Avoid hot pages being replaced by others to remarkably decrease cache
misses
Sample results with the test program which quote from xbzrle.txt ran in
vm:(migrate bandwidth:1GE and xbzrle cache size 8MB)
the test program:
include <stdlib.h>
include <stdio.h>
int main()
{
char *buf = (char *) calloc(4096, 4096);
while (1) {
int i;
for (i = 0; i < 4096 * 4; i++) {
buf[i * 4096 / 4]++;
}
printf(".");
}
}
before this patch:
virsh qemu-monitor-command test_vm '{"execute": "query-migrate"}'
{"return":{"expected-downtime":1020,"xbzrle-cache":{"bytes":1108284,
"cache-size":8388608,"cache-miss-rate":0.987013,"pages":18297,"overflow":8,
"cache-miss":1228737},"status":"active","setup-time":10,"total-time":52398,
"ram":{"total":12466991104,"remaining":1695744,"mbps":935.559472,
"transferred":5780760580,"dirty-sync-counter":271,"duplicate":2878530,
"dirty-pages-rate":29130,"skipped":0,"normal-bytes":5748592640,
"normal":1403465}},"id":"libvirt-706"}
18k pages sent compressed in 52 seconds.
cache-miss-rate is 98.7%, totally miss.
after optimizing:
virsh qemu-monitor-command test_vm '{"execute": "query-migrate"}'
{"return":{"expected-downtime":2054,"xbzrle-cache":{"bytes":5066763,
"cache-size":8388608,"cache-miss-rate":0.485924,"pages":194823,"overflow":0,
"cache-miss":210653},"status":"active","setup-time":11,"total-time":18729,
"ram":{"total":12466991104,"remaining":3895296,"mbps":937.663549,
"transferred":1615042219,"dirty-sync-counter":98,"duplicate":2869840,
"dirty-pages-rate":58781,"skipped":0,"normal-bytes":1588404224,
"normal":387794}},"id":"libvirt-266"}
194k pages sent compressed in 18 seconds.
The value of cache-miss-rate decrease to 48.59%.
Signed-off-by: ChenLiang <chenliang88@huawei.com>
Signed-off-by: Gonglei <arei.gonglei@huawei.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>
If block used_length does not match, try to resize it.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
This patch allows us to distinguish between two
length values for each block:
max_length - length of memory block that was allocated
used_length - length of block used by QEMU/guest
Currently, we set used_length - max_length, unconditionally.
Follow-up patches allow used_length <= max_length.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
The static variables in migration_bitmap_sync will not be reset in
the case of a second attempted migration.
Signed-off-by: ChenLiang <chenliang88@huawei.com>
Signed-off-by: Gonglei <arei.gonglei@huawei.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>
During migration, the values read from migration stream during ram load
are not validated. Especially offset in host_from_stream_offset() and
also the length of the writes in the callers of said function.
To fix this, we need to make sure that the [offset, offset + length]
range fits into one of the allocated memory regions.
Validating addr < len should be sufficient since data seems to always be
managed in TARGET_PAGE_SIZE chunks.
Fixes: CVE-2014-7840
Note: follow-up patches add extra checks on each block->host access.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>
this patch extends commit db80fac by not only checking
for unknown flags, but also filtering out unknown flag
combinations.
Suggested-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Juan Quintela <quintela@redhat.com>
As the function always return 1, it is not needed anymore.
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
When live migrate fails due to a section length mismatch we currently
see an error message like:
Length mismatch: 0000:00:03.0/virtio-net-pci.rom: 10000 in != 20000
The section lengths are in fact in hex, so this should read
Length mismatch: 0000:00:03.0/virtio-net-pci.rom: 0x10000 in != 0x20000
Correct the error string to reflect this.
Signed-off-by: Alex Bligh <alex@alex.org.uk>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
if a saved vm has unknown flags in the memory data qemu
currently simply ignores this flag and continues which
yields in an unpredictable result.
This patch catches all unknown flags and aborts the
loading of the vm. Additionally error reports are thrown
if the migration aborts abnormally.
Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Juan Quintela <quintela@redhat.com>
We call g_free() after cache_fini() in migration_end(), but we don't
call it after cache_fini() in xbzrle_cache_resize(), leaking the
memory.
cache_init() and cache_fini() are a pair. Since cache_init()
allocates the cache, let cache_fini() free it. This plugs the leak.
Signed-off-by: Chen Gang <gang.chen.5i5j@gmail.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Replace fprintf(stderr,...) with error_report() in the file
arch_init.c. The trailing "\n"s of the @fmt argument have been removed
because @fmt of error_report() should not contain newline.
Signed-off-by: Le Tan <tamlokveer@gmail.com>
Reviewed-by: Andreas Färber <afaerber@suse.de>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
ram_save_block is getting a bit too complicated, and does two separate
things:
1) Finds a page to send
2) Sends the page (dealing with compression etc)
Split into 'ram_save_page' to send the page and deal with compression (2)
Rename remaining function to 'ram_find_and_save_block'
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
For xbzrle_decode_buffer(), when decoding contents will exceed writing
buffer, it will return -1, so need not check the return value whether
large than writing buffer.
And when failure occurs within load_xbzrle(), it always return -1
without any resources which need release.
So can remove the related checking statements, and also can remove 'rc'
and 'ret' local variables,
Signed-off-by: Chen Gang <gang.chen.5i5j@gmail.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
When DPRINTF() has effect, the original author wants to print all
ram_load() calling results. So need use 'goto' instead of 'return'
within ram_load(), just like other areas have done.
Signed-off-by: Chen Gang <gang.chen.5i5j@gmail.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
expose the count that logs the times of updating the dirty bitmap to
end user.
Signed-off-by: ChenLiang <chenliang88@huawei.com>
Signed-off-by: Gonglei <arei.gonglei@huawei.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Add counts to log the times of updating the dirty bitmap.
Signed-off-by: ChenLiang <chenliang88@huawei.com>
Signed-off-by: Gonglei <arei.gonglei@huawei.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
The page may not be inserted into cache after executing save_xbzrle_page.
In case of failure to insert, the original page should be sent rather
than the page in the cache.
Signed-off-by: ChenLiang <chenliang88@huawei.com>
Signed-off-by: Gonglei <arei.gonglei@huawei.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
version_id is checked twice in the ram_load.
Signed-off-by: ChenLiang <chenliang88@huawei.com>
Signed-off-by: Gonglei <arei.gonglei@huawei.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Initialising the XBZRLE.lock earlier simplifies the lock use.
Based on Markus's patch in:
http://lists.gnu.org/archive/html/qemu-devel/2014-03/msg03879.html
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Gonglei <arei.gonglei@huawei.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Provide ram_mig_init (like blk_mig_init) for vl.c to initialise stuff
to do with ram migration (currently in arch_init.c).
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Gonglei <arei.gonglei@huawei.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
This is a fix for a bug* triggered by a migration after hot unplugging
a few virtio-net NICs, that caused migration never to converge, because
'migration_dirty_pages' is incorrectly initialised.
'migration_dirty_pages' is used as a tally of the number of outstanding
dirty pages, to give the migration code an idea of how much more data
will need to be transferred, and thus whether it can end the iterative
phase.
It was initialised to the total size of the RAMBlock address space,
however hotunplug can leave this space sparse, and hence
migration_dirty_pages ended up too large.
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
(* https://bugzilla.redhat.com/show_bug.cgi?id=1074913 )
Signed-off-by: Juan Quintela <quintela@redhat.com>
Resizing the xbzrle cache during migration causes qemu-crash,
because the main-thread and migration-thread modify the xbzrle
cache size concurrently without lock-protection.
Signed-off-by: ChenLiang <chenliang88@huawei.com>
Signed-off-by: Gonglei <arei.gonglei@huawei.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Push zero'd pages into the XBZRLE cache
A page that was cached by XBZRLE, zero'd and then XBZRLE'd again
was being compared against a stale cache value
Don't use 'qemu_put_buffer_async' to put pages from the XBZRLE cache
Since the cache might change before the data hits the wire
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
It is better to fail migration in case of failure to
allocate new cache item
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
When qemu do live migration with xbzrle, qemu malloc decoded_buf
at destination end but free it at source end. It will crash qemu
by double free error in some scenarios. Splitting the XBZRLE structure
for clear logic distinguishing src/dst side.
