytes it
reserved (2KB for now) so QEMU needs to reserve at least this much;
SLOF will be fixed separately to check that rtas-size from QEMU is
enough for those 20 bytes for the H_RTAS hcall.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-Id: <20200316011841.99970-1-aik@ozlabs.ru>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
At the moment "pseries" starts in SLOF which only expects the FDT blob
pointer in r3. As we are going to introduce a OpenFirmware support in
QEMU, we will be booting OF clients directly and these expect a stack
pointer in r1, Linux looks at r3/r4 for the initramdisk location
(although vmlinux can find this from the device tree but zImage from
distro kernels cannot).
This extends spapr_cpu_set_entry_state() to take more registers. This
should cause no behavioral change.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-Id: <20200310050733.29805-2-aik@ozlabs.ru>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Move the calculation of the Real Mode Area (RMA) size into a helper
function. While we're there clean it up and correct it in a few ways:
* Add comments making it clearer where the various constraints come from
* Remove a pointless check that the RMA fits within Node 0 (we've just
clamped it so that it does)
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
In spapr_machine_init() we clamp the size of the RMA to 16GiB and the
comment saying why doesn't make a whole lot of sense. In fact, this was
done because the real mode handling code elsewhere limited the RMA in TCG
mode to the maximum value configurable in LPCR[RMLS], 16GiB.
But,
* Actually LPCR[RMLS] has been able to encode a 256GiB size for a very
long time, we just didn't implement it properly in the softmmu
* LPCR[RMLS] shouldn't really be relevant anyway, it only was because we
used to abuse the RMOR based translation mode in order to handle the
fact that we're not modelling the hypervisor parts of the cpu
We've now removed those limitations in the modelling so the 16GiB clamp no
longer serves a function. However, we can't just remove the limit
universally: that would break migration to earlier qemu versions, where
the 16GiB RMLS limit still applies, no matter how bad the reasons for it
are.
So, we replace the 16GiB clamp, with a clamp to a limit defined in the
machine type class. We set it to 16 GiB for machine types 4.2 and earlier,
but set it to 0 meaning unlimited for the new 5.0 machine type.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
The Real Mode Area (RMA) is the part of memory which a guest can access
when in real (MMU off) mode. Of course, for a guest under KVM, the MMU
isn't really turned off, it's just in a special translation mode - Virtual
Real Mode Area (VRMA) - which looks like real mode in guest mode.
The mechanics of how this works when using the hash MMU (HPT) put a
constraint on the size of the RMA, which depends on the size of the
HPT. So, the latter part of spapr_setup_hpt_and_vrma() clamps the RMA
we advertise to the guest based on this VRMA limit.
There are several things wrong with this:
1) spapr_setup_hpt_and_vrma() doesn't actually clamp, it takes the minimum
of Node 0 memory size and the VRMA limit. That will *often* work the
same as clamping, but there can be other constraints on RMA size which
supersede Node 0 memory size. We have real bugs caused by this
(currently worked around in the guest kernel)
2) Some callers of spapr_setup_hpt_and_vrma() are in a situation where
we're past the point that we can actually advertise an RMA limit to the
guest
3) But most fundamentally, the VRMA limit depends on host configuration
(page size) which shouldn't be visible to the guest, but this partially
exposes it. This can cause problems with migration in certain edge
cases, although we will mostly get away with it.
In practice, this clamping is almost never applied anyway. With 64kiB
pages and the normal rules for sizing of the HPT, the theoretical VRMA
limit will be 4x(guest memory size) and so never hit. It will hit with
4kiB pages, where it will be (guest memory size)/4. However all mainstream
distro kernels for POWER have used a 64kiB page size for at least 10 years.
So, simply replace this logic with a check that the RMA we've calculated
based only on guest visible configuration will fit within the host implied
VRMA limit. This can break if running HPT guests on a host kernel with
4kiB page size. As noted that's very rare. There also exist several
possible workarounds:
* Change the host kernel to use 64kiB pages
* Use radix MMU (RPT) guests instead of HPT
* Use 64kiB hugepages on the host to back guest memory
* Increase the guest memory size so that the RMA hits one of the fixed
limits before the RMA limit. This is relatively easy on POWER8 which
has a 16GiB limit, harder on POWER9 which has a 1TiB limit.
