Peter Maydell [Tue, 27 Oct 2015 16:17:55 +0000 (16:17 +0000)]
Merge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20151027-1' into staging
target-arm queue:
* more EL2 preparation: handling for stage 2 translations
* standardize debug macros in i.MX devices
* improve error message in a corner case for virt board
* disable live migration of KVM GIC if the kernel can't handle it
* add SPSR_(ABT|UND|IRQ|FIQ) registers
* handle non-executable page-straddling Thumb instructions
* fix a "no 64-bit EL2" assumption in arm_excp_unmasked()
Introduce ARMMMUFaultInfo to propagate MMU Fault information
across the MMU translation code path. This is in preparation for
adding Stage-2 translation.
Andrew Jones [Tue, 27 Oct 2015 12:00:50 +0000 (12:00 +0000)]
hw/arm/virt: don't use a15memmap directly
We should always go through VirtBoardInfo when we need the memmap.
To avoid using a15memmap directly, in this case, we need to defer
the max-cpus check from class init time to instance init time. In
class init we now use MAX_CPUMASK_BITS for max_cpus initialization,
which is the maximum QEMU supports, and also, incidentally, the
maximum KVM/gicv3 currently supports. Also, a nice side-effect of
delaying the max-cpus check is that we now get more appropriate
error messages for gicv2 machines that try to configure more than
123 cpus. Before this patch it would complain that the requested
number of cpus was greater than 123, but for gicv2 configs, it
should complain that the number is greater than 8.
Pavel Fedin [Tue, 27 Oct 2015 12:00:50 +0000 (12:00 +0000)]
arm_gic_kvm: Disable live migration if not supported
Currently, if the kernel does not have live migration API, the migration
will still be attempted, but vGIC save/restore functions will just not do
anything. This will result in a broken machine state.
This patch fixes the problem by adding migration blocker if kernel API is
not supported.
Peter Maydell [Tue, 27 Oct 2015 12:00:50 +0000 (12:00 +0000)]
target-arm/translate.c: Handle non-executable page-straddling Thumb insns
When the memory we're trying to translate code from is not executable we have
to turn this into a guest fault. In order to report the correct PC for this
fault, and to make sure it is not reported until after any other possible
faults for instructions earlier in execution, we must terminate TBs at
the end of a page, in case the next instruction is in a non-executable page.
This is simple for T16, A32 and A64 instructions, which are always aligned
to their size. However T32 instructions may be 32-bits but only 16-aligned,
so they can straddle a page boundary.
Correct the condition that checks whether the next instruction will touch
the following page, to ensure that if we're 2 bytes before the boundary
and this insn is T32 then we end the TB.
Peter Maydell [Tue, 27 Oct 2015 12:00:50 +0000 (12:00 +0000)]
target-arm: Fix "no 64-bit EL2" assumption in arm_excp_unmasked()
The code in arm_excp_unmasked() suppresses the ability of PSTATE.AIF
to mask exceptions from a lower EL targeting EL2 or EL3 if the
CPU is 64-bit. This is correct for a target of EL3, but not correct
for targeting EL2. Further, we go to some effort to calculate
scr and hcr values which are not used at all for the 64-bit CPU
case.
Rearrange the code to correctly implement the 64-bit CPU logic
and keep the hcr/scr calculations in the 32-bit CPU codepath.
Peter Maydell [Tue, 27 Oct 2015 10:10:46 +0000 (10:10 +0000)]
Merge remote-tracking branch 'remotes/jasowang/tags/net-pull-request' into staging
# gpg: Signature made Tue 27 Oct 2015 05:47:28 GMT using RSA key ID 398D6211
# gpg: Good signature from "Jason Wang (Jason Wang on RedHat) <[email protected]>"
# gpg: WARNING: This key is not certified with sufficiently trusted signatures!
# gpg: It is not certain that the signature belongs to the owner.
# Primary key fingerprint: 215D 46F4 8246 689E C77F 3562 EF04 965B 398D 6211
* remotes/jasowang/tags/net-pull-request:
net: free the string returned by object_get_canonical_path_component
net: make iov_to_buf take right size argument in nc_sendv_compat()
net: Remove duplicate data from query-rx-filter on multiqueue net devices
vmxnet3: Do not fill stats if device is inactive
options: Add documentation for filter-dump
net/dump: Provide the dumping facility as a net-filter
net/dump: Separate the NetClientState from the DumpState
net/dump: Rework net-dump init functions
net/dump: Add support for receive_iov function
net: cadence_gem: Set initial MAC address
net: Remove duplicate data from query-rx-filter on multiqueue net devices
When responding to a query-rx-filter command on a multiqueue
netdev, qemu reports the data for each queue. The data, however,
is not per-queue, but per device and the same data is reported
multiple times. This causes confusion and may also cause extra
unnecessary processing when looking at the data.
