Using MgrationStats as type for ram_counters mean that we didn't have
to re-declare each value in another struct. The need of atomic
counters have make us to create MigrationAtomicStats for this atomic
counters.
Create RAMStats type which is a merge of MigrationStats and
MigrationAtomicStats removing unused members.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
---
Fix typos found by David Edmondson
I removed that bit on commit:
commit c8df4a7aef
Author: Juan Quintela <quintela@redhat.com>
Date: Mon Oct 3 02:00:03 2022 +0200
migration: Split save_live_pending() into state_pending_*
Fixes: c8df4a7aef
Suggested-by: Nina Schoetterl-Glausch <nsg@linux.ibm.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Since ec6f3ab9, migration with compress enabled was broken, because
the compress threads use a dummy QEMUFile which just acts as a
buffer and that commit accidentally changed it to use the outgoing
migration channel instead.
Fix this by using the dummy file again in the compress threads.
Signed-off-by: Lukas Straub <lukasstraub2@web.de>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Once that res_compatible is removed, they don't make sense anymore.
We remove the _only preffix. And to make things clearer we rename
them to must_precopy and can_postcopy.
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Nothing assigns to it after previous commit.
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Signed-off-by: Juan Quintela <quintela@redhat.com>
So remove last assignation of res_compatible.
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Signed-off-by: Juan Quintela <quintela@redhat.com>
0x80 is RAM_SAVE_FLAG_HOOK, it is in qemu-file now.
Bigger usable flag is 0x200, noticing that.
We can reuse RAM_SAVe_FLAG_FULL.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
This commit is the same with [PATCH v6 1/2], and provides avx512 support for xbzrle_encode_buffer
function to accelerate xbzrle encoding speed. Runtime check of avx512
support and benchmark for this feature are added. Compared with C
version of xbzrle_encode_buffer function, avx512 version can achieve
50%-70% performance improvement on benchmarking. In addition, if dirty
data is randomly located in 4K page, the avx512 version can achieve
almost 140% performance gain.
Signed-off-by: ling xu <ling1.xu@intel.com>
Co-authored-by: Zhou Zhao <zhou.zhao@intel.com>
Co-authored-by: Jun Jin <jun.i.jin@intel.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
We are going to create a new function for multifd latest in the series.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
We are recalculating ram size continously, when we know that it don't
change during migration. Create a field in RAMState to track it.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
It is just a big if in the middle of the function, and we need two
functions anways.
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Juan Quintela <quintela@redhat.com>
---
Reindent to make Phillipe happy (and CODING_STYLE)
We used to return two bools, just return a single int with the
following meaning:
old return / again / new return
false false PAGE_ALL_CLEAN
false true PAGE_TRY_AGAIN
true true PAGE_DIRTY_FOUND /* We don't care about again at all */
Signed-off-by: Juan Quintela <quintela@redhat.com>
We will need later that find_dirty_block() return errors, so
simplify the loop.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Let's factor out this check, to be used in virtio-mem context next.
While at it, fix a spelling error in a related comment.
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>S
Signed-off-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
ram_block_populate_read() already optimizes for RamDiscardManager.
However, ram_write_tracking_start() will still try protecting discarded
memory ranges.
Let's optimize, because discarded ranges don't map any pages and
(1) For anonymous memory, trying to protect using uffd-wp without a mapped
page is ignored by the kernel and consequently a NOP.
(2) For shared/file-backed memory, we will fill present page tables in the
range with PTE markers. However, we will even allocate page tables
just to fill them with unnecessary PTE markers and effectively
waste memory.
So let's exclude these ranges, just like ram_block_populate_read()
already does.
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
ram_mig_ram_block_resized() will abort migration (including background
snapshots) when resizing a RAMBlock. ram_block_populate_read() will only
populate RAM up to used_length, so at least for anonymous memory
protecting everything between used_length and max_length won't
actually be protected and is just a NOP.
So let's only protect everything up to used_length.
Note: it still makes sense to register uffd-wp for max_length, such
that RAM_UF_WRITEPROTECT is independent of a changing used_length.
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
When unregistering uffd-wp, older kernels before commit f369b07c86143
("mm/uffd:reset write protection when unregister with wp-mode") won't
clear the uffd-wp PTE bit. When re-registering uffd-wp, the previous
uffd-wp PTE bits would trigger again. With above commit, the kernel will
clear the uffd-wp PTE bits when unregistering itself.
Consequently, we'll clear the uffd-wp PTE bits now twice -- whereby we
don't care about clearing them at all: a new background snapshot will
re-register uffd-wp and re-protect all memory either way.
So let's skip the manual clearing of uffd-wp. If ever relevant, we
could clear conditionally in uffd_unregister_memory() -- we just need a
way to figure out more recent kernels.
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
If something goes wrong during uffd_change_protection(), we would miss
to unregister uffd-wp and not release our reference. Fix it by
performing the uffd_change_protection(true) last.
Note that a uffd_change_protection(false) on the recovery path without a
prior uffd_change_protection(false) is fine.
Fixes: 278e2f551a ("migration: support UFFD write fault processing in ram_save_iterate()")
Cc: qemu-stable@nongnu.org
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Unfortunately, commit f7b9dcfbcf broke populate_read_range(): the loop
end condition is very wrong, resulting in that function not populating the
full range. Lets' fix that.
