Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20210315184615.1985590-16-richard.henderson@linaro.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20210315184615.1985590-15-richard.henderson@linaro.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
On Mac99 and newer machines, the Uninorth PCI host bridge maps
the PCI hole region at 2GiB, so the RAM area beside 2GiB is not
accessible by the CPU. Restrict the memory to 2GiB to avoid
problems such the one reported in the buglink.
Buglink: https://bugs.launchpad.net/qemu/+bug/1922391
Reported-by: Håvard Eidnes <he@NetBSD.org>
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20210406084842.2859664-1-f4bug@amsat.org>
Reviewed-by: BALATON Zoltan <balaton@eik.bme.hu>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Commit 47c8c915b1 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 4515a5f786
"qemu_timer.c: add timer_deadline_ms() helper"
This reverts commit d1c2e3ce3d
"spapr_drc.c: add hotunplug timeout for CPUs"
This reverts commit 51254ffb32
"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>
The e500plat machine device plug callback currently calls
platform_bus_link_device() for any sysbus device. This is overly
broad, because platform_bus_link_device() will unconditionally grab
the IRQs and MMIOs of the device it is passed, whether it was
intended for the platform bus or not. Restrict hotpluggability of
sysbus devices to only those devices on the dynamic sysbus allowlist.
We were mostly getting away with this because the board creates the
platform bus as the last device it creates, and so the hotplug
callback did not do anything for all the sysbus devices created by
the board itself. However if the user plugged in a device which
itself uses a sysbus device internally we would have mishandled this
and probably asserted. An example of this is:
qemu-system-ppc64 -M ppce500 -device macio-oldworld
This isn't a sensible command because the macio-oldworld device
is really specific to the 'g3beige' machine, but we now fail
with a reasonable error message rather than asserting:
qemu-system-ppc64: Device heathrow is not supported by this machine yet.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Message-id: 20210325153310.9131-5-peter.maydell@linaro.org
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>
Per devicetree spec v0.3 [1] chapter 2.3.5:
The #address-cells and #size-cells properties are not inherited
from ancestors in the devicetree. They shall be explicitly defined.
If missing, a client program should assume a default value of 2
for #address-cells, and a value of 1 for #size-cells.
These properties are currently missing, causing the <reg> property
of the queue-group subnode to be incorrectly parsed using default
values.
[1] https://github.com/devicetree-org/devicetree-specification/releases/download/v0.3/devicetree-specification-v0.3.pdf
Fixes: fdfb7f2cdb ("e500: Add support for eTSEC in device tree")
Signed-off-by: Bin Meng <bin.meng@windriver.com>
Message-Id: <20210311081608.66891-1-bmeng.cn@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
The 'ide-hd' and 'ide-cd' devices provide suitable alternatives.
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Previous work on dev-iotlb message broke spapr_iommu/vhost integration
as it did for SMMU and virtio-iommu. The spapr_iommu currently
only sends IOMMU_NOTIFIER_UNMAP notifications. Since commit
958ec334bc ("vhost: Unbreak SMMU and virtio-iommu on dev-iotlb support"),
VHOST first tries to register IOMMU_NOTIFIER_DEVIOTLB_UNMAP notifier
and if it fails, falls back to legacy IOMMU_NOTIFIER_UNMAP. So
spapr_iommu must fail on the IOMMU_NOTIFIER_DEVIOTLB_UNMAP
registration.
