Rearrange the bitmap initialization and the first sync. Since at it,
make sure the locks are taken/released in correct order (I moved RCU
unlock upper - though it may not affect much).
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Let's further simplify ram_init_all() and ram_save_cleanup() by abstract
all the XBZRLE related codes into their own functions.
When allocating xbzrle cache, we are always very careful on -ENOMEM;
which makes sense. Replacing the last g_malloc0() with g_try_malloc0(),
then refactor the logic a bit.
This patch should be fixing some memory leaks when some memory
allocation failed for XBZRLE in the past.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
There are two Mutexes that are created but not yet destroyed for
RAMState. Fix that.
Since we are at it, provide helper function to clean up RAMState.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
The old ram_state_init() is not really initializing the RAMState only,
but including lots of other stuff that is RAM-related. Renaming it to
ram_init_all(). Instead, provide a real ram_state_init().
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
This patch let address_space_get_iotlb_entry() to use the newly
introduced page_mask parameter in flatview_do_translate(). Then we
will be sure the IOTLB can be aligned to page mask, also we should
nicely support huge pages now when introducing a764040.
Fixes: a764040 ("exec: abstract address_space_do_translate()")
Signed-off-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Message-Id: <20171010094247.10173-3-maxime.coquelin@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The function is originally used for flatview_space_translate() and what
we care about most is (xlat, plen) range. However for iotlb requests, we
don't really care about "plen", but the size of the page that "xlat" is
located on. While, plen cannot really contain this information.
A simple example to show why "plen" is not good for IOTLB translations:
E.g., for huge pages, it is possible that guest mapped 1G huge page on
device side that used this GPA range:
0x100000000 - 0x13fffffff
Then let's say we want to translate one IOVA that finally mapped to GPA
0x13ffffe00 (which is located on this 1G huge page). Then here we'll
get:
(xlat, plen) = (0x13fffe00, 0x200)
So the IOTLB would be only covering a very small range since from
"plen" (which is 0x200 bytes) we cannot tell the size of the page.
Actually we can really know that this is a huge page - we just throw the
information away in flatview_do_translate().
This patch introduced "page_mask" optional parameter to capture that
page mask info. Also, I made "plen" an optional parameter as well, with
some comments for the whole function.
No functional change yet.
Signed-off-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Message-Id: <20171010094247.10173-2-maxime.coquelin@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
When gcontext is used with iothread, the context will be destroyed
during iothread_stop(). That's not good since sometimes we would like
to keep the resources until iothread is destroyed, but we may want to
stop the thread before that point.
Delay the destruction of gcontext to iothread finalize. Then we can do:
iothread_stop(thread);
some_cleanup_on_resources();
iothread_destroy(thread);
We may need this patch if we want to run chardev IOs in iothreads and
hopefully clean them up correctly. For more specific information,
please see 2b316774f6 ("qemu-char: do not operate on sources from
finalize callbacks").
Reviewed-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-id: 20170928025958.1420-5-peterx@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
So that internal iothread users can explicitly stop one iothread without
destroying it.
Since at it, fix iothread_stop() to allow it to be called multiple
times. Before this patch we may call iothread_stop() more than once on
single iothread, while that may not be correct since qemu_thread_join()
is not allowed to run twice. From manual of pthread_join():
Joining with a thread that has previously been joined results in
undefined behavior.
Reviewed-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-id: 20170928025958.1420-4-peterx@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
IOThread is a general framework that contains IO loop environment and a
real thread behind. It's also good to be used internally inside qemu.
Provide some helpers for it to create iothreads to be used internally.
Put all the internal used iothreads into the internal object container.
Reviewed-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-id: 20170928025958.1420-3-peterx@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
We had a per-chardev cache for context, then we don't need this
parameter to be passed in every time when chr_update_read_handler()
called. As long as we are calling chr_update_read_handler() using
qemu_chr_be_update_read_handlers() we'll be fine.
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1505975754-21555-5-git-send-email-peterx@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
It was only passed in by chr_update_read_handlers(). However when
reconnect, we'll lose that context information. So if a chardev was
running on another context (rather than the default context, the NULL
pointer), it'll switch back to the default context if reconnection
happens. But, it should really stick to the old context.