Signed-off-by: ChenLiang <chenliang88@huawei.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Orit Wasserman <owasserm@redhat.com>
Signed-off-by: GongLei <arei.gonglei@huawei.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
We use the old code if the bitmaps are not aligned
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Orit Wasserman <owasserm@redhat.com>
This function is the only bit where we care about speed.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Orit Wasserman <owasserm@redhat.com>
The madvise for zeroed out pages was introduced when every transferred
zero page was memset to zero and thus allocated. Since commit
211ea740 we check for zeroness of a target page before we memset
it to zero. Additionally we memmap target memory so it is essentially
zero initialized (except for e.g. option roms and bios which are loaded
into target memory although they shouldn't).
It was reported recently that this madvise causes a performance degradation
in some situations. As the madvise should only be called rarely and if it's called
it is likely on a busy page (it was non-zero and changed to zero during migration)
drop it completely.
Reported-By: Zhang Haoyu <haoyu.zhang@huawei.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Juan Quintela <quintela@redhat.com>
This includes pc and pci cleanups and enhancements,
and a virtio-net bugfix related to softmac programming.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.14 (GNU/Linux)
iQEcBAABAgAGBQJSR83kAAoJECgfDbjSjVRpX08H/jKgYBNJaChev1TROIVHEGbu
IzvkjfocvKO+6wmhOf5x+xwFmzrijUMa1CPvOkCp8c2A3Iek7rmnedknlhXYh7dM
z5mXcvFGjnu7ST38ydF/Emk9+Z6rRg5Y/hkmlDyr+9lNcoiCDLXXcUrKjeIHNoWl
e8w3yiPCJ528QyrLwQ890XetJphv67pMlsjMgLQ2betMk++Ac/ctUf1D2p1X4NeQ
Q2drbo5Z4yDk0i6QMA3iLq1Bh/AhE10bCDq9rCzfZGIKVyncL6ne2pSi/xDvpLrF
dmxoiJ5QrK6xLnagCcn5T6SB9DkwbEPdL7qCqlxZ8USr7cVyPdzYtHtGSBWdeXY=
=xF01
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'mst/tags/for_anthony' into staging
pc,pci,virtio fixes and cleanups
This includes pc and pci cleanups and enhancements,
and a virtio-net bugfix related to softmac programming.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
# gpg: Signature made Sun 29 Sep 2013 01:51:16 AM CDT using RSA key ID D28D5469
# gpg: Can't check signature: public key not found
# By Michael S. Tsirkin (8) and others
# Via Michael S. Tsirkin
* mst/tags/for_anthony:
smbios: Factor out smbios_maybe_add_str()
smbios: Make multiple -smbios type= accumulate sanely
smbios: Improve diagnostics for conflicting entries
smbios: Convert to QemuOpts
smbios: Normalize smbios_entry_add()'s error handling to exit(1)
virtio-net: fix up HMP NIC info string on reset
pci: remove explicit check to 64K ioport size
piix4: disable io on reset
piix: use 64 bit window programmed by guest
q35: use 64 bit window programmed by guest
pci: add helper to retrieve the 64-bit range
range: add min/max operations on ranges
range: add Range to typedefs
q35: make pci window address/size match guest cfg
Message-id: 1380437951-21788-1-git-send-email-mst@redhat.com
Currently, -smbios type=T,NAME=VAL,... adds one field (T,NAME) with
value VAL to fw_cfg for each unique NAME. If NAME occurs multiple
times, the last one's VAL is used (before the QemuOpts conversion, the
first one was used).
Multiple -smbios can add multiple fields with the same (T, NAME).
SeaBIOS reads all of them from fw_cfg, but uses only the first field
(T, NAME). The others are ignored.
"First one wins, subsequent ones get ignored silently" isn't nice. We
commonly let the last option win. Useful, because it lets you
-readconfig first, then selectively override with command line
options.
Clean up -smbios to work the common way. Accumulate the settings,
with later ones overwriting earlier ones. Put the result into fw_cfg
(no more useless duplicates).
Bonus cleanup: qemu_uuid_parse() no longer sets SMBIOS system uuid by
side effect.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
So that it can be set in config file for -readconfig.