Reported-by: Peter Xu <peterx@redhat.com>
Fixes: b68ba1ca57 ("memory: Add IOMMU_NOTIFIER_DEVIOTLB_UNMAP IOMMUTLBNotificationType")
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Message-Id: <20210209213233.40985-3-eric.auger@redhat.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Acked-by: Jason Wang <jasowang@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Next batch of patches for the ppc target and machine types. Includes:
* Several cleanups for sm501 from Peter Maydell
* An update to the SLOF guest firmware
* Improved handling of hotplug failures in spapr, associated cleanups
to the hotplug handling code
* Several etsec fixes and cleanups from Bin Meng
* Assorted other fixes and cleanups
-----BEGIN PGP SIGNATURE-----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=/uyx
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'remotes/dg-gitlab/tags/ppc-for-6.0-20210310' into staging
ppc patch queue for 2021-03-10
Next batch of patches for the ppc target and machine types. Includes:
* Several cleanups for sm501 from Peter Maydell
* An update to the SLOF guest firmware
* Improved handling of hotplug failures in spapr, associated cleanups
to the hotplug handling code
* Several etsec fixes and cleanups from Bin Meng
* Assorted other fixes and cleanups
# gpg: Signature made Wed 10 Mar 2021 04:08:53 GMT
# 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-6.0-20210310:
spapr.c: send QAPI event when memory hotunplug fails
spapr.c: remove duplicated assert in spapr_memory_unplug_request()
target/ppc: fix icount support on Book-e vms accessing SPRs
qemu_timer.c: add timer_deadline_ms() helper
spapr_pci.c: add 'unplug already in progress' message for PCI unplug
spapr.c: add 'unplug already in progress' message for PHB unplug
hw/ppc: e500: Add missing <ranges> in the eTSEC node
hw/net: fsl_etsec: Fix build error when HEX_DUMP is on
spapr_drc.c: use DRC reconfiguration to cleanup DIMM unplug state
spapr_drc.c: add hotunplug timeout for CPUs
spapr_drc.c: introduce unplug_timeout_timer
target/ppc: Fix bcdsub. emulation when result overflows
docs/system: Extend PPC section
spapr: rename spapr_drc_detach() to spapr_drc_unplug_request()
spapr_drc.c: use spapr_drc_release() in isolate_physical/set_unusable
pseries: Update SLOF firmware image
spapr_drc.c: do not call spapr_drc_detach() in drc_isolate_logical()
hw/display/sm501: Inline template header into C file
hw/display/sm501: Expand out macros in template header
hw/display/sm501: Remove dead code for non-32-bit RGB surfaces
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
The 'running' argument from VMChangeStateHandler does not require
other value than 0 / 1. Make it a plain boolean.
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <20210111152020.1422021-3-philmd@redhat.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
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>
The pSeries machine is using QEMUTimer internals to return the timeout
in seconds for a timer object, in hw/ppc/spapr.c, function
spapr_drc_unplug_timeout_remaining_sec().
Create a helper in qemu-timer.c to retrieve the deadline for a QEMUTimer
object, in ms, to avoid exposing timer internals to the PPC code.
CC: Paolo Bonzini <pbonzini@redhat.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210301124133.23800-2-danielhb413@gmail.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Hotunplug for all other devices are warning the user when the hotunplug
is already in progress. Do the same for PCI devices in
spapr_pci_unplug_request().
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210226163301.419727-5-danielhb413@gmail.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
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>
The eTSEC node should provide an empty <ranges> property in the
eTSEC node, otherwise of_translate_address() in the Linux kernel
fails to get the eTSEC register base, reporting:
OF: ** translation for device /platform@f00000000/ethernet@0/queue-group **
OF: bus is default (na=1, ns=1) on /platform@f00000000/ethernet@0
OF: translating address: 00000000
OF: parent bus is default (na=1, ns=1) on /platform@f00000000
OF: no ranges; cannot translate
Per devicetree spec v0.3 [1] chapter 2.3.8:
If the property is not present in a bus node, it is assumed that
no mapping exists between children of the node and the parent
address space.
This is why of_translate_address() aborts the address translation.
Apparently U-Boot devicetree parser seems to be tolerant with
missing <ranges> as this was not noticed when testing with U-Boot.
The empty <ranges> property is present in all kernel shipped dtsi
files for eTSEC, Let's add it to conform with the spec.