Fixes: f7b9dcfbcf ("migration/ram: Factor out populating pages readable in ram_block_populate_pages()")
Cc: qemu-stable@nongnu.org
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Until previous commit, save_live_pending() was used for ram. Now with
the split into state_pending_estimate() and state_pending_exact() it
is not needed anymore, so remove them.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
We split the function into to:
- state_pending_estimate: We estimate the remaining state size without
stopping the machine.
- state pending_exact: We calculate the exact amount of remaining
state.
The only "device" that implements different functions for _estimate()
and _exact() is ram.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Commit d9e474ea56 overlooked the case where the target psize is even larger
than the host psize. One example is Alpha has 8K page size and migration
will start to crash the source QEMU when running Alpha migration on x86.
Fix it by detecting that case and set host start/end just to cover the
single page to be migrated.
This will slightly optimize the common case where host psize equals to
guest psize so we don't even need to do the roundups, but that's trivial.
Cc: qemu-stable@nongnu.org
Reported-by: Thomas Huth <thuth@redhat.com>
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/1456
Fixes: d9e474ea56 ("migration: Teach PSS about host page")
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Now with rs->pss we can already cache channels in pss->pss_channels. That
pss_channel contains more infromation than rs->f because it's per-channel.
So rs->f could be replaced by rss->pss[RAM_CHANNEL_PRECOPY].pss_channel,
while rs->f itself is a bit vague now.
Note that vanilla postcopy still send pages via pss[RAM_CHANNEL_PRECOPY],
that's slightly confusing but it reflects the reality.
Then, after the replacement we can safely drop rs->f.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
With the new code to send pages in rp-return thread, there's little help to
keep lots of the old code on maintaining the preempt state in migration
thread, because the new way should always be faster..
Then if we'll always send pages in the rp-return thread anyway, we don't
need those logic to maintain preempt state anymore because now we serialize
things using the mutex directly instead of using those fields.
It's very unfortunate to have those code for a short period, but that's
still one intermediate step that we noticed the next bottleneck on the
migration thread. Now what we can do best is to drop unnecessary code as
long as the new code is stable to reduce the burden. It's actually a good
thing because the new "sending page in rp-return thread" model is (IMHO)
even cleaner and with better performance.
Remove the old code that was responsible for maintaining preempt states, at
the meantime also remove x-postcopy-preempt-break-huge parameter because
with concurrent sender threads we don't really need to break-huge anymore.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
With all the facilities ready, send the requested page directly in the
rp-return thread rather than queuing it in the request queue, if and only
if postcopy preempt is enabled. It can achieve so because it uses separate
channel for sending urgent pages. The only shared data is bitmap and it's
protected by the bitmap_mutex.
Note that since we're moving the ownership of the urgent channel from the
migration thread to rp thread it also means the rp thread is responsible
for managing the qemufile, e.g. properly close it when pausing migration
happens. For this, let migration_release_from_dst_file to cover shutdown
of the urgent channel too, renaming it as migration_release_dst_files() to
better show what it does.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Since we use PageSearchStatus to represent a channel, it makes perfect
sense to keep last_sent_block (aka, leverage RAM_SAVE_FLAG_CONTINUE) to be
per-channel rather than global because each channel can be sending
different pages on ramblocks.
Hence move it from RAMState into PageSearchStatus.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
We used to allocate PSS structure on the stack for precopy when sending
pages. Make it static, so as to describe per-channel ram migration status.
Here we declared RAM_CHANNEL_MAX instances, preparing for postcopy to use
it, even though this patch has not yet to start using the 2nd instance.
This should not have any functional change per se, but it already starts to
export PSS information via the RAMState, so that e.g. one PSS channel can
start to reference the other PSS channel.
Always protect PSS access using the same RAMState.bitmap_mutex. We already
do so, so no code change needed, just some comment update. Maybe we should
consider renaming bitmap_mutex some day as it's going to be a more commonly
and big mutex we use for ram states, but just leave it for later.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Helper to init PSS structures.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Introduce pss_channel for PageSearchStatus, define it as "the migration
channel to be used to transfer this host page".
We used to have rs->f, which is a mirror to MigrationState.to_dst_file.
After postcopy preempt initial version, rs->f can be dynamically changed
depending on which channel we want to use.
But that later work still doesn't grant full concurrency of sending pages
in e.g. different threads, because rs->f can either be the PRECOPY channel
or POSTCOPY channel. This needs to be per-thread too.
PageSearchStatus is actually a good piece of struct which we can leverage
if we want to have multiple threads sending pages. Sending a single guest
page may not make sense, so we make the granule to be "host page", and in
the PSS structure we allow specify a QEMUFile* to migrate a specific host
page. Then we open the possibility to specify different channels in
different threads with different PSS structures.
The PSS prefix can be slightly misleading here because e.g. for the
upcoming usage of postcopy channel/thread it's not "searching" (or,
scanning) at all but sending the explicit page that was requested. However
since PSS existed for some years keep it as-is until someone complains.