This tightens parsing of -smbios, and makes it more consistent with
other options: unknown parameters are rejected, numbers with trailing
junk are rejected, when a parameter is given multiple times, last
rather than first wins, ...
MST: drop one chunk to fix build errors
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
It exits on all error conditions but one, where it returns -1.
Normalize, and return void.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
ram_handle_compressed() should be aware of size > TARGET_PAGE_SIZE.
migration-rdma can call it with larger size.
Signed-off-by: Isaku Yamahata <yamahata@private.email.ne.jp>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Later is_zero_page will be used for non TARGET_PAGE_SIZE
range.
And rename it to is_zero_range as it isn't page size any more.
Signed-off-by: Isaku Yamahata <yamahata@private.email.ne.jp>
Signed-off-by: Juan Quintela <quintela@redhat.com>
qemu_file_rate_limit() never return negative value since the refactor
by Commit 1964a39, this patch gets rid of the negative check for it,
adjust bytes_transferred and return value correspondingly in
ram_save_iterate().
Signed-off-by: Lei Li <lilei@linux.vnet.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
It was introduced to loop over CPUs from target-independent code, but
since commit 182735efaf target-independent
CPUState is used.
A loop can be considered more efficient than function calls in a loop,
and CPU_FOREACH() hides implementation details just as well, so use that
instead.
Suggested-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Andreas Färber <afaerber@suse.de>
This includes pc and pci cleanups, future-proofing of ROM files,
and a virtio bugfix correcting splice on virtio console.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.14 (GNU/Linux)
iQEcBAABAgAGBQJSGvbsAAoJECgfDbjSjVRp0qsH/2N/b/5bNhL6gx5Kjigzv7Lt
bmCUsSsUm/Rnhke2SpDPVdIOU9W5WIX2zte8qiXuBJi3j/RS9procbgUZrCetpUI
GHQOn+n8e2TX/P9CIcHaN/bA1b+Kx+bvaVbDL/6P2PJvodbDGwDLp/9y+Tisv5Ye
kLow8Y4ZJcaTlmPf/Mh1AnhNa3gel231A3qwugVUNDSyITM6pG/0M07xk8YBj+JJ
6DYrlK0bKMNDxPu5St+YP94D1ODv2zM1aio/TdMUaNfqTZM1iqGTj3zKkBS2PjT0
RhuU2x4N91lY/uCdudtWSj5dYtfRT/xw7qwNAz9IHNz6RlGeX0n++ClMC8z9Y8A=
=Pc09
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'mst/tags/for_anthony' into stable-1.5
pc,pci,virtio fixes and cleanups
This includes pc and pci cleanups, future-proofing of ROM files,
and a virtio bugfix correcting splice on virtio console.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
# gpg: Signature made Mon 26 Aug 2013 01:34:20 AM CDT using RSA key ID D28D5469
# gpg: Can't check signature: public key not found
# By Markus Armbruster (5) and others
# Via Michael S. Tsirkin
* mst/tags/for_anthony:
virtio: virtqueue_get_avail_bytes: fix desc_pa when loop over the indirect descriptor table
pc_piix: Kill pc_init1() memory region args
pc: pc_compat_1_4() now can call pc_compat_1_5()
pc: Create pc_compat_*() functions
pc: Kill pc_init_pci_1_0()
pc: Don't explode QEMUMachineInitArgs into local variables needlessly
pc: Don't prematurely explode QEMUMachineInitArgs
ppc: Don't duplicate QEMUMachineInitArgs in PPCE500Params
ppc: Don't explode QEMUMachineInitArgs into local variables needlessly
sun4: Don't prematurely explode QEMUMachineInitArgs
q35: Add PCIe switch to example q35 configuration
loader: store FW CFG ROM files in RAM
arch_init: align MR size to target page size
pc: cleanup 1.4 compat support
Message-id: 1377535318-30491-1-git-send-email-mst@redhat.com
# By Alex Bligh (32) and others
# Via Stefan Hajnoczi
* stefanha/block: (42 commits)
win32-aio: drop win32_aio_flush_cb()
aio-win32: replace incorrect AioHandler->opaque usage with ->e
aio / timers: remove dummy_io_handler_flush from tests/test-aio.c
aio / timers: Remove legacy interface
aio / timers: Switch entire codebase to the new timer API
aio / timers: Add scripts/switch-timer-api
aio / timers: Add test harness for AioContext timers
aio / timers: convert block_job_sleep_ns and co_sleep_ns to new API
aio / timers: Convert rtc_clock to be a QEMUClockType
aio / timers: Remove main_loop_timerlist
aio / timers: Rearrange timer.h & make legacy functions call non-legacy
aio / timers: Add qemu_clock_get_ms and qemu_clock_get_ms
aio / timers: Remove legacy qemu_clock_deadline & qemu_timerlist_deadline
aio / timers: Remove alarm timers
aio / timers: Add documentation and new format calls
aio / timers: Use all timerlists in icount warp calculations
aio / timers: Introduce new API timer_new and friends
aio / timers: On timer modification, qemu_notify or aio_notify
aio / timers: Convert mainloop to use timeout
aio / timers: Convert aio_poll to use AioContext timers' deadline
...
