| Age | Commit message (Collapse) | Author |
|---|
|
tl;dr: This allows Big Endian zImage booting via -kernel + x-vof=on.
QEMU loads the kernel at 0x400000 by default which works most of
the time as Linux kernels are relocatable, 64bit and compiled with "-pie"
(position independent code). This works for a little endian zImage too.
However a big endian zImage is compiled without -pie, is 32bit, linked to
0x4000000 so current QEMU ends up loading it at
0x4400000 but keeps spapr->kernel_addr unchanged so booting fails.
This uses the kernel address returned from load_elf().
If the default kernel_addr is used, there is no change in behavior (as
translate_kernel_address() takes care of this), which is:
LE/BE vmlinux and LE zImage boot, BE zImage does not.
If the VM created with "-machine kernel-addr=0,x-vof=on", then QEMU
prints a warning and BE zImage boots.
Note #1: SLOF (x-vof=off) still cannot boot a big endian zImage as
SLOF enables MSR_SF for everything loaded by QEMU and this leads to early
crash of 32bit zImage.
Note #2: BE/LE vmlinux images set MSR_SF in early boot so these just work;
a LE zImage restores MSR_SF after every CI call and we are lucky enough
not to crash before the first CI call.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Tested-by: Joel Stanley <joel@jms.id.au>
Reviewed-by: Fabiano Rosas <farosas@linux.ibm.com>
Message-Id: <20220504065536.3534488-1-aik@ozlabs.ru>
[danielhb: use PRIx64 instead of lx in warn_report]
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
|
|
-machine graphics=off is the usual way to tell the firmware or the OS that the
user wants a serial console. The pseries machine however does not support
this, and never adds the stdout-path node to the device tree if a VGA device
is provided. This is in addition to the other magic behavior of VGA devices,
which is to add a keyboard and mouse to the default USB bus.
Split spapr->has_graphics in two variables so that the two behaviors can be
separated: the USB devices remains the same, but the stdout-path is added
even with "-device VGA -machine graphics=off".
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220507054826.124936-1-pbonzini@redhat.com>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
|
|
As part of converting -boot to a property with a QAPI type, define
the struct and use it throughout QEMU to access boot configuration.
machine_boot_parse takes care of doing the QemuOpts->QAPI conversion by
hand, for now.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220414165300.555321-2-pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
A global boolean variable "vga_interface_created"(declared in softmmu/globals.c)
has been used to track the creation of vga interface. If the vga flag is passed
in the command line "default_vga"(declared in softmmu/vl.c) variable is set to 0.
To warn user, the condition checks if vga_interface_created is false
and default_vga is equal to 0. If "-vga none" is passed, this patch will not warn the
user regarding the creation of VGA device.
The warning "A -vga option was passed but this
machine type does not use that option; no VGA device has been created"
is logged if vga flag is passed but no vga device is created.
This patch has been tested for x86_64, i386, sparc, sparc64 and arm boards.
Signed-off-by: Gautam Agrawal <gautamnagrawal@gmail.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/581
Message-Id: <20220501122505.29202-1-gautamnagrawal@gmail.com>
[thuth: Fix wrong warning with "-device" in some cases as reported by Paolo]
Signed-off-by: Thomas Huth <thuth@redhat.com>
|
|
msr_pr macro hides the usage of env->msr, which is a bad behavior
Substitute it with FIELD_EX64 calls that explicitly use env->msr
as a parameter.
Suggested-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Víctor Colombo <victor.colombo@eldorado.org.br>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20220504210541.115256-4-victor.colombo@eldorado.org.br>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
|
|
Add 7.1 machine types for arm/i440fx/m68k/q35/s390x/spapr.
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Message-Id: <20220316145521.1224083-1-cohuck@redhat.com>
Reviewed-by: Eric Farman <farman@linux.ibm.com>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Acked-by: Laurent Vivier <laurent@vivier.eu>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Yanan Wang <wangyanan55@huawei.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
|
|
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20220323155743.1585078-33-marcandre.lureau@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
g_new(T, n) is neater than g_malloc(sizeof(T) * n). It's also safer,
for two reasons. One, it catches multiplication overflowing size_t.
Two, it returns T * rather than void *, which lets the compiler catch
more type errors.
This commit only touches allocations with size arguments of the form
sizeof(T).
Patch created mechanically with:
$ spatch --in-place --sp-file scripts/coccinelle/use-g_new-etc.cocci \
--macro-file scripts/cocci-macro-file.h FILES...
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Acked-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Message-Id: <20220315144156.1595462-4-armbru@redhat.com>
Reviewed-by: Pavel Dovgalyuk <Pavel.Dovgalyuk@ispras.ru>
|
|
Move the various memalign-related functions out of osdep.h and into
their own header, which we include only where they are used.
While we're doing this, add some brief documentation comments.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-id: 20220226180723.1706285-10-peter.maydell@linaro.org
|
|
The firmware check consists on a file search (qemu_find_file) and load
it via load_imag_targphys(). This validation is not dependent on any
other machine state but it currently being done at the end of
spapr_machine_init(). This means that we can do a lot of stuff and end
up failing at the end for something that we can verify right out of the
gate.
Move this validation to the start of spapr_machine_init() to fail
earlier. While we're at it, use g_autofree in the 'filename' pointer.