* Use a guest NUMA configuration which artificially constrains the RMA
within the VRMA limit (the RMA must always fit within Node 0).
Previously, on KVM, we also temporarily reduced the rma_size to 256M so
that the we'd load the kernel and initrd safely, regardless of the VRMA
limit. This was a) confusing, b) could significantly limit the size of
images we could load and c) introduced a behavioural difference between
KVM and TCG. So we remove that as well.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: Greg Kurz <groug@kaod.org>
This function calculates the maximum size of the RMA as implied by the
host's page size of structure of the VRMA (there are a number of other
constraints on the RMA size which will supersede this one in many
circumstances).
The current interface takes the current RMA size estimate, and clamps it
to the VRMA derived size. The only current caller passes in an arguably
wrong value (it will match the current RMA estimate in some but not all
cases).
We want to fix that, but for now just keep concerns separated by having the
KVM helper function just return the VRMA derived limit, and let the caller
combine it with other constraints. We call the new function
kvmppc_vrma_limit() to more clearly indicate its limited responsibility.
The helper should only ever be called in the KVM enabled case, so replace
its !CONFIG_KVM stub with an assert() rather than a dummy value.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Cedric Le Goater <clg@fr.ibm.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
MIN_RMA_SLOF records the minimum about of RMA that the SLOF firmware
requires. It lets us give a meaningful error if the RMA ends up too small,
rather than just letting SLOF crash.
It's currently stored as a number of megabytes, which is strange for global
constants. Move that megabyte scaling into the definition of the constant
like most other things use.
Change from M to MiB in the associated message while we're at it.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Reviewed-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
There's no good reason for it to be type int, change it to bool.
Suggested-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200207161948.15972-3-philmd@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
This series removes ad hoc RAM allocation API (memory_region_allocate_system_memory)
and consolidates it around hostmem backend. It allows to
* resolve conflicts between global -mem-prealloc and hostmem's "policy" option,
fixing premature allocation before binding policy is applied
* simplify complicated memory allocation routines which had to deal with 2 ways
to allocate RAM.
* reuse hostmem backends of a choice for main RAM without adding extra CLI
options to duplicate hostmem features. A recent case was -mem-shared, to
enable vhost-user on targets that don't support hostmem backends [1] (ex: s390)
* move RAM allocation from individual boards into generic machine code and
provide them with prepared MemoryRegion.
* clean up deprecated NUMA features which were tied to the old API (see patches)
- "numa: remove deprecated -mem-path fallback to anonymous RAM"
- (POSTPONED, waiting on libvirt side) "forbid '-numa node,mem' for 5.0 and newer machine types"
- (POSTPONED) "numa: remove deprecated implicit RAM distribution between nodes"
Introduce a new machine.memory-backend property and wrapper code that aliases
global -mem-path and -mem-alloc into automatically created hostmem backend
properties (provided memory-backend was not set explicitly given by user).
A bulk of trivial patches then follow to incrementally convert individual
boards to using machine.memory-backend provided MemoryRegion.
Board conversion typically involves:
* providing MachineClass::default_ram_size and MachineClass::default_ram_id
so generic code could create default backend if user didn't explicitly provide
memory-backend or -m options
* dropping memory_region_allocate_system_memory() call
* using convenience MachineState::ram MemoryRegion, which points to MemoryRegion
allocated by ram-memdev
On top of that for some boards:
* missing ram_size checks are added (typically it were boards with fixed ram size)
* ram_size fixups are replaced by checks and hard errors, forcing user to
provide correct "-m" values instead of ignoring it and continuing running.
After all boards are converted, the old API is removed and memory allocation
routines are cleaned up.
This allows moving the kernel in the guest memory. The option is useful
for step debugging (as Linux is linked at 0x0); it also allows loading
grub which is normally linked to run at 0x20000.
