Paolo Bonzini [Wed, 31 Oct 2012 09:09:11 +0000 (10:09 +0100)]
threadpool: do not take lock in event_notifier_ready
The ordering is:
worker thread consumer thread
-------------------------------------------------------------------
write ret event_notifier_test_and_clear
wmb() read state
write state rmb()
event_notifier_set read ret
Paolo Bonzini [Thu, 23 Feb 2012 12:23:34 +0000 (13:23 +0100)]
aio: add generic thread-pool facility
Add a generic thread-pool. The code is roughly based on posix-aio-compat.c,
with some changes, especially the following:
- use QemuSemaphore instead of QemuCond;
- separate the state of the thread from the return code of the worker
function. The return code is totally opaque for the thread pool;
- do not busy wait when doing cancellation.
A more generic threadpool (but still specific to I/O so that in the future
it can use special scheduling classes or PI mutexes) can have many uses:
it allows more flexibility in raw-posix.c and can more easily be extended
to Win32, and it will also be used to do an msync of the persistent bitmap.
Paolo Bonzini [Mon, 8 Aug 2011 12:36:41 +0000 (14:36 +0200)]
qemu-thread: add QemuSemaphore
The new thread pool will use semaphores instead of condition
variables, because QemuCond does not have qemu_cond_timedwait.
(I also like it more this way).
Paolo Bonzini [Fri, 24 Feb 2012 07:39:02 +0000 (08:39 +0100)]
linux-aio: use event notifiers
Since linux-aio already uses an eventfd, converting it to use the
EventNotifier-based API simplifies the code even though it is not
meant to be portable.
Paolo Bonzini [Tue, 25 Sep 2012 08:22:39 +0000 (10:22 +0200)]
aio: call aio_notify after setting I/O handlers
In the current code, this is done by qemu_set_fd_handler2, which is
called by qemu_aio_set_fd_handler. We need to keep the same behavior
even after removing the call to qemu_set_fd_handler2.
Paolo Bonzini [Sat, 9 Jun 2012 02:01:51 +0000 (04:01 +0200)]
aio: add Win32 implementation
The Win32 implementation will only accept EventNotifiers, thus a few
drivers are disabled under Windows. EventNotifiers are a good match
for the GSource implementation, too, because the Win32 port of glib
allows to place their HANDLEs in a GPollFD.
Paolo Bonzini [Mon, 24 Sep 2012 12:57:22 +0000 (14:57 +0200)]
aio: prepare for introducing GSource-based dispatch
This adds a GPollFD to each AioHandler. It will then be possible to
attach these GPollFDs to a GSource, and from there to the main loop.
aio_wait examines the GPollFDs and avoids calling select() if any
is set (similar to what it does if bottom halves are available).
Paolo Bonzini [Mon, 29 Oct 2012 22:45:23 +0000 (23:45 +0100)]
aio: introduce AioContext, move bottom halves there
Start introducing AioContext, which will let us remove globals from
aio.c/async.c, and introduce multiple I/O threads.
The bottom half functions now take an additional AioContext argument.
A bottom half is created with a specific AioContext that remains the
same throughout the lifetime. qemu_bh_new is just a wrapper that
uses a global context.
Paolo Bonzini [Mon, 29 Oct 2012 14:15:15 +0000 (15:15 +0100)]
sockets: use weak aliases instead of qemu-tool.c
qemu-tool.c has its own (largeish) set of dependencies. Weak aliases
can be placed directly where people use them, and do not contribute
to increasing the dependencies of generic utility files.
Luiz Capitulino [Wed, 24 Oct 2012 16:12:15 +0000 (14:12 -0200)]
win32: fix broken build due to missing QEMU_MADV_HUGEPAGE
Commit ad0b5321f1f797274603ebbe20108b0750baee94 forgot to add
QEMU_MADV_HUGEPAGE macros for when CONFIG_MADVISE is not defined.
This broke the build for Windows. Fix it.
Anthony Liguori [Wed, 24 Oct 2012 14:39:49 +0000 (09:39 -0500)]
Merge remote-tracking branch 'bonzini/nbd-next' into staging
* bonzini/nbd-next: (30 commits)
qmp: add NBD server commands
block: add close notifiers
block: prepare code for adding block notifiers
qemu-sockets: add socket_listen, socket_connect, socket_parse
tests: do not include tools-obj-y Signed-off-by: Paolo Bonzini <[email protected]>
qemu-sockets: return InetSocketAddress from inet_parse
qapi: add socket address types
build: add QAPI files to the tools
vnc: drop QERR_VNC_SERVER_FAILED
qemu-sockets: add error propagation to Unix socket functions
qemu-sockets: add error propagation to inet_parse
qemu-sockets: add error propagation to inet_dgram_opts
qemu-sockets: add error propagation to inet_connect_addr
qemu-sockets: include strerror or gai_strerror output in error messages
vnc: add error propagation to vnc_display_open
vnc: reorganize code for reverse mode
vnc: introduce a single label for error returns
vnc: avoid Yoda conditionals
qemu-ga: ask and print error information from qemu-sockets
nbd: ask and print error information from qemu-sockets
...