This patch mostly (simply) replace rs->f with pss->pss_channel only. No
functional change intended for this patch yet. But it does prepare to
finally drop rs->f, and make ram_save_guest_page() thread safe.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Migration code has a lot to do with host pages. Teaching PSS core about
the idea of host page helps a lot and makes the code clean. Meanwhile,
this prepares for the future changes that can leverage the new PSS helpers
that this patch introduces to send host page in another thread.
Three more fields are introduced for this:
(1) host_page_sending: this is set to true when QEMU is sending a host
page, false otherwise.
(2) host_page_{start|end}: these point to the start/end of host page
we're sending, and it's only valid when host_page_sending==true.
For example, when we look up the next dirty page on the ramblock, with
host_page_sending==true, we'll not try to look for anything beyond the
current host page boundary. This can be slightly efficient than current
code because currently we'll set pss->page to next dirty bit (which can be
over current host page boundary) and reset it to host page boundary if we
found it goes beyond that.
With above, we can easily make migration_bitmap_find_dirty() self contained
by updating pss->page properly. rs* parameter is removed because it's not
even used in old code.
When sending a host page, we should use the pss helpers like this:
- pss_host_page_prepare(pss): called before sending host page
- pss_within_range(pss): whether we're still working on the cur host page?
- pss_host_page_finish(pss): called after sending a host page
Then we can use ram_save_target_page() to save one small page.
Currently ram_save_host_page() is still the only user. If there'll be
another function to send host page (e.g. in return path thread) in the
future, it should follow the same style.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
To prepare for thread-safety on page accountings, at least below counters
need to be accessed only atomically, they are:
ram_counters.transferred
ram_counters.duplicate
ram_counters.normal
ram_counters.postcopy_bytes
There are a lot of other counters but they won't be accessed outside
migration thread, then they're still safe to be accessed without atomic
ops.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Don't take the bitmap mutex when sending pages, or when being throttled by
migration_rate_limit() (which is a bit tricky to call it here in ram code,
but seems still helpful).
It prepares for the possibility of concurrently sending pages in >1 threads
using the function ram_save_host_page() because all threads may need the
bitmap_mutex to operate on bitmaps, so that either sendmsg() or any kind of
qemu_sem_wait() blocking for one thread will not block the other from
progressing.
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>
Removing referencing to RAMState.f in compress_page_with_multi_thread() and
flush_compressed_data().
Compression code by default isn't compatible with having >1 channels (or it
won't currently know which channel to flush the compressed data), so to
make it simple we always flush on the default to_dst_file port until
someone wants to add >1 ports support, as rs->f right now can really
change (after postcopy preempt is introduced).
There should be no functional change at all after patch applied, since as
long as rs->f referenced in compression code, it must be to_dst_file.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
The 2nd check on RAM_SAVE_FLAG_CONTINUE is a bit redundant. Use a boolean
to be clearer.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
The major change is to replace "!save_page_use_compression()" with
"xbzrle_enabled" to make it clear.
Reasonings:
(1) When compression enabled, "!save_page_use_compression()" is exactly the
same as checking "xbzrle_enabled".
(2) When compression disabled, "!save_page_use_compression()" always return
true. We used to try calling the xbzrle code, but after this change we
won't, and we shouldn't need to.
Since at it, drop the xbzrle_enabled check in xbzrle_cache_zero_page()
because with this change it's not needed anymore.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Add the helper to show that postcopy preempt enabled, meanwhile active.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Any call to ram_find_and_save_block() needs to take the bitmap mutex. We
used to not take it for most of ram_save_complete() because we thought
we're the only one left using the bitmap, but it's not true after the
preempt full patchset applied, since the return path can be taking it too.
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>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: David Edmondson <david.edmondson@oracle.com>
Reviewed-by: Leonardo Bras <leobras@redhat.com>
Multifd thread model does not work for compression, explicitly disable it.
Note that previuosly even we can enable both of them, nothing will go
wrong, because the compression code has higher priority so multifd feature
will just be ignored. Now we'll fail even earlier at config time so the
user should be aware of the consequence better.
Note that there can be a slight chance of breaking existing users, but
let's assume they're not majority and not serious users, or they should
have found that multifd is not working already.
With that, we can safely drop the check in ram_save_target_page() for using
multifd, because when multifd=on then compression=off, then the removed
check on save_page_use_compression() will also always return false too.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
When starting ram saving procedure (especially at the completion phase),
always set last_seen_block to non-NULL to make sure we can always correctly
detect the case where "we've migrated all the dirty pages".
Then we'll guarantee both last_seen_block and pss.block will be valid
always before the loop starts.
See the comment in the code for some details.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
This reverts commit cfd66f30fb.
The simplification of unqueue_page() introduced a bug that sometimes
breaks migration on s390x hosts.
The problem is not fully understood yet, but since we are already in
the freeze for QEMU 7.1 and we need something working there, let's
revert this patch for the upcoming release. The optimization can be
redone later again in a proper way if necessary.
Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=2099934
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20220802061949.331576-1-thuth@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Some errors, like the lack of Scatter-Gather support by the network
interface(NETIF_F_SG) may cause sendmsg(...,MSG_ZEROCOPY) to fail on using
zero-copy, which causes it to fall back to the default copying mechanism.