Convert all the callers of io_add_watch_poll() to use the internally
cached gcontext. Then the context should be able to survive even after
reconnections.
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1505975754-21555-4-git-send-email-peterx@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
It caches the gcontext that is used to poll the chardev IO. Before this
patch, we only passed it in via chr_update_read_handlers(). However
that may not be enough if the char backend is disconnected and
reconnected afterward. There are chardev codes that still assumed the
context be NULL (which is the main context). Will fix that up in
following up patches.
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1505975754-21555-3-git-send-email-peterx@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Provide helpers to convert bitmaps to little endian format. It can be
used when we want to send one bitmap via network to some other hosts.
One thing to mention is that, these helpers only solve the problem of
endianess, but it does not solve the problem of different word size on
machines (the bitmaps managing same count of bits may contains different
size when malloced). So we need to take care of the size alignment issue
on the callers for now.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
In vtd_switch_address_space() we did the memory region switch, however
it's possible that the caller of it has not taken the BQL at all. Make
sure we have it.
CC: Paolo Bonzini <pbonzini@redhat.com>
CC: Jason Wang <jasowang@redhat.com>
CC: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
When accept failed, we should setup errp with the reason. More
importantly, the caller may assume errp be non-NULL when error happens,
and not setting the errp may crash QEMU.
At the same time, move the trace_qio_channel_socket_accept_fail() after
the if check on EINTR. Two reasons:
1. when EINTR happened, it's not really a fault (we should just try
again), so we should not log with an "accept failure".
2. trace_*() functions may overwrite errno, then the old errno will be
missing. We need to either check errno before trace_*() calls, or
reserve the errno.
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1501666880-10159-3-git-send-email-peterx@redhat.com>
Reviewed-by: Daniel P. Berrange <berrange@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
It was cached by read/write separately. Let's merge them.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
IOMMUTLBEntry.iova is returned incorrectly on one PT path (though mostly
we cannot really trigger this path, even if we do, we are mostly
disgarding this value, so it didn't break anything). Fix it by
converting the VTD_PAGE_MASK into the correct definition
VTD_PAGE_MASK_4K, then remove VTD_PAGE_MASK.
Fixes: b93130 ("intel_iommu: cleanup vtd_{do_}iommu_translate()")
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Abstract helper function to check migration capabilities (from the old
qmp_migrate_set_capabilities). Prepare to be used somewhere else.
There is side effect on the change: when applying the capabilities, we
were skipping the invalid ones, but still applying the valid ones (if
they are provided in the same QMP request). After this refactoring,
we'll ignore all the capabilities if we detected invalid setup along the
way. However, I don't think it is a problem since general users should
not provide anything invalid after all.
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1500349150-13240-9-git-send-email-peterx@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Export migration parameters to qdev properties. Then we can use, for
example:
-global migration.x-cpu-throttle-initial=xxx
To specify migration parameters during init.
Prefix "x-" is appended for each parameter exported to show that this is
not a stable interface, and only for debugging/testing purpose.
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1500349150-13240-3-git-send-email-peterx@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Currently drive_init_func() may call migrate_get_current() while the
migrate object is still not ready yet at that time. Move the migration
object init earlier, along with the global properties, right after
acceleration init.
This fixes a breakage for iotest 055, which caused an assertion failure.
Reported-by: Max Reitz <mreitz@redhat.com>
Reported-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>
Tested-by: QingFeng Hao <haoqf@linux.vnet.ibm.com>
Fixes: 3df663 ("migration: move only_migratable to MigrationState")
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1499242883-2184-3-git-send-email-peterx@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Migration is broken after the vfio integration work:
qemu-kvm: AHCI: Failed to start FIS receive engine: bad FIS receive buffer address
qemu-kvm: Failed to load ich9_ahci:ahci
qemu-kvm: error while loading state for instance 0x0 of device '0000:00:1f.2/ich9_ahci'
qemu-kvm: load of migration failed: Operation not permitted
The problem is that vfio work introduced dynamic memory region
switching (actually it is also used for future PT mode), and this memory
region layout is not properly delivered to destination when migration
happens. Solution is to rebuild the layout in post_load.