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20220228175004.8862-3-danielhb413@gmail.com>
Signed-off-by: Cédric Le Goater <clg@kaod.org>
|
|
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20220228175004.8862-2-danielhb413@gmail.com>
Signed-off-by: Cédric Le Goater <clg@kaod.org>
|
|
This implements the Nested KVM HV hcall API for spapr under TCG.
The L2 is switched in when the H_ENTER_NESTED hcall is made, and the
L1 is switched back in returned from the hcall when a HV exception
is sent to the vhyp. Register state is copied in and out according to
the nested KVM HV hcall API specification.
The hdecr timer is started when the L2 is switched in, and it provides
the HDEC / 0x980 return to L1.
The MMU re-uses the bare metal radix 2-level page table walker by
using the get_pate method to point the MMU to the nested partition
table entry. MMU faults due to partition scope errors raise HV
exceptions and accordingly are routed back to the L1.
The MMU does not tag translations for the L1 (direct) vs L2 (nested)
guests, so the TLB is flushed on any L1<->L2 transition (hcall entry
and exit).
Reviewed-by: Fabiano Rosas <farosas@linux.ibm.com>
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
[ clg: checkpatch fixes ]
Message-Id: <20220216102545.1808018-10-npiggin@gmail.com>
Signed-off-by: Cédric Le Goater <clg@kaod.org>
|
|
Introduce virtual hypervisor methods that can support a "Nested KVM HV"
implementation using the bare metal 2-level radix MMU, and using HV
exceptions to return from H_ENTER_NESTED (rather than cause interrupts).
HV exceptions can now be raised in the TCG spapr machine when running a
nested KVM HV guest. The main ones are the lev==1 syscall, the hdecr,
hdsi and hisi, hv fu, and hv emu, and h_virt external interrupts.
HV exceptions are intercepted in the exception handler code and instead
of causing interrupts in the guest and switching the machine to HV mode,
they go to the vhyp where it may exit the H_ENTER_NESTED hcall with the
interrupt vector numer as return value as required by the hcall API.
Address translation is provided by the 2-level page table walker that is
implemented for the bare metal radix MMU. The partition scope page table
is pointed to the L1's partition scope by the get_pate vhc method.
Reviewed-by: Fabiano Rosas <farosas@linux.ibm.com>
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Message-Id: <20220216102545.1808018-9-npiggin@gmail.com>
Signed-off-by: Cédric Le Goater <clg@kaod.org>
|
|
In prepartion for implementing a full partition table option for
vhyp, update the get_pate method to take an lpid and return a
success/fail indicator.
The spapr implementation currently just asserts lpid is always 0
and always return success.
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
[ clg: checkpatch fixes ]
Message-Id: <20220216102545.1808018-6-npiggin@gmail.com>
Signed-off-by: Cédric Le Goater <clg@kaod.org>
|
|
The patch adds support for the SCM flush hcall for the nvdimm devices.
To be available for exploitation by guest through the next patch. The
hcall is applicable only for new SPAPR specific device class which is
also introduced in this patch.
The hcall expects the semantics such that the flush to return with
H_LONG_BUSY_ORDER_10_MSEC when the operation is expected to take longer
time along with a continue_token. The hcall to be called again by providing
the continue_token to get the status. So, all fresh requests are put into
a 'pending' list and flush worker is submitted to the thread pool. The
thread pool completion callbacks move the requests to 'completed' list,
which are cleaned up after collecting the return status for the guest
in subsequent hcall from the guest.
The semantics makes it necessary to preserve the continue_tokens and
their return status across migrations. So, the completed flush states
are forwarded to the destination and the pending ones are restarted
at the destination in post_load. The necessary nvdimm flush specific
vmstate structures are also introduced in this patch which are to be
saved in the new SPAPR specific nvdimm device to be introduced in the
following patch.
Signed-off-by: Shivaprasad G Bhat <sbhat@linux.ibm.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <164396254862.109112.16675611182159105748.stgit@ltczzess4.aus.stglabs.ibm.com>
Signed-off-by: Cédric Le Goater <clg@kaod.org>
|
|
spapr_get_fw_dev_path() is an impl of
FWPathProviderClass::get_dev_path(). This interface is used by
hw/core/qdev-fw.c via fw_path_provider_try_get_dev_path() in two
functions:
- static char *qdev_get_fw_dev_path_from_handler(), which is used only in
qdev_get_fw_dev_path_helper() and it's guarded by "if (dev &&
dev->parent_bus)";
- char *qdev_get_own_fw_dev_path_from_handler(), which is used in
softmmu/bootdevice.c in get_boot_device_path() like this:
if (dev) {
d = qdev_get_own_fw_dev_path_from_handler(dev->parent_bus, dev);
This means that, when called via softmmu/bootdevice.c, there's no check
of 'dev->parent_bus' being not NULL. The result is that the "BusState
*bus" arg of spapr_get_fw_dev_path() can potentially be NULL and if, at
the same time, "SCSIDevice *d" is not NULL, we'll hit this line:
void *spapr = CAST(void, bus->parent, "spapr-vscsi");
And we'll SIGINT because 'bus' is NULL and we're accessing bus->parent.
Adding a simple 'bus != NULL' check to guard the instances where we
access 'bus->parent' can avoid this altogether.
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Message-Id: <20220121213852.30243-1-danielhb413@gmail.com>
Signed-off-by: Cédric Le Goater <clg@kaod.org>
|
|
POWER5+ (ISA v2.03) processors are supported by the pseries machine
but they do not have Altivec instructions. Do not advertise support
for it in the DT.