Commit 638fb14169 (net: Make qmp_query_rx_filter() with name argument
more obvious) partially addresses this issue, by limiting the output
when the name is specified. However, when the name is not specified,
the issue still persists.
Thomas Huth [Tue, 13 Oct 2015 10:40:00 +0000 (12:40 +0200)]
net/dump: Separate the NetClientState from the DumpState
With the upcoming dumping-via-netfilter patch, the DumpState
should not be related to NetClientState anymore, so move the
related information to a new struct called DumpNetClient.
Thomas Huth [Tue, 13 Oct 2015 10:39:59 +0000 (12:39 +0200)]
net/dump: Rework net-dump init functions
Move the creation of the dump client from net_dump_init() into
net_init_dump(), so we can later use the former function for
dump via netfilter, too. Also rename net_dump_init() to
net_dump_state_init() to make it easier distinguishable from
net_init_dump().
Thomas Huth [Tue, 13 Oct 2015 10:39:58 +0000 (12:39 +0200)]
net/dump: Add support for receive_iov function
Adding a proper receive_iov function to the net dump module.
This will make it easier to support the dump filter feature for
the -netdev option in later patches.
Eduardo Habkost [Wed, 21 Oct 2015 15:46:50 +0000 (13:46 -0200)]
xen-platform: Replace assert() with appropriate error reporting
Commit dbb7405d8caad0814ceddd568cb49f163a847561 made it possible to
trigger an assert using "-device xen-platform". Replace it with
appropriate error reporting.
Peter Maydell [Mon, 26 Oct 2015 11:32:20 +0000 (11:32 +0000)]
Merge remote-tracking branch 'remotes/elmarco/tags/ivshmem-pull-request' into staging
ivshmem series
# gpg: Signature made Mon 26 Oct 2015 09:27:46 GMT using RSA key ID 75969CE5
# gpg: Good signature from "Marc-André Lureau <[email protected]>"
# gpg: aka "Marc-André Lureau <[email protected]>"
# gpg: WARNING: This key is not certified with sufficiently trusted signatures!
# gpg: It is not certain that the signature belongs to the owner.
# Primary key fingerprint: 87A9 BD93 3F87 C606 D276 F62D DAE8 E109 7596 9CE5
* remotes/elmarco/tags/ivshmem-pull-request: (51 commits)
doc: document ivshmem & hugepages
ivshmem: use little-endian int64_t for the protocol
ivshmem: use kvm irqfd for msi notifications
ivshmem: rename MSI eventfd_table
ivshmem: remove EventfdEntry.vector
ivshmem: add hostmem backend
ivshmem: use qemu_strtosz()
ivshmem: do not keep shm_fd open
tests: add ivshmem qtest
qtest: add qtest_add_abrt_handler()
msix: implement pba write (but read-only)
contrib: remove unnecessary strdup()
ivshmem: add check on protocol version in QEMU
docs: update ivshmem device spec
ivshmem-server: fix hugetlbfs support
ivshmem-server: use a uint16 for client ID
ivshmem-client: check the number of vectors
contrib: add ivshmem client and server
util: const event_notifier_get_fd() argument
ivshmem: reset mask on device reset
...
Lan Tianyu [Sun, 11 Oct 2015 15:19:24 +0000 (23:19 +0800)]
Qemu/Xen: Fix early freeing MSIX MMIO memory region
msix->mmio is added to XenPCIPassthroughState's object as property.
object_finalize_child_property is called for XenPCIPassthroughState's
object, which calls object_property_del_all, which is going to try to
delete msix->mmio. object_finalize_child_property() will access
msix->mmio's obj. But the whole msix struct has already been freed
by xen_pt_msix_delete. This will cause segment fault when msix->mmio
has been overwritten.
ivshmem: use little-endian int64_t for the protocol
The current ivshmem protocol uses 'long' for integers. But the
sizeof(long) depends on the host and the endianess is not defined, which
may cause portability troubles.
Instead, switch to using little-endian int64_t. This breaks the
protocol, except on x64 little-endian host where this change
should be compatible.
Adds 4 ivshmemtests:
- single qemu instance and basic IO
- pair of instances, check memory sharing
- pair of instances with server, and MSIX
- hot plug/unplug
A temporary shm is created as well as a directory to place server
socket, both should be clear on exit and abort.