After each full dirty-bitmap scan there should be a zero-copy flush
happening, which checks for errors each of the previous calls to
sendmsg(...,MSG_ZEROCOPY). If all of them failed to use zero-copy, then
increment dirty_sync_missed_zero_copy migration stat to let the user know
about it.
Signed-off-by: Leonardo Bras <leobras@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Acked-by: Peter Xu <peterx@redhat.com>
Message-Id: <20220711211112.18951-4-leobras@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
With preemption mode on, when we see a postcopy request that was requesting
for exactly the page that we have preempted before (so we've partially sent
the page already via PRECOPY channel and it got preempted by another
postcopy request), currently we drop the request so that after all the
other postcopy requests are serviced then we'll go back to precopy stream
and start to handle that.
We dropped the request because we can't send it via postcopy channel since
the precopy channel already contains partial of the data, and we can only
send a huge page via one channel as a whole. We can't split a huge page
into two channels.
That's a very corner case and that works, but there's a change on the order
of postcopy requests that we handle since we're postponing this (unlucky)
postcopy request to be later than the other queued postcopy requests. The
problem is there's a possibility that when the guest was very busy, the
postcopy queue can be always non-empty, it means this dropped request will
never be handled until the end of postcopy migration. So, there's a chance
that there's one dest QEMU vcpu thread waiting for a page fault for an
extremely long time just because it's unluckily accessing the specific page
that was preempted before.
The worst case time it needs can be as long as the whole postcopy migration
procedure. It's extremely unlikely to happen, but when it happens it's not
good.
The root cause of this problem is because we treat pss->postcopy_requested
variable as with two meanings bound together, as the variable shows:
1. Whether this page request is urgent, and,
2. Which channel we should use for this page request.
With the old code, when we set postcopy_requested it means either both (1)
and (2) are true, or both (1) and (2) are false. We can never have (1)
and (2) to have different values.
However it doesn't necessarily need to be like that. It's very legal that
there's one request that has (1) very high urgency, but (2) we'd like to
use the precopy channel. Just like the corner case we were discussing
above.
To differenciate the two meanings better, introduce a new field called
postcopy_target_channel, showing which channel we should use for this page
request, so as to cover the old meaning (2) only. Then we leave the
postcopy_requested variable to stand only for meaning (1), which is the
urgency of this page request.
With this change, we can easily boost priority of a preempted precopy page
as long as we know that page is also requested as a postcopy page. So with
the new approach in get_queued_page() instead of dropping that request, we
send it right away with the precopy channel so we get back the ordering of
the page faults just like how they're requested on dest.
Reported-by: Manish Mishra <manish.mishra@nutanix.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Manish Mishra <manish.mishra@nutanix.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20220707185520.27583-1-peterx@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Add a property field that can conditionally disable the "break sending huge
page" behavior in postcopy preemption. By default it's enabled.
It should only be used for debugging purposes, and we should never remove
the "x-" prefix.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Manish Mishra <manish.mishra@nutanix.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20220707185511.27366-1-peterx@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
This patch enables postcopy-preempt feature.
It contains two major changes to the migration logic:
(1) Postcopy requests are now sent via a different socket from precopy
background migration stream, so as to be isolated from very high page
request delays.
(2) For huge page enabled hosts: when there's postcopy requests, they can now
intercept a partial sending of huge host pages on src QEMU.
After this patch, we'll live migrate a VM with two channels for postcopy: (1)
PRECOPY channel, which is the default channel that transfers background pages;
and (2) POSTCOPY channel, which only transfers requested pages.
There's no strict rule of which channel to use, e.g., if a requested page is
already being transferred on precopy channel, then we will keep using the same
precopy channel to transfer the page even if it's explicitly requested. In 99%
of the cases we'll prioritize the channels so we send requested page via the
postcopy channel as long as possible.
On the source QEMU, when we found a postcopy request, we'll interrupt the
PRECOPY channel sending process and quickly switch to the POSTCOPY channel.
After we serviced all the high priority postcopy pages, we'll switch back to
PRECOPY channel so that we'll continue to send the interrupted huge page again.
There's no new thread introduced on src QEMU.
On the destination QEMU, one new thread is introduced to receive page data from
the postcopy specific socket (done in the preparation patch).
This patch has a side effect: after sending postcopy pages, previously we'll
assume the guest will access follow up pages so we'll keep sending from there.
Now it's changed. Instead of going on with a postcopy requested page, we'll go
back and continue sending the precopy huge page (which can be intercepted by a
postcopy request so the huge page can be sent partially before).
Whether that's a problem is debatable, because "assuming the guest will
continue to access the next page" may not really suite when huge pages are
used, especially if the huge page is large (e.g. 1GB pages). So that locality
hint is much meaningless if huge pages are used.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20220707185504.27203-1-peterx@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Create a new socket for postcopy to be prepared to send postcopy requested
pages via this specific channel, so as to not get blocked by precopy pages.
A new thread is also created on dest qemu to receive data from this new channel
based on the ram_load_postcopy() routine.
The ram_load_postcopy(POSTCOPY) branch and the thread has not started to
function, and that'll be done in follow up patches.
Cleanup the new sockets on both src/dst QEMUs, meanwhile look after the new
thread too to make sure it'll be recycled properly.