[1] https://github.com/devicetree-org/devicetree-specification/releases/download/v0.3/devicetree-specification-v0.3.pdf
Fixes: fdfb7f2cdb ("e500: Add support for eTSEC in device tree")
Signed-off-by: Bin Meng <bin.meng@windriver.com>
Message-Id: <1614158919-9473-1-git-send-email-bmeng.cn@gmail.com>
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>
The LoPAR spec provides no way for the guest kernel to report failure of
hotplug/hotunplug events. This wouldn't be bad if those operations were
granted to always succeed, but that's far for the reality.
What ends up happening is that, in the case of a failed hotunplug,
regardless of whether it was a QEMU error or a guest misbehavior, the
pSeries machine is retaining the unplug state of the device in the
running guest. This state is cleanup in machine reset, where it is
assumed that this state represents a device that is pending unplug, and
the device is hotunpluged from the board. Until the reset occurs, any
hotunplug operation of the same device is forbid because there is a
pending unplug state.
This behavior has at least one undesirable side effect. A long standing
pending unplug state is, more often than not, the result of a hotunplug
error. The user had to dealt with it, since retrying to unplug the
device is noy allowed, and then in the machine reset we're removing the
device from the guest. This means that we're failing the user twice -
failed to hotunplug when asked, then hotunplugged without notice.
Solutions to this problem range between trying to predict when the
hotunplug will fail and forbid the operation from the QEMU layer, from
opening up the IRQ queue to allow for multiple hotunplug attempts, from
telling the users to 'reboot the machine if something goes wrong'. The
first solution is flawed because we can't fully predict guest behavior
from QEMU, the second solution is a trial and error remediation that
counts on a hope that the unplug will eventually succeed, and the third
is ... well.
This patch introduces a crude, but effective solution to hotunplug
errors in the pSeries machine. For each unplug done, we'll timeout after
some time. If a certain amount of time passes, we'll cleanup the
hotunplug state from the machine. During the timeout period, any unplug
operations in the same device will still be blocked. After that, we'll
assume that the guest failed the operation, and allow the user to try
again. If the timeout is too short we'll prevent legitimate hotunplug
situations to occur, so we'll need to overestimate the regular time an
unplug operation takes to succeed to account that.
The true solution for the hotunplug errors in the pSeries machines is a
PAPR change to allow for the guest to warn the platform about it. For
now, the work done in this timeout design can be used for the new PAPR
'abort hcall' in the future, given that for both cases we'll need code
to cleanup the existing unplug states of the DRCs.
At this moment we're adding the basic wiring of the timer into the DRC.
Next patch will use the timer to timeout failed CPU hotunplugs.
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210222194531.62717-4-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
spapr_drc_detach() is not the best name for what the function does. The
function does not detach the DRC, it makes an uncommited attempt to do
it. It'll mark the DRC as pending unplug, via the 'unplug_request'
flag, and only if the DRC state is drck->empty_state it will detach the
DRC, via spapr_drc_release().
This is a contrast with its pair spapr_drc_attach(), where the function
is indeed creating the DRC QOM object. If you know what
spapr_drc_attach() does, you can be misled into thinking that
spapr_drc_detach() is removing the DRC from QEMU internal state, which
isn't true.
The current role of this function is better described as a request for
detach, since there's no guarantee that we're going to detach the DRC in
the end. Rename the function to spapr_drc_unplug_request to reflect
what is is doing.
The initial idea was to change the name to spapr_drc_detach_request(),
and later on change the unplug_request flag to detach_request. However,
unplug_request is a migratable boolean for a long time now and renaming
it is not worth the trouble. spapr_drc_unplug_request() setting
drc->unplug_request is more natural than spapr_drc_detach_request
setting drc->unplug_request.
Reviewed-by: Greg Kurz <groug@kaod.org>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210222194531.62717-3-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
When moving a physical DRC to "Available", drc_isolate_physical() will
move the DRC state to STATE_PHYSICAL_POWERON and, if the DRC is marked
for unplug, call spapr_drc_detach(). For physical DRCs,
drck->empty_state is STATE_PHYSICAL_POWERON, meaning that we're sure
that spapr_drc_detach() will end up calling spapr_drc_release() in the
end.