Paolo Bonzini [Wed, 22 Aug 2012 14:43:07 +0000 (16:43 +0200)]
qmp: add NBD server commands
Adding an NBD server inside QEMU is trivial, since all the logic is
in nbd.c and can be shared easily between qemu-nbd and QEMU itself.
The main difference is that qemu-nbd serves a single unnamed export,
while QEMU serves named exports.
Paolo Bonzini [Thu, 23 Aug 2012 09:20:36 +0000 (11:20 +0200)]
block: add close notifiers
The first user of close notifiers will be the embedded NBD server.
It would be possible to use them to do some of the ad hoc processing
(e.g. for block jobs and I/O limits) that is currently done by
bdrv_close.
Paolo Bonzini [Fri, 19 Oct 2012 09:36:48 +0000 (11:36 +0200)]
block: prepare code for adding block notifiers
There is no reason in principle to skip job cancellation and draining
of pending I/O when there is no medium in the disk. Do these unconditionally,
which also prepares the code for the next patch.
These are QAPI-friendly versions of the qemu-sockets functions. They
support IP sockets, Unix sockets, and named file descriptors, using a
QAPI union to dispatch to the correct function.
Avi Kivity [Tue, 23 Oct 2012 10:30:10 +0000 (12:30 +0200)]
Rename target_phys_addr_t to hwaddr
target_phys_addr_t is unwieldly, violates the C standard (_t suffixes are
reserved) and its purpose doesn't match the name (most target_phys_addr_t
addresses are not target specific). Replace it with a finger-friendly,
standards conformant hwaddr.
Outstanding patchsets can be fixed up with the command
Paolo Bonzini [Wed, 19 Sep 2012 11:54:39 +0000 (13:54 +0200)]
qemu-sockets: add error propagation to Unix socket functions
Before:
$ qemu-system-x86_64 -monitor unix:/vvv,server=off
connect(unix:/vvv): No such file or directory
chardev: opening backend "socket" failed
After:
$ x86_64-softmmu/qemu-system-x86_64 -monitor unix:/vvv,server=off
qemu-system-x86_64: -monitor unix:/vvv,server=off: Failed to connect to socket: No such file or directory
chardev: opening backend "socket" failed
Paolo Bonzini [Tue, 2 Oct 2012 07:19:01 +0000 (09:19 +0200)]
qemu-sockets: add error propagation to inet_connect_addr
perror and fprintf can be removed because all clients can now consume
Errors properly. However, we'll need to change the non-blocking connect
handlers to take an Error, in order to improve error handling for
migration with the TCP protocol.
This is a minor degradation in error reporting for outgoing migration.
However, until 1.2 this case just failed without even attempting to
connect, so it is still an improvement as far as overall QoI is
concerned.
Paolo Bonzini [Tue, 2 Oct 2012 08:17:21 +0000 (10:17 +0200)]
vnc: add error propagation to vnc_display_open
Before:
$ qemu-system-x86_64 -vnc foo.bar:12345
getaddrinfo(foo.bar,18245): Name or service not known
Failed to start VNC server on `foo.bar:12345'
$ qemu-system-x86_64 -vnc localhost:12345,reverse=on
inet_connect_opts: connect(ipv4,yakj.usersys.redhat.com,127.0.0.1,12345): Connection refused
Failed to start VNC server on `localhost:12345,reverse=on'
After:
$ x86_64-softmmu/qemu-system-x86_64 -vnc foo.bar:12345
Failed to start VNC server on `foo.bar:12345': address resolution failed for foo.bar:18245: Name or service not known
$ x86_64-softmmu/qemu-system-x86_64 -vnc localhost:12345,reverse=on
Failed to start VNC server on `localhost:12345,reverse=on': Failed to connect to socket: Connection refused
Paolo Bonzini [Tue, 2 Oct 2012 08:07:21 +0000 (10:07 +0200)]
nbd: ask and print error information from qemu-sockets
Before:
$ qemu-system-x86_64 nbd:localhost:12345
inet_connect_opts: connect(ipv4,yakj.usersys.redhat.com,127.0.0.1,12345): Connection refused
qemu-system-x86_64: could not open disk image nbd:localhost:12345: Connection refused
After:
$ x86_64-softmmu/qemu-system-x86_64 nbd:localhost:12345
qemu-system-x86_64: Failed to connect to socket: Connection refused
qemu-system-x86_64: could not open disk image nbd:localhost:12345: Connection refused
$ x86_64-softmmu/qemu-system-x86_64 -monitor tcp:localhost:6000
x86_64-softmmu/qemu-system-x86_64: -monitor tcp:localhost:6000: Failed to connect to socket: Connection refused
chardev: opening backend "socket" failed
$ x86_64-softmmu/qemu-system-x86_64 -monitor tcp:foo.bar:12345
qemu-system-x86_64: -monitor tcp:foo.bar:12345: address resolution failed for foo.bar:12345: Name or service not known
chardev: opening backend "socket" failed
Paolo Bonzini [Tue, 2 Oct 2012 08:02:46 +0000 (10:02 +0200)]
migration (outgoing): add error propagation for all protocols
Error propagation is already there for socket backends. Add it to other
protocols, simplifying code that tests for errors that will never happen.
