Paolo Bonzini [Fri, 3 Jul 2015 22:24:51 +0000 (00:24 +0200)]
exec: skip MMIO regions correctly in cpu_physical_memory_write_rom_internal
Loading the BIOS in the mac99 machine is interesting, because there is a
PROM in the middle of the BIOS region (from 16K to 32K). Before memory
region accesses were clamped, when QEMU was asked to load a BIOS from
0xfff00000 to 0xffffffff it would put even those 16K from the BIOS file
into the region. This is weird because those 16K were not actually
visible between 0xfff04000 and 0xfff07fff. However, it worked.
After clamping was added, this also worked. In this case, the
cpu_physical_memory_write_rom_internal function split the write in
three parts: the first 16K were copied, the PROM area (second 16K) were
ignored, then the rest was copied.
Problems then started with commit 965eb2f (exec: do not clamp accesses
to MMIO regions, 2015-06-17). Clamping accesses is not done for MMIO
regions because they can overlap wildly, and MMIO registers can be
expected to perform full-width accesses based only on their address
(with no respect for adjacent registers that could decode to completely
different MemoryRegions). However, this lack of clamping also applied
to the PROM area! cpu_physical_memory_write_rom_internal thus failed
to copy the third range above, i.e. only copied the first 16K of the BIOS.
In effect, address_space_translate is expecting _something else_ to do
the clamping for MMIO regions if the incoming length is large. This
"something else" is memory_access_size in the case of address_space_rw,
so use the same logic in cpu_physical_memory_write_rom_internal.
Reported-by: Alexander Graf <agraf@redhat.com> Reviewed-by: Laurent Vivier <lvivier@redhat.com> Tested-by: Laurent Vivier <lvivier@redhat.com> Fixes: 965eb2f Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Peter Maydell [Fri, 3 Jul 2015 14:18:24 +0000 (15:18 +0100)]
Stop including qemu-common.h in memory.h
Including qemu-common.h from other header files is generally a bad
idea, because it means it's very easy to end up with a circular
dependency. For instance, if we wanted to include memory.h from
qom/cpu.h we'd end up with this loop:
memory.h -> qemu-common.h -> cpu.h -> cpu-qom.h -> qom/cpu.h -> memory.h
Remove the include from memory.h. This requires us to fix up a few
other files which were inadvertently getting declarations indirectly
through memory.h.
The biggest change is splitting the fprintf_function typedef out
into its own header so other headers can get at it without having
to include qemu-common.h.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <1435933104-15216-1-git-send-email-peter.maydell@linaro.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Paolo Bonzini [Thu, 18 Jun 2015 16:47:26 +0000 (18:47 +0200)]
kvm: Switch to unlocked MMIO
Do not take the BQL before dispatching MMIO requests of KVM VCPUs.
Instead, address_space_rw will do it if necessary. This enables completely
BQL-free MMIO handling in KVM mode for upcoming devices with fine-grained
locking.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <1434646046-27150-10-git-send-email-pbonzini@redhat.com>
Jan Kiszka [Thu, 18 Jun 2015 16:47:24 +0000 (18:47 +0200)]
kvm: Switch to unlocked PIO
Do not take the BQL before dispatching PIO requests of KVM VCPUs.
Instead, address_space_rw will do it if necessary. This enables
completely BQL-free PIO handling in KVM mode for upcoming devices with
fine-grained locking.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <1434646046-27150-8-git-send-email-pbonzini@redhat.com>
Jan Kiszka [Thu, 18 Jun 2015 16:47:23 +0000 (18:47 +0200)]
kvm: First step to push iothread lock out of inner run loop
This opens the path to get rid of the iothread lock on vmexits in KVM
mode. On x86, the in-kernel irqchips has to be used because we otherwise
need to synchronize APIC and other per-cpu state accesses that could be
changed concurrently.
Regarding pre/post-run callbacks, s390x and ARM should be fine without
specific locking as the callbacks are empty. MIPS and POWER require
locking for the pre-run callback.
For the handle_exit callback, it is non-empty in x86, POWER and s390.