To be noted that this test is in contradiction with the assert in
cap_vsx_apply().
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Tested-by: Fabiano Rosas <farosas@linux.ibm.com>
Message-Id: <20220105095142.3990430-3-clg@kaod.org>
Signed-off-by: Cédric Le Goater <clg@kaod.org>
|
|
Add 7.0 machine types for arm/i440fx/q35/s390x/spapr.
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Christian Borntraeger <borntraeger@de.ibm.com>
Acked-by: Cédric Le Goater <clg@kaod.org>
Message-Id: <20211217143948.289995-1-cohuck@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
|
|
When updating the R bit of a PTE, the Hash64 MMU was using a wrong byte
offset, causing the first byte of the adjacent PTE to be corrupted.
This caused a panic when booting FreeBSD, using the Hash MMU.
Fixes: a2dd4e83e76b ("ppc/hash64: Rework R and C bit updates")
Signed-off-by: Leandro Lupori <leandro.lupori@eldorado.org.br>
Signed-off-by: Cédric Le Goater <clg@kaod.org>
|
|
Now we have a common structure SMPCompatProps used to store information
about SMP compatibility stuff, so we can also move smp_prefer_sockets
there for cleaner code.
No functional change intended.
Signed-off-by: Yanan Wang <wangyanan55@huawei.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20210929025816.21076-15-wangyanan55@huawei.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
In the real SMP hardware topology world, it's much more likely that
we have high cores-per-socket counts and few sockets totally. While
the current preference of sockets over cores in smp parsing results
in a virtual cpu topology with low cores-per-sockets counts and a
large number of sockets, which is just contrary to the real world.
Given that it is better to make the virtual cpu topology be more
reflective of the real world and also for the sake of compatibility,
we start to prefer cores over sockets over threads in smp parsing
since machine type 6.2 for different arches.
In this patch, a boolean "smp_prefer_sockets" is added, and we only
enable the old preference on older machines and enable the new one
since type 6.2 for all arches by using the machine compat mechanism.
Suggested-by: Daniel P. Berrange <berrange@redhat.com>
Signed-off-by: Yanan Wang <wangyanan55@huawei.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Acked-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Reviewed-by: Pankaj Gupta <pankaj.gupta@ionos.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20210929025816.21076-10-wangyanan55@huawei.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
The main feature of FORM2 affinity support is the separation of NUMA
distances from ibm,associativity information. This allows for a more
flexible and straightforward NUMA distance assignment without relying on
complex associations between several levels of NUMA via
ibm,associativity matches. Another feature is its extensibility. This base
support contains the facilities for NUMA distance assignment, but in the
future more facilities will be added for latency, performance, bandwidth
and so on.
This patch implements the base FORM2 affinity support as follows:
- the use of FORM2 associativity is indicated by using bit 2 of byte 5
of ibm,architecture-vec-5. A FORM2 aware guest can choose to use FORM1
or FORM2 affinity. Setting both forms will default to FORM2. We're not
advertising FORM2 for pseries-6.1 and older machine versions to prevent
guest visible changes in those;
- ibm,associativity-reference-points has a new semantic. Instead of
being used to calculate distances via NUMA levels, it's now used to
indicate the primary domain index in the ibm,associativity domain of
each resource. In our case it's set to {0x4}, matching the position
where we already place logical_domain_id;
- two new RTAS DT artifacts are introduced: ibm,numa-lookup-index-table
and ibm,numa-distance-table. The index table is used to list all the
NUMA logical domains of the platform, in ascending order, and allows for
spartial NUMA configurations (although QEMU ATM doesn't support that).
ibm,numa-distance-table is an array that contains all the distances from
the first NUMA node to all other nodes, then the second NUMA node
distances to all other nodes and so on;
- get_max_dist_ref_points(), get_numa_assoc_size() and get_associativity()
now checks for OV5_FORM2_AFFINITY and returns FORM2 values if the guest
selected FORM2 affinity during CAS.
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210920174947.556324-7-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
|
FORM2 NUMA affinity is prepared to deal with empty (memory/cpu less)
NUMA nodes. This is used by the DAX KMEM driver to locate a PAPR SCM
device that has a different latency than the original NUMA node from the
regular memory. FORM2 is also able to deal with asymmetric NUMA
distances gracefully, something that our FORM1 implementation doesn't
do.
Move these FORM1 verifications to a new function and wait until after
CAS, when we're sure that we're sticking with FORM1, to enforce them.
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210920174947.556324-6-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
|
Linux Kernel 5.12 is now unisolating CPU DRCs in the device_removal
error path, signalling that the hotunplug process wasn't successful.
This allow us to send a DEVICE_UNPLUG_GUEST_ERROR in drc_unisolate_logical()
to signal this error to the management layer.
We also have another error path in spapr_memory_unplug_rollback() for
configured LMB DRCs. Kernels older than 5.13 will not unisolate the LMBs
in the hotunplug error path, but it will reconfigure them. Let's send
the DEVICE_UNPLUG_GUEST_ERROR event in that code path as well to cover the
case of older kernels.
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210907004755.424931-7-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
|
As done in hw/acpi/memory_hotplug.c, pass an empty string if dev->id
is NULL to qapi_event_send_mem_unplug_error() to avoid relying on
a behavior that can be changed in the future.