Bug: https://bugzilla.redhat.com/show_bug.cgi?id=1459906
Fixes: 558e0024 ("intel_iommu: allow dynamic switch of IOMMU region")
Reviewed-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
When this capability is enabled, QEMU will use the return path even for
precopy migration. This is helpful at least in one case when destination
failed to load the image while source quited without confirmation. With
return path, source will wait for the last response from destination,
and if destination fails, it'll fail the migration on source, then the
guest can be run again on the source (rather than assuming to be good,
then the guest will be lost after source quits).
It needs to be enabled explicitly on source, otherwise disabled.
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1498472935-14461-1-git-send-email-peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Now we have some globals that can be configured for migration. Dump them
in HMP info migration for better debugging.
(we can also use this to monitor whether COMPAT fields are applied
correctly on compatible machines)
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1498536619-14548-11-git-send-email-peterx@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
These two parameters:
- MachineState::enforce_config_section
- MigrationState::send_configuration
are playing similar role here. This patch merges the first one into
second, then we'll have a single place to reference whether we need to
send the configuration section.
I didn't remove the MachineState.enforce_config_section field since when
applying that machine property (in machine_set_property()) we haven't
yet initialized global properties and migration object. Then, it's
still not easy to pass that boolean to MigrationState at such an early
time.
A natural benefit for current patch is that now we kept the meaning of
"enforce-config-section" since it'll still have the highest
priority (that's what "enforce" mean I guess).
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1498536619-14548-10-git-send-email-peterx@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
It was in SaveState but now moved to MigrationState altogether, reverted
its meaning, then renamed to "send_configuration". Again, using
HW_COMPAT_2_3 for old PC/SPAPR machines, and accel_register_prop() for
xen_init().
Removing savevm_skip_configuration().
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1498536619-14548-8-git-send-email-peterx@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
One less global variable, and it does only matter with migration.
We keep the old "--only-migratable" option, but also now we support:
-global migration.only-migratable=true
Currently still keep the old interface.
Hmm, now vl.c has no way to access migrate_get_current(). Export a
function for it to setup only_migratable.
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1498536619-14548-7-git-send-email-peterx@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Put it into MigrationState then we can use the properties to specify
whether to enable storing global state.
Removing global_state_set_optional() since now we can use HW_COMPAT_2_3
for x86/power, and AccelClass.global_props for Xen.
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1498536619-14548-6-git-send-email-peterx@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Let the old man "MigrationState" join the object family. Direct benefit
is that we can start to use all the property features derived from
current QDev, like: HW_COMPAT_* bits, command line setup for migration
parameters (so will never need to set them up each time using HMP/QMP,
this is really, really attractive for test writters), etc.
I see no reason to disallow this happen yet. So let's start from this
one, to see whether it would be anything good.
Now we init the MigrationState struct statically in main() to make sure
it's initialized after global properties are applied, since we'll use
them during creation of the object.
No functional change at all.
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1498536619-14548-5-git-send-email-peterx@redhat.com>
Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
It's not that clear on how the global properties are registered to
global_props (and also its priority relationship). Let's provide a
single function to be called in main() for that, with comment to explain
it a bit.
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1498536619-14548-4-git-send-email-peterx@redhat.com>
Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Introduce this new field for the accelerator classes so that each
specific accelerator in the future can register its own global
properties to be used further by the system. It works just like how the
old machine compatible properties do, but only tailored for
accelerators.
Introduce register_compat_props_array() for it. Export it so that it may
be used in other codes as well in the future.
Suggested-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1498536619-14548-3-git-send-email-peterx@redhat.com>
Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
We have HW_COMPAT_*, however that's only bound to machines, not other
things (like accelerators). Behind it, it was register_compat_prop()
that played the trick. Let's export the function for further use
outside HW_COMPAT_* magic.
Meanwhile, move it to qdev-properties.c where seems more proper (since
it'll be used not only in machine codes).
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1498536619-14548-2-git-send-email-peterx@redhat.com>
Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Move the memcpy upper into where needed, then share the trace so that we
trace every correct remapping.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
First, let vtd_do_iommu_translate() return a status, so that we
explicitly knows whether error occured. Meanwhile, we make sure that
IOMMUTLBEntry is filled in in that.
