Stefan Weil [Tue, 29 Nov 2011 05:34:48 +0000 (06:34 +0100)]
malta: Fix regression (i8259 interrupts did not work)
Commit 5632ae46d5bda798e971dae48ebb318ac2c3686a passes the address
of i8259 to qemu_irq_proxy. i8259 is an auto variable with undefined
value outside of mips_malta_init.
This made the interrupt proxy unusable: either QEMU crashes, or
the interrupt handler was not called.
Ethernet for example no longer worked with MIPS Malta.
v2:
While v1 used a static variable for i8259, this patch introduces
a qdev for the malta machine. i8259 is now part of the device status.
This is a minimal qdev implementation to keep the patch small.
Signed-off-by: Stefan Weil <sw@weilnetz.de> Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Anthony Liguori [Fri, 6 Jan 2012 18:51:21 +0000 (12:51 -0600)]
Merge remote-tracking branch 'stefanha/trivial-patches' into staging
* stefanha/trivial-patches:
qemu-nbd: drop loop which can never loop
Make python mandatory
net/socket.c: Fix fd leak in net_socket_listen_init() error paths
gdbstub: Fix fd leak in gdbserver_open() error path
configure: Fix test for supported host CPU type
configure: CONFIG_QEMU_INTERP_PREFIX only for user mode
scsi virtio-blk usb-msd: Clean up device init error messages
Strip trailing '\n' from error_report()'s first argument (again)
qemu-options.hx: fix tls-channel help text
Peter Maydell [Fri, 6 Jan 2012 18:47:21 +0000 (18:47 +0000)]
virtio-9p-proxy: Fix typo causing compile failure on 32 bit hosts
Fix a compile failure on 32 bit hosts (integer constant is too large
for 'unsigned long' type) by correcting a typo where the mask used
for filling in the second f_fsid word had too many 'F's in it.
Also drop the 'L' suffix that allowed this typo to go undetected on
64 bit hosts.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Stefan Hajnoczi [Thu, 5 Jan 2012 13:16:07 +0000 (13:16 +0000)]
qemu-nbd: drop loop which can never loop
For some reason nbd_client_thread() has a do..while loop which can never
loop, the condition is bogus because we would take a goto instead. Drop
the loop.
Reported-by: Dr David Alan Gilbert <davidagilbert@uk.ibm.com> Reviewed-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
The QEMU build depends on Python so make it an explicit requirement.
Reviewed-by: Stefan Weil <sw@weilnetz.de> Signed-off-by: Sebastian Herbszt <herbszt@gmx.de> Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Peter Maydell [Sat, 24 Dec 2011 13:07:25 +0000 (13:07 +0000)]
configure: Fix test for supported host CPU type
The test for whether the host CPU is supported had several problems:
* the attempt to fall back to TCI was done as a duplicate
test, very late (so "--cpu foo" would fail early but "--cpu unicore32"
would fail late, differently, and after configure had already
printed a lot of output)
* a number of CPUs only supported as guests were included in the
list of CPUs we would accept as valid hosts, which would result
in a late compile failure on those systems rather than a
configure failure or fallback to TCI
* bailing out for an unsupported CPU happened before the main
option parsing, so "configure --help" wouldn't work
Fix these by folding the setting of ARCH into the first test for
supported host CPU, removing spurious guest-only CPU names from it,
and moving the "fall back to TCI" code earlier.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
scsi virtio-blk usb-msd: Clean up device init error messages
Replace
error_report("DEVICE-NAME: MESSAGE");
by just
error_report("MESSAGE");
in block device init functions.
DEVICE-NAME is bogus in some cases: it's "scsi-disk" for device
scsi-hd and scsi-cd, "virtio-blk-pci" for virtio-blk-s390, and
"usb-msd" for usb-storage.
There is no real need to put a device name in the message, because
error_report() points to the offending command line option already:
$ qemu-system-x86_64 --nodefaults --enable-kvm -vnc :0 -S -monitor stdio -usb -device virtio-blk-pci
upstream-qemu: -device virtio-blk-pci: virtio-blk-pci: drive property not set
upstream-qemu: -device virtio-blk-pci: Device 'virtio-blk-pci' could not be initialized
And for a monitor command, it's obvious anyway:
$ qemu-system-x86_64 --nodefaults --enable-kvm -vnc :0 -S -monitor stdio -usb
(qemu) device_add virtio-blk-pci
virtio-blk-pci: drive property not set
Device 'virtio-blk-pci' could not be initialized
Reported-by: Amit Shah <amit.shah@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Alon Levy [Tue, 20 Dec 2011 11:05:18 +0000 (13:05 +0200)]
qemu-options.hx: fix tls-channel help text
Remove the default compiled out tunnel channel, add the always available
cursor channel. Optimally the man page would depend on compiled in
options, but that's harder to do.