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20220707185502.27149-1-peterx@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
dgilbert: With Peter's fix to quieten compiler warning on
start_migration
Now that all QEMUFile callbacks are removed, the entire concept can be
deleted.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
The qemu_update_position method name gives the misleading impression
that it is changing the current file offset. Most of the files are
just streams, however, so there's no concept of a file offset in the
general case.
What this method is actually used for is to report on the number of
bytes that have been transferred out of band from the main I/O methods.
This new name better reflects this purpose.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
This removes one further custom impl of QEMUFile, in favour of a
QIOChannel based impl.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Even though multifd_send_sync_main() currently emits error_reports, it's
callers don't really check it before continuing.
Change multifd_send_sync_main() to return -1 on error and 0 on success.
Also change all it's callers to make use of this change and possibly fail
earlier.
(This change is important to next patch on multifd zero copy
implementation, to make it sure an error in zero-copy flush does not go
unnoticed.
Signed-off-by: Leonardo Bras <leobras@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Message-Id: <20220513062836.965425-7-leobras@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Clang spotted an & that should have been an &&; fix it.
Reported by: David Binderman / https://gitlab.com/dcb
Fixes: 65dacaa04f ("migration: introduce save_normal_page()")
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/963
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Message-Id: <20220406102515.96320-1-dgilbert@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Will be reused in postcopy fast load thread.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20220331150857.74406-6-peterx@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
This boolean flag shows whether the current page during migration is triggered
by postcopy or not. Then in ram_save_host_page() and deeper stack we'll be
able to have a reference on the priority of this page.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20220331150857.74406-4-peterx@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
g_new(T, n) is neater than g_malloc(sizeof(T) * n). It's also safer,
for two reasons. One, it catches multiplication overflowing size_t.
Two, it returns T * rather than void *, which lets the compiler catch
more type errors.
This commit only touches allocations with size arguments of the form
sizeof(T).
Patch created mechanically with:
$ spatch --in-place --sp-file scripts/coccinelle/use-g_new-etc.cocci \
--macro-file scripts/cocci-macro-file.h FILES...
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Acked-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Message-Id: <20220315144156.1595462-4-armbru@redhat.com>
Reviewed-by: Pavel Dovgalyuk <Pavel.Dovgalyuk@ispras.ru>
Static variable is very unfriendly to threading of ram_block_from_stream().
Move it into MigrationIncomingState.
Make the incoming state pointer to be passed over to ram_block_from_stream() on
both caller sites.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20220301083925.33483-8-peterx@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
In ram_load_postcopy() we'll try to detect non-same-page case and dump error.
This error is very helpful for debugging. Adding ramblock & offset into the
error log too.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20220301083925.33483-6-peterx@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
dgilbert: Fix up long line
Postcopy handles huge pages in a special way that currently we can only have
one "channel" to transfer the page.
It's because when we install pages using UFFDIO_COPY, we need to have the whole
huge page ready, it also means we need to have a temp huge page when trying to
receive the whole content of the page.
Currently all maintainance around this tmp page is global: firstly we'll
allocate a temp huge page, then we maintain its status mostly within
ram_load_postcopy().
To enable multiple channels for postcopy, the first thing we need to do is to
prepare N temp huge pages as caching, one for each channel.
Meanwhile we need to maintain the tmp huge page status per-channel too.
To give some example, some local variables maintained in ram_load_postcopy()
are listed; they are responsible for maintaining temp huge page status:
- all_zero: this keeps whether this huge page contains all zeros
- target_pages: this counts how many target pages have been copied
- host_page: this keeps the host ptr for the page to install
Move all these fields to be together with the temp huge pages to form a new
structure called PostcopyTmpPage. Then for each (future) postcopy channel, we
need one structure to keep the state around.
For vanilla postcopy, obviously there's only one channel. It contains both
precopy and postcopy pages.
This patch teaches the dest migration node to start realize the possible number
of postcopy channels by introducing the "postcopy_channels" variable. Its
value is calculated when setup postcopy on dest node (during POSTCOPY_LISTEN
phase).
Vanilla postcopy will have channels=1, but when postcopy-preempt capability is
enabled (in the future), we will boost it to 2 because even during partial
sending of a precopy huge page we still want to preempt it and start sending
the postcopy requested page right away (so we start to keep two temp huge
pages; more if we want to enable multifd). In this patch there's a TODO marked
for that; so far the channels is always set to 1.
We need to send one "host huge page" on one channel only and we cannot split
them, because otherwise the data upon the same huge page can locate on more
than one channel so we need more complicated logic to manage. One temp host
huge page for each channel will be enough for us for now.
Postcopy will still always use the index=0 huge page even after this patch.
However it prepares for the latter patches where it can start to use multiple
channels (which needs src intervention, because only src knows which channel we
should use).
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20220301083925.33483-5-peterx@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
dgilbert: Fixed up long line
The function qemu_madvise() and the QEMU_MADV_* constants associated
with it are used in only 10 files. Move them out of osdep.h to a new
qemu/madvise.h header that is included where it is needed.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20220208200856.3558249-2-peter.maydell@linaro.org
This patch simplifies unqueue_page() on both sides of it (itself, and caller).