This uses the existing kernel address by default.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-Id: <20200203032943.121178-6-aik@ozlabs.ru>
Reviewed-by: Fabiano Rosas <farosas@linux.ibm.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Add support for NVDIMM devices for sPAPR. Piggyback on existing nvdimm
device interface in QEMU to support virtual NVDIMM devices for Power.
Create the required DT entries for the device (some entries have
dummy values right now).
The patch creates the required DT node and sends a hotplug
interrupt to the guest. Guest is expected to undertake the normal
DR resource add path in response and start issuing PAPR SCM hcalls.
The device support is verified based on the machine version unlike x86.
This is how it can be used ..
Ex :
For coldplug, the device to be added in qemu command line as shown below
-object memory-backend-file,id=memnvdimm0,prealloc=yes,mem-path=/tmp/nvdimm0,share=yes,size=1073872896
-device nvdimm,label-size=128k,uuid=75a3cdd7-6a2f-4791-8d15-fe0a920e8e9e,memdev=memnvdimm0,id=nvdimm0,slot=0
For hotplug, the device to be added from monitor as below
object_add memory-backend-file,id=memnvdimm0,prealloc=yes,mem-path=/tmp/nvdimm0,share=yes,size=1073872896
device_add nvdimm,label-size=128k,uuid=75a3cdd7-6a2f-4791-8d15-fe0a920e8e9e,memdev=memnvdimm0,id=nvdimm0,slot=0
Signed-off-by: Shivaprasad G Bhat <sbhat@linux.ibm.com>
Signed-off-by: Bharata B Rao <bharata@linux.ibm.com>
[Early implementation]
Message-Id: <158131058078.2897.12767731856697459923.stgit@lep8c.aus.stglabs.ibm.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
We are going to add more init for the latest machine, so move the setup
to a function so we don't have to change the DEFINE_SPAPR_MACHINE macro
each time.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-Id: <20200207064628.1196095-1-mst@redhat.com>
Reviewed-by: Laurent Vivier <lvivier@redhat.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
memory_region_allocate_system_memory() API is going away, so
replace it with memdev allocated MemoryRegion. The later is
initialized by generic code, so board only needs to opt in
to memdev scheme by providing
MachineClass::default_ram_id
and using MachineState::ram instead of manually initializing
RAM memory region.
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200219160953.13771-68-imammedo@redhat.com>
This patch sets the default value of SPAPR_CAP_FWNMI_MCE
to SPAPR_CAP_ON for machine type 5.0.
Signed-off-by: Aravinda Prasad <arawinda.p@gmail.com>
Signed-off-by: Ganesh Goudar <ganeshgr@linux.ibm.com>
Message-Id: <20200130184423.20519-8-ganeshgr@linux.ibm.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
This patch includes migration support for machine check
handling. Especially this patch blocks VM migration
requests until the machine check error handling is
complete as these errors are specific to the source
hardware and is irrelevant on the target hardware.
Signed-off-by: Aravinda Prasad <arawinda.p@gmail.com>
[Do not set FWNMI cap in post_load, now its done in .apply hook]
Signed-off-by: Ganesh Goudar <ganeshgr@linux.ibm.com>
Message-Id: <20200130184423.20519-7-ganeshgr@linux.ibm.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Memory error such as bit flips that cannot be corrected
by hardware are passed on to the kernel for handling.
If the memory address in error belongs to guest then
the guest kernel is responsible for taking suitable action.
Patch [1] enhances KVM to exit guest with exit reason
set to KVM_EXIT_NMI in such cases. This patch handles
KVM_EXIT_NMI exit.
[1] https://www.spinics.net/lists/kvm-ppc/msg12637.html
(e20bbd3d and related commits)
Signed-off-by: Aravinda Prasad <arawinda.p@gmail.com>
Signed-off-by: Ganesh Goudar <ganeshgr@linux.ibm.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Greg Kurz <groug@kaod.org>
Message-Id: <20200130184423.20519-4-ganeshgr@linux.ibm.com>
[dwg: #ifdefs to fix compile for 32-bit target]
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Introduce fwnmi an spapr capability and add a helper function
which tries to enable it, which would be used by following patch
of the series. This patch by itself does not change the existing
behavior.