David Marchand [Tue, 16 Jun 2015 15:43:34 +0000 (17:43 +0200)]
ivshmem: add check on protocol version in QEMU
Send a protocol version as the first message from server, clients must
close communication if they don't support this protocol version. Older
QEMUs should be fine with this change in the protocol since they
overrides their own vm_id on reception of an id associated to no
eventfd.
David Marchand [Mon, 8 Sep 2014 09:17:49 +0000 (11:17 +0200)]
docs: update ivshmem device spec
Add some notes on the parts needed to use ivshmem devices: more specifically,
explain the purpose of an ivshmem server and the basic concept to use the
ivshmem devices in guests.
Move some parts of the documentation and re-organise it.
As pointed out on the ML by Andrew Jones, glibc no longer permits
creating POSIX shm on hugetlbfs directly. When given a hugetlbfs path,
create a shareable file there.
David Marchand [Mon, 8 Sep 2014 09:17:48 +0000 (11:17 +0200)]
contrib: add ivshmem client and server
When using ivshmem devices, notifications between guests can be sent as
interrupts using a ivshmem-server (typical use described in documentation).
The client is provided as a debug tool.
Signed-off-by: Olivier Matz <[email protected]> Signed-off-by: David Marchand <[email protected]>
[fix a valgrind warning, option and server_close() segvs, extra server
headers includes, getopt() return type, out-of-tree build, use qemu
event_notifier instead of eventfd, fix x86/osx warnings - Marc-André] Signed-off-by: Marc-André Lureau <[email protected]>
The terms 'guest' and 'peer' are used sometime interchangeably which may
be confusing. Instead, use 'peer' for the remote instances of ivshmem
clients, and 'guest' for the local VM.
The test whether the chardev is an AF_UNIX socket rejects
"-chardev socket,id=chr0,path=/tmp/foo,server,nowait -device
ivshmem,chardev=chr0", but fails to explain why.
Use an explicit error on why a chardev may be rejected.
max_peer isn't really useful, it tracks the maximum received VM id, but
that quickly matches nb_peers, the size of the peers array. Since VM
come and go, there might be sparse peers so it doesn't help much in
general to have this value around.
Use the latest qemu device modeling API, in particular, convert to
realize to fix the error handling; right now a botched device_add
ivhsmem command kills the VM.
If a chardev is allowed to be created outside of QMP, then it must be
also possible to free it. This is useful for ivshmem that creates
chardev anonymously and must be able to free them.
Peter Maydell [Fri, 23 Oct 2015 17:14:42 +0000 (18:14 +0100)]
Merge remote-tracking branch 'remotes/kevin/tags/for-upstream' into staging
Block layer patches
# gpg: Signature made Fri 23 Oct 2015 17:59:56 BST using RSA key ID C88F2FD6
# gpg: Good signature from "Kevin Wolf <[email protected]>"
* remotes/kevin/tags/for-upstream: (37 commits)
tests: Add test case for aio_disable_external
block: Add "drained begin/end" for internal snapshot
block: Add "drained begin/end" for transactional blockdev-backup
block: Add "drained begin/end" for transactional backup
block: Add "drained begin/end" for transactional external snapshot
block: Introduce "drained begin/end" API
aio: introduce aio_{disable,enable}_external
dataplane: Mark host notifiers' client type as "external"
nbd: Mark fd handlers client type as "external"
aio: Add "is_external" flag for event handlers
throttle: Remove throttle_group_lock/unlock()
blockdev: Allow more options for BB-less BDS tree
blockdev: Pull out blockdev option extraction
blockdev: Do not create BDS for empty drive
block: Prepare for NULL BDS
block: Add blk_insert_bs()
block: Prepare remaining BB functions for NULL BDS
block: Fail requests to empty BlockBackend
block: Make some BB functions fall back to BBRS
block: Add BlockBackendRootState
...
Fam Zheng [Fri, 23 Oct 2015 03:08:13 +0000 (11:08 +0800)]
block: Add "drained begin/end" for internal snapshot
This ensures the atomicity of the transaction by avoiding processing of
external requests such as those from ioeventfd.
state->bs is assigned right after bdrv_drained_begin. Because it was
used as the flag for deletion or not in abort, now we need a separate
flag - InternalSnapshotState.created.
Fam Zheng [Fri, 23 Oct 2015 03:08:11 +0000 (11:08 +0800)]
block: Add "drained begin/end" for transactional backup
This ensures the atomicity of the transaction by avoiding processing of
external requests such as those from ioeventfd.
Move the assignment to state->bs up right after bdrv_drained_begin, so
that we can use it in the clean callback. The abort callback will still
check bs->job and state->job, so it's OK.
projects / qemu.git / log