Some POWER cases could do without the locking, but it is left in
place for now.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <1434646046-27150-7-git-send-email-pbonzini@redhat.com>
Jan Kiszka [Thu, 18 Jun 2015 16:47:22 +0000 (18:47 +0200)]
memory: let address_space_rw/ld*/st* run outside the BQL
The MMIO case is further broken up in two cases: if the caller does not
hold the BQL on invocation, the unlocked one takes or avoids BQL depending
on the locking strategy of the target memory region and its coalesced
MMIO handling. In this case, the caller should not hold _any_ lock
(a friendly suggestion which is disregarded by virtio-scsi-dataplane).
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com> Cc: Frederic Konrad <fred.konrad@greensocs.com>
Message-Id: <1434646046-27150-6-git-send-email-pbonzini@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Paolo Bonzini [Thu, 18 Jun 2015 16:47:21 +0000 (18:47 +0200)]
exec: pull qemu_flush_coalesced_mmio_buffer() into address_space_rw/ld*/st*
As memory_region_read/write_accessor will now be run also without BQL held,
we need to move coalesced MMIO flushing earlier in the dispatch process.
Cc: Frederic Konrad <fred.konrad@greensocs.com>
Message-Id: <1434646046-27150-5-git-send-email-pbonzini@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Jan Kiszka [Thu, 18 Jun 2015 16:47:20 +0000 (18:47 +0200)]
memory: Add global-locking property to memory regions
This introduces the memory region property "global_locking". It is true
by default. By setting it to false, a device model can request BQL-free
dispatching of region accesses to its r/w handlers. The actual BQL
break-up will be provided in a separate patch.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com> Cc: Frederic Konrad <fred.konrad@greensocs.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <1434646046-27150-4-git-send-email-pbonzini@redhat.com>
Paolo Bonzini [Thu, 18 Jun 2015 16:47:19 +0000 (18:47 +0200)]
main-loop: introduce qemu_mutex_iothread_locked
This function will be used to avoid recursive locking of the iothread lock
whenever address_space_rw/ld*/st* are called with the BQL held, which is
almost always the case.
Tracking whether the iothread is owned is very cheap (just use a TLS
variable) but requires some care because now the lock must always be
taken with qemu_mutex_lock_iothread(). Previously this wasn't the case.
Outside TCG mode this is not a problem. In TCG mode, we need to be
careful and avoid the "prod out of compiled code" step if already
in a VCPU thread. This is easily done with a check on current_cpu,
i.e. qemu_in_vcpu_thread().
Hopefully, multithreaded TCG will get rid of the whole logic to kick
VCPUs whenever an I/O event occurs!
Cc: Frederic Konrad <fred.konrad@greensocs.com>
Message-Id: <1434646046-27150-3-git-send-email-pbonzini@redhat.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Paolo Bonzini [Thu, 18 Jun 2015 16:47:18 +0000 (18:47 +0200)]
main-loop: use qemu_mutex_lock_iothread consistently
The next patch will require the BQL to be always taken with
qemu_mutex_lock_iothread(), while right now this isn't the case.
Outside TCG mode this is not a problem. In TCG mode, we need to be
careful and avoid the "prod out of compiled code" step if already
in a VCPU thread. This is easily done with a check on current_cpu,
i.e. qemu_in_vcpu_thread().
Hopefully, multithreaded TCG will get rid of the whole logic to kick
VCPUs whenever an I/O event occurs!
Cc: Frederic Konrad <fred.konrad@greensocs.com>
Message-Id: <1434646046-27150-2-git-send-email-pbonzini@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
After analysis and verification, we can confirm it's irq-balance
daemon(in guest) leads to the assertion failure. Start a 8 core guest with
two disks, execute the following scripts will reproduce the BUG quickly:
vda_irq_num=25
vdb_irq_num=27
while [ 1 ]
do
for irq in {1,2,4,8,10,20,40,80}
do
echo $irq > /proc/irq/$vda_irq_num/smp_affinity
echo $irq > /proc/irq/$vdb_irq_num/smp_affinity
dd if=/dev/vda of=/dev/zero bs=4K count=100 iflag=direct
dd if=/dev/vdb of=/dev/zero bs=4K count=100 iflag=direct
done
done
========================================================================
QEMU setup static irq route entries in kvm_pc_setup_irq_routing(), PIC and
IOAPIC share the first 15 GSI numbers, take up 23 GSI numbers, but take up
38 irq route entries. When change irq smp_affinity in guest, a dynamic route
entry may be setup, the current logic is: if allocate GSI number succeeds,
a new route entry can be added. The available dynamic GSI numbers is
1021(KVM_MAX_IRQ_ROUTES-23), but available irq route entries is only
986(KVM_MAX_IRQ_ROUTES-38), GSI numbers greater than route entries.