RHBZ: 688586
Acked-by: Gerd Hoffmann <kraxel@redhat.com> Signed-off-by: Alon Levy <alevy@redhat.com> Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Anthony Liguori [Fri, 6 Jan 2012 14:14:28 +0000 (08:14 -0600)]
Merge remote-tracking branch 'aneesh/for-upstream' into staging
* aneesh/for-upstream:
hw/9pfs: Add support to use named socket for proxy FS
hw/9pfs: man page for proxy helper
hw/9pfs: Documentation changes related to proxy fs
hw/9pfs: Proxy getversion
hw/9pfs: xattr interfaces in proxy filesystem driver
hw/9pfs: File ownership and others
hw/9pfs: Add stat/readlink/statfs for proxy FS
hw/9pfs: Create other filesystem objects
hw/9pfs: Open and create files
hw/9pfs: File system helper process for qemu 9p proxy FS
hw/9pfs: Add new proxy filesystem driver
hw/9pfs: Add validation to {un}marshal code
hw/9pfs: Move pdu_marshal/unmarshal code to a seperate file
hw/9pfs: Move opt validation to FsDriver callback
Anthony Liguori [Fri, 6 Jan 2012 14:11:41 +0000 (08:11 -0600)]
Merge remote-tracking branch 'kraxel/usb.33' into staging
* kraxel/usb.33:
usb-ohci: td.cbp incorrectly updated near page end
usb-host: properly release port on unplug & exit
usb-storage: cancel I/O on reset
Fix parse of usb device description with multiple configurations
Andriy Gapon [Thu, 22 Dec 2011 09:34:30 +0000 (11:34 +0200)]
usb-ohci: td.cbp incorrectly updated near page end
The current code that updates the cbp value after a transfer looks like this:
td.cbp += ret;
if ((td.cbp & 0xfff) + ret > 0xfff) {
<handle page overflow>
because the 'ret' value is effectively added twice the check may fire too early
when the overflow hasn't happened yet.
Below is one of the possible changes that correct the behavior:
Gerd Hoffmann [Thu, 5 Jan 2012 14:49:18 +0000 (15:49 +0100)]
usb-host: properly release port on unplug & exit
Factor out port release into a separate function. Call release function
in exit notifier too. Add explicit call the USBDEVFS_RELEASE_PORT
ioctl, just closing the hub file handle seems not to be enougth. Make
sure we release the port before resetting the device, otherwise host
drivers will not re-attach.
Gerd Hoffmann [Wed, 4 Jan 2012 17:13:54 +0000 (18:13 +0100)]
usb-storage: cancel I/O on reset
When resetting the usb-storage device we'll have to carefully cancel
and clear any requests which might be in flight, otherwise we'll confuse
the state machine.
Anthony PERARD [Thu, 5 Jan 2012 12:47:08 +0000 (12:47 +0000)]
xen: Fix after recent change in dirty bitmap tracking.
A recent patch set from Avi break the dirty bitmap support of Xen. But
this is because xen_sync_dirty_bitmap will return an error for an
unhandled memory range (a0000 - bffff). However this is not a fatal
error, so we should just continue instead of aborting.
There is now an error printed when the Xen call failed.
Signed-off-by: Anthony PERARD <anthony.perard@citrix.com> Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Anthony Liguori [Wed, 4 Jan 2012 16:06:25 +0000 (10:06 -0600)]
Merge remote-tracking branch 'pmaydell/arm-devs.for-upstream' into staging
* pmaydell/arm-devs.for-upstream:
add L2x0/PL310 cache controller device
arm: add dummy gic security registers
arm: Set frequencies for arm_timer
arm: add missing scu registers
hw/omap_gpmc: Fix region map/unmap when configuring prefetch engine
hw/omap1.c: Drop unused includes
hw/omap1.c: Separate dpll_ctl from omap_mpu_state
hw/omap1.c: Separate PWT from omap_mpu_state
hw/omap1.c: Separate PWL from omap_mpu_state
hw/omap1.c: omap_mpuio_init() need not be public
hw/pl110.c: Add post-load hook to invalidate display
hw/pl181.c: Add save/load support
M. Mohan Kumar [Wed, 14 Dec 2011 08:28:47 +0000 (13:58 +0530)]
hw/9pfs: Add support to use named socket for proxy FS
Add option to use named socket for communicating between proxy helper
and qemu proxy FS. Access to socket can be given by using command line
options -u and -g.
