This introduces a failed request queue and links all the inflight
requests to the list after network error happens. After QEMU
reconnects to the sheepdog server successfully, the sheepdog block
driver will retry all the requests in the failed queue.
Signed-off-by: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Tested-by: Liu Yuan <namei.unix@gmail.com>
Reviewed-by: Liu Yuan <namei.unix@gmail.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This helper function behaves similarly to co_sleep_ns(), but the
sleeping coroutine will be resumed when using qemu_aio_wait().
Signed-off-by: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Tested-by: Liu Yuan <namei.unix@gmail.com>
Reviewed-by: Liu Yuan <namei.unix@gmail.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This prepares for using resend_aioreq() after reconnecting to the
sheepdog server.
Signed-off-by: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Tested-by: Liu Yuan <namei.unix@gmail.com>
Reviewed-by: Liu Yuan <namei.unix@gmail.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The current resend_aio_req() doesn't work when the request is against
vdi objects. This fixes the problem.
Signed-off-by: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Tested-by: Liu Yuan <namei.unix@gmail.com>
Reviewed-by: Liu Yuan <namei.unix@gmail.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
If qemu_co_recv/send doesn't return the specified length, it means
that an error happened.
Signed-off-by: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Tested-by: Liu Yuan <namei.unix@gmail.com>
Reviewed-by: Liu Yuan <namei.unix@gmail.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Add a test case for trying to open an image file where it is impossible
to open its backing file (in this case, because it was deleted). When
doing this, qemu (or qemu-io in this case) should not crash but rather
print an appropriate error message.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The block layer generally keeps the size of an image cached in
bs->total_sectors so that it doesn't have to perform expensive
operations to get the size whenever it needs it.
This doesn't work however when using a backend that can change its size
without qemu being aware of it, i.e. passthrough of removable media like
CD-ROMs or floppy disks. For this reason, the caching is disabled when a
removable device is used.
It is obvious that checking whether the _guest_ device has removable
media isn't the right thing to do when we want to know whether the size
of the host backend can change. To make things worse, non-top-level
BlockDriverStates never have any device attached, which makes qemu
assume they are removable, so drv->bdrv_getlength() is always called on
the protocol layer. In the case of raw-posix, this causes unnecessary
lseek() system calls, which turned out to be rather expensive.
This patch completely changes the logic and disables bs->total_sectors
caching only for certain block driver types, for which a size change is
expected: host_cdrom and host_floppy on POSIX, host_device on win32; also
the raw format in case it sits on top of one of these protocols, but in
the common case the nested bdrv_getlength() call on the protocol driver
will use the cache again and avoid an expensive drv->bdrv_getlength()
call.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Since commit 0ebd24e0a2,
bdrv_open_common will throw an error when trying to open a file
read-only with the BDRV_O_COPY_ON_READ flag set.
Although BDRV_O_RDWR is unset for the backing files,
BDRV_O_COPY_ON_READ is still passed on if copy-on-read was requested
for the drive. Let's unset this flag too before opening the backing
file, or bdrv_open_common will fail.
Signed-off-by: Thibaut LAURENT <thibaut.laurent@gmail.com>
Reviewed-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
When AHCI executes an asynchronous IDE command, it checked DRDY without
checking either DRQ or BSY. This sometimes caused interrupt to be sent
before command is actually completed.
This resulted in a race condition: if guest then managed to access the
device before command has completed, it would hang waiting for an
interrupt.
This was observed with windows 7 guests.
To fix, check for DRQ or BSY in additiona to DRDY, if set,
the command is asynchronous so delay the interrupt until
asynchronous done callback is invoked.
Reported-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Tested-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Currently copy_policy isn't used. Recent sheepdog supports erasure coding, which
make use of copy_policy internally, but require client explicitly passing
copy_policy from base inode to newly creately inode for snapshot related
operations.
If connected sheep daemon doesn't utilize copy_policy, passing it to sheep
daemon is just one extra null effect operation. So no compatibility problem.