Then, cleanup vtd_iommu_translate a bit. So even with PT we'll get a log
now. Also, remove useless assignments.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
We have converted many of the DPRINTF() into traces. This patch does the
last 100+ ones.
To debug VT-d when error happens, let's try enable:
-trace enable="vtd_err*"
This should works just like the old GENERAL but of course better, since
we don't need to recompile.
Similar rules apply to the other modules. I was trying to make the
prefix good enough for sub-module debugging.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
0425dc9 is actually v1 of that patch, but it was accidentally
merged (while there was a v2). That will cause problem when we try to
migrate to some old QEMUs when return path is not really there. Let's
fix it, then squashing this patch with 0425dc9 will be exactly patch
content of v2.
Fixes: 0425dc9 ("migration: isolate return path on src")
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
There are some places that binded "return path" with postcopy. Let's be
prepared for its usage even without postcopy. This patch mainly did this
on source side.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
It's possible that one device kept its irqfd/virq there even when
MSI/MSIX was disabled globally for that device. One example is
virtio-net-pci (see commit f1d0f15a6 and virtio_pci_vq_vector_mask()).
It is used as a fast path to avoid allocate/release irqfd/virq
frequently when guest enables/disables MSIX.
However, this fast path brought a problem to msi_route_list, that the
device MSIRouteEntry is still dangling there even if MSIX disabled -
then we cannot know which message to fetch, even if we can, the messages
are meaningless. In this case, we can just simply ignore this entry.
It's safe, since when MSIX is enabled again, we'll rebuild them no
matter what.
Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1448813
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1494309644-18743-4-git-send-email-peterx@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
It'll be nice to know which virq belongs to which device/vector when
adding msi routes, so adding two more parameters for the add trace.
Meanwhile, releasing virq has no tracing before. Add one for it.
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1494309644-18743-2-git-send-email-peterx@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
We were do the shutting off only for postcopy. Now we do this as long as
the source return path is there.
Moving the cleanup of from_src_file there too.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Hardware support for VT-d device passthrough. Although current Linux can
live with iommu=pt even without this, but this is faster than when using
software passthrough.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Liu, Yi L <yi.l.liu@linux.intel.com>
Reviewed-by: Jason Wang <jasowang@redhat.com>
When device-iotlb is not specified, we should fail this check. A new
function vtd_ce_type_check() is introduced.
While I'm at it, clean up the vtd_dev_to_context_entry() a bit - replace
many "else if" usage into direct if check. That'll make the logic more
clear.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Jason Wang <jasowang@redhat.com>
The old names are too long and less ordered. Let's start to use
vtd_ce_*() as a pattern.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Jason Wang <jasowang@redhat.com>
No reason to keep tens of lines if we can do it actually far shorter.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Jason Wang <jasowang@redhat.com>
We were always passing in that one as "false" to assume that's an read
operation, and we also assume that IOMMU translation would always have
that read permission. A better permission would be IOMMU_NONE since the
replay is after all not a real read operation, but just a page table
rebuilding process.
CC: David Gibson <david@gibson.dropbear.id.au>
CC: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Jason Wang <jasowang@redhat.com>
This patch converts the old "is_write" bool into IOMMUAccessFlags. The
difference is that "is_write" can only express either read/write, but
sometimes what we really want is "none" here (neither read nor write).
Replay is an good example - during replay, we should not check any RW
permission bits since thats not an actual IO at all.
CC: Paolo Bonzini <pbonzini@redhat.com>
CC: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Jason Wang <jasowang@redhat.com>
This function is an abstraction helper for address_space_translate() and
address_space_get_iotlb_entry(). It does the lookup of address into
memory region section, then does proper IOMMU translation if necessary.
Refactor the two existing functions to use it.
This fixes vhost when IOMMU is disabled by guest.