Signed-off-by: M. Mohan Kumar <mohan@in.ibm.com> Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
M. Mohan Kumar [Wed, 14 Dec 2011 08:19:06 +0000 (13:49 +0530)]
hw/9pfs: Move pdu_marshal/unmarshal code to a seperate file
Move p9 marshaling/unmarshaling code to a separate file so that
proxy filesytem driver can use these calls. Also made marshaling
code generic to accept "struct iovec" instead of V9fsPDU.
Signed-off-by: M. Mohan Kumar <mohan@in.ibm.com> Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Rob Herring [Thu, 29 Dec 2011 06:19:54 +0000 (06:19 +0000)]
add L2x0/PL310 cache controller device
This is just a dummy device for ARM L2 cache controllers, based on the
pl310. The cache type parameter can be defined by a property value
and has a meaningful default.
Signed-off-by: Rob Herring <rob.herring@calxeda.com> Signed-off-by: Mark Langsdorf <mark.langsdorf@calxeda.com>
[Peter Maydell: removed stray blank line at end] Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Rob Herring [Thu, 29 Dec 2011 06:19:53 +0000 (06:19 +0000)]
arm: add dummy gic security registers
Implement handling for the RAZ/WI gic security registers.
Signed-off-by: Rob Herring <rob.herring@calxeda.com> Signed-off-by: Mark Langsdorf <mark.langsdorf@calxeda.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Mark Langsdorf [Thu, 29 Dec 2011 06:19:51 +0000 (06:19 +0000)]
arm: Set frequencies for arm_timer
Use qdev properties to allow board modelers to set the frequencies
for the sp804 timer. Each of the sp804's timers can have an
individual frequency. The timers default to 1MHz.
Signed-off-by: Mark Langsdorf <mark.langsdorf@calxeda.com> Reviewed-by: Andreas Färber <afaerber@suse.de> Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Rob Herring [Thu, 29 Dec 2011 06:19:50 +0000 (06:19 +0000)]
arm: add missing scu registers
Add power control register to a9mpcore
Signed-off-by: Rob Herring <rob.herring@calxeda.com> Signed-off-by: Mark Langsdorf <mark.langsdorf@calxeda.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Peter Maydell [Tue, 20 Dec 2011 00:21:56 +0000 (00:21 +0000)]
hw/omap_gpmc: Fix region map/unmap when configuring prefetch engine
When configuring the prefetch engine (and also when resetting from
a state where the prefetch engine was enabled) be careful to adhere
to the "unmap/change config fields/map" ordering, to avoid trying
to delete the wrong MemoryRegions. This fixes an assertion failure
in some cases.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reported-by: Alexander Graf <agraf@suse.de> Tested-by: Alexander Graf <agraf@suse.de>
Juha Riihimäki [Tue, 20 Dec 2011 08:11:34 +0000 (08:11 +0000)]
hw/omap1.c: Separate dpll_ctl from omap_mpu_state
Signed-off-by: Juha Riihimäki <juha.riihimaki@nokia.com>
[Riku Voipio: Fixes and restructuring patchset] Signed-off-by: Riku Voipio <riku.voipio@iki.fi>
[Peter Maydell: More fixes and cleanups for upstream submission] Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Juha Riihimäki [Tue, 20 Dec 2011 08:11:33 +0000 (08:11 +0000)]
hw/omap1.c: Separate PWT from omap_mpu_state
Signed-off-by: Juha Riihimäki <juha.riihimaki@nokia.com>
[Riku Voipio: Fixes and restructuring patchset] Signed-off-by: Riku Voipio <riku.voipio@iki.fi>
[Peter Maydell: More fixes and cleanups for upstream submission] Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Avi Kivity [Mon, 2 Jan 2012 15:21:07 +0000 (17:21 +0200)]
Remove IO_MEM_SHIFT
We no longer use any of the lower bits of a ram_addr, so we might as well
use them for the io table index. This increases the number of potential
I/O handlers by a factor of 8.
Signed-off-by: Avi Kivity <avi@redhat.com> Reviewed-by: Richard Henderson <rth@twiddle.net>
Avi Kivity [Mon, 2 Jan 2012 11:12:08 +0000 (13:12 +0200)]
Direct dispatch through MemoryRegion
Now that all mmio goes through MemoryRegions, we can convert
io_mem_opaque to be a MemoryRegion pointer, and remove the thunks
that convert from old-style CPU{Read,Write}MemoryFunc to MemoryRegionOps.