Likewise, for logical DRCs, drc_set_unusable will move the DRC to
"Unusable" state, setting drc->state to STATE_LOGICAL_UNUSABLE, which is
the drck->empty_state for logical DRCs. spapr_drc_detach() will call
spapr_drc_release() in this case as well.
In both scenarios, spapr_drc_detach() is being used as a
spapr_drc_release(), wrapper, where we also set unplug_requested (which
is already true, otherwise spapr_drc_detach() wouldn't be called in the
first place) and check if drc->state == drck->empty_state, which we also
know it's guaranteed to be true because we just set it.
Just use spapr_drc_release() in these functions to be clear of our
intentions in both these functions.
Reviewed-by: Greg Kurz <groug@kaod.org>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210222194531.62717-2-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
drc_isolate_logical() is used to move the DRC from the "Configured" to
the "Available" state, erroring out if the DRC is in the unexpected
"Unisolate" state and doing nothing (with RTAS_OUT_SUCCESS) if the DRC
is already in "Available" or in "Unusable" state.
When moving from "Configured" to "Available", the DRC is moved to the
LOGICAL_AVAILABLE state, a drc->unplug_requested check is done and, if
true, spapr_drc_detach() is called.
What spapr_drc_detach() does then is:
- set drc->unplug_requested to true. In fact, this is the only place
where unplug_request is set to true;
- does nothing else if drc->state != drck->empty_state. If the DRC
state is equal to drck->empty_state, spapr_drc_release() is
called. For logical DRCs, drck->empty_state = LOGICAL_UNUSABLE.
In short, calling spapr_drc_detach() in drc_isolate_logical() does
nothing. It'll set unplug_request to true again ('again' since it was
already true - otherwise the function wouldn't be called), and will
return without calling spapr_drc_release() because the DRC is not in
LOGICAL_UNUSABLE, since drc_isolate_logical() just moved it to
LOGICAL_AVAILABLE. The only place where the logical DRC is released is
when called from drc_set_unusable(), when it is moved to the
"Unusable" state. As it should, according to PAPR.
Even though calling spapr_drc_detach() in drc_isolate_logical() is
benign, removing it will avoid further thought about the matter. So
let's go ahead and do that.
As a note, this logic was introduced in commit bbf5c878ab. Since
then, the DRC handling code was refactored and enhanced, and PAPR
itself went through some changes in the DRC area as well. It is
expected that some assumptions we had back then are now deprecated.
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210211225246.17315-2-danielhb413@gmail.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
We forward-declare Object typedef in "qemu/typedefs.h" since commit
ca27b5eb7c ("qom/object: Move Object typedef to 'qemu/typedefs.h'").
Use it everywhere to make the code simpler.
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Message-Id: <20210225182003.3629342-1-philmd@redhat.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
An assorted set of spelling fixes in various places.
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Reviewed-by: Stefan Weil <sw@weilnetz.de>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20210309111510.79495-1-mjt@msgid.tls.msk.ru>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
There are 23 files that include the "sysemu/qtest.h",
but they do not use any qtest functions.
Signed-off-by: Chen Qun <kuhn.chenqun@huawei.com>
Acked-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20210226081414.205946-1-kuhn.chenqun@huawei.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
At present the <clock-frequency> property of the serial node is
populated with value zero. U-Boot's ns16550 driver is not happy
about this, so let's fill in a meaningful value.
Signed-off-by: Bin Meng <bin.meng@windriver.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <1612362288-22216-2-git-send-email-bmeng.cn@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
At present the platform clock frequency is using a magic number.
Convert it to a macro and use it everywhere.