Message-id: 1377202298-22896-1-git-send-email-stefanha@redhat.com
Signed-off-by: Anthony Liguori <anthony@codemonkey.ws>
This is an autogenerated patch using scripts/switch-timer-api.
Switch the entire code base to using the new timer API.
Note this patch may introduce some line length issues.
Signed-off-by: Alex Bligh <alex@alex.org.uk>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Migration code assumes that each MR is a multiple of TARGET_PAGE_SIZE:
MR size is divided by TARGET_PAGE_SIZE, so if it isn't migration
never completes.
But this isn't really required for regions set up with
memory_region_init_ram, since that calls qemu_ram_alloc
which aligns size up using TARGET_PAGE_ALIGN.
Align MR size up to full target page sizes, this way
migration completes even if we create a RAM MR
which is not a full target page size.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Convert stderr messages calling error_get_pretty()
to error_report().
Timestamp is prepended by -msg timstamp option with it.
Per Markus's comment below, A conversion from fprintf() to
error_report() is always an improvement, regardless of
error_get_pretty().
http://marc.info/?l=qemu-devel&m=137513283408601&w=2
But, it is not reasonable to convert them at one time
because fprintf() is used everwhere in qemu.
So, it should be done step by step with avoiding regression.
Signed-off-by: Seiji Aguchi <seiji.aguchi@hds.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
This takes advantages of the previous patches:
1. use the new QEMUFileOps hook 'save_page'
2. call out to the right accessor methods to invoke
the iteration hooks defined in QEMUFileOps
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Chegu Vinod <chegu_vinod@hp.com>
Tested-by: Chegu Vinod <chegu_vinod@hp.com>
Tested-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
This gives RDMA shared access to madvise() on the destination side
when an entire chunk is found to be zero.
Reviewed-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Chegu Vinod <chegu_vinod@hp.com>
Tested-by: Chegu Vinod <chegu_vinod@hp.com>
Tested-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
If a user chooses to turn on the auto-converge migration capability
these changes detect the lack of convergence and throttle down the
guest. i.e. force the VCPUs out of the guest for some duration
and let the migration thread catchup and help converge.
Verified the convergence using the following :
- Java Warehouse workload running on a 20VCPU/256G guest(~80% busy)
- OLTP like workload running on a 80VCPU/512G guest (~80% busy)
Sample results with Java warehouse workload : (migrate speed set to 20Gb and
migrate downtime set to 4seconds).
(qemu) info migrate
capabilities: xbzrle: off auto-converge: off <----
Migration status: active
total time: 1487503 milliseconds
expected downtime: 519 milliseconds
transferred ram: 383749347 kbytes
remaining ram: 2753372 kbytes
total ram: 268444224 kbytes
duplicate: 65461532 pages
skipped: 64901568 pages
normal: 95750218 pages
normal bytes: 383000872 kbytes
dirty pages rate: 67551 pages
---
(qemu) info migrate
capabilities: xbzrle: off auto-converge: on <----
Migration status: completed
total time: 241161 milliseconds
downtime: 6373 milliseconds
transferred ram: 28235307 kbytes
remaining ram: 0 kbytes
total ram: 268444224 kbytes
duplicate: 64946416 pages
skipped: 64903523 pages
normal: 7044971 pages
normal bytes: 28179884 kbytes
Signed-off-by: Chegu Vinod <chegu_vinod@hp.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
We have stayed at 800x600x15 as default graphics mode for the last 9 years.