Signed-off-by: Avi Kivity <avi@redhat.com> Reviewed-by: Richard Henderson <rth@twiddle.net>
Avi Kivity [Sun, 1 Jan 2012 22:32:15 +0000 (00:32 +0200)]
Convert IO_MEM_{RAM,ROM,UNASSIGNED,NOTDIRTY} to MemoryRegions
Convert the fixed-address IO_MEM_RAM, IO_MEM_ROM, IO_MEM_UNASSIGNED,
and IO_MEM_NOTDIRTY io handlers to MemoryRegions. These aren't real
regions, since they are never added to the memory hierarchy, but they
allow reuse of the dispatch functionality.
Signed-off-by: Avi Kivity <avi@redhat.com> Reviewed-by: Richard Henderson <rth@twiddle.net>
Avi Kivity [Sun, 1 Jan 2012 16:24:24 +0000 (18:24 +0200)]
Fix wrong region_offset when overlaying a page with another
cpu_register_physical_memory_log() does not update region_offset
if a page was previously registered for the same address. This
could cause mmio accesses going to the wrong place, by using the
old region_offset.
Signed-off-by: Avi Kivity <avi@redhat.com> Acked-by: Andreas Färber <afaerber@suse.de> Reviewed-by: Richard Henderson <rth@twiddle.net>
Avi Kivity [Sun, 1 Jan 2012 16:02:31 +0000 (18:02 +0200)]
memory: remove MemoryRegion::backend_registered
backend_registered was used to lazify the process of registering an
mmio region, since the it is different for the I/O address space and
the memory address space. However, it also makes registration dependent
on the region being visible in the address space. This is not the case
for "fake" regions, like watchpoints or IO_MEM_UNASSIGNED.
Remove backend_registered and always initialize the region. If it turns
out to be part of the I/O address space, we've wasted an I/O slot, but
that's not too bad. In any case this will be optimized later on.
Signed-off-by: Avi Kivity <avi@redhat.com> Reviewed-by: Richard Henderson <rth@twiddle.net>
Avi Kivity [Sun, 20 Nov 2011 15:52:22 +0000 (17:52 +0200)]
exec: make phys_page_find() return a temporary
Instead of returning a PhysPageDesc pointer, return a temporary.
This lets us move away from actually storing PhysPageDesc's, and
instead sythesising them when needed.
Signed-off-by: Avi Kivity <avi@redhat.com> Reviewed-by: Richard Henderson <rth@twiddle.net>
Avi Kivity [Wed, 21 Dec 2011 11:37:56 +0000 (13:37 +0200)]
Remove support for version 3 ram_load
Version 3 ram_load depends on ram_addrs, which are not stable. Version 4
was introduced in 0.13 (and RHEL 6), so this means live migration from 0.12
and earlier to 1.1 or later will not work.
Reviewed-by: Anthony Liguori <aliguori@us.ibm.com> Signed-off-by: Avi Kivity <avi@redhat.com>
Avi Kivity [Wed, 21 Dec 2011 11:22:16 +0000 (13:22 +0200)]
Sort RAMBlocks by ID for migration, not by ram_addr
ram_addr is (a) unstable (b) going away. Sort by idstr instead.
Commit b2e0a138e initially introduced the sorting for the purpose
of improving debuggability. After this patch, the order is still
stable, but perhaps less usable by a human.
Reviewed-by: Anthony Liguori <aliguori@us.ibm.com> Signed-off-by: Avi Kivity <avi@redhat.com>
Avi Kivity [Tue, 20 Dec 2011 13:59:12 +0000 (15:59 +0200)]
vmstate, memory: decouple vmstate from memory API
Currently creating a memory region automatically registers it for
live migration. This differs from other state (which is enumerated
in a VMStateDescription structure) and ties the live migration code
into the memory core.
Decouple the two by introducing a separate API, vmstate_register_ram(),
for registering a RAM block for migration. Currently the same
implementation is reused, but later it can be moved into a separate list,
and registrations can be moved to VMStateDescription blocks.
Cao,Bing Bu [Tue, 13 Dec 2011 01:22:20 +0000 (09:22 +0800)]
Fix parse of usb device description with multiple configurations
Changed From V1:
Use DPRINTF instead of fprintf,because it is not an error.