Signed-off-by: Aravinda Prasad <arawinda.p@gmail.com>
[eliminate cap_ppc_fwnmi, add fwnmi cap to migration state
and reprhase the commit message]
Signed-off-by: Ganesh Goudar <ganeshgr@linux.ibm.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <20200130184423.20519-3-ganeshgr@linux.ibm.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
For POWER9 DD2.2 cpus, the best current Spectre v2 indirect branch
mitigation is "count cache disabled", which is configured with:
-machine cap-ibs=fixed-ccd
However, this option isn't available on DD2.3 CPUs with KVM, because they
don't have the count cache disabled.
For POWER9 DD2.3 cpus, it is "count cache flush with assist", configured
with:
-machine cap-ibs=workaround,cap-ccf-assist=on
However this option isn't available on DD2.2 CPUs with KVM, because they
don't have the special CCF assist instruction this relies on.
On current machine types, we default to "count cache flush w/o assist",
that is:
-machine cap-ibs=workaround,cap-ccf-assist=off
This runs, with mitigation on both DD2.2 and DD2.3 host cpus, but has a
fairly significant performance impact.
It turns out we can do better. The special instruction that CCF assist
uses to trigger a count cache flush is a no-op on earlier CPUs, rather than
trapping or causing other badness. It doesn't, of itself, implement the
mitigation, but *if* we have count-cache-disabled, then the count cache
flush is unnecessary, and so using the count cache flush mitigation is
harmless.
Therefore for the new pseries-5.0 machine type, enable cap-ccf-assist by
default. Along with that, suppress throwing an error if cap-ccf-assist
is selected but KVM doesn't support it, as long as KVM *is* giving us
count-cache-disabled. To allow TCG to work out of the box, even though it
doesn't implement the ccf flush assist, downgrade the error in that case to
a warning. This matches several Spectre mitigations where we allow TCG
to operate for debugging, since we don't really make guarantees about TCG
security properties anyway.
While we're there, make the TCG warning for this case match that for other
mitigations.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Tested-by: Michael Ellerman <mpe@ellerman.id.au>
While loading the executable, some platforms (like AVR) need to
detect CPU type that executable is built for - and, with this patch,
this is enabled by reading the field 'e_flags' of the ELF header of
the executable in question. The change expands functionality of
the following functions:
- load_elf()
- load_elf_as()
- load_elf_ram()
- load_elf_ram_sym()
The argument added to these functions is called 'pflags' and is of
type 'uint32_t*' (that matches 'pointer to 'elf_word'', 'elf_word'
being the type of the field 'e_flags', in both 32-bit and 64-bit
variants of ELF header). Callers are allowed to pass NULL as that
argument, and in such case no lookup to the field 'e_flags' will
happen, and no information will be returned, of course.