Suggested-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210907004755.424931-3-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
|
Add 6.2 machine types for arm/i440fx/q35/s390x/spapr.
Signed-off-by: Yanan Wang <wangyanan55@huawei.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Pankaj Gupta <pankaj.gupta@ionos.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
|
|
staging
* More SVM fixes (Lara)
* Module annotation database (Gerd)
* Memory leak fixes (myself)
* Build fixes (myself)
* --with-devices-* support (Alex)
# gpg: Signature made Fri 09 Jul 2021 17:23:52 BST
# gpg: using RSA key F13338574B662389866C7682BFFBD25F78C7AE83
# gpg: issuer "pbonzini@redhat.com"
# gpg: Good signature from "Paolo Bonzini <bonzini@gnu.org>" [full]
# gpg: aka "Paolo Bonzini <pbonzini@redhat.com>" [full]
# Primary key fingerprint: 46F5 9FBD 57D6 12E7 BFD4 E2F7 7E15 100C CD36 69B1
# Subkey fingerprint: F133 3857 4B66 2389 866C 7682 BFFB D25F 78C7 AE83
* remotes/bonzini-gitlab/tags/for-upstream: (48 commits)
meson: Use input/output for entitlements target
configure: allow the selection of alternate config in the build
configs: rename default-configs to configs and reorganise
hw/arm: move CONFIG_V7M out of default-devices
hw/arm: add dependency on OR_IRQ for XLNX_VERSAL
meson: Introduce target-specific Kconfig
meson: switch function tests from compilation to linking
vl: fix leak of qdict_crumple return value
target/i386: fix exceptions for MOV to DR
target/i386: Added DR6 and DR7 consistency checks
target/i386: Added MSRPM and IOPM size check
monitor/tcg: move tcg hmp commands to accel/tcg, register them dynamically
usb: build usb-host as module
monitor/usb: register 'info usbhost' dynamically
usb: drop usb_host_dev_is_scsi_storage hook
monitor: allow register hmp commands
accel: build tcg modular
accel: add tcg module annotations
accel: build qtest modular
accel: add qtest module annotations
...
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
|
|
Introduce an usb device flag instead, set it when usb-host looks at the
device descriptors anyway. Also set it for emulated storage devices,
for consistency. Add an inline helper function to check the flag.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Jose R. Ziviani <jziviani@suse.de>
Message-Id: <20210624103836.2382472-32-kraxel@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
If KVM_CAP_RPT_INVALIDATE KVM capability is enabled, then
- indicate the availability of H_RPT_INVALIDATE hcall to the guest via
ibm,hypertas-functions property.
- Enable the hcall
Both the above are done only if the new sPAPR machine capability
cap-rpt-invalidate is set.
Signed-off-by: Bharata B Rao <bharata@linux.ibm.com>
Message-Id: <20210706112440.1449562-3-bharata@linux.ibm.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
|
This addresses the comments from v22.
The functional changes are (the VOF ones need retesting with Pegasos2):
(VOF) setprop will start failing if the machine class callback
did not handle it;
(VOF) unit addresses are lowered in path_offset();
(SPAPR) /chosen/bootargs is initialized from kernel_cmdline if
the client did not change it.
Fixes: 5c991e5d4378 ("spapr: Implement Open Firmware client interface")
Cc: BALATON Zoltan <balaton@eik.bme.hu>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-Id: <20210708065625.548396-1-aik@ozlabs.ru>
Tested-by: BALATON Zoltan <balaton@eik.bme.hu>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
|
The PAPR platform describes an OS environment that's presented by
a combination of a hypervisor and firmware. The features it specifies
require collaboration between the firmware and the hypervisor.
Since the beginning, the runtime component of the firmware (RTAS) has
been implemented as a 20 byte shim which simply forwards it to
a hypercall implemented in qemu. The boot time firmware component is
SLOF - but a build that's specific to qemu, and has always needed to be
updated in sync with it. Even though we've managed to limit the amount
of runtime communication we need between qemu and SLOF, there's some,
and it has become increasingly awkward to handle as we've implemented
new features.
This implements a boot time OF client interface (CI) which is
enabled by a new "x-vof" pseries machine option (stands for "Virtual Open
Firmware). When enabled, QEMU implements the custom H_OF_CLIENT hcall
which implements Open Firmware Client Interface (OF CI). This allows
using a smaller stateless firmware which does not have to manage
the device tree.
The new "vof.bin" firmware image is included with source code under
pc-bios/. It also includes RTAS blob.
This implements a handful of CI methods just to get -kernel/-initrd
working. In particular, this implements the device tree fetching and
simple memory allocator - "claim" (an OF CI memory allocator) and updates
"/memory@0/available" to report the client about available memory.
This implements changing some device tree properties which we know how
to deal with, the rest is ignored. To allow changes, this skips
fdt_pack() when x-vof=on as not packing the blob leaves some room for
appending.
In absence of SLOF, this assigns phandles to device tree nodes to make
device tree traversing work.
When x-vof=on, this adds "/chosen" every time QEMU (re)builds a tree.
This adds basic instances support which are managed by a hash map
ihandle -> [phandle].
Before the guest started, the used memory is:
0..e60 - the initial firmware
8000..10000 - stack
400000.. - kernel
3ea0000.. - initramdisk
This OF CI does not implement "interpret".