irq-balance's behavior will eventually leads to total irq route entries
exceed KVM_MAX_IRQ_ROUTES, ioctl(KVM_SET_GSI_ROUTING) fail and
kvm_irqchip_commit_routes() trigger assertion failure.
This patch fix the BUG.
Signed-off-by: Wenshuang Ma <kevinnma@tencent.com> Cc: qemu-stable@nongnu.org Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
These are not Architecture specific in any way so move them out of
cpu-defs.h. tb-hash.h is an appropriate place as a leading user and
their strong relationship to TB hashing and caching.
Reviewed-by: Richard Henderson <rth@redhat.com> Signed-off-by: Peter Crosthwaite <crosthwaite.peter@gmail.com>
Message-Id: <43ceca65a3fa240efac49aa0bf604ad0442e1710.1433052532.git.crosthwaite.peter@gmail.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This is one of very few things in exec-all with a genuine CPU
architecture dependency. Move these hashing helpers to a new
header to trim exec-all.h down to a near architecture-agnostic
header.
The defs are only used by cpu-exec and translate-all which are both
arch-obj's so the new tb-hash.h has no core code usage.
Reviewed-by: Richard Henderson <rth@redhat.com> Signed-off-by: Peter Crosthwaite <crosthwaite.peter@gmail.com>
Message-Id: <9d048b96f7cfa64a4d9c0b88e0dd2877fac51d41.1433052532.git.crosthwaite.peter@gmail.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
include/exec: Move standard exceptions to cpu-all.h
These exception indicies are generic and don't have any reliance on the
per-arch cpu.h defs. Move them to cpu-all.h so they can be used by core
code that does not have access to cpu-defs.h.
Reviewed-by: Richard Henderson <rth@redhat.com> Signed-off-by: Peter Crosthwaite <crosthwaite.peter@gmail.com>
Message-Id: <dbebd3062c7cd4332240891a3564e73f374ddfcd.1433052532.git.crosthwaite.peter@gmail.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The usages of this define are pure TCG and there is no architecture
specific variation of the value. Localise it to the TCG engine to
remove another architecture agnostic piece from cpu-defs.h.
This follows on from a28177820a868eafda8fab007561cc19f41941f4 where
temp_buf was moved out of the CPU_COMMON obsoleting the need for
the super early definition.
Reviewed-by: Richard Henderson <rth@twiddle.net> Signed-off-by: Peter Crosthwaite <crosthwaite.peter@gmail.com>
Message-Id: <498e8e5325c1a1aff79e5bcfc28cb760ef6b214e.1433052532.git.crosthwaite.peter@gmail.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This makes it more consistent with all other core code files, which
either just rely on qemu-common.h inclusion or precede cpu.h with
qemu-common.h.
cpu-all.h should not be included in addition to cpu.h. Remove it.