CC: Richard Henderson <rth@twiddle.net>
CC: Peter Maydell <peter.maydell@linaro.org>
CC: Edgar E. Iglesias <edgar.iglesias@gmail.com>
CC: Michael Walle <michael@walle.cc>
CC: Thomas Huth <huth@tuxfamily.org>
CC: Laurent Vivier <laurent@vivier.eu>
CC: Philippe Mathieu-Daudé <f4bug@amsat.org>
CC: Aleksandar Rikalo <aleksandar.rikalo@rt-rk.com>
CC: Aurelien Jarno <aurelien@aurel32.net>
CC: Jia Liu <proljc@gmail.com>
CC: David Gibson <david@gibson.dropbear.id.au>
CC: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
CC: BALATON Zoltan <balaton@eik.bme.hu>
CC: Christian Borntraeger <borntraeger@de.ibm.com>
CC: Thomas Huth <thuth@redhat.com>
CC: Artyom Tarasenko <atar4qemu@gmail.com>
CC: Fabien Chouteau <chouteau@adacore.com>
CC: KONRAD Frederic <frederic.konrad@adacore.com>
CC: Max Filippov <jcmvbkbc@gmail.com>
Reviewed-by: Aleksandar Rikalo <aleksandar.rikalo@rt-rk.com>
Signed-off-by: Michael Rolnik <mrolnik@gmail.com>
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Aleksandar Markovic <amarkovic@wavecomp.com>
Message-Id: <1580079311-20447-24-git-send-email-aleksandar.markovic@rt-rk.com>
Define the new macro VMSTATE_INSTANCE_ID_ANY for callers who wants to
auto-generate the vmstate instance ID. Previously it was hard coded
as -1 instead of this macro. It helps to change this default value in
the follow up patches. No functional change.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
spapr_match_nvt() is a XIVE operation and is used by the machine to
look for a matching target when an event notification is being
delivered. An assert checks that spapr_match_nvt() is called only when
the machine has selected the XIVE interrupt mode but it is redundant
with the XIVE_PRESENTER() dynamic cast.
Apply the cast to spapr->active_intc and remove the assert.
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Message-Id: <20200106163207.4608-1-clg@kaod.org>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
'out' can be replaced by 'return' with the appropriate
return value.
CC: David Gibson <david@gibson.dropbear.id.au>
CC: qemu-ppc@nongnu.org
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20200106182425.20312-2-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
A pseries guest can be run as a secure guest on Ultravisor-enabled
POWER platforms. When such a secure guest is reset, we need to
release/reset a few resources both on ultravisor and hypervisor side.
This is achieved by invoking this new ioctl KVM_PPC_SVM_OFF from the
machine reset path.
As part of this ioctl, the secure guest is essentially transitioned
back to normal mode so that it can reboot like a regular guest and
become secure again.
This ioctl has no effect when invoked for a normal guest. If this ioctl
fails for a secure guest, the guest is terminated.
Signed-off-by: Bharata B Rao <bharata@linux.ibm.com>
Message-Id: <20191219031445.8949-3-bharata@linux.ibm.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
spapr_ovec_diff(ov, old, new) has somewhat complex semantics. ov is set
to those bits which are in new but not old, and it returns as a boolean
whether or not there are any bits in old but not new.
It turns out that both callers only care about the second, not the first.
This is basically equivalent to a bitmap subset operation, which is easier
to understand and implement. So replace spapr_ovec_diff() with
spapr_ovec_subset().
Cc: Mike Roth <mdroth@linux.vnet.ibm.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Cedric Le Goater <clg@fr.ibm.com>
spapr_h_cas_compose_response() handles the last piece of the PAPR feature
negotiation process invoked via the ibm,client-architecture-support OF
call. Its only caller is h_client_architecture_support() which handles
most of the rest of that process.
I believe it was placed in a separate file originally to handle some
fiddly dependencies between functions, but mostly it's just confusing
to have the CAS process split into two pieces like this. Now that
compose response is simplified (by just generating the whole device
tree anew), it's cleaner to just fold it into
h_client_architecture_support().
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Cedric Le Goater <clg@fr.ibm.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Previously, spapr_build_fdt() constructed the device tree in a fixed
buffer of size FDT_MAX_SIZE. This is a bit inflexible, but more
importantly it's awkward for the case where we use it during CAS. In
that case the guest firmware supplies a buffer and we have to
awkwardly check that what we generated fits into it afterwards, after
doing a lot of size checks during spapr_build_fdt().
Simplify this by having spapr_build_fdt() take a 'space' parameter.
For the CAS case, we pass in the buffer size provided by SLOF, for the
machine init case, we continue to pass FDT_MAX_SIZE.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Cedric Le Goater <clg@fr.ibm.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Make kvmppc_hint_smt_possible hint append helper well formed:
rename errp to errp_in, as it is IN-parameter here (which is unusual
for errp), rename function to be kvmppc_error_append_*_hint.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20191127191434.20945-1-vsementsov@virtuozzo.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
The CAM line matching sequence in the pseries machine does not change
much apart from the use of the new QOM interfaces. There is an extra
indirection because of the sPAPR IRQ backend of the machine. Only the
XIVE backend implements the new 'match_nvt' handler.