Tested-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
To dump information about ramblocks. It looks like:
(qemu) info ramblock
Block Name PSize Offset Used Total
/objects/mem 2 MiB 0x0000000000000000 0x0000000080000000 0x0000000080000000
vga.vram 4 KiB 0x0000000080060000 0x0000000001000000 0x0000000001000000
/rom@etc/acpi/tables 4 KiB 0x00000000810b0000 0x0000000000020000 0x0000000000200000
pc.bios 4 KiB 0x0000000080000000 0x0000000000040000 0x0000000000040000
0000:00:03.0/e1000.rom 4 KiB 0x0000000081070000 0x0000000000040000 0x0000000000040000
pc.rom 4 KiB 0x0000000080040000 0x0000000000020000 0x0000000000020000
0000:00:02.0/vga.rom 4 KiB 0x0000000081060000 0x0000000000010000 0x0000000000010000
/rom@etc/table-loader 4 KiB 0x00000000812b0000 0x0000000000001000 0x0000000000001000
/rom@etc/acpi/rsdp 4 KiB 0x00000000812b1000 0x0000000000001000 0x0000000000001000
Ramblock is something hidden internally in QEMU implementation, and this
command should only be used by mostly QEMU developers on RAM stuff. It
is not a command suitable for QMP interface. So only HMP interface is
provided for it.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1494562661-9063-4-git-send-email-peterx@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Moving the algorithm from print_type_size() into size_to_str() so that
other component can also leverage it. With that, refactor
print_type_size().
The assert() in that logic is removed though, since even UINT64_MAX
would not overflow.
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1494562661-9063-3-git-send-email-peterx@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Dump the info in a single line is hard to read. Do it one per line.
Also, the first "capabilities:" didn't help much. Let's remove it.
CC: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
In migration codes (especially in migration_thread()), max_size is used
in many place for the threshold value that we will start to do the final
flush and jump to the next stage to dump the whole rest things to
destination. However its name is confusing to first readers. Let's
rename it to "threshold_size" when proper and add a comment for it. No
functional change is made.
CC: Juan Quintela <quintela@redhat.com>
CC: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
This patch is based on Aviv Ben-David (<bd.aviv@gmail.com>)'s patch
upstream:
"IOMMU: enable intel_iommu map and unmap notifiers"
https://lists.gnu.org/archive/html/qemu-devel/2016-11/msg01453.html
However I removed/fixed some content, and added my own codes.
Instead of translate() every page for iotlb invalidations (which is
slower), we walk the pages when needed and notify in a hook function.
This patch enables vfio devices for VT-d emulation.
And, since we already have vhost DMAR support via device-iotlb, a
natural benefit that this patch brings is that vt-d enabled vhost can
live even without ATS capability now. Though more tests are needed.
Signed-off-by: Aviv Ben-David <bdaviv@cs.technion.ac.il>
Reviewed-by: Jason Wang <jasowang@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: \"Michael S. Tsirkin\" <mst@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1491562755-23867-10-git-send-email-peterx@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
This is preparation work to finally enabled dynamic switching ON/OFF for
VT-d protection. The old VT-d codes is using static IOMMU address space,
and that won't satisfy vfio-pci device listeners.
Let me explain.
vfio-pci devices depend on the memory region listener and IOMMU replay
mechanism to make sure the device mapping is coherent with the guest
even if there are domain switches. And there are two kinds of domain
switches:
(1) switch from domain A -> B
(2) switch from domain A -> no domain (e.g., turn DMAR off)
Case (1) is handled by the context entry invalidation handling by the
VT-d replay logic. What the replay function should do here is to replay
the existing page mappings in domain B.
However for case (2), we don't want to replay any domain mappings - we
just need the default GPA->HPA mappings (the address_space_memory
mapping). And this patch helps on case (2) to build up the mapping
automatically by leveraging the vfio-pci memory listeners.
Another important thing that this patch does is to seperate
IR (Interrupt Remapping) from DMAR (DMA Remapping). IR region should not
depend on the DMAR region (like before this patch). It should be a
standalone region, and it should be able to be activated without
DMAR (which is a common behavior of Linux kernel - by default it enables
IR while disabled DMAR).
Reviewed-by: Jason Wang <jasowang@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: \"Michael S. Tsirkin\" <mst@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1491562755-23867-9-git-send-email-peterx@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
The default replay() don't work for VT-d since vt-d will have a huge
default memory region which covers address range 0-(2^64-1). This will
normally consumes a lot of time (which looks like a dead loop).