With this patch, sheepdog can provide erasure coded volume for QEMU VM.
Cc: Kevin Wolf <kwolf@redhat.com>
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Liu Yuan <namei.unix@gmail.com>
Acked-by: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
'copies' is actually uint8_t since day one, but request headers and some helper
functions parameterize it as uint32_t for unknown reasons and effectively
reserve 24 bytes for possible future use. This patch explicitly set the correct
for copies and reserve the left bytes.
This is a preparation patch that allow passing copy_policy in request header.
Cc: Kevin Wolf <kwolf@redhat.com>
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Liu Yuan <namei.unix@gmail.com>
Acked-by: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
bdrv_open_backing_file() tries to copy the backing file name using
pstrcpy directly after calling bdrv_open() to open the backing file
without checking whether that was actually successful. If it was not,
ps->backing_hd->file will probably be NULL and qemu will crash.
Fix this by moving pstrcpy after checking whether bdrv_open() succeeded.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Amos Kong <kongjianjun@gmail.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This adds a test case for Multiboot memory map in the tests/multiboot
directory, where future i386 test kernels can be dropped. Because this
requires an x86 build host and an installed 32 bit libgcc, the test is
not part of a regular 'make check'.
The reference output for the test is verified against test runs of the
same multiboot kernel booted by some GRUB 0.97.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This fixes a regression introduced by commit e3127ae0c, which kept the
allocation size of the bounce buffer limited to one page in order to
avoid unbounded allocations (as explained in the commit message of
6d16c2f88), but broke the reporting of the shortened bounce buffer to
the caller. The caller therefore assumes that the full requested size
was provided and causes memory corruption when writing beyond the end of
the actually allocated buffer.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Opening the qcow2 image with BDRV_O_NO_FLUSH prevents any flushes during
the image creation. This means that the image has not yet been flushed
to disk when qemu-img create exits. This flush is delayed until the next
operation on the image involving opening it without BDRV_O_NO_FLUSH and
closing (or directly flushing) it. For large images and/or images with a
small cluster size and preallocated metadata, this flush may take a
significant amount of time and may occur unexpectedly.
Reopening the image without BDRV_O_NO_FLUSH right before the end of
qcow2_create2() results in hoisting the potentially costly flush into
the image creation, which is expected to take some time (whereas
successive image operations may be not).
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Benoit Canet <benoit@irqsave.net>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Fix error: ‘inline’ is not at beginning of declaration
[-Werror=old-style-declaration]
Signed-off-by: Alex Bligh <alex@alex.org.uk>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
* it's -> its
* grammar fix in ui/vnc-enc-zywrle.h
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Reviewed-by: Don Koch <dkoch@verizon.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
The QMP wire format uses "", not '', around strings.
* docs/qapi-code-gen.txt: Fix typo.
Signed-off-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
This file is moved out from QMP/ to BUILD dir, change the ignore file
too.
Signed-off-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
compatiblity -> compatibility
continously -> continuously
existance -> existence
usefull -> useful
shoudl -> should
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
In some cases when building with parallelism (make -jN),
build fails because the directory where output files are
supposed to be does not exist. In particular, when make
decides to build virtfs-proxy-helper.1 before other files
in fsdev/, build will fail with the following error:
perl -Ww -- BUILDDIR/scripts/texi2pod.pl BUILDDIR/fsdev/virtfs-proxy-helper.texi fsdev/virtfs-proxy-helper.pod && pod2man --utf8 --section=1 --center=" " --release=" " fsdev/virtfs-proxy-helper.pod > fsdev/virtfs-proxy-helper.1
opening "fsdev/virtfs-proxy-helper.pod": No such file or directory
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Instead of relying on cpu_model, obtain the device tree node label
per CPU. Use DeviceClass::fw_name as source.
Whenever DeviceClass::fw_name is unknown, default to "PowerPC,UNKNOWN".