Firstly, due to the fact that right after unqueue_page() returned true, we'll
definitely send a huge page (see ram_save_huge_page() call - it will _never_
exit before finish sending that huge page), so unqueue_page() does not need to
jump in small page size if huge page is enabled on the ramblock. IOW, it's
destined that only the 1st 4K page will be valid, when unqueue the 2nd+ time
we'll notice the whole huge page has already been sent anyway. Switching to
operating on huge page reduces a lot of the loops of redundant unqueue_page().
Meanwhile, drop the dirty check. It's not helpful to call test_bit() every
time to jump over clean pages, as ram_save_host_page() has already done so,
while in a faster way (see commit ba1b7c812c ("migration/ram: Optimize
ram_save_host_page()", 2021-05-13)). So that's not necessary too.
Drop the two tracepoints along the way - based on above analysis it's very
possible that no one is really using it..
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Add a helper to detect whether postcopy has pending request.
Since at it, cleanup the code a bit, e.g. in unqueue_page() we shouldn't need
to check it again on queue empty because we're the only one (besides cleanup
code, which should never run during this process) that will take a request off
the list, so the request list can only grow but not shrink under the hood.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
We used to do off-by-one fixup for pss->page when finished one host huge page
transfer. That seems to be unnecesary at all. Drop it.
Cc: Keqian Zhu <zhukeqian1@huawei.com>
Cc: Kunkun Jiang <jiangkunkun@huawei.com>
Cc: Andrey Gruzdev <andrey.gruzdev@virtuozzo.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Provide information on the number of bytes copied in the pre-copy,
downtime and post-copy phases of migration.
Signed-off-by: David Edmondson <david.edmondson@oracle.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Replace direct manipulation of ram_counters.transferred with a
function.
Signed-off-by: David Edmondson <david.edmondson@oracle.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
postcopy_send_discard_bm_ram() always return zero. Since it can't
fail, simplify and do not return anything.
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: David Edmondson <david.edmondson@oracle.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
It will just never fail. Drop those return values where they're constantly
zeros.
A tiny touch-up on the tracepoint so trace_ram_postcopy_send_discard_bitmap()
is called after the logic itself (which sounds more reasonable).
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Right now we loop ramblocks for twice, the 1st time chunk the dirty bits with
huge page information; the 2nd time we send the discard ranges. That's not
necessary - we can do them in a single loop.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
This function calls three functions:
- postcopy_discard_send_init(ms, block->idstr);
- postcopy_chunk_hostpages_pass(ms, block);
- postcopy_discard_send_finish(ms);
However only the 2nd function call is meaningful. It's major role is to make
sure dirty bits are applied in host-page-size granule, so there will be no
partial dirty bits set for a whole host page if huge pages are used.
The 1st/3rd call are for latter when we want to send the disgard ranges.
They're mostly no-op here besides some tracepoints (which are misleading!).
Drop them, then we can directly drop postcopy_chunk_hostpages() as a whole
because we can call postcopy_chunk_hostpages_pass() directly.
There're still some nice comments above postcopy_chunk_hostpages() that explain
what it does. Copy it over to the caller's site.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
It always return zero, because it just can't go wrong so far. Simplify the
code with no functional change.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
I planned to add "#ifdef DEBUG_POSTCOPY" around the function too because
otherwise it'll be compiled into qemu binary even if it'll never be used. Then
I found that maybe it's easier to just drop it for good..
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Just a removal of an unused comment.
a0a8aa147a did many fixes and removed the parameter named "ms", but forget to remove the corresponding comment in function named "ram_save_host_page".
Signed-off-by: Xu Zheng <xuzheng@cmss.chinamobile.com>
Signed-off-by: Mao Zhongyi <maozhongyi@cmss.chinamobile.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
We always need to call it when we find a zero page, so put it in a
single place.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Remove the mask in the call to ram_release_pages(). Nothing else does
it, and if the offset has that bits set, we have a lot of trouble.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Remove the pages argument. And s/pages/page/
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
---
- Use 1LL instead of casts (philmd)
- Change the whole 1ULL for TARGET_PAGE_SIZE
We only need last_stage in two places and we are passing it all
around. Just add a field to RAMState that passes it.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
---
Repeat subject (philmd suggestion)
It just calls buffer_is_zero(). Just change the callers.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
The code to acquire bitmap_mutex is added in the commit of
"63268c4970a5f126cc9af75f3ccb8057abef5ec0". There is no
need to acquire bitmap_mutex in colo_flush_ram_cache(). This
is because the colo_flush_ram_cache only be called on the COLO
secondary VM, which is the destination side.
On the COLO secondary VM, only the COLO thread will touch
the bitmap of ram cache.
Signed-off-by: Lei Rao <lei.rao@intel.com>
Reviewed-by: Zhang Chen <chen.zhang@intel.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
if we don't reset the auto-converge counter,
it will continue to run with COLO running,
and eventually the system will hang due to the
CPU throttle reaching DEFAULT_MIGRATE_MAX_CPU_THROTTLE.
Signed-off-by: Lei Rao <lei.rao@intel.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Lukas Straub <lukasstraub2@web.de>
Tested-by: Lukas Straub <lukasstraub2@web.de>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
When flushing memory from ram cache to ram during every checkpoint
on secondary VM, we can copy continuous chunks of memory instead of
4096 bytes per time to reduce the time of VM stop during checkpoint.