Unlike SLOF, this does not format uninitialized nvram. Instead, this
includes a disk image with pre-formatted nvram.
With this basic support, this can only boot into kernel directly.
However this is just enough for the petitboot kernel and initradmdisk to
boot from any possible source. Note this requires reasonably recent guest
kernel with:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=df5be5be8735
The immediate benefit is much faster booting time which especially
crucial with fully emulated early CPU bring up environments. Also this
may come handy when/if GRUB-in-the-userspace sees light of the day.
This separates VOF and sPAPR in a hope that VOF bits may be reused by
other POWERPC boards which do not support pSeries.
This assumes potential support for booting from QEMU backends
such as blockdev or netdev without devices/drivers used.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-Id: <20210625055155.2252896-1-aik@ozlabs.ru>
Reviewed-by: BALATON Zoltan <balaton@eik.bme.hu>
[dwg: Adjusted some includes which broke compile in some more obscure
compilation setups]
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
|
QEMU reserves space for RTAS via /rtas/rtas-size which tells the client
how much space the RTAS requires to work which includes the RTAS binary
blob implementing RTAS runtime. Because pseries supports FWNMI which
requires plenty of space, QEMU reserves more than 2KB which is
enough for the RTAS blob as it is just 20 bytes (under QEMU).
Since FWNMI reset delivery was added, RTAS_SIZE macro is not used anymore.
This replaces RTAS_SIZE with RTAS_MIN_SIZE and uses it in
the /rtas/rtas-size calculation to account for the RTAS blob.
Fixes: 0e236d347790 ("ppc/spapr: Implement FWNMI System Reset delivery")
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-Id: <20210622070336.1463250-1-aik@ozlabs.ru>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
|
QEMU 6.0 moved all the -boot variables to the machine. Especially, the
removal of the boot_order static changed the handling of '-boot once'
from:
if (boot_once) {
qemu_boot_set(boot_once, &error_fatal);
qemu_register_reset(restore_boot_order, g_strdup(boot_order));
}
to
if (current_machine->boot_once) {
qemu_boot_set(current_machine->boot_once, &error_fatal);
qemu_register_reset(restore_boot_order,
g_strdup(current_machine->boot_order));
}
This means that we now register as subsequent boot order a copy
of current_machine->boot_once that was just set with the previous
call to qemu_boot_set(), i.e. we never transition away from the
once boot order.
It is certainly fragile^Wwrong for the spapr code to hijack a
field of the base machine type object like that. The boot order
rework simply turned this software boundary violation into an
actual bug.
Have the spapr code to handle that with its own field in
SpaprMachineState. Also kfree() the initial boot device
string when "once" was used.
Fixes: 4b7acd2ac821 ("vl: clean up -boot variables")
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1960119
Cc: pbonzini@redhat.com
Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <20210521160735.1901914-1-groug@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
|
Moved has_spr to cpu.h as ppc_has_spr and turned it into an inline function.
Change spr verification in pnv.c and spapr.c to a version that can
compile in a !TCG environment.
Signed-off-by: Lucas Mateus Castro (alqotel) <lucas.araujo@eldorado.org.br>
Message-Id: <20210507164146.67086-1-lucas.araujo@eldorado.org.br>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
|
Starting with Linux kernel v5.12 we dropped support[1] in KVM for
hosts that can't have their threads running in different MMU modes
(POWER9 < DD2.2). In these hosts, KVM will no longer report the
KVM_CAP_PPC_MMU_HASH_V3 capability[2] when the host is running Radix.
For guests that support both MMU modes, the negotiation during CAS
will make sure it selects the correct one.
For guests that only support Hash, such as P8 compat mode guests, the
following error is currently thrown:
$ ~/qemu-system-ppc64 -machine pseries,accel=kvm,max-cpu-compat=power8 ...
error: kvm run failed Invalid argument
NIP 0000000000000100 LR 0000000000000000 CTR 0000000000000000 XER 0000000000000000 CPU#0
MSR 8000000000001000 HID0 0000000000000000 HF 8000000000000000 iidx 3 didx 3
TB 00000000 00000000 DECR 0
GPR00 0000000000000000 0000000000000000 0000000000000000 000000007ff00000
GPR04 0000000000000000 0000000000000000 0000000000000000 0000000000000000
GPR08 0000000000000000 0000000000000000 0000000000000000 0000000000000000
GPR12 0000000000000000 0000000000000000 0000000000000000 0000000000000000
GPR16 0000000000000000 0000000000000000 0000000000000000 0000000000000000
GPR20 0000000000000000 0000000000000000 0000000000000000 0000000000000000
GPR24 0000000000000000 0000000000000000 0000000000000000 0000000000000000
GPR28 0000000000000000 0000000000000000 0000000000000000 0000000000000000
CR 00000000 [ - - - - - - - - ] RES ffffffffffffffff
SRR0 0000000000000000 SRR1 0000000000000000 PVR 00000000004e1201 VRSAVE 0000000000000000
SPRG0 0000000000000000 SPRG1 0000000000000000 SPRG2 0000000000000000 SPRG3 0000000000000000
SPRG4 0000000000000000 SPRG5 0000000000000000 SPRG6 0000000000000000 SPRG7 0000000000000000
HSRR0 0000000000000000 HSRR1 0000000000000000
CFAR 0000000000000000
LPCR 000000000004f01f
PTCR 0000000000000000 DAR 0000000000000000 DSISR 0000000000000000
This patch adds a verification during the writing of the platform
support vector so that we error out as soon as we determine this guest
only supports Hash and the host doesn't.