As a consequence, spapr_fixup_cpu_dt() can operate on each CPU's fw_name,
obsoleting sPAPREnvironment::cpu_model, and spapr_create_fdt_skel() can
drop its cpu_model argument.
Signed-off-by: Prerna Saxena <prerna@linux.vnet.ibm.com>
Signed-off-by: Andreas Färber <afaerber@suse.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
Set the expected values for POWER7, POWER7+, POWER8 and POWER5+.
Note that POWER5+ and POWER7+ are intentionally lacking the '+', so the
lack of a POWER7P family constitutes no problem.
Signed-off-by: Andreas Färber <afaerber@suse.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
This patch add support for dumping guest memory using dump-guest-memory
monitor command.
Before patch:
(qemu) dump-guest-memory testcrash
this feature or command is not currently supported
(qemu)
After patch:
(qemu) dump-guest-memory testcrash
(qemu)
crash was able to read the file
crash> bt
PID: 0 TASK: c000000000c0d0d0 CPU: 0 COMMAND: "swapper/0"
R0: 0000000028000084 R1: c000000000cafa50 R2: c000000000cb05b0
R3: 0000000000000000 R4: c000000000bc4cb0 R5: 0000000000000000
R6: 001efe93b8000000 R7: 0000000000000000 R8: 0000000000000000
R9: b000000000001032 R10: 0000000000000001 R11: 0001eb2117e00d55
....
...
NOTE: Currently crash tools doesn't look at ELF notes in the dump on ppc64.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: Alexander Graf <agraf@suse.de>
We should check for error with s->note_size
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: Alexander Graf <agraf@suse.de>
Instead of opencoding 64 use MAX_SLB_ENTRIES. We don't update the kernel
header here.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: Alexander Graf <agraf@suse.de>
When we translate the virtual address to physical check for error.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: Alexander Graf <agraf@suse.de>
Without this, a value of rb=0 and rs=0 results in replacing the 0th
index. This can be observed when using gdb remote debugging support.
(gdb) x/10i do_fork
0xc000000000085330 <do_fork>: Cannot access memory at address 0xc000000000085330
(gdb)
This is because when we do the slb sync via kvm_cpu_synchronize_state,
we overwrite the slb entry (0th entry) for 0xc000000000085330
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: Alexander Graf <agraf@suse.de>
This enables IRQFD for LSI (level triggered INTx interrupts) by adding
a spapr_route_intx_pin_to_irq() callback to the sPAPR PCI host bus. This
callback is called to know the global interrupt number to link resampling fd
with IRQFD's fd in KVM.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Alexander Graf <agraf@suse.de>
This enables IRQFD support for sPAPR. The feature decreases the latency
of interrupt handling.
To enable IRQFD for MSI, this sets kvm_gsi_direct_mapping to true which
enables direct MSI mapping.
To enable IRQFD for LSI (level triggered INTx interrupts), a PCI host bus
callback is required. The patch for that is coming next.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Alexander Graf <agraf@suse.de>
This implements H_XIRR_X hypercall in addition to H_XIRR as
it is mandatory for PAPR+ and there is no way for the guest to
detect whether it is supported or not so just add it.
As the Partition Adjunct Option is not supported at the moment,
the CPPR parameter of the hypercall is ignored.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Alexander Graf <agraf@suse.de>
This adds support for the H_IPOLL hypercall which the guest
uses to poll for a pending interrupt. This hypercall is
mandatory for PAPR+ and there is no way for the guest to
detect whether it is supported or not so just add it.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Acked-by: Alexander Graf <agraf@suse.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
Recent (host) kernels support emulating the PAPR defined "XICS" interrupt
controller system within KVM. This patch allows qemu to initialize and
configure the in-kernel XICS, and keep its state in sync with qemu's XICS
state as necessary.