With all protocols understanding Error, the code can be simplified
further by removing the return value.
Unfortunately, the quality of error messages varies depending
on where the error is detected, because no Error is passed to the
NonBlockingConnectHandler. Thus, the exact error message still cannot
be sent to the user if the OS reports it asynchronously via SO_ERROR.
If NonBlockingConnectHandler received an Error**, we could for
example report the error class and/or message via a new field of the
query-migration command even if it is reported asynchronously.
Before:
(qemu) migrate fd:ffff
migrate: An undefined error has occurred
(qemu) info migrate
(qemu)
After:
(qemu) migrate fd:ffff
migrate: File descriptor named 'ffff' has not been found
(qemu) info migrate
capabilities: xbzrle: off
Migration status: failed
total time: 0 milliseconds
Paolo Bonzini [Tue, 2 Oct 2012 07:59:38 +0000 (09:59 +0200)]
migration: centralize call to migrate_fd_error()
The call to migrate_fd_error() was missing for non-socket backends, so
centralize it in qmp_migrate().
Before:
(qemu) migrate fd:ffff
migrate: An undefined error has occurred
(qemu) info migrate
(qemu)
After:
(qemu) migrate fd:ffff
migrate: An undefined error has occurred
(qemu) info migrate
capabilities: xbzrle: off
Migration status: failed
total time: 0 milliseconds
(The awful error message will be fixed later in the series).
Paolo Bonzini [Wed, 3 Oct 2012 12:34:33 +0000 (14:34 +0200)]
migration: avoid using error_is_set and thus relying on errp != NULL
The migration code is using errp to detect "internal" errors, this means
that it relies on errp being non-NULL.
No impact so far because our only QMP clients (the QMP marshaller and HMP)
never pass a NULL Error **. But if we had others, this patch would make
sure that migration can work with a NULL Error **.
Anthony Liguori [Mon, 22 Oct 2012 19:49:18 +0000 (14:49 -0500)]
Merge remote-tracking branch 'qemu-kvm/memory/urgent' into staging
* qemu-kvm/memory/urgent:
memory: abort if a memory region is destroyed during a transaction
i440fx: avoid destroying memory regions within a transaction
memory: Make eventfd adhere to device endianness
Gerd Hoffmann [Wed, 17 Oct 2012 07:54:19 +0000 (09:54 +0200)]
serial: split serial.c
Split serial.c into serial.c, serial.h and serial-isa.c. While being at
creating a serial.h header file move the serial prototypes from pc.h to
the new serial.h. The latter leads to s/pc.h/serial.h/ in tons of
boards which just want the serial bits from pc.h
Luiz Capitulino [Fri, 5 Oct 2012 19:47:57 +0000 (16:47 -0300)]
Call MADV_HUGEPAGE for guest RAM allocations
This makes it possible for QEMU to use transparent huge pages (THP)
when transparent_hugepage/enabled=madvise. Otherwise THP is only
used when it's enabled system wide.