~/qemu-system-ppc64 -machine pseries,accel=kvm,max-cpu-compat=power8 ...
qemu-system-ppc64: Guest requested unavailable MMU mode (hash).
1- https://git.kernel.org/torvalds/p/b1b1697ae0cc8
2- https://git.kernel.org/torvalds/p/a722076e94702
Signed-off-by: Fabiano Rosas <farosas@linux.ibm.com>
Message-Id: <20210505001130.3999968-3-farosas@linux.ibm.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
|
A following patch will make use of it.
Signed-off-by: Fabiano Rosas <farosas@linux.ibm.com>
Message-Id: <20210505001130.3999968-2-farosas@linux.ibm.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
|
into staging
ppc patch queue 2021-05-04
Here's the first ppc pull request for qemu-6.1. It has a wide variety
of stuff accumulated during the 6.0 freeze. Highlights are:
* Multi-phase reset cleanups for PAPR
* Preliminary cleanups towards allowing !CONFIG_TCG for the ppc target
* Cleanup of AIL logic and extension to POWER10
* Further improvements to handling of hot unplug failures on PAPR
* Allow much larger numbers of CPU on pseries
* Support for the H_SCM_HEALTH hypercall
* Add support for the Pegasos II board
* Substantial cleanup to hflag handling
* Assorted minor fixes and cleanups
# gpg: Signature made Tue 04 May 2021 06:52:39 BST
# gpg: using RSA key 75F46586AE61A66CC44E87DC6C38CACA20D9B392
# gpg: Good signature from "David Gibson <david@gibson.dropbear.id.au>" [full]
# gpg: aka "David Gibson (Red Hat) <dgibson@redhat.com>" [full]
# gpg: aka "David Gibson (ozlabs.org) <dgibson@ozlabs.org>" [full]
# gpg: aka "David Gibson (kernel.org) <dwg@kernel.org>" [unknown]
# Primary key fingerprint: 75F4 6586 AE61 A66C C44E 87DC 6C38 CACA 20D9 B392
* remotes/dg-gitlab/tags/ppc-for-6.1-20210504: (46 commits)
hw/ppc/pnv_psi: Use device_cold_reset() instead of device_legacy_reset()
hw/ppc/spapr_vio: Reset TCE table object with device_cold_reset()
hw/intc/spapr_xive: Use device_cold_reset() instead of device_legacy_reset()
target/ppc: removed VSCR from SPR registration
target/ppc: Reduce the size of ppc_spr_t
target/ppc: Clean up _spr_register et al
target/ppc: Add POWER10 exception model
target/ppc: rework AIL logic in interrupt delivery
target/ppc: move opcode table logic to translate.c
target/ppc: code motion from translate_init.c.inc to gdbstub.c
spapr_drc.c: handle hotunplug errors in drc_unisolate_logical()
spapr.h: increase FDT_MAX_SIZE
spapr.c: do not use MachineClass::max_cpus to limit CPUs
ppc: Rename current DAWR macros and variables
target/ppc: POWER10 supports scv
target/ppc: Fix POWER9 radix guest HV interrupt AIL behaviour
docs/system: ppc: Add documentation for ppce500 machine
roms/u-boot: Bump ppce500 u-boot to v2021.04 to fix broken pci support
roms/Makefile: Update ppce500 u-boot build directory name
ppc/spapr: Add support for implement support for H_SCM_HEALTH
...
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
|
|
Up to this patch, 'max_cpus' value is hardcoded to 1024 (commit
6244bb7e5811). In theory this patch would simply bump it to 2048, since
it's the default NR_CPUS kernel setting for ppc64 servers nowadays, but
the whole mechanic of MachineClass:max_cpus is flawed for the pSeries
machine. The two supported accelerators, KVM and TCG, can live without
it.
TCG guests don't have a theoretical limit. The user must be free to
emulate as many CPUs as the hardware is capable of. And even if there
were a limit, max_cpus is not the proper way to report it since it's a
common value checked by SMP code in machine_smp_parse() for KVM as well.
For KVM guests, the proper way to limit KVM CPUs is by host
configuration via NR_CPUS, not a QEMU hardcoded value. There is no
technical reason for a pSeries QEMU guest to forcefully stay below
NR_CPUS.
This hardcoded value also disregard hosts that might have a lower
NR_CPUS limit, say 512. In this case, machine.c:machine_smp_parse() will
allow a 1024 value to pass, but then kvm_init() will complain about it
because it will exceed NR_CPUS:
Number of SMP cpus requested (1024) exceeds the maximum cpus supported
by KVM (512)
A better 'max_cpus' value would consider host settings, but
MachineClass::max_cpus is defined well before machine_init() and
kvm_init(). We can't check for KVM limits because it's too soon, so we
end up making a guess.
This patch makes MachineClass:max_cpus settings innocuous by setting it
to INT32_MAX. machine.c:machine_smp_parse() will not fail the
verification based on max_cpus, letting kvm_init() do the checking with
actual host settings. And TCG guests get to do whatever the hardware is
capable of emulating.
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210408204049.221802-2-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
|
SLOF instantiates RTAS since
744a928ccee9 ("spapr: Stop providing RTAS blob")
so the max address applies to the FDT only.