Signed-off-by: Bin Meng <bin.meng@windriver.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <1612362288-22216-1-git-send-email-bmeng.cn@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
The current logic for calculating 'maxdomain' making it a sum of
numa_state->num_nodes with spapr->gpu_numa_id. spapr->gpu_numa_id is
used as a index to determine the next available NUMA id that a
given NVGPU can use.
The problem is that the initial value of gpu_numa_id, for any topology
that has more than one NUMA node, is equal to numa_state->num_nodes.
This means that our maxdomain will always be, at least, twice the
amount of existing NUMA nodes. This means that a guest with 4 NUMA
nodes will end up with the following max-associativity-domains:
rtas/ibm,max-associativity-domains
00000004 00000008 00000008 00000008 00000008
This overtuning of maxdomains doesn't go unnoticed in the guest, being
detected in SLUB during boot:
dmesg | grep SLUB
[ 0.000000] SLUB: HWalign=128, Order=0-3, MinObjects=0, CPUs=4, Nodes=8
SLUB is detecting 8 total nodes, with 4 nodes being online.
This patch fixes ibm,max-associativity-domains by considering the amount
of NVGPUs NUMA nodes presented in the guest, instead of just
spapr->gpu_numa_id.
Reported-by: Cédric Le Goater <clg@kaod.org>
Tested-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210128174213.1349181-4-danielhb413@gmail.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
We'll need to check the initial value given to spapr->gpu_numa_id when
building the rtas DT, so put it in a helper for easier access and to
avoid repetition.
Tested-by: Cédric Le Goater <clg@kaod.org>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210128174213.1349181-3-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
This function is used only in spapr_numa.c.
Tested-by: Cédric Le Goater <clg@kaod.org>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210128174213.1349181-2-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
This to map the PNOR from the machine init handler directly and finish
the cleanup of the LPC model.
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Message-Id: <20210126171059.307867-8-clg@kaod.org>
Reviewed-by: Joel Stanley <joel@jms.id.au>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
On PowerNV systems, the BMC is in charge of mapping the PNOR contents
on the LPC FW address space using the HIOMAP protocol. Under QEMU, we
emulate this behavior and we also add an extra control on the flash
accesses by letting the HIOMAP command handler decide whether the
memory region is accessible or not depending on the firmware requests.
However, this behavior is not compatible with hostboot like firmwares
which need this mapping to be always available. For this reason, the
PNOR memory region is initially disabled for skiboot mode only.
This is badly placed under the LPC model and requires the use of the
machine. Since it doesn't add much, simply remove the initial setting.
The extra control in the HIOMAP command handler will still be performed.
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Message-Id: <20210126171059.307867-7-clg@kaod.org>
Reviewed-by: Joel Stanley <joel@jms.id.au>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
The PowerNV machine can be run with an external IPMI BMC device
connected to a remote QEMU machine acting as BMC, using these options :
-chardev socket,id=ipmi0,host=localhost,port=9002,reconnect=10 \
-device ipmi-bmc-extern,id=bmc0,chardev=ipmi0 \
-device isa-ipmi-bt,bmc=bmc0,irq=10 \
-nodefaults
In that case, some aspects of the BMC initialization should be
skipped, since they rely on the simulator interface.
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Message-Id: <20210126171059.307867-6-clg@kaod.org>
Reviewed-by: Joel Stanley <joel@jms.id.au>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
and reuse pnv_bmc_set_pnor() to share the setting of the PNOR.
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Message-Id: <20210126171059.307867-5-clg@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
The current settings are useful to load large kernels (with debug) but
it moves the initrd image in a memory region not protected by
skiboot. If skiboot is compiled with DEBUG=1, memory poisoning will
corrupt the initrd.