Signed-off-by: Peter Crosthwaite <crosthwaite.peter@gmail.com>
Message-Id: <1433714349-7262-1-git-send-email-crosthwaite.peter@gmail.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Peter Maydell [Thu, 25 Jun 2015 13:03:55 +0000 (14:03 +0100)]
Merge remote-tracking branch 'remotes/stefanha/tags/net-pull-request' into staging
# gpg: Signature made Wed Jun 24 16:37:23 2015 BST using RSA key ID 81AB73C8
# gpg: Good signature from "Stefan Hajnoczi <stefanha@redhat.com>"
# gpg: aka "Stefan Hajnoczi <stefanha@gmail.com>"
* remotes/stefanha/tags/net-pull-request:
net: simplify net_client_init1()
net: drop if expression that is always true
net: raise an error if -net type is invalid
net: replace net_client_init1() netdev whitelist with blacklist
net: add missing "netmap" to host_net_devices[]
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Peter Maydell [Thu, 25 Jun 2015 10:19:46 +0000 (11:19 +0100)]
Merge remote-tracking branch 'remotes/stefanha/tags/block-pull-request' into staging
# gpg: Signature made Wed Jun 24 16:27:53 2015 BST using RSA key ID 81AB73C8
# gpg: Good signature from "Stefan Hajnoczi <stefanha@redhat.com>"
# gpg: aka "Stefan Hajnoczi <stefanha@gmail.com>"
* remotes/stefanha/tags/block-pull-request:
virito-blk: drop duplicate check
qemu-iotests: fix 051.out after qdev error message change
iov: don't touch iov in iov_send_recv()
raw-posix: Introduce hdev_is_sg()
raw-posix: Use DPRINTF for DEBUG_FLOPPY
raw-posix: DPRINTF instead of DEBUG_BLOCK_PRINT
Fix migration in case of scsi-generic
block: Use bdrv_is_sg() everywhere
nvme: Fix memleak in nvme_dma_read_prp
vvfat: add a label option
util/hbitmap: Add an API to reset all set bits in hbitmap
virtio-blk: Use blk_drain() to drain IO requests
block-backend: Introduce blk_drain()
throttle: Check current timers before updating any_timer_armed[]
block: Let bdrv_drain_all() to call aio_poll() for each AioContext
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Stefan Hajnoczi [Wed, 27 May 2015 16:16:52 +0000 (17:16 +0100)]
net: simplify net_client_init1()
Drop the union and move the hubport creation into the !is_netdev case.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-id: 1432743412-15943-6-git-send-email-stefanha@redhat.com
Stefan Hajnoczi [Wed, 27 May 2015 16:16:51 +0000 (17:16 +0100)]
net: drop if expression that is always true
Both is_netdev and !is_netdev paths already check that
net_client_init_func[opts->kind] is non-NULL so there is no need for the
if statement.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-id: 1432743412-15943-5-git-send-email-stefanha@redhat.com
Stefan Hajnoczi [Wed, 27 May 2015 16:16:50 +0000 (17:16 +0100)]
net: raise an error if -net type is invalid
When a -net type is used that was not compiled into the binary there
should be an error message.
Note the special case for -net none, which is a no-op.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-id: 1432743412-15943-4-git-send-email-stefanha@redhat.com
Stefan Hajnoczi [Wed, 27 May 2015 16:16:49 +0000 (17:16 +0100)]
net: replace net_client_init1() netdev whitelist with blacklist
It's cumbersome to keep the whitelist up-to-date. New netdev backends
should most likely be allowed so a blacklist makes more sense than a
whitelist.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-id: 1432743412-15943-3-git-send-email-stefanha@redhat.com
Stefan Hajnoczi [Tue, 23 Jun 2015 14:56:09 +0000 (15:56 +0100)]
qemu-iotests: fix 051.out after qdev error message change
Commit f006cf7fa9a63ba8e4ccf57d46231ce594301727 ("qdev-monitor:
Propagate errors through qdev_device_add()") dropped a meaningless error
message. This change in output caused qemu-iotests 051 to fail:
QEMU_PROG: -device ide-drive,drive=disk: Device initialization failed.
-QEMU_PROG: -device ide-drive,drive=disk: Device 'ide-drive' could not be initialized
Update 051.out so the test passes again.
Cc: Markus Armbruster <armbru@redhat.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Fam Zheng <famz@redhat.com>
Message-id: 1435071369-30936-1-git-send-email-stefanha@redhat.com
util/qemu-sockets: improve ai_flag hints for ipv6 hosts
*) Do not use AI_ADDRCONFIG on listening sockets, because this flag
makes it impossible to explicitly listen on '127.0.0.1' if no global
ipv4 address is configured additionally, making this a very
uncomfortable option.
*) Add AI_V4MAPPED hint for connecting sockets.
If your system is globally only connected via ipv6 you often still want
to be able to use '127.0.0.1' and 'localhost' (even if localhost doesn't
also have an ipv6 entry).
For example, PVE - unless explicitly asking for insecure mode - uses
ipv4 loopback addresses with QEMU for live migrations tunneled over SSH.
These fail to start because AI_ADDRCONFIG makes getaddrinfo refuse to
work with '127.0.0.1'.