When testing ipod on QEMU by He Jie Xu<xuhj@linux.vnet.ibm.com>,qemu made a assertion.
We found that the ipod with 2 configurations,and the usb-linux did not parse the descriptor correctly.
The descr_len returned is the total length of the all configurations,not one configuration.
The older version will through the other configurations instead of skip,continue parsing the descriptor of interfaces/endpoints in other configurations,then went wrong.
This patch will put the configuration descriptor parse in loop outside and dispel the other configurations not requested.
Signed-off-by: Cao,Bing Bu <mars@linux.vnet.ibm.com> Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Juha Riihimäki [Tue, 20 Dec 2011 08:11:32 +0000 (08:11 +0000)]
hw/omap1.c: Separate PWL from omap_mpu_state
Signed-off-by: Juha Riihimäki <juha.riihimaki@nokia.com>
[Riku Voipio: Fixes and restructuring patchset] Signed-off-by: Riku Voipio <riku.voipio@iki.fi>
[Peter Maydell: More fixes and cleanups for upstream submission] Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Peter Maydell [Mon, 19 Dec 2011 12:01:58 +0000 (12:01 +0000)]
hw/pl110.c: Add post-load hook to invalidate display
Add a post-load hook which invalidates the display. In particular, if we
don't do this and the display size we've just reloaded is larger than
the default then we will segfault trying to read off the end of the buffer.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Anthony Liguori [Tue, 3 Jan 2012 20:39:05 +0000 (14:39 -0600)]
Merge remote-tracking branch 'qemu-kvm/memory/page_desc' into staging
* qemu-kvm/memory/page_desc: (22 commits)
Remove cpu_get_physical_page_desc()
sparc: avoid cpu_get_physical_page_desc()
virtio-balloon: avoid cpu_get_physical_page_desc()
vhost: avoid cpu_get_physical_page_desc()
kvm: avoid cpu_get_physical_page_desc()
memory: remove CPUPhysMemoryClient
xen: convert to MemoryListener API
memory: temporarily add memory_region_get_ram_addr()
xen, vga: add API for registering the framebuffer
vhost: convert to MemoryListener API
kvm: convert to MemoryListener API
kvm: switch kvm slots to use host virtual address instead of ram_addr_t
memory: add API for observing updates to the physical memory map
memory: replace cpu_physical_sync_dirty_bitmap() with a memory API
framebuffer: drop use of cpu_physical_sync_dirty_bitmap()
loader: remove calls to cpu_get_physical_page_desc()
framebuffer: drop use of cpu_get_physical_page_desc()
memory: introduce memory_region_find()
memory: add memory_region_is_logging()
memory: add memory_region_is_rom()
...
when I tried qemu with -virtio-console pty the guest hangs and attaching
on /dev/pts/<x> does not return anything if the attachment is too late.
This results in pty_chr_write() returning 0, which causes the port to
get throttled. This results in the guest getting frozen as the
guest->host virtio_console writes don't return until the host releases
the vq element back to the guest.
For the virtio-serial use case we don't want to lose data but for the
console case we better drop data instead of "killing" the guest
console. If we get chardev->frontend notification and a better behaving
virtio-console we can revert this fix.
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com> Signed-off-by: Amit Shah <amit.shah@redhat.com>
Avi Kivity [Tue, 27 Dec 2011 14:02:16 +0000 (16:02 +0200)]
Fix qapi code generation wrt parallel build
Make's multiple output syntax
x.c x.h: x.template
gen < x.template
actually invokes the command once for x.c and once for x.h (with differing $@
in each invocation). During a parallel build, the two commands may be invoked
in parallel; this opens up a race, where the second invocation trashes a file
supposedly produced during the first, and now in use by a dependent command.
The various qapi code generators are susceptible to this; fix by making them
generate just one file per invocation.
Signed-off-by: Avi Kivity <avi@redhat.com> Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Anthony Liguori [Tue, 27 Dec 2011 14:53:35 +0000 (08:53 -0600)]
Merge remote-tracking branch 'aneesh/for-upstream' into staging
* aneesh/for-upstream:
scripts/analyse-9p-simpletrace.py: Add symbolic names for 9p operations.
hw/9pfs: iattr_valid flags are kernel internal flags map them to 9p values.