Cc: Murilo Opsfelder Araujo <muriloo@linux.ibm.com>
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Message-Id: <20210126171059.307867-4-clg@kaod.org>
Reviewed-by: Murilo Opsfelder Araujo <muriloo@linux.ibm.com>
Reviewed-by: Joel Stanley <joel@jms.id.au>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
It is currently not possible to perform a strict boot from USB storage:
$ qemu-system-ppc64 -accel kvm -nodefaults -nographic -serial stdio \
-boot strict=on \
-device qemu-xhci \
-device usb-storage,drive=disk,bootindex=0 \
-blockdev driver=file,node-name=disk,filename=fedora-ppc64le.qcow2
SLOF **********************************************************************
QEMU Starting
Build Date = Jul 17 2020 11:15:24
FW Version = git-e18ddad8516ff2cf
Press "s" to enter Open Firmware.
Populating /vdevice methods
Populating /vdevice/vty@71000000
Populating /vdevice/nvram@71000001
Populating /pci@800000020000000
00 0000 (D) : 1b36 000d serial bus [ usb-xhci ]
No NVRAM common partition, re-initializing...
Scanning USB
XHCI: Initializing
USB Storage
SCSI: Looking for devices
101000000000000 DISK : "QEMU QEMU HARDDISK 2.5+"
Using default console: /vdevice/vty@71000000
Welcome to Open Firmware
Copyright (c) 2004, 2017 IBM Corporation All rights reserved.
This program and the accompanying materials are made available
under the terms of the BSD License available at
http://www.opensource.org/licenses/bsd-license.php
Trying to load: from: /pci@800000020000000/usb@0/storage@1/disk@101000000000000 ...
E3405: No such device
E3407: Load failed
Type 'boot' and press return to continue booting the system.
Type 'reset-all' and press return to reboot the system.
Ready!
0 >
The device tree handed over by QEMU to SLOF indeed contains:
qemu,boot-list =
"/pci@800000020000000/usb@0/storage@1/disk@101000000000000 HALT";
but the device node is named usb-xhci@0, not usb@0.
This happens because the firmware names of PCI devices returned
by get_boot_devices_list() come from pcibus_get_fw_dev_path(),
while the sPAPR PHB code uses a different naming scheme for
device nodes. This inconsistency has always been there but it was
hidden for a long time because SLOF used to rename USB device
nodes, until this commit, merged in QEMU 4.2.0 :
commit 85164ad4ed
Author: Alexey Kardashevskiy <aik@ozlabs.ru>
Date: Wed Sep 11 16:24:32 2019 +1000
pseries: Update SLOF firmware image
This fixes USB host bus adapter name in the device tree to match QEMU's
one.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Fortunately, sPAPR implements the firmware path provider interface.
This provides a way to override the default firmware paths.
Just factor out the sPAPR PHB naming logic from spapr_dt_pci_device()
to a helper, and use it in the sPAPR firmware path provider hook.
Fixes: 85164ad4ed ("pseries: Update SLOF firmware image")
Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <20210122170157.246374-1-groug@kaod.org>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
In the CPU hotunplug bug [1] the guest kernel throws a scary
message in dmesg:
pseries-hotplug-cpu: Failed to offline CPU <NULL>, rc: -16
The reason isn't related to the bug though. This happens because the
kernel file arch/powerpc/platform/pseries/hotplug-cpu.c, function
dlpar_cpu_remove(), is not finding the device_node.name of the offending
CPU.
We're not populating the 'name' property for hotplugged CPUs. Since the
kernel relies on device_node.name for identifying CPU nodes, and the
CPUs that are coldplugged has the 'name' property filled by SLOF, this
is creating an unneeded inconsistency between hotplug and coldplug CPUs
in the kernel.
Let's fill the 'name' property for hotplugged CPUs as well. This will
make the guest dmesg throws a less intimidating message when we try to
unplug the last online CPU:
pseries-hotplug-cpu: Failed to offline CPU PowerPC,POWER9@1, rc: -16
[1] https://bugzilla.redhat.com/1911414
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210120232305.241521-3-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Next patch will use the 'nodename' string in spapr_core_dt_populate()
after the point it's being freed today.
Instead of moving 'g_free(nodename)' around, let's do a QoL change in
both CPU DT functions where 'nodename' is being freed, and use
g_autofree to avoid the 'g_free()' call altogether.