If there ever was a reason to be there, surely nobody remembers it.
However, recently non-Linux PPC guests started to show bad effects on 15 bit
color mode. They do work just fine with 32 bits however.
So let's switch to 32 bit color as the default graphic mode.
Reported-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Signed-off-by: Alexander Graf <agraf@suse.de>
length is a ram_addr_t, so RAM_ADDR_FMT must be used instead of %ld.
This fixes a recently introduced regression for w64 builds.
Using RAM_ADDR_FMT also changes decimal output to sedecimal.
This is good here because length and block->length should both
use the same base in the error message.
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Message-id: 1372359606-2759-1-git-send-email-sw@weilnetz.de
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
RDMA writes happen asynchronously, and thus the performance accounting
also needs to be able to occur asynchronously. This allows anybody
to call into savevm.c to update both f->pos as well as into arch_init.c
to update the acct_info structure with up-to-date values when
the RDMA transfer actually completes.
Reviewed-by: Juan Quintela <quintela@redhat.com>
Tested-by: Chegu Vinod <chegu_vinod@hp.com>
Tested-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
on incoming migration do not memset pages to zero if they already read as zero.
this will allocate a new zero page and consume memory unnecessarily. even
if we madvise a MADV_DONTNEED later this will only deallocate the memory
asynchronously.
Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Not sending zero pages breaks migration if a page is zero
at the source but not at the destination. This can e.g. happen
if different BIOS versions are used at source and destination.
It has also been reported that migration on pseries is completely
broken with this patch.
This effectively reverts commit f1c72795af.
Conflicts:
arch_init.c
Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Makes it easier to debug situations where the source and target have
different ram blocks in a device and migration fails due to that, for
instance a BAR size change on a PCI device.
Signed-off-by: Alon Levy <alevy@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
# By Paolo Bonzini (4) and others
# Via Peter Maydell
* pmaydell/configury.next:
ppc: Remove CONFIG_FDT conditionals
microblaze: Remove CONFIG_FDT conditionals
arm: Remove CONFIG_FDT conditionals
configure: Require libfdt for arm, ppc, microblaze softmmu targets
configure: dtc: Probe for libfdt_env.h
build: drop TARGET_TYPE
main: use TARGET_ARCH only for the target-specific #define
build: do not use TARGET_ARCH
build: rename TARGET_ARCH2 to TARGET_NAME
Add a stp file for usage from build directory
Message-id: 1371221594-11556-1-git-send-email-peter.maydell@linaro.org
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Just use the TARGET_NAME free string.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 1370349928-20419-6-git-send-email-pbonzini@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Everything else needs to match the executable name, which is
TARGET_NAME.
Before:
$ sh4eb-linux-user/qemu-sh4eb --help
usage: qemu-sh4 [options] program [arguments...]
Linux CPU emulator (compiled for sh4 emulation)
After:
$ sh4eb-linux-user/qemu-sh4eb --help
usage: qemu-sh4eb [options] program [arguments...]
Linux CPU emulator (compiled for sh4eb emulation)
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 1370349928-20419-5-git-send-email-pbonzini@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Having size precede the associated pointer is odd. Swap them, and fix
up the types.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Laszlo "ever the optimist" Ersek <lersek@redhat.com>
Message-id: 1370610036-10577-5-git-send-email-armbru@redhat.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Improves diagnistics from ad hoc messages like
Invalid SMBIOS UUID string
to
qemu-system-x86_64: -smbios type=1,uuid=gaga: Invalid UUID
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Laszlo "ever the optimist" Ersek <lersek@redhat.com>
Message-id: 1370610036-10577-4-git-send-email-armbru@redhat.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Since this is a MemoryListener operation, it only makes sense
on an AddressSpace granularity.
Suggested-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Several targets can have wavcapture/-soundhw support via PCI cards.