Signed-off-by: Lei Rao <lei.rao@intel.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Lukas Straub <lukasstraub2@web.de>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Tested-by: Lukas Straub <lukasstraub2@web.de>
Signed-off-by: Juan Quintela <quintela@redhat.com>
One might set dump-guest-core=off to make coredumps smaller and
still allow to debug many qemu bugs. Extend this option to the colo
cache.
Signed-off-by: Lukas Straub <lukasstraub2@web.de>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
This avoids to call migrate_get_current() in the caller function
whereas migration_cancel() already needs the pointer to the current
migration state.
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
We already don't ever migrate memory that corresponds to discarded ranges
as managed by a RamDiscardManager responsible for the mapped memory region
of the RAMBlock.
virtio-mem uses this mechanism to logically unplug parts of a RAMBlock.
Right now, we still populate zeropages for the whole usable part of the
RAMBlock, which is undesired because:
1. Even populating the shared zeropage will result in memory getting
consumed for page tables.
2. Memory backends without a shared zeropage (like hugetlbfs and shmem)
will populate an actual, fresh page, resulting in an unintended
memory consumption.
Discarded ("logically unplugged") parts have to remain discarded. As
these pages are never part of the migration stream, there is no need to
track modifications via userfaultfd WP reliably for these parts.
Further, any writes to these ranges by the VM are invalid and the
behavior is undefined.
Note that Linux only supports userfaultfd WP on private anonymous memory
for now, which usually results in the shared zeropage getting populated.
The issue will become more relevant once userfaultfd WP supports shmem
and hugetlb.
Acked-by: Peter Xu <peterx@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Let's factor out prefaulting/populating to make further changes easier to
review and add a comment what we are actually expecting to happen. While at
it, use the actual page size of the ramblock, which defaults to
qemu_real_host_page_size for anonymous memory. Further, rename
ram_block_populate_pages() to ram_block_populate_read() as well, to make
it clearer what we are doing.
In the future, we might want to use MADV_POPULATE_READ to speed up
population.
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Let's use QEMU_ALIGN_DOWN() and friends to make the code a bit easier to
read.
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Currently, when someone (i.e., the VM) accesses discarded parts inside a
RAMBlock with a RamDiscardManager managing the corresponding mapped memory
region, postcopy will request migration of the corresponding page from the
source. The source, however, will never answer, because it refuses to
migrate such pages with undefined content ("logically unplugged"): the
pages are never dirty, and get_queued_page() will consequently skip
processing these postcopy requests.
Especially reading discarded ("logically unplugged") ranges is supposed to
work in some setups (for example with current virtio-mem), although it
barely ever happens: still, not placing a page would currently stall the
VM, as it cannot make forward progress.
Let's check the state via the RamDiscardManager (the state e.g.,
of virtio-mem is migrated during precopy) and avoid sending a request
that will never get answered. Place a fresh zero page instead to keep
the VM working. This is the same behavior that would happen
automatically without userfaultfd being active, when accessing virtual
memory regions without populated pages -- "populate on demand".
For now, there are valid cases (as documented in the virtio-mem spec) where
a VM might read discarded memory; in the future, we will disallow that.
Then, we might want to handle that case differently, e.g., warning the
user that the VM seems to be mis-behaving.
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
We don't want to migrate memory that corresponds to discarded ranges as
managed by a RamDiscardManager responsible for the mapped memory region of
the RAMBlock. The content of these pages is essentially stale and
without any guarantees for the VM ("logically unplugged").
Depending on the underlying memory type, even reading memory might populate
memory on the source, resulting in an undesired memory consumption. Of
course, on the destination, even writing a zeropage consumes memory,
which we also want to avoid (similar to free page hinting).
Currently, virtio-mem tries achieving that goal (not migrating "unplugged"
memory that was discarded) by going via qemu_guest_free_page_hint() - but
it's hackish and incomplete.
For example, background snapshots still end up reading all memory, as
they don't do bitmap syncs. Postcopy recovery code will re-add
previously cleared bits to the dirty bitmap and migrate them.
Let's consult the RamDiscardManager after setting up our dirty bitmap
initially and when postcopy recovery code reinitializes it: clear
corresponding bits in the dirty bitmaps (e.g., of the RAMBlock and inside
KVM). It's important to fixup the dirty bitmap *after* our initial bitmap
sync, such that the corresponding dirty bits in KVM are actually cleared.
As colo is incompatible with discarding of RAM and inhibits it, we don't
have to bother.
Note: if a misbehaving guest would use discarded ranges after migration
started we would still migrate that memory: however, then we already
populated that memory on the migration source.
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
since dirty ring has been introduced, there are two methods
to track dirty pages of vm. it seems that "logging" has
a hint on the method, so rename the global_dirty_log to
global_dirty_tracking would make description more accurate.
dirty rate measurement may start or stop dirty tracking during
calculation. this conflict with migration because stop dirty
tracking make migration leave dirty pages out then that'll be
a problem.
make global_dirty_tracking a bitmask can let both migration and
dirty rate measurement work fine. introduce GLOBAL_DIRTY_MIGRATION
and GLOBAL_DIRTY_DIRTY_RATE to distinguish what current dirty
tracking aims for, migration or dirty rate.