The solution is simple - we don't walk over all the regions. Instead, we
jump over the regions when we found that the page directories are empty.
It'll greatly reduce the time to walk the whole region.
To achieve this, we provided a page walk helper to do that, invoking
corresponding hook function when we found an page we are interested in.
vtd_page_walk_level() is the core logic for the page walking. It's
interface is designed to suite further use case, e.g., to invalidate a
range of addresses.
Reviewed-by: Jason Wang <jasowang@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: \"Michael S. Tsirkin\" <mst@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1491562755-23867-8-git-send-email-peterx@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Originally we have one memory_region_iommu_replay() function, which is
the default behavior to replay the translations of the whole IOMMU
region. However, on some platform like x86, we may want our own replay
logic for IOMMU regions. This patch adds one more hook for IOMMUOps for
the callback, and it'll override the default if set.
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: \"Michael S. Tsirkin\" <mst@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1491562755-23867-6-git-send-email-peterx@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
In this patch, IOMMUNotifier.{start|end} are introduced to store section
information for a specific notifier. When notification occurs, we not
only check the notification type (MAP|UNMAP), but also check whether the
notified iova range overlaps with the range of specific IOMMU notifier,
and skip those notifiers if not in the listened range.
When removing an region, we need to make sure we removed the correct
VFIOGuestIOMMU by checking the IOMMUNotifier.start address as well.
This patch is solving the problem that vfio-pci devices receive
duplicated UNMAP notification on x86 platform when vIOMMU is there. The
issue is that x86 IOMMU has a (0, 2^64-1) IOMMU region, which is
splitted by the (0xfee00000, 0xfeefffff) IRQ region. AFAIK
this (splitted IOMMU region) is only happening on x86.
This patch also helps vhost to leverage the new interface as well, so
that vhost won't get duplicated cache flushes. In that sense, it's an
slight performance improvement.
Suggested-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <1491562755-23867-2-git-send-email-peterx@redhat.com>
[ehabkost: included extra vhost_iommu_region_del() change from Peter Xu]
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
VT-d codes are still using static DEBUG_INTEL_IOMMU macro. That's not
good, and we should end the day when we need to recompile the code
before getting useful debugging information for vt-d. Time to switch to
the trace system. This is the first patch to do it.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Jason Wang <jasowang@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
There are lots of places in current intel_iommu.c codes that named
"iova" as "gpa". It is really confusing to use a name "gpa" in these
places (which is very easily to be understood as "Guest Physical
Address", while it's not). To make the codes (much) easier to be read, I
decided to do this once and for all.
No functional change is made. Only literal ones.
Reviewed-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Now we have a standalone memory region for MSI, all the irq region
requests should be redirected there. Cleaning up the block with an
assertion instead.
Reviewed-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Linux vfio driver supports to do VFIO_IOMMU_UNMAP_DMA for a very big
region. This can be leveraged by QEMU IOMMU implementation to cleanup
existing page mappings for an entire iova address space (by notifying
with an IOTLB with extremely huge addr_mask). However current
vfio_iommu_map_notify() does not allow that. It make sure that all the
translated address in IOTLB is falling into RAM range.
The check makes sense, but it should only be a sensible checker for
mapping operations, and mean little for unmap operations.
This patch moves this check into map logic only, so that we'll get
faster unmap handling (no need to translate again), and also we can then
better support unmapping a very big region when it covers non-ram ranges
or even not-existing ranges.
Acked-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
A cleanup for vfio_iommu_map_notify(). Now we will fetch vaddr even if
the operation is unmap, but it won't hurt much.
One thing to mention is that we need the RCU read lock to protect the
whole translation and map/unmap procedure.
Acked-by: Alex Williamson <alex.williamson@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
When we add PCIe extended capabilities, we should be following the rule
that we add the head extended cap (at offset 0x100) first, then the rest
of them. Meanwhile, we are always adding new capability bits at the end
of the list. Here the "next" looks meaningless in all cases since it
should always be zero (along with the "header").
Simplify the function a bit, and it looks more readable now.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Peter Xu