This renames the macro and fixes up the comment.
This should not cause any behavioral change.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-Id: <20210331025123.29310-1-aik@ozlabs.ru>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
|
Stop including exec/address-spaces.h in files that don't need it.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20210416171314.2074665-5-thuth@redhat.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
|
|
Stop including hw/boards.h in files that don't need it.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20210416171314.2074665-3-thuth@redhat.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
|
|
Add 6.1 machine types for arm/i440fx/q35/s390x/spapr.
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Acked-by: Greg Kurz <groug@kaod.org>
Message-id: 20210331111900.118274-1-cohuck@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
|
|
Commit 47c8c915b162 fixed a problem where multiple spapr_drc_detach()
requests were breaking QEMU. The solution was to just spapr_drc_detach()
once, and use spapr_drc_unplug_requested() to filter whether we already
detached it or not. The commit also tied the hotplug request to the
guest in the same condition.
Turns out that there is a reliable way for a CPU hotunplug to fail. If a
guest with one CPU hotplugs a CPU1, then offline CPU0s via 'echo 0 >
/sys/devices/system/cpu/cpu0/online', then attempts to hotunplug CPU1,
the kernel will refuse it because it's the last online CPU of the
system. Given that we're pulsing the IRQ only in the first try, in a
failed attempt, all other CPU1 hotunplug attempts will fail, regardless
of the online state of CPU1 in the kernel, because we're simply not
letting the guest know that we want to hotunplug the device.
Let's move spapr_hotplug_req_remove_by_index() back out of the "if
(!spapr_drc_unplug_requested(drc))" conditional, allowing for multiple
'device_del' requests to the same CPU core to reach the guest, in case
the CPU core didn't fully hotunplugged previously.
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210401000437.131140-3-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
|
The pseries machines introduced the concept of 'unplug timeout' for CPU
hotunplugs. The idea was to circunvent a deficiency in the pSeries
specification (PAPR), that currently does not define a proper way for
the hotunplug to fail. If the guest refuses to release the CPU (see [1]
for an example) there is no way for QEMU to detect the failure.
Further discussions about how to send a QAPI event to inform about the
hotunplug timeout [2] exposed problems that weren't predicted back when
the idea was developed. Other QEMU machines don't have any type of
hotunplug timeout mechanism for any device, e.g. ACPI based machines
have a way to make hotunplug errors visible to the hypervisor. This
would make this timeout mechanism exclusive to pSeries, which is not
ideal.
The real problem is that a QAPI event that reports hotunplug timeouts
puts the management layer (namely Libvirt) in a weird spot. We're not
telling that the hotunplug failed, because we can't be 100% sure of
that, and yet we're resetting the unplug state back, preventing any
DEVICE_DEL events to reach out in case the guest decides to release the
device. Libvirt would need to inspect the guest itself to see if the
device was released or not, otherwise the internal domain states will be
inconsistent. Moreover, Libvirt already has an 'unplug timeout'
concept, and a QEMU side timeout would need to be juggled together with
the existing Libvirt timeout.
All this considered, this solution ended up creating more trouble than
it solved. This patch reverts the 3 commits that introduced the timeout
mechanism for CPU hotplugs in pSeries machines.
This reverts commit 4515a5f786024fabf0bef4cf3d28adf5647e6e82
"qemu_timer.c: add timer_deadline_ms() helper"
This reverts commit d1c2e3ce3d5a5424651967bce1cf1f4caa0c6d91
"spapr_drc.c: add hotunplug timeout for CPUs"
This reverts commit 51254ffb320183a4636635840c23ee0e3a1efffa
"spapr_drc.c: introduce unplug_timeout_timer"
[1] https://bugzilla.redhat.com/show_bug.cgi?id=1911414
[2] https://lists.gnu.org/archive/html/qemu-devel/2021-03/msg04682.html
CC: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210401000437.131140-2-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
|
spapr_memory_unplug() is the last step of the hot unplug sequence.
It is indirectly called by:
spapr_lmb_release()
hotplug_handler_unplug()
and spapr_lmb_release() already buys us that DIMM unplug state is
present : it gets restored with spapr_recover_pending_dimm_state()
if missing.
g_assert() that spapr_pending_dimm_unplugs_find() cannot return NULL
in spapr_memory_unplug() to make this clear and silence Coverity.
Fixes: Coverity CID 1450767
Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <161562021166.948373.15092876234470478331.stgit@bahia.lan>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
|
Recent changes allowed the pSeries machine to rollback the hotunplug
process for the DIMM when the guest kernel signals, via a
reconfiguration of the DR connector, that it's not going to release the
LMBs.
Let's also warn QAPI listerners about it. One place to do it would be
right after the unplug state is cleaned up,
spapr_clear_pending_dimm_unplug_state(). This would mean that the
function is now doing more than cleaning up the pending dimm state
though.
This patch does the following changes in spapr.c:
- send a QAPI event to inform that we experienced a failure in the
hotunplug of the DIMM;
- rename spapr_clear_pending_dimm_unplug_state() to
spapr_memory_unplug_rollback(). This is a better fit for what the
function is now doing, and it makes callers care more about what the
function goal is and less about spapr.c internals such as clearing
the pending dimm unplug state.
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210302141019.153729-3-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
|
We are asserting the existence of the first DRC LMB after sending unplug
requests to all LMBs of the DIMM, where every DRC is being asserted
inside the loop. This means that the first DRC is being asserted twice.