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210120232305.241521-2-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
v2
Dropped vmstate: Fix memory leak in vmstate_handle_alloc
Broke on Power
Added migration: only check page size match if RAM postcopy is enabled
-----BEGIN PGP SIGNATURE-----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=4tbe
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'remotes/dgilbert/tags/pull-migration-20210208a' into staging
Migration pull 2021-02-08
v2
Dropped vmstate: Fix memory leak in vmstate_handle_alloc
Broke on Power
Added migration: only check page size match if RAM postcopy is enabled
# gpg: Signature made Mon 08 Feb 2021 11:28:14 GMT
# gpg: using RSA key 45F5C71B4A0CB7FB977A9FA90516331EBC5BFDE7
# gpg: Good signature from "Dr. David Alan Gilbert (RH2) <dgilbert@redhat.com>" [full]
# Primary key fingerprint: 45F5 C71B 4A0C B7FB 977A 9FA9 0516 331E BC5B FDE7
* remotes/dgilbert/tags/pull-migration-20210208a: (27 commits)
migration: only check page size match if RAM postcopy is enabled
migration: introduce snapshot-{save, load, delete} QMP commands
iotests: fix loading of common.config from tests/ subdir
iotests: add support for capturing and matching QMP events
migration: introduce a delete_snapshot wrapper
migration: wire up support for snapshot device selection
migration: control whether snapshots are ovewritten
block: rename and alter bdrv_all_find_snapshot semantics
block: allow specifying name of block device for vmstate storage
block: add ability to specify list of blockdevs during snapshot
migration: stop returning errno from load_snapshot()
migration: Make save_snapshot() return bool, not 0/-1
block: push error reporting into bdrv_all_*_snapshot functions
migration: Display the migration blockers
migration: Add blocker information
migration: Fix a few absurdly defective error messages
migration: Fix cache_init()'s "Failed to allocate" error messages
migration: Clean up signed vs. unsigned XBZRLE cache-size
migration: Fix migrate-set-parameters argument validation
migration: introduce 'userfaultfd-wrlat.py' script
...
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
When VM migrate VMState of spapr_pci, the field(msi_devs) of spapr_pci
having a flag of VMS_ALLOC need to allocate memory. If the src doesn't free
memory of msi_devs in SaveStateEntry of spapr_pci after QEMUFile save
VMState of spapr_pci, it may result in memory leak of msi_devs. We add the
post_save func to free memory, which prevents memory leak.
Reported-by: Euler Robot <euler.robot@huawei.com>
Signed-off-by: Jinhao Gao <gaojinhao@huawei.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Message-Id: <20201231061020.828-2-gaojinhao@huawei.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
We haven't yet implemented the fairly involved handshaking that will be
needed to migrate PEF protected guests. For now, just use a migration
blocker so we get a meaningful error if someone attempts this (this is the
same approach used by AMD SEV).
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Some upcoming POWER machines have a system called PEF (Protected
Execution Facility) which uses a small ultravisor to allow guests to
run in a way that they can't be eavesdropped by the hypervisor. The
effect is roughly similar to AMD SEV, although the mechanisms are
quite different.
Most of the work of this is done between the guest, KVM and the
ultravisor, with little need for involvement by qemu. However qemu
does need to tell KVM to allow secure VMs.
Because the availability of secure mode is a guest visible difference
which depends on having the right hardware and firmware, we don't
enable this by default. In order to run a secure guest you need to
create a "pef-guest" object and set the confidential-guest-support
property to point to it.
Note that this just *allows* secure guests, the architecture of PEF is
such that the guest still needs to talk to the ultravisor to enter
secure mode. Qemu has no direct way of knowing if the guest is in
secure mode, and certainly can't know until well after machine
creation time.
To start a PEF-capable guest, use the command line options:
-object pef-guest,id=pef0 -machine confidential-guest-support=pef0
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Greg Kurz <groug@kaod.org>