HAS_AUDIO is a useless limitation, remove it.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 1366303444-24620-4-git-send-email-pbonzini@redhat.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Headers in include/exec/ are for the deepest innards of QEMU,
they should almost never be included directly.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Functions defined in acpi/ should be declared in
acpi.h
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Many of these should be cleaned up with proper qdev-/QOM-ification.
Right now there are many catch-all headers in include/hw/ARCH depending
on cpu.h, and this makes it necessary to compile these files per-target.
However, fixing this does not belong in these patches.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
As one consequence, strtok() -- which modifies its argument -- is replaced
with g_strsplit().
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Message-id: 1363821803-3380-6-git-send-email-lersek@redhat.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This will remove an unneeded copy of guest memory pages.
For the page header and device state we still copy the data to the
static buffer the other option is to allocate the memory on demand
which is more expensive.
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
at the beginning of migration all pages are marked dirty and
in the first round a bulk migration of all pages is performed.
currently all these pages are copied to the page cache regardless
of whether they are frequently updated or not. this doesn't make sense
since most of these pages are never transferred again.
this patch changes the XBZRLE transfer to only be used after
the bulk stage has been completed. that means a page is added
to the page cache the second time it is transferred and XBZRLE
can benefit from the third time of transfer.
since the page cache is likely smaller than the number of pages
it's also likely that in the second round the page is missing in the
cache due to collisions in the bulk phase.
on the other hand a lot of unnecessary mallocs, memdups and frees
are saved.
the following results have been taken earlier while executing
the test program from docs/xbzrle.txt. (+) with the patch and (-)
without. (thanks to Eric Blake for reformatting and comments)
+ total time: 22185 milliseconds
- total time: 22410 milliseconds
Shaved 0.3 seconds, better than 1%!
+ downtime: 29 milliseconds
- downtime: 21 milliseconds
Not sure why downtime seemed worse, but probably not the end of the world.
+ transferred ram: 706034 kbytes
- transferred ram: 721318 kbytes
Fewer bytes sent - good.
+ remaining ram: 0 kbytes
- remaining ram: 0 kbytes
+ total ram: 1057216 kbytes
- total ram: 1057216 kbytes
+ duplicate: 108556 pages
- duplicate: 105553 pages
+ normal: 175146 pages
- normal: 179589 pages
+ normal bytes: 700584 kbytes
- normal bytes: 718356 kbytes
Fewer normal bytes...
+ cache size: 67108864 bytes
- cache size: 67108864 bytes
+ xbzrle transferred: 3127 kbytes
- xbzrle transferred: 630 kbytes
...and more compressed pages sent - good.
+ xbzrle pages: 117811 pages
- xbzrle pages: 21527 pages
+ xbzrle cache miss: 18750
- xbzrle cache miss: 179589
And very good improvement on the cache miss rate.
+ xbzrle overflow : 0
- xbzrle overflow : 0
Signed-off-by: Peter Lieven <pl@kamp.de>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Orit Wasserman <owasserm@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
avoid searching for dirty pages just increment the
page offset. all pages are dirty anyway.
Signed-off-by: Peter Lieven <pl@kamp.de>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Orit Wasserman <owasserm@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
during bulk stage of ram migration if a page is a
zero page do not send it at all.
the memory at the destination reads as zero anyway.
even if there is an madvise with QEMU_MADV_DONTNEED
at the target upon receipt of a zero page I have observed
that the target starts swapping if the memory is overcommitted.
it seems that the pages are dropped asynchronously.
this patch also updates QMP to return the number of
skipped pages in MigrationStats.
Signed-off-by: Peter Lieven <pl@kamp.de>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
the first round of ram transfer is special since all pages
are dirty and thus all memory pages are transferred to
the target. this patch adds a boolean variable to track
this stage.
Signed-off-by: Peter Lieven <pl@kamp.de>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Orit Wasserman <owasserm@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
virtually all dup pages are zero pages. remove
the special is_dup_page() function and use the
optimized buffer_find_nonzero_offset() function
instead.
here buffer_find_nonzero_offset() is used directly
to avoid the unnecssary additional checks in
buffer_is_zero().
raw performace gain checking 1 GByte zeroed memory
over is_dup_page() is approx. 10-12% with SSE2
and 8-10% with unsigned long arithmedtic.