Remove the duplicated assert.
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210302141019.153729-2-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
|
Both CPU hotunplug and PC_DIMM unplug reports an user warning,
mentioning that the hotunplug is in progress, if consecutive
'device_del' are issued in quick succession.
Do the same for PHBs in spapr_phb_unplug_request().
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210226163301.419727-4-danielhb413@gmail.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
|
Handling errors in memory hotunplug in the pSeries machine is more
complex than any other device type, because there are all the
complications that other devices has, and more.
For instance, determining a timeout for a DIMM hotunplug must consider
if it's a Hash-MMU or a Radix-MMU guest, because Hash guests takes
longer to hotunplug DIMMs. The size of the DIMM is also a factor, given
that longer DIMMs naturally takes longer to be hotunplugged from the
kernel. And there's also the guest memory usage to be considered: if
there's a process that is consuming memory that would be lost by the
DIMM unplug, the kernel will postpone the unplug process until the
process finishes, and then initiate the regular hotunplug process. The
first two considerations are manageable, but the last one is a deal
breaker.
There is no sane way for the pSeries machine to determine the memory
load in the guest when attempting a DIMM hotunplug - and even if there
was a way, the guest can start using all the RAM in the middle of the
unplug process and invalidate our previous assumptions - and in result
we can't even begin to calculate a timeout for the operation. This means
that we can't implement a viable timeout mechanism for memory unplug in
pSeries.
Going back to why we would consider an unplug timeout, the reason is
that we can't know if the kernel is giving up the unplug. Turns out
that, sometimes, we can. Consider a failed memory hotunplug attempt
where the kernel will error out with the following message:
'pseries-hotplug-mem: Memory indexed-count-remove failed, adding any
removed LMBs'
This happens when there is a LMB that the kernel gave up in removing,
and the LMBs previously marked for removal are now being added back.
This happens in the pseries kernel in [1], dlpar_memory_remove_by_ic()
into dlpar_add_lmb(), and after that update_lmb_associativity_index().
In this function, the kernel is configuring the LMB DRC connector again.
Note that this is a valid usage in LOPAR, as stated in section
"ibm,configure-connector RTAS Call":
'A subsequent sequence of calls to ibm,configure-connector with the same
entry from the “ibm,drc-indexes” or “ibm,drc-info” property will restart
the configuration of devices which were not completely configured.'
We can use this kernel behavior in our favor. If a DRC connector
reconfiguration for a LMB that we marked as unplug pending happens, this
indicates that the kernel changed its mind about the unplug and is
reasserting that it will keep using all the LMBs of the DIMM. In this
case, it's safe to assume that the whole DIMM device unplug was
cancelled.
This patch hops into rtas_ibm_configure_connector() and, in the scenario
described above, clear the unplug state for the DIMM device. This will
not solve all the problems we still have with memory unplug, but it will
cover this case where the kernel reconfigures LMBs after a failed
unplug. We are a bit more resilient, without using an unreliable
timeout, and we didn't make the remaining error cases any worse.
[1] arch/powerpc/platforms/pseries/hotplug-memory.c
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210222194531.62717-6-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
|
There is a reliable way to make a CPU hotunplug fail in the pseries
machine. Hotplug a CPU A, then offline all other CPUs inside the guest
but A. When trying to hotunplug A the guest kernel will refuse to do it,
because A is now the last online CPU of the guest. PAPR has no 'error
callback' in this situation to report back to the platform, so the guest
kernel will deny the unplug in silent and QEMU will never know what
happened. The unplug pending state of A will remain until the guest is
shutdown or rebooted.
Previous attempts of fixing it (see [1] and [2]) were aimed at trying to
mitigate the effects of the problem. In [1] we were trying to guess
which guest CPUs were online to forbid hotunplug of the last online CPU
in the QEMU layer, avoiding the scenario described above because QEMU is
now failing in behalf of the guest. This is not robust because the last
online CPU of the guest can change while we're in the middle of the
unplug process, and our initial assumptions are now invalid. In [2] we
were accepting that our unplug process is uncertain and the user should
be allowed to spam the IRQ hotunplug queue of the guest in case the CPU
hotunplug fails.
This patch presents another alternative, using the timeout
infrastructure introduced in the previous patch. CPU hotunplugs in the
pSeries machine will now timeout after 15 seconds. This is a long time
for a single CPU unplug to occur, regardless of guest load - although
the user is *strongly* encouraged to *not* hotunplug devices from a
guest under high load - and we can be sure that something went wrong if
it takes longer than that for the guest to release the CPU (the same
can't be said about memory hotunplug - more on that in the next patch).
Timing out the unplug operation will reset the unplug state of the CPU
and allow the user to try it again, regardless of the error situation
that prevented the hotunplug to occur. Of all the not so pretty
fixes/mitigations for CPU hotunplug errors in pSeries, timing out the
operation is an admission that we have no control in the process, and
must assume the worst case if the operation doesn't succeed in a
sensible time frame.
[1] https://lists.gnu.org/archive/html/qemu-devel/2021-01/msg03353.html
[2] https://lists.gnu.org/archive/html/qemu-devel/2021-01/msg04400.html
Reported-by: Xujun Ma <xuma@redhat.com>
Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1911414
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210222194531.62717-5-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
|