Signed-off-by: Hyman Huang(黄勇) <huangy81@chinatelecom.cn>
Message-Id: <9c9388657cfa0301bd2c1cfa36e7cf6da4aeca19.1624040308.git.huangy81@chinatelecom.cn>
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
The parameter is unused, let's drop it.
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
When skipping free pages to send, their corresponding dirty bits in the
memory region dirty bitmap need to be cleared. Otherwise the skipped
pages will be sent in the next round after the migration thread syncs
dirty bits from the memory region dirty bitmap.
Cc: David Hildenbrand <david@redhat.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Reported-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Wei Wang <wei.w.wang@intel.com>
Message-Id: <20210722083055.23352-1-wei.w.wang@intel.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
migration uses QIOChannel typed qemufiles. In follow up patches, we'll need
the capability to identify this fact, so that we can get the backing QIOChannel
from a QEMUFile.
We can also define types for QEMUFile but so far since we only need to be able
to identify QIOChannel, introduce a boolean which is simpler.
Introduce another helper qemu_file_get_ioc() to return the ioc backend of a
qemufile if has_ioc is set.
No functional change.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20210722175841.938739-5-peterx@redhat.com>
Reviewed-by: Lukas Straub <lukasstraub2@web.de>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Accessing from_dst_file is potentially racy in current code base like below:
if (s->from_dst_file)
do_something(s->from_dst_file);
Because from_dst_file can be reset right after the check in another
thread (rp_thread). One example is migrate_fd_cancel().
Use the same qemu_file_lock to protect it too, just like to_dst_file.
When it's safe to access without lock, comment it.
There's one special reference in migration_thread() that can be replaced by
the newly introduced rp_thread_created flag.
Reported-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Lukas Straub <lukasstraub2@web.de>
Message-Id: <20210722175841.938739-3-peterx@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
with Peter's fixup
Taking the mutex every time for each dirty bit to clear is too slow, especially
we'll take/release even if the dirty bit is cleared. So far it's only used to
sync with special cases with qemu_guest_free_page_hint() against migration
thread, nothing really that serious yet. Let's move the lock to be upper.
There're two callers of migration_bitmap_clear_dirty().
For migration, move it into ram_save_iterate(). With the help of MAX_WAIT
logic, we'll only run ram_save_iterate() for no more than 50ms-ish time, so
taking the lock once there at the entry. It also means any call sites to
qemu_guest_free_page_hint() can be delayed; but it should be very rare, only
during migration, and I don't see a problem with it.
For COLO, move it up to colo_flush_ram_cache(). I think COLO forgot to take
that lock even when calling ramblock_sync_dirty_bitmap(), where another example
is migration_bitmap_sync() who took it right. So let the mutex cover both the
ramblock_sync_dirty_bitmap() and migration_bitmap_clear_dirty() calls.
It's even possible to drop the lock so we use atomic operations upon rb->bmap
and the variable migration_dirty_pages. I didn't do it just to still be safe,
also not predictable whether the frequent atomic ops could bring overhead too
e.g. on huge vms when it happens very often. When that really comes, we can
keep a local counter and periodically call atomic ops. Keep it simple for now.
Cc: Wei Wang <wei.w.wang@intel.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Hailiang Zhang <zhang.zhanghailiang@huawei.com>
Cc: Dr. David Alan Gilbert <dgilbert@redhat.com>
Cc: Juan Quintela <quintela@redhat.com>
Cc: Leonardo Bras Soares Passos <lsoaresp@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20210630200805.280905-1-peterx@redhat.com>
Reviewed-by: Wei Wang <wei.w.wang@intel.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Fixes commit 3d0684b2ad ("ram: Update
all functions comments")
Signed-off-by: Olaf Hering <olaf@aepfle.de>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20210708162159.18045-1-olaf@aepfle.de>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Let's introduce RAM_NORESERVE, allowing mmap'ing with MAP_NORESERVE. The
new flag has the following semantics:
"
RAM is mmap-ed with MAP_NORESERVE. When set, reserving swap space (or huge
pages if applicable) is skipped: will bail out if not supported. When not
set, the OS will do the reservation, if supported for the memory type.
"
Allow passing it into:
- memory_region_init_ram_nomigrate()
- memory_region_init_resizeable_ram()
- memory_region_init_ram_from_file()
... and teach qemu_ram_mmap() and qemu_anon_ram_alloc() about the flag.
Bail out if the flag is not supported, which is the case right now for
both, POSIX and win32. We will add Linux support next and allow specifying
RAM_NORESERVE via memory backends.
The target use case is virtio-mem, which dynamically exposes memory
inside a large, sparse memory area to the VM.
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Acked-by: Eduardo Habkost <ehabkost@redhat.com> for memory backend and machine core
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20210510114328.21835-9-david@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Juan Quintela
李皆俊
Lukas Straub
ling xu
David Hildenbrand
Peter Xu
Thomas Huth
Leonardo Bras
Daniel P. Berrangé
Dr. David Alan Gilbert
Markus Armbruster
Peter Maydell
David Edmondson
Philippe Mathieu-Daudé
Xu Zheng
Rao, Lei
Laurent Vivier
Hyman Huang(黄勇)
Wei Wang
Olaf Hering