As for the AI_V4MAPPED flag: glibc uses it by default, and providing
non-0 flags removes it. I think it makes sense to use it.
I also want to point out that glibc explicitly sidesteps POSIX standards
when passing 0 as hints by then assuming both AI_V4MAPPED and
AI_ADDRCONFIG (the latter being a rather weird choice IMO), while
according to POSIX.1-2001 it should be assumed 0. (glibc considers its
choice an improvement.)
Since either AI_CANONNAME or AI_PASSIVE are passed in our cases, glibc's
default flags in turn are disabled again unless explicitly added, which
I do with this patch.
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com> Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Fam Zheng [Fri, 22 May 2015 05:35:07 +0000 (13:35 +0800)]
Makefile: Fix "make cscope TAGS"
Cscope and TAGS files work in source directory rather than the build
directory, also, don't ask users to run configure first, because they
may have an out of tree build.
Signed-off-by: Fam Zheng <famz@redhat.com> Reviewed-by: John Snow <jsnow@redhat.com> Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Recent commit 3751d7c "vl: allow full-blown QemuOpts syntax for
-global" overloaded its existing argument syntax DRIVER.PROP=VALUE
with QemuOpts syntax. Unambigious as long as no DRIVER contains '='.
Its documentation claims that "the two syntaxes are equivalent."
Improve it to spell out how exactly the old syntax gets desugared into
the new one.
Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Michael Tokarev [Wed, 17 Jun 2015 18:02:03 +0000 (21:02 +0300)]
libcacard: pkgconfig: tidy dependent libs
libcacard.pc file lists only one package in Requires
field, which is nss, while glib-2.0 is also a requiriment.
Furthermore, for libraries used internally by the library
(this is the way nss and glib are used by libcacard),
Requires.private shold be used instead of Requires.
Fix both issues.
This does not affect linking of qemu because it links
with objects from libcacard directly.
When loading migration fails due to a disagreement about
PCI config data we don't currently get any errors explaining
that was the cause of the problem or which byte in the config
data was at fault.
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com> Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Michael Tokarev [Wed, 3 Jun 2015 14:37:27 +0000 (17:37 +0300)]
remove libdecnumber/dpd/decimal128Local.h
Commit 72ac97cdfc added two equivalent versions of decimal128Local.h,
one in libdecnumber/dpd/ and another in include/libdecnumber/dpd/.
Being identical by the code, the two files however differs in the
licensing terms. The one in libdecnumber/dpd/ (which is being
removed by this patch) is licensed as GPL3.1 (plus gcc runtime
exception), which, as far as I know, is not compatible with GPL-2.
This file is not used (it is included from
include/libdecnumber/dpd/decimal128.h, so version in include/ is
used).
More, the version in include/ can also be removed, since none
of the 3 defines from that file are actually used by the code.
Even more, one of the defines from there, decimal128SetSign,
is redefined (to equivalent value) in libdecnumber/dpd/decimal128.c,
but again, never used.
What a mess...
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru> Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Peter Maydell [Tue, 23 Jun 2015 16:46:20 +0000 (17:46 +0100)]
Merge remote-tracking branch 'remotes/sstabellini/tags/xen-220615-3' into staging
xen-220615, more SOB lines
# gpg: Signature made Tue Jun 23 17:19:08 2015 BST using RSA key ID 70E1AE90
# gpg: Good signature from "Stefano Stabellini <stefano.stabellini@eu.citrix.com>"
* remotes/sstabellini/tags/xen-220615-3:
Revert "xen-hvm: increase maxmem before calling xc_domain_populate_physmap"
xen/pass-through: constify some static data
xen/pass-through: log errno values rather than function return ones
xen/pass-through: ROM BAR handling adjustments
xen/pass-through: fold host PCI command register writes
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
The original commit fixes a bug when assigning a large number of
devices which require option roms to a guest. (One known
configuration that needs extra memory is having more than 4 emulated
NICs assigned. Three or fewer NICs seems to work without this
functionality.)
However, by unilaterally increasing maxmem, it introduces two
problems.
First, now libxl's calculation of the required maxmem during migration
is broken -- any guest which exercised this functionality will fail on
migration. (Guests which have the default number of devices are not
affected.)