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Message-Id: <20191125065820.927-11-clg@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Add 5.0 machine types for arm/i440fx/q35/s390x/spapr.
For i440fx and q35, unversioned cpu models are still translated
to -v1; I'll leave changing this (if desired) to the respective
maintainers.
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Message-Id: <20191112104811.30323-1-cohuck@redhat.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>
Virtio spec 1.1 (and earlier), 5.2.5.2 Driver Requirements: Device
Initialization:
"Devices SHOULD always offer VIRTIO_BLK_F_FLUSH, and MUST offer it if
they offer VIRTIO_BLK_F_CONFIG_WCE"
Currently F_CONFIG_WCE and F_WCE are not connected to each other.
Qemu will advertise F_CONFIG_WCE if config-wce argument is
set for virtio-blk device. And F_WCE is advertised only if
underlying block backend actually has it's caching enabled.
Fix this by advertising F_WCE if F_CONFIG_WCE is also advertised.
To preserve backwards compatibility with newer machine types make this
behaviour governed by "x-enable-wce-if-config-wce" virtio-blk-device
property and introduce hw_compat_4_2 with new property being off by
default for all machine types <= 4.2 (but don't introduce 4.3
machine type itself yet).
Signed-off-by: Evgeny Yakovlev <wrfsh@yandex-team.ru>
Message-Id: <1572978137-189218-1-git-send-email-wrfsh@yandex-team.ru>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Since "spapr: Render full FDT on ibm,client-architecture-support" we build
the entire flatten device tree (FDT) twice - at the reset time and
when "ibm,client-architecture-support" (CAS) is called. The full FDT from
CAS is then applied on top of the SLOF internal device tree.
This is mostly ok, however there is a case when the QEMU is started with
-initrd and for some reason the guest decided to move/unpack the init RAM
disk image - the guest correctly notifies SLOF about the change but
at CAS it is overridden with the QEMU initial location addresses and
the guest may fail to boot if the original initrd memory was changed.
This fixes the problem by only adding the /chosen node at the reset time
to prevent the original QEMU's linux,initrd-start/linux,initrd-end to
override the updated addresses.
This only treats /chosen differently as we know there is a special case
already and it is unlikely anything else will need to change /chosen at CAS
we are better off not touching /chosen after we handed it over to SLOF.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-Id: <20191024041308.5673-1-aik@ozlabs.ru>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
We must not call spapr_drc_detach() on a detached DRC otherwise bad things
can happen, ie. QEMU hangs or crashes. This is easily demonstrated with
a CPU hotplug/unplug loop using QMP.
Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <157185826035.3073024.1664101000438499392.stgit@bahia.lan>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
For the benefit of peripheral device allocation, the number of available
irqs really wants to be the same on a given machine type version,
regardless of what irq backends we are using. That's the case now, but
only because we make sure the different SpaprIrq instances have the same
value except for the special legacy one.
Since this really only depends on machine type version, move the value to
SpaprMachineClass instead of SpaprIrq. This also puts the code to set it
to the lower value on old machine types right next to setting
legacy_irq_allocation, which needs to go hand in hand.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
The nr_msis value we use here has to line up with whether we're using
legacy or modern irq allocation. Therefore it's safer to derive it based
on legacy_irq_allocation rather than having SpaprIrq contain a canned
value.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
This method depends only on the active irq controller. Now that we've
formalized the notion of active controller we can dispatch directly
through that, rather than dispatching via SpaprIrq with the dual
version having to do a second conditional dispatch.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
This method depends only on the active irq controller. Now that we've
formalized the notion of active controller we can dispatch directly
through that, rather than dispatching via SpaprIrq with the dual
version having to do a second conditional dispatch.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Support for setting VSMT is available in KVM since linux-4.13. Most distros
that support KVM on POWER already have it. It thus seem reasonable enough
to have the default machine to set VSMT to smp_threads.