Signed-off-by: Peter Lieven <pl@kamp.de>
Reviewed-by: Orit Wasserman <owasserm@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
vector optimizations will now be used at various places
not just in is_dup_page() in arch_init.c
Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Juan Quintela <quintela@redhat.com>
The page cache frees all data on finish, on resize and
if there is collision on insert. So it should be the caches
responsibility to dup the data that is stored in the cache.
Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Only the migration_bitmap_sync() call needs the iothread lock.
Reviewed-by: Orit Wasserman <owasserm@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
This makes it possible to do blocking writes directly to the socket,
with no buffer in the middle. For RAM, only the migration_bitmap_sync()
call needs the iothread lock. For block migration, it is needed by
the block layer (including bdrv_drain_all and dirty bitmap access),
but because some code is shared between iterate and complete, all of
mig_save_device_dirty is run with the lock taken.
In the savevm case, the iterate callback runs within the big lock.
This is annoying because it complicates the rules. Luckily we do not
need to do anything about it: the RAM iterate callback does not need
the iothread lock, and block migration never runs during savevm.
Reviewed-by: Orit Wasserman <owasserm@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
We removed the calculation in commit e4ed1541ac
Now we add it back. We need to create dirty_bytes_rate because we
can't include cpu-all.h from migration.c, and there is no other way to
include TARGET_PAGE_SIZE.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Orit Wasserman <owasserm@redhat.com>
Before this fix we couldn't load a guest from
XBZRLE compressed file.
For example:
The user activated the XBZRLE capability
The user run migrate -d "exec:gzip -c > vm.gz"
The user won't be able to load vm.gz and get an error.
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Reviewed-by: Eric Blake <eblake@redhat.com>
It could only return 0 if we only found dirty xbzrle pages that hadn't
changed (i.e. they were written with the same content). We don't care
about that case, it is the same than nothing dirty.
So now the return of the function is how much have it written, nothing
else. Adjust callers.
And we also made ram_save_iterate() return the number of transferred
bytes, not the number of transferred pages.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Instead of testing each page individually, we search what is the next
dirty page with a bitmap operation. We have to reorganize the code to
move from a "for" loop, to a while(dirty) loop.
Signed-off-by: Juan Quintela <quintela@redhat.com>
This avoids having to do two walks over the dirty bitmap, once reading
the dirty bits, and anthoer cleaning them.
Signed-off-by: Juan Quintela <quintela@redhat.com>
This is the last block from where we have sent data.
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Code just now does (simplified for clarity)
if (qemu_savevm_state_iterate(s->file) == 1) {
vm_stop_force_state(RUN_STATE_FINISH_MIGRATE);
qemu_savevm_state_complete(s->file);
}
Problem here is that qemu_savevm_state_iterate() returns 1 when it
knows that remaining memory to sent takes less than max downtime.
But this means that we could end spending 2x max_downtime, one
downtime in qemu_savevm_iterate, and the other in
qemu_savevm_state_complete.
Changed code to:
pending_size = qemu_savevm_state_pending(s->file, max_size);
DPRINTF("pending size %lu max %lu\n", pending_size, max_size);
if (pending_size >= max_size) {
ret = qemu_savevm_state_iterate(s->file);
} else {
vm_stop_force_state(RUN_STATE_FINISH_MIGRATE);
qemu_savevm_state_complete(s->file);
}
So what we do is: at current network speed, we calculate the maximum
number of bytes we can sent: max_size.
Then we ask every save_live section how much they have pending. If
they are less than max_size, we move to complete phase, otherwise we
do an iterate one.
This makes things much simpler, because now individual sections don't
have to caluclate the bandwidth (it was implossible to do right from
there).
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Add the new mutex that protects shared state between ram_save_live
and the iothread. If the iothread mutex has to be taken together
with the ramlist mutex, the iothread shall always be _outside_.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Umesh Deshpande <udeshpan@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Orit Wasserman <owasserm@redhat.com>
This will be used to detect if last_block might have become invalid
across different calls to ram_save_live.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Umesh Deshpande <udeshpan@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Orit Wasserman <owasserm@redhat.com>
Most of the time, only 2 items will be active (from/to for a string operation,
or code/data). But TCG guests likely won't have gigabytes of memory, so
this actually goes down to 1 item.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
ChenLiang