Secondly, it makes it impossible for a higher-level toolstack or
administer to predict how much memory a VM will actually use, making
it much more difficult to effectively use all of the memory on a
machine.
Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com> Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Jan Beulich [Fri, 5 Jun 2015 12:04:55 +0000 (13:04 +0100)]
xen/pass-through: constify some static data
This is done indirectly by adjusting two typedefs and helps emphasizing
that the respective tables aren't supposed to be modified at runtime
(as they may be shared between devices).
Signed-off-by: Jan Beulich <jbeulich@suse.com> Reviewed-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Jan Beulich [Mon, 8 Jun 2015 13:11:51 +0000 (14:11 +0100)]
xen/pass-through: ROM BAR handling adjustments
Expecting the ROM BAR to be written with an all ones value when sizing
the region is wrong - the low bit has another meaning (enable/disable)
and bits 1..10 are reserved. The PCI spec also mandates writing all
ones to just the address portion of the register.
Use suitable constants also for initializing the ROM BAR register field
description.
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com> Signed-off-by: Jan Beulich <jbeulich@suse.com>
The code introduced to address XSA-126 allows simplification of other
code in xen_pt_initfn(): All we need to do is update "cmd" suitably,
as it'll be written back to the host register near the end of the
function anyway.
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com> Signed-off-by: Jan Beulich <jbeulich@suse.com>
Until now, an SG device was identified only by checking if its path
started with "/dev/sg". Then, hdev_open() would set the bs->sg flag
accordingly. The patch relies on the actual properties of the device
instead of the specified file path.
To this end, test for an SG device (e.g. /dev/sg0) by ensuring that
all of the following holds:
- The specified file name corresponds to a character device
- The device supports the SG_GET_VERSION_NUM ioctl
- The device supports the SG_GET_SCSI_ID ioctl
Signed-off-by: Dimitris Aragiorgis <dimara@arrikto.com>
Message-id: 1435056300-14924-6-git-send-email-dimara@arrikto.com Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Building the QEMU tools fails if we #define DEBUG_BLOCK inside
block/raw-posix.c. Here instead of adding qemu-log.o in block-obj-y
so that DEBUG_BLOCK_PRINT can be used, we substitute the latter with
a simple DPRINTF() (that does not cause bit-rot).
Signed-off-by: Dimitris Aragiorgis <dimara@arrikto.com> Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 1435056300-14924-4-git-send-email-dimara@arrikto.com Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
During migration, QEMU uses fsync()/fdatasync() on the open file
descriptor for read-write block devices to flush data just before
stopping the VM.
However, fsync() on a scsi-generic device returns -EINVAL which
causes the migration to fail. This patch skips flushing data in case
of an SG device, since submitting SCSI commands directly via an SG
character device (e.g. /dev/sg0) bypasses the page cache completely,
anyway.
Note that fsync() not only flushes the page cache but also the disk
cache. The scsi-generic device never sends flushes, and for
migration it assumes that the same SCSI device is used by the
destination host, so it does not issue any SCSI SYNCHRONIZE CACHE
(10) command.
Finally, remove the bdrv_is_sg() test from iscsi_co_flush() since
this is now redundant (we flush the underlying protocol at the end
of bdrv_co_flush() which, with this patch, we never reach).
Signed-off-by: Dimitris Aragiorgis <dimara@arrikto.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 1435056300-14924-3-git-send-email-dimara@arrikto.com Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Lu Lina [Fri, 19 Jun 2015 06:27:34 +0000 (14:27 +0800)]
nvme: Fix memleak in nvme_dma_read_prp
Signed-off-by: Lu Lina <lina.lulina@huawei.com> Acked-by: Keith Busch <keith.busch@intel.com>
Message-id: 1434695254-69808-1-git-send-email-kathy.wangting@huawei.com Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Until now the vvfat volume label was hardcoded to be
"QEMU VVFAT", now you can pass a file.label=labelname option
to the -drive to change it.
The FAT structure defines the volume label to be limited to
11 bytes and is filled up spaces when shorter than that. The
trailing spaces however aren't exposed to the user by
operating systems.
[Added missing comment '#' characters in block-core.json to fix build
errors.