hw/9pfs: Use the correct signed type for different variables
hw/9pfs: replace iovec manipulation with QEMUIOVector
Anthony Liguori [Tue, 27 Dec 2011 14:52:42 +0000 (08:52 -0600)]
Merge remote-tracking branch 'bonzini/nbd-for-anthony' into staging
* bonzini/nbd-for-anthony: (26 commits)
nbd: add myself as maintainer
qemu-nbd: throttle requests
qemu-nbd: asynchronous operation
qemu-nbd: add client pointer to NBDRequest
qemu-nbd: move client handling to nbd.c
qemu-nbd: use common main loop
link the main loop and its dependencies into the tools
qemu-nbd: introduce NBDRequest
qemu-nbd: introduce NBDExport
qemu-nbd: introduce nbd_do_receive_request
qemu-nbd: more robust handling of invalid requests
qemu-nbd: introduce nbd_do_send_reply
qemu-nbd: simplify nbd_trip
move corking functions to osdep.c
qemu-nbd: remove data_size argument to nbd_trip
qemu-nbd: remove offset argument to nbd_trip
Update ioctl order in nbd_init() to detect EBUSY
nbd: add support for NBD_CMD_TRIM
nbd: add support for NBD_CMD_FLUSH
nbd: add support for NBD_CMD_FLAG_FUA
...
qemu-kvm passes numa/SRAT topology information for smp_cpus to SeaBIOS. However
SeaBIOS always expects to setup max_cpus number of SRAT cpu entries
(MaxCountCPUs variable in build_srat function of Seabios). When qemu-kvm runs
with smp_cpus != max_cpus (e.g. -smp 2,maxcpus=4), Seabios will mistakenly use
memory SRAT info for setting up CPU SRAT entries for the offline CPUs. Wrong
SRAT memory entries are also created. This breaks NUMA in a guest.
Fix by setting up SRAT info for max_cpus in qemu-kvm.
Jan Kiszka [Wed, 26 Oct 2011 11:09:45 +0000 (13:09 +0200)]
kvm: x86: Drop redundant apic base and tpr update from kvm_get_sregs
The latter was already commented out, the former is redundant as well.
We always get the latest changes after return from the guest via
kvm_arch_post_run.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com> Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Paolo Bonzini [Mon, 19 Sep 2011 13:25:40 +0000 (15:25 +0200)]
qemu-nbd: throttle requests
Limiting the number of in-flight requests is implemented very simply
with a can_read callback. It does not require a semaphore, unlike the
client side in block/nbd.c, because we can throttle directly the creation
of coroutines. The client side can have a coroutine created at any time
when an I/O request is made.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Paolo Bonzini [Mon, 19 Sep 2011 13:19:27 +0000 (15:19 +0200)]
qemu-nbd: asynchronous operation
Using coroutines enable asynchronous operation on both the network and
the block side. Network can be owned by two coroutines at the same time,
one writing and one reading. On the send side, mutual exclusion is
guaranteed by a CoMutex. On the receive side, mutual exclusion is
guaranteed because new coroutines immediately start receiving data,
and no new coroutines are created as long as the previous one is receiving.
Between receive and send, qemu-nbd can have an arbitrary number of
in-flight block transfers. Throttling is implemented by the next
patch.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Paolo Bonzini [Fri, 7 Oct 2011 14:47:56 +0000 (16:47 +0200)]
qemu-nbd: add client pointer to NBDRequest
By attaching a client to an NBDRequest, we can avoid passing around the
socket descriptor and data buffer.
Also, we can now manage the reference count for the client in
nbd_request_get/put request instead of having to do it ourselved in
nbd_read. This simplifies things when coroutines are used.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Paolo Bonzini [Mon, 19 Sep 2011 12:33:23 +0000 (14:33 +0200)]
qemu-nbd: move client handling to nbd.c
This patch sets up the fd handler in nbd.c instead of qemu-nbd.c. It
introduces NBDClient, which wraps the arguments to nbd_trip in a single
structure, so that we can add a notifier to it. This way, qemu-nbd can
know about disconnections.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Paolo Bonzini [Mon, 12 Sep 2011 14:20:11 +0000 (16:20 +0200)]
link the main loop and its dependencies into the tools
Using the main loop code from QEMU enables tools to operate fully
asynchronously. Advantages include better Windows portability (for some
definition of portability) over glib's.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Paolo Bonzini [Mon, 19 Sep 2011 12:18:33 +0000 (14:18 +0200)]
qemu-nbd: introduce NBDRequest
Move the buffer from NBDExport to a new structure, so that it will be
possible to have multiple in-flight requests for the same export
(and for the same client too---we get that for free).
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
projects / qemu-xen-4.5-testing.git / log