Anthony Liguori [Mon, 22 Oct 2012 18:26:23 +0000 (13:26 -0500)]
Merge remote-tracking branch 'quintela/migration-next-20121017' into staging
* quintela/migration-next-20121017: (41 commits)
cpus: create qemu_in_vcpu_thread()
savevm: make qemu_file_put_notify() return errors
savevm: un-export qemu_file_set_error()
block-migration: handle errors with the return codes correctly
block-migration: Switch meaning of return value
block-migration: make flush_blks() return errors
buffered_file: buffered_put_buffer() don't need to set last_error
savevm: Only qemu_fflush() can generate errors
savevm: make qemu_fill_buffer() be consistent
savevm: unexport qemu_ftell()
savevm: unfold qemu_fclose_internal()
savevm: make qemu_fflush() return an error code
savevm: Remove qemu_fseek()
virtio-net: use qemu_get_buffer() in a temp buffer
savevm: unexport qemu_fflush
migration: make migrate_fd_wait_for_unfreeze() return errors
buffered_file: make buffered_flush return the error code
buffered_file: callers of buffered_flush() already check for errors
buffered_file: We can access directly to bandwidth_limit
buffered_file: unfold migrate_fd_close
...
Anthony Liguori [Mon, 22 Oct 2012 18:26:07 +0000 (13:26 -0500)]
Merge remote-tracking branch 'qemu-kvm/memory/dma' into staging
* qemu-kvm/memory/dma: (23 commits)
pci: honor PCI_COMMAND_MASTER
pci: give each device its own address space
memory: add address_space_destroy()
dma: make dma access its own address space
memory: per-AddressSpace dispatch
s390: avoid reaching into memory core internals
memory: use AddressSpace for MemoryListener filtering
memory: move tcg flush into a tcg memory listener
memory: move address_space_memory and address_space_io out of memory core
memory: manage coalesced mmio via a MemoryListener
xen: drop no-op MemoryListener callbacks
kvm: drop no-op MemoryListener callbacks
xen_pt: drop no-op MemoryListener callbacks
vfio: drop no-op MemoryListener callbacks
memory: drop no-op MemoryListener callbacks
memory: provide defaults for MemoryListener operations
memory: maintain a list of address spaces
memory: export AddressSpace
memory: prepare AddressSpace for exporting
xen_pt: use separate MemoryListeners for memory and I/O
...
Avi Kivity [Wed, 3 Oct 2012 15:17:27 +0000 (17:17 +0200)]
pci: give each device its own address space
Accesses from different devices can resolve differently
(depending on bridge settings, iommus, and PCI_COMMAND_MASTER), so
set up an address space for each device.
Currently iommus are expressed outside the memory API, so this doesn't
work if an iommu is present.
Avi Kivity [Wed, 3 Oct 2012 14:22:53 +0000 (16:22 +0200)]
memory: per-AddressSpace dispatch
Currently we use a global radix tree to dispatch memory access. This only
works with a single address space; to support multiple address spaces we
make the radix tree a member of AddressSpace (via an intermediate structure
AddressSpaceDispatch to avoid exposing too many internals).
A side effect is that address_space_io also gains a dispatch table. When
we remove all the pre-memory-API I/O registrations, we can use that for
dispatching I/O and get rid of the original I/O dispatch.
Avi Kivity [Tue, 2 Oct 2012 16:54:45 +0000 (18:54 +0200)]
memory: move tcg flush into a tcg memory listener
We plan to make the core listener listen to all address spaces; this
will cause many more flushes than necessary. Prepare for that by
moving the flush into a tcg-specific listener.
Later we can avoid registering the listener if tcg is disabled.
Avi Kivity [Tue, 2 Oct 2012 16:21:54 +0000 (18:21 +0200)]
memory: manage coalesced mmio via a MemoryListener
Instead of calling a global function on coalesced mmio changes, which
routes the call to kvm if enabled, add coalesced mmio hooks to
MemoryListener and make kvm use that instead.
The motivation is support for multiple address spaces (which means we
we need to filter the call on the right address space) but the result
is cleaner as well.
Michael Tokarev [Sun, 21 Oct 2012 18:52:54 +0000 (22:52 +0400)]
fix CONFIG_QEMU_HELPERDIR generation again
commit 38f419f35225 fixed a breakage with CONFIG_QEMU_HELPERDIR
which has been introduced by 8bf188aa18ef7a8. But while techinically
that fix has been correct, all other similar variables are handled
differently. Make it consistent, and let scripts/create_config
expand and capitalize the variable properly like for all other
qemu_*dir variables.
Peter Maydell [Thu, 18 Oct 2012 13:11:35 +0000 (14:11 +0100)]
qemu-log: Add new log category for guest bugs
Add a new category for device models to log guest behaviour
which is likely to be a guest bug of some kind (accessing
nonexistent registers, reading 32 bit wide registers with
a byte access, etc). Making this its own log category allows
those who care (mostly guest OS authors) to see the complaints
without bothering most users.
Subroutines do their own local temporary management.
Within disas_sparc_insn we limit the existance of the variable
to OP=2 insns, and delay initialization as late as is reasonable
for the specific XOP.
projects / qemu.git / log