This brings contiguous VCPU ids and thus brings their upper bound down to
the machine's max_cpus. This is especially useful for XIVE KVM devices,
which may thus allocate only one VP descriptor per VCPU.
Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <157010411885.246126.12610015369068227139.stgit@bahia.lan>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Add MachineClass::auto_enable_numa field. When it is true, a NUMA node
is expected to be created implicitly.
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Suggested-by: Igor Mammedov <imammedo@redhat.com>
Suggested-by: Eduardo Habkost <ehabkost@redhat.com>
Reviewed-by: Igor Mammedov <imammedo@redhat.com>
Signed-off-by: Tao Xu <tao3.xu@intel.com>
Message-Id: <20190905083238.1799-1-tao3.xu@intel.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
SpaprIrq::ov5 stores the value for a particular byte in PAPR option vector
5 which indicates whether XICS, XIVE or both interrupt controllers are
available. As usual for PAPR, the encoding is kind of overly complicated
and confusing (though to be fair there are some backwards compat things it
has to handle).
But to make our internal code clearer, have SpaprIrq encode more directly
which backends are available as two booleans, and derive the OV5 value from
that at the point we need it.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Reviewed-by: Greg Kurz <groug@kaod.org>
The ibm,client-architecture-support call is a way for the guest to
negotiate capabilities with a hypervisor. It is implemented as:
- the guest calls SLOF via client interface;
- SLOF calls QEMU (H_CAS hypercall) with an options vector from the guest;
- QEMU returns a device tree diff (which uses FDT format with
an additional header before it);
- SLOF walks through the partial diff tree and updates its internal tree
with the values from the diff.
This changes QEMU to simply re-render the entire tree and send it as
an update. SLOF can handle this already mostly, [1] is needed before this
can be applied. This stores the resulting tree in the spapr machine to have
the latest valid FDT copy possible (this should not matter much as
H_UPDATE_DT happens right after that but nevertheless).
The benefit is reduced code size as there is no need for another set of
DT rendering helpers such as spapr_fixup_cpu_dt().
The downside is that the updates are bigger now (as they include all
nodes and properties) but the difference on a '-smp 256,threads=1' system
before/after is 2.35s vs. 2.5s.
[1] https://patchwork.ozlabs.org/patch/1152915/
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
SLOF implements one itself so let's remove it from QEMU. It is one less
image and simpler setup as the RTAS blob never stays in its initial place
anyway as the guest OS always decides where to put it.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
We are going to use spapr_build_fdt() for the boot time FDT and as an
update for SLOF during handling of H_CAS. SLOF will apply all properties
from the QEMU's FDT which is usually ok unless there are properties
changed by grub or guest kernel. The properties are:
bootargs, linux,initrd-start, linux,initrd-end, linux,stdout-path,
linux,rtas-base, linux,rtas-entry. Resetting those during CAS will most
likely cause grub failure.
Don't create such properties if we're booting without "-kernel" and
"-initrd" so they won't get included into the DT update blob and
therefore the guest is more likely to boot successfully.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
[dwg: Tweaked commit message based on Greg Kurz's input]
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
The device tree build by QEMU at the machine reset time is used by SLOF
to build its internal device tree but the node names are not preserved
exactly so when QEMU provides a device tree update in response to H_CAS,
it might become tricky to match a node from the update blob to
the actual node in SLOF.
This removed leading zeroes from "memory@" nodes and makes
the DTC checker happy.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Greg Kurz <groug@kaod.org>
Add a missing g_free(fdt) if the resulting tree is bigger
than the space allocated by SLOF.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
The number of NUMA nodes in the system is fixed from the command line.
Therefore, there's no need to recalculate it at reset time, and we can
determine the special gpu_numa_id value used for NVLink2 devices at init
time.
This simplifies the reset path a bit which will make further improvements
easier.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Reviewed-by: Greg Kurz <groug@kaod.org>
Tested-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru>