Perform the atomic update for hardware management of the dirty bit.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20221024051851.3074715-14-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Perform the atomic update for hardware management of the access flag.
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20221024051851.3074715-13-richard.henderson@linaro.org
[PMM: Fix accidental PROT_WRITE to PAGE_WRITE; add missing
main-loop.h include]
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Replace some gotos with some nested if statements.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-id: 20221024051851.3074715-12-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Both GP and DBM are in the upper attribute block.
Extend the computation of attrs to include them,
then simplify the setting of guarded.
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-id: 20221024051851.3074715-11-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Leave the upper and lower attributes in the place they originate
from in the descriptor. Shifting them around is confusing, since
one cannot read the bit numbers out of the manual. Also, new
attributes have been added which would alter the shifts.
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-id: 20221024051851.3074715-10-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Always overriding fi->type was incorrect, as we would not properly
propagate the fault type from S1_ptw_translate, or arm_ldq_ptw.
Simplify things by providing a new label for a translation fault.
For other faults, store into fi directly.
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-id: 20221024051851.3074715-9-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
The unconditional loop was used both to iterate over levels
and to control parsing of attributes. Use an explicit goto
in both cases.
While this appears less clean for iterating over levels, we
will need to jump back into the middle of this loop for
atomic updates, which is even uglier.
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20221024051851.3074715-8-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
This fault type is to be used with FEAT_HAFDBS when
the guest enables hw updates, but places the tables
in memory where atomic updates are unsupported.
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-id: 20221024051851.3074715-7-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Separate S1 translation from the actual lookup.
Will enable lpae hardware updates.
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20221024051851.3074715-6-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-id: 20221024051851.3074715-5-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
The MMFR1 field may indicate support for hardware update of
access flag alone, or access flag and dirty bit.
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20221024051851.3074715-4-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Hoist the computation of the mmu_idx for the ptw up to
get_phys_addr_with_struct and get_phys_addr_twostage.
This removes the duplicate check for stage2 disabled
from the middle of the walk, performing it only once.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Tested-by: Alex Bennée <alex.bennee@linaro.org>
Message-id: 20221024051851.3074715-3-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reduce the amount of typing required for this check.
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20221024051851.3074715-2-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
The semantic difference between the deprecated device_legacy_reset()
function and the newer device_cold_reset() function is that the new
function resets both the device itself and any qbuses it owns,
whereas the legacy function resets just the device itself and nothing
else. In hyperv_synic_reset() we reset a SynICState, which has no
qbuses, so for this purpose the two functions behave identically and
we can stop using the deprecated one.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Maciej S. Szmigiero <maciej.szmigiero@oracle.com>
Message-id: 20221013171817.1447562-1-peter.maydell@linaro.org
The code for handling the reset level count in the Resettable code
has two issues:
The reset count is only decremented for the 1->0 case. This means
that if there's ever a nested reset that takes the count to 2 then it
will never again be decremented. Eventually the count will exceed
the '50' limit in resettable_phase_enter() and QEMU will trip over
the assertion failure. The repro case in issue 1266 is an example of
this that happens now the SCSI subsystem uses three-phase reset.
Secondly, the count is decremented only after the exit phase handler
is called. Moving the reset count decrement from "just after" to
"just before" calling the exit phase handler allows
resettable_is_in_reset() to return false during the handler
execution.
This simplifies reset handling in resettable devices. Typically, a
function that updates the device state will just need to read the
current reset state and not anymore treat the "in a reset-exit
transition" as a special case.
Note that the semantics change to the *_is_in_reset() functions
will have no effect on the current codebase, because only two
devices (hw/char/cadence_uart.c and hw/misc/zynq_sclr.c) currently
call those functions, and in neither case do they do it from the
device's exit phase methed.
Fixes: 4a5fc890 ("scsi: Use device_cold_reset() and bus_cold_reset()")
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/1266
Signed-off-by: Damien Hedde <damien.hedde@greensocs.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reported-by: Michael Peter <michael.peter@hensoldt-cyber.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-id: 20221020142749.3357951-1-peter.maydell@linaro.org
Buglink: https://bugs.launchpad.net/qemu/+bug/1905297
Reported-by: Michael Peter <michael.peter@hensoldt-cyber.com>
[PMM: adjust the docs paragraph changed to get the name of the
'enter' phase right and to clarify exactly when the count is
adjusted; rewrite the commit message]
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
An exception targeting EL2 from lower EL is actually maskable when
HCR_E2H and HCR_TGE are both set. This applies to both secure and
non-secure Security state.
We can remove the conditions that try to suppress masking of
interrupts when we are Secure and the exception targets EL2 and
Secure EL2 is disabled. This is OK because in that situation
arm_phys_excp_target_el() will never return 2 as the target EL. The
'not if secure' check in this function was originally written before
arm_hcr_el2_eff(), and back then the target EL returned by
arm_phys_excp_target_el() could be 2 even if we were in Secure
EL0/EL1; but it is no longer needed.
Signed-off-by: Ake Koomsin <ake@igel.co.jp>
Message-id: 20221017092432.546881-1-ake@igel.co.jp
[PMM: Add commit message paragraph explaining why it's OK to
remove the checks on secure and SCR_EEL2]
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
The "PCI Bus Binding to: IEEE Std 1275-1994" defines the compatible
string for a PCIe bus or endpoint as "pci<vendorid>,<deviceid>" or
similar. Since the initial binding for PCI virtio-iommu didn't follow
this rule, it was modified to accept both strings and ensure backward
compatibility. Also, the unit-name for the node should be
"device,function".
Fix corresponding dt-validate and dtc warnings:
pcie@10000000: virtio_iommu@16:compatible: ['virtio,pci-iommu'] does not contain items matching the given schema
pcie@10000000: Unevaluated properties are not allowed (... 'virtio_iommu@16' were unexpected)
From schema: linux/Documentation/devicetree/bindings/pci/host-generic-pci.yaml
virtio_iommu@16: compatible: 'oneOf' conditional failed, one must be fixed:
['virtio,pci-iommu'] is too short
'pci1af4,1057' was expected
From schema: dtschema/schemas/pci/pci-bus.yaml
Warning (pci_device_reg): /pcie@10000000/virtio_iommu@16: PCI unit address format error, expected "2,0"
Signed-off-by: Jean-Philippe Brucker <jean-philippe@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
FEAT_E0PD adds new bits E0PD0 and E0PD1 to TCR_EL1, which allow the
OS to forbid EL0 access to half of the address space. Since this is
an EL0-specific variation on the existing TCR_ELx.{EPD0,EPD1}, we can
implement it entirely in aa64_va_parameters().
This requires moving the existing regime_is_user() to internals.h
so that the code in helper.c can get at it.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20221021160131.3531787-1-peter.maydell@linaro.org
Currently, there is no way to configure a CPU affinity inside QEMU when
the sandbox option disables it for QEMU as a whole, for example, via:
-sandbox enable=on,resourcecontrol=deny
While ThreadContext objects can be created on the QEMU commandline and
the CPU affinity can be configured externally via the thread-id, this is
insufficient if a ThreadContext with a certain CPU affinity is already
required during QEMU startup, before we can intercept QEMU and
configure the CPU affinity.
Blocking sched_setaffinity() was introduced in 24f8cdc572 ("seccomp:
add resourcecontrol argument to command line"), "to avoid any bigger of the
process". However, we only care about once QEMU is running, not when
the instance starting QEMU explicitly requests a certain CPU affinity
on the QEMU comandline.
Right now, for NUMA-aware preallocation of memory backends used for initial
machine RAM, one has to:
1) Start QEMU with the memory-backend with "prealloc=off"
2) Pause QEMU before it starts the guest (-S)
3) Create ThreadContext, configure the CPU affinity using the thread-id
4) Configure the ThreadContext as "prealloc-context" of the memory
backend
5) Trigger preallocation by setting "prealloc=on"
To simplify this handling especially for initial machine RAM,
allow creation of ThreadContext objects before parsing sandbox options,
such that the CPU affinity requested on the QEMU commandline alongside the
sandbox option can be set. As ThreadContext objects essentially only create
a persistent context thread and set the CPU affinity, this is easily
possible.
With this change, we can create a ThreadContext with a CPU affinity on
the QEMU commandline and use it for preallocation of memory backends
glued to the machine (simplified example):
To make "-name debug-threads=on" keep working as expected for the
context threads, perform earlier parsing of "-name".
qemu-system-x86_64 -m 1G \
-object thread-context,id=tc1,cpu-affinity=3-4 \
-object memory-backend-ram,id=pc.ram,size=1G,prealloc=on,prealloc-threads=2,prealloc-context=tc1 \
-machine memory-backend=pc.ram \
-S -monitor stdio -sandbox enable=on,resourcecontrol=deny
And while we can query the current CPU affinity:
(qemu) qom-get tc1 cpu-affinity
[
3,
4
]
We can no longer change it from QEMU directly:
(qemu) qom-set tc1 cpu-affinity 1-2
Error: Setting CPU affinity failed: Operation not permitted
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Message-Id: <20221014134720.168738-8-david@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Let's allow for specifying a thread context via the "prealloc-context"
property. When set, preallcoation threads will be crated via the
thread context -- inheriting the same CPU affinity as the thread
context.
Pinning preallcoation threads to CPUs can heavily increase performance
in NUMA setups, because, preallocation from a CPU close to the target
NUMA node(s) is faster then preallocation from a CPU further remote,
simply because of memory bandwidth for initializing memory with zeroes.
This is especially relevant for very large VMs backed by huge/gigantic
pages, whereby preallocation is mandatory.
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Message-Id: <20221014134720.168738-7-david@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
... and implement it under POSIX. When a ThreadContext is provided,
create new threads via the context such that these new threads obtain a
properly configured CPU affinity.
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Message-Id: <20221014134720.168738-6-david@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Let's make it easier to pin threads created via a ThreadContext to
all host CPUs currently belonging to a given set of host NUMA nodes --
which is the common case.
"node-affinity" is simply a shortcut for setting "cpu-affinity" manually
to the list of host CPUs belonging to the set of host nodes. This property
can only be written.
A simple QEMU example to set the CPU affinity to host node 1 on a system
with two nodes, 24 CPUs each, whereby odd-numbered host CPUs belong to
host node 1:
qemu-system-x86_64 -S \
-object thread-context,id=tc1,node-affinity=1
And we can query the cpu-affinity via HMP/QMP:
(qemu) qom-get tc1 cpu-affinity
[
1,
3,
5,
7,
9,
11,
13,
15,
17,
19,
21,
23,
25,
27,
29,
31,
33,
35,
37,
39,
41,
43,
45,
47
]
We cannot query the node-affinity:
(qemu) qom-get tc1 node-affinity
Error: Insufficient permission to perform this operation
But note that due to dynamic library loading this example will not work
before we actually make use of thread_context_create_thread() in QEMU
code, because the type will otherwise not get registered. We'll wire
this up next to make it work.
Note that if the host CPUs for a host node change due do CPU hot(un)plug
CPU onlining/offlining (i.e., lscpu output changes) after the ThreadContext
was started, the CPU affinity will not get updated.
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Acked-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20221014134720.168738-5-david@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Setting the CPU affinity of QEMU threads is a bit problematic, because
QEMU doesn't always have permissions to set the CPU affinity itself,
for example, with seccomp after initialized by QEMU:
-sandbox enable=on,resourcecontrol=deny
General information about CPU affinities can be found in the man page of
taskset:
CPU affinity is a scheduler property that "bonds" a process to a given
set of CPUs on the system. The Linux scheduler will honor the given CPU
affinity and the process will not run on any other CPUs.
While upper layers are already aware of how to handle CPU affinities for
long-lived threads like iothreads or vcpu threads, especially short-lived
threads, as used for memory-backend preallocation, are more involved to
handle. These threads are created on demand and upper layers are not even
able to identify and configure them.
Introduce the concept of a ThreadContext, that is essentially a thread
used for creating new threads. All threads created via that context
thread inherit the configured CPU affinity. Consequently, it's
sufficient to create a ThreadContext and configure it once, and have all
threads created via that ThreadContext inherit the same CPU affinity.
The CPU affinity of a ThreadContext can be configured two ways:
(1) Obtaining the thread id via the "thread-id" property and setting the
CPU affinity manually (e.g., via taskset).
(2) Setting the "cpu-affinity" property and letting QEMU try set the
CPU affinity itself. This will fail if QEMU doesn't have permissions
to do so anymore after seccomp was initialized.
A simple QEMU example to set the CPU affinity to host CPU 0,1,6,7 would be:
qemu-system-x86_64 -S \
-object thread-context,id=tc1,cpu-affinity=0-1,cpu-affinity=6-7
And we can query it via HMP/QMP:
(qemu) qom-get tc1 cpu-affinity
[
0,
1,
6,
7
]
But note that due to dynamic library loading this example will not work
before we actually make use of thread_context_create_thread() in QEMU
code, because the type will otherwise not get registered. We'll wire
this up next to make it work.
In general, the interface behaves like pthread_setaffinity_np(): host
CPU numbers that are currently not available are ignored; only host CPU
numbers that are impossible with the current kernel will fail. If the
list of host CPU numbers does not include a single CPU that is
available, setting the CPU affinity will fail.
A ThreadContext can be reused, simply by reconfiguring the CPU affinity.
Note that the CPU affinity of previously created threads will not get
adjusted.
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Acked-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20221014134720.168738-4-david@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Usually, we let upper layers handle CPU pinning, because
pthread_setaffinity_np() (-> sched_setaffinity()) is blocked via
seccomp when starting QEMU with
-sandbox enable=on,resourcecontrol=deny
However, we want to configure and observe the CPU affinity of threads
from QEMU directly in some cases when the sandbox option is either not
enabled or not active yet.
So let's add a way to configure CPU pinning via
qemu_thread_set_affinity() and obtain CPU affinity via
qemu_thread_get_affinity() and implement them under POSIX using
pthread_setaffinity_np() + pthread_getaffinity_np().
Implementation under Windows is possible using SetProcessAffinityMask()
+ GetProcessAffinityMask(), however, that is left as future work.
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Message-Id: <20221014134720.168738-3-david@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Let's
* give the function a "qemu_*" style name
* make sure the parameters in the implementation match the prototype
* rename smp_cpus to max_threads, which makes the semantics of that
parameter clearer
... and add a function documentation.
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Message-Id: <20221014134720.168738-2-david@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
The element size is encoded in the M3 field, not in the M4
field.
Fixes: be6324c6b7 ("s390x/tcg: Implement VECTOR ISOLATE STRING")
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/1248
Message-Id: <20221012182755.1014853-3-thuth@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
This is common practice, see the Makefile.target in the aarch64
folder for example.
Suggested-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20221012182755.1014853-2-thuth@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Under PV, the guest's TOD clock is under control of the ultravisor and the
hypervisor cannot change it.
With upcoming kernel changes[1], the Linux kernel will reject QEMU's
request to adjust the guest's clock in this case, so don't attempt to set
the clock.
This avoids the following warning message on save/restore of a PV guest:
warning: Unable to set KVM guest TOD clock: Operation not supported
[1] https://lore.kernel.org/all/20221011160712.928239-2-nrb@linux.ibm.com/
Fixes: c3347ed0d2 ("s390x: protvirt: Support unpack facility")
Signed-off-by: Nico Boehr <nrb@linux.ibm.com>
Message-Id: <20221012123229.1196007-1-nrb@linux.ibm.com>
[thuth: Add curly braces]
Signed-off-by: Thomas Huth <thuth@redhat.com>
I haven't really been working on s390x for some time now, and in
practice, I don't have time for it, either. So let's remove myself
from this entry.
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Message-Id: <20221010160957.40779-1-cohuck@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Remove spurious semicolon at the end of the macro s390_pv_cmd
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Acked-by: Cornelia Huck <cohuck@redhat.com>
Message-Id: <20221010151041.89071-1-imbrenda@linux.ibm.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
When you try to set qtest property "log" while the qtest object is
active, the error message blames "insufficient permission":
$ qemu-system-x86_64 -S -display none -nodefaults -monitor stdio -chardev socket,id=chrqt0,path=qtest.socket,server=on,wait=off -object qtest,id=qt0,chardev=chrqt0,log=/dev/null
QEMU 7.1.50 monitor - type 'help' for more information
(qemu) qom-set /objects/qt0 log qtest.log
Error: Insufficient permission to perform this operation
This implies it could work with "sufficient permission". It can't.
Change the error message to:
Error: Property 'log' can not be set now
Same for property "chardev".
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20221012153801.2604340-4-armbru@redhat.com>
When you try to set virtio-rng property "filename" after the backend
has been completed with user_creatable_complete(), the error message
blames "insufficient permission":
$ qemu-system-x86_64 -S -display none -nodefaults -monitor stdio -object rng-random,id=rng0 -device virtio-rng,id=vrng0,rng=rng0
QEMU 7.1.50 monitor - type 'help' for more information
(qemu) qom-set /objects/rng0 filename /dev/random
Error: Insufficient permission to perform this operation
This implies it could work with "sufficient permission". It can't.
Change the error message to:
Error: Property 'filename' can no longer be set
Same for cryptodev-vhost-user property "chardev", rng-egd property
"chardev", and vhost-user-backend property "chardev".
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20221012153801.2604340-3-armbru@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
[Commit message tidied up]
When you try to set a property that has no setter, the error message
blames "insufficient permission":
$ qemu-system-x86_64 -S -display none -nodefaults -monitor stdio
QEMU 7.1.50 monitor - type 'help' for more information
(qemu) qom-set /machine type q35
Error: Insufficient permission to perform this operation
This implies it could work with "sufficient permission". It can't.
Change the error message to:
Error: Property 'pc-i440fx-7.2-machine.type' is not writable
Do the same for getting a property that has no getter.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20221012153801.2604340-2-armbru@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Some versions of FreeBSD now require sys/param.h for machine/pmap.h on
x86. Include them here to meet that requirement. It does no harm on
older versions, so there's no need to #ifdef it.
Signed-off-by: Muhammad Moinur Rahman <bofh@FreeBSD.org>
Reviewed-by: John Baldwin <jhb@FreeBSD.org>
Signed-off-by: Warner Losh <imp@bsdimp.com>
Register guest RAM using BlockRAMRegistrar and set the
BDRV_REQ_REGISTERED_BUF flag so block drivers can optimize memory
accesses in I/O requests.
This is for vdpa-blk, vhost-user-blk, and other I/O interfaces that rely
on DMA mapping/unmapping.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20221013185908.1297568-14-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Avoid bounce buffers when QEMUIOVector elements are within previously
registered bdrv_register_buf() buffers.
The idea is that emulated storage controllers will register guest RAM
using bdrv_register_buf() and set the BDRV_REQ_REGISTERED_BUF on I/O
requests. Therefore no blkio_map_mem_region() calls are necessary in the
performance-critical I/O code path.
This optimization doesn't apply if the I/O buffer is internally
allocated by QEMU (e.g. qcow2 metadata). There we still take the slow
path because BDRV_REQ_REGISTERED_BUF is not set.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20221013185908.1297568-13-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
The blkio block driver will need to look up the file descriptor for a
given pointer. This is possible in softmmu builds where the RAMBlock API
is available for querying guest RAM.
Add stubs so tools like qemu-img that link the block layer still build
successfully. In this case there is no guest RAM but that is fine.
Bounce buffers and their file descriptors will be allocated with
libblkio's blkio_alloc_mem_region() so we won't rely on QEMU's
qemu_ram_get_fd() in that case.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20221013185908.1297568-12-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Add a function to get the file descriptor for a RAMBlock. Device
emulation code typically uses the MemoryRegion APIs but vhost-style code
may use RAMBlock directly for sharing guest memory with another process.
This new API will be used by the libblkio block driver so it can share
guest memory via .bdrv_register_buf().
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20221013185908.1297568-11-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Emulated devices and other BlockBackend users wishing to take advantage
of blk_register_buf() all have the same repetitive job: register
RAMBlocks with the BlockBackend using RAMBlockNotifier.
Add a BlockRAMRegistrar API to do this. A later commit will use this
from hw/block/virtio-blk.c.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20221013185908.1297568-10-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Make list traversal work when a callback removes a notifier
mid-traversal. This is a cleanup to prevent bugs in the future.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Message-id: 20221013185908.1297568-9-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Registering an I/O buffer is only a performance optimization hint but it
is still necessary to return errors when it fails.
Later patches will need to detect errors when registering buffers but an
immediate advantage is that error_report() calls are no longer needed in
block driver .bdrv_register_buf() functions.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20221013185908.1297568-8-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Block drivers may optimize I/O requests accessing buffers previously
registered with bdrv_register_buf(). Checking whether all elements of a
request's QEMUIOVector are within previously registered buffers is
expensive, so we need a hint from the user to avoid costly checks.
Add a BDRV_REQ_REGISTERED_BUF request flag to indicate that all
QEMUIOVector elements in an I/O request are known to be within
previously registered buffers.
Always pass the flag through to driver read/write functions. There is
little harm in passing the flag to a driver that does not use it.
Passing the flag to drivers avoids changes across many block drivers.
Filter drivers would need to explicitly support the flag and pass
through to their children when the children support it. That's a lot of
code changes and it's hard to remember to do that everywhere, leading to
silent reduced performance when the flag is accidentally dropped.
The only problematic scenario with the approach in this patch is when a
driver passes the flag through to internal I/O requests that don't use
the same I/O buffer. In that case the hint may be set when it should
actually be clear. This is a rare case though so the risk is low.
Some drivers have assert(!flags), which no longer works when
BDRV_REQ_REGISTERED_BUF is passed in. These assertions aren't very
useful anyway since the functions are called almost exclusively by
bdrv_driver_preadv/pwritev() so if we get flags handling right there
then the assertion is not needed.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20221013185908.1297568-7-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Use the enum type so GDB displays the enum members instead of printing a
numeric constant.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20221013185908.1297568-6-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
The only implementor of bdrv_register_buf() is block/nvme.c, where the
size is not needed when unregistering a buffer. This is because
util/vfio-helpers.c can look up mappings by address.
Future block drivers that implement bdrv_register_buf() may not be able
to do their job given only the buffer address. Add a size argument to
bdrv_unregister_buf().
Also document the assumptions about
bdrv_register_buf()/bdrv_unregister_buf() calls. The same <host, size>
values that were given to bdrv_register_buf() must be given to
bdrv_unregister_buf().
gcc 11.2.1 emits a spurious warning that img_bench()'s buf_size local
variable might be uninitialized, so it's necessary to silence the
compiler.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20221013185908.1297568-5-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
When a RAMBlockNotifier is added, ->ram_block_added() is called with all
existing RAMBlocks. There is no equivalent ->ram_block_removed() call
when a RAMBlockNotifier is removed.
The util/vfio-helpers.c code (the sole user of RAMBlockNotifier) is fine
with this asymmetry because it does not rely on RAMBlockNotifier for
cleanup. It walks its internal list of DMA mappings and unmaps them by
itself.
Future users of RAMBlockNotifier may not have an internal data structure
that records added RAMBlocks so they will need ->ram_block_removed()
callbacks.
This patch makes ram_block_notifier_remove() symmetric with respect to
callbacks. Now util/vfio-helpers.c needs to unmap remaining DMA mappings
after ram_block_notifier_remove() has been called. This is necessary
since users like block/nvme.c may create additional DMA mappings that do
not originate from the RAMBlockNotifier.
Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20221013185908.1297568-4-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
libblkio (https://gitlab.com/libblkio/libblkio/) is a library for
high-performance disk I/O. It currently supports io_uring,
virtio-blk-vhost-user, and virtio-blk-vhost-vdpa with additional drivers
under development.
One of the reasons for developing libblkio is that other applications
besides QEMU can use it. This will be particularly useful for
virtio-blk-vhost-user which applications may wish to use for connecting
to qemu-storage-daemon.
libblkio also gives us an opportunity to develop in Rust behind a C API
that is easy to consume from QEMU.
This commit adds io_uring, nvme-io_uring, virtio-blk-vhost-user, and
virtio-blk-vhost-vdpa BlockDrivers to QEMU using libblkio. It will be
easy to add other libblkio drivers since they will share the majority of
code.
For now I/O buffers are copied through bounce buffers if the libblkio
driver requires it. Later commits add an optimization for
pre-registering guest RAM to avoid bounce buffers.
The syntax is:
--blockdev io_uring,node-name=drive0,filename=test.img,readonly=on|off,cache.direct=on|off
--blockdev nvme-io_uring,node-name=drive0,filename=/dev/ng0n1,readonly=on|off,cache.direct=on
--blockdev virtio-blk-vhost-vdpa,node-name=drive0,path=/dev/vdpa...,readonly=on|off,cache.direct=on
--blockdev virtio-blk-vhost-user,node-name=drive0,path=vhost-user-blk.sock,readonly=on|off,cache.direct=on
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20221013185908.1297568-3-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
When a coroutine wakes up it may determine that it must re-queue.
Normally coroutines are pushed onto the back of the CoQueue, but for
fairness it may be necessary to push it onto the front of the CoQueue.
Add a flag to specify that the coroutine should be pushed onto the front
of the CoQueue. A later patch will use this to ensure fairness in the
bounce buffer CoQueue used by the blkio BlockDriver.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20221013185908.1297568-2-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
It's useful to know which device was used if/when it fails.
channel-win32.c had this since 2015, with
c69403fcd4 ("qemu-ga: debug printouts to
help troubleshoot installation"), this brings channel-posix.c up to
speed.
Signed-off-by: Bjørn Forsman <bjorn.forsman@gmail.com>
Reviewed-by: Konstantin Kostiuk <kkostiuk@redhat.com>
Signed-off-by: Konstantin Kostiuk <kkostiuk@redhat.com>
Richard Henderson