This should give considerable performance improvements. e.g. on a simple
IPI ping-pong test between hardware threads, using qemu XICS gives us
around 5,000 irqs/second, whereas the in-kernel XICS gives us around
70,000 irqs/s on the same hardware configuration.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
[Mike Qiu <qiudayu@linux.vnet.ibm.com>: fixed mistype which caused ics_set_kvm_state() to fail]
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: Alexander Graf <agraf@suse.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
This adds a cpu_setup callback to the XICS device class (as XICS-KVM
will do it different), xics_cpu_setup() will call it if it is set.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Alexander Graf <agraf@suse.de>
The upcoming XICS-KVM support will use bits of emulated XICS code.
So this introduces new level of hierarchy - "xics-common" class. Both
emulated XICS and XICS-KVM will inherit from it and override class
callbacks when required.
The new "xics-common" class implements:
1. replaces static "nr_irqs" and "nr_servers" properties with
the dynamic ones and adds callbacks to be executed when properties
are set.
2. xics_cpu_setup() callback renamed to xics_common_cpu_setup() as
it is a common part for both XICS'es
3. xics_reset() renamed to xics_common_reset() for the same reason.
The emulated XICS changes:
1. the part of xics_realize() which creates ICPs is moved to
the "nr_servers" property callback as realize() is too late to
create/initialize devices and instance_init() is too early to create
devices as the number of child devices comes via the "nr_servers"
property.
2. added ics_initfn() which does a little part of what xics_realize() did.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: Alexander Graf <agraf@suse.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
This adds missing const specifiers to ICS and ICP TypeInfo's.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: Andreas Färber <afaerber@suse.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
This fixes XICS according new QOM rules.
This converts ICS's init() callbacks to realize().
This converts legacy qdev_init_nofail() to property_set(realized).
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: Andreas Färber <afaerber@suse.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
The upcoming support of in-kernel XICS will redefine migration callbacks
for both ICS and ICP so classes and callback pointers are added.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Alexander Graf <agraf@suse.de>
This replaces old-style fprintf with new style error_report.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: Andreas Färber <afaerber@suse.de>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Alexander Graf <agraf@suse.de>
This moves the xics_cpu_setup() call after kvmppc_set_papr()
in order to get VCPUs initialized as this is required by upcoming
XICS-KVM.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Alexander Graf <agraf@suse.de>
This simple change makes following patches nicer.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Alexander Graf <agraf@suse.de>
Recent PowerKVM allows the kernel to intercept some RTAS calls from the
guest directly. This is used to implement the more efficient in-kernel
XICS for example. qemu is still responsible for assigning the RTAS token
numbers however, and needs to tell the kernel which RTAS function name is
assigned to a given token value. This patch adds a convenience wrapper for
the KVM_PPC_RTAS_DEFINE_TOKEN ioctl() which is used for this purpose.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Alexander Graf <agraf@suse.de>
On the real hardware, RTAS is called in real mode and therefore
top 4 bits of the address passed in the call are ignored.
So does the patch.
This converts h_rtas() to use existing rtas_ld() handlers.
This fixed rtas_ld()/rtas_st() to ignore top 4 bits.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Alexander Graf <agraf@suse.de>
PAPR+ says that no "ibm,purr" tells the guest that H_PURR is not
supported. However some guests still try calling H_PURR on POWER7 unless
the property is present and equal to 0. This adds the property for CPUs
supporting the PURR special register.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Alexander Graf <agraf@suse.de>
At the moment the size of the buffer is set to 64K which is
enough for approximately 150 VCPUs which is not the limit.
This increases the buffer up to 256K which allows having
a tree for approximately 600 VCPUs which is way beyond the real
number we need.
As only the real size of the tree is copied to the guest, there
will be no impact on existing configurations.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Alexander Graf <agraf@suse.de>
Commit 2345f1c01 was supposed to render L2CR writes into noops. Instead,
it made them illegal instruction traps which apparently didn't confuse
XNU, but can easily confuse other OSs.
Fix it up by actually doing nothing when we write to L2CR.
Reported-by: Julio Guerra <guerr@julio.in>
Signed-off-by: Alexander Graf <agraf@suse.de>
Tested-by: Julio Guerra <guerr@julio.in>
MORITA Kazutaka