--Stefan]
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
Message-id: 1434706529-13895-2-git-send-email-w.bumiller@proxmox.com Reviewed-by: Eric Blake <eblake@redhat.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Each call of the virtio_blk_reset() function calls blk_drain_all(),
which works for all existing BlockDriverStates, while draining only
one is needed.
This patch replaces blk_drain_all() by blk_drain() in
virtio_blk_reset(). virtio_blk_data_plane_stop() should be called
after draining because it restores vblk->complete_request.
Cc: "Michael S. Tsirkin" <mst@redhat.com> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Cornelia Huck <cornelia.huck@de.ibm.com> Cc: Kevin Wolf <kwolf@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Alexander Yarygin <yarygin@linux.vnet.ibm.com>
Message-id: 1434537440-28236-3-git-send-email-yarygin@linux.vnet.ibm.com Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This patch introduces the blk_drain() function which allows to replace
blk_drain_all() when only one BlockDriverState needs to be drained.
Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Cornelia Huck <cornelia.huck@de.ibm.com> Cc: Kevin Wolf <kwolf@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Alexander Yarygin <yarygin@linux.vnet.ibm.com> Reviewed-by: Paolo Bonzini <pbonzini@redhat.com> Acked-by: Kevin Wolf <kwolf@redhat.com>
Message-id: 1434537440-28236-2-git-send-email-yarygin@linux.vnet.ibm.com Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Alberto Garcia [Mon, 15 Jun 2015 15:41:15 +0000 (18:41 +0300)]
throttle: Check current timers before updating any_timer_armed[]
Calling throttle_group_config() cancels all timers from a particular
BlockDriverState, so any_timer_armed[] should be updated accordingly.
However, with the current code it may happen that a timer is armed in
a different BlockDriverState from the same group, so any_timer_armed[]
would be set to false in a situation where there is still a timer
armed.
The consequence is that we might end up with two timers armed. This
should not have any noticeable impact however, since all accesses to
the ThrottleGroup are protected by a lock, and the situation would
become normal again shortly thereafter as soon as all timers have been
fired.
The correct way to solve this is to check that we're actually
cancelling a timer before updating any_timer_armed[].
Signed-off-by: Alberto Garcia <berto@igalia.com>
Message-id: 1434382875-3998-1-git-send-email-berto@igalia.com Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
block: Let bdrv_drain_all() to call aio_poll() for each AioContext
After the commit 9b536adc ("block: acquire AioContext in
bdrv_drain_all()") the aio_poll() function got called for every
BlockDriverState, in assumption that every device may have its own
AioContext. If we have thousands of disks attached, there are a lot of
BlockDriverStates but only a few AioContexts, leading to tons of
unnecessary aio_poll() calls.
This patch changes the bdrv_drain_all() function allowing it find shared
AioContexts and to call aio_poll() only for unique ones.
Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Cornelia Huck <cornelia.huck@de.ibm.com> Cc: Kevin Wolf <kwolf@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Alexander Yarygin <yarygin@linux.vnet.ibm.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Tested-by: Christian Borntraeger <borntraeger@de.ibm.com>
Message-id: 1433936297-7098-4-git-send-email-yarygin@linux.vnet.ibm.com Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Peter Maydell [Tue, 23 Jun 2015 09:38:00 +0000 (10:38 +0100)]
Merge remote-tracking branch 'remotes/armbru/tags/pull-monitor-2015-06-22' into staging
Monitor patches
# gpg: Signature made Mon Jun 22 18:56:18 2015 BST using RSA key ID EB918653
# gpg: Good signature from "Markus Armbruster <armbru@redhat.com>"
# gpg: aka "Markus Armbruster <armbru@pond.sub.org>"
* remotes/armbru/tags/pull-monitor-2015-06-22: (24 commits)
Include monitor/monitor.h exactly where needed
Include qapi/qmp/qerror.h exactly where needed
qerror: Move #include out of qerror.h
qerror: Finally unused, clean up
qmp: Wean off qerror_report()
tpm: Avoid qerror_report() outside QMP command handlers
qerror: Clean up QERR_ macros to expand into a single string
qerror: Eliminate QERR_DEVICE_NOT_FOUND
vl: Use error_report() for --display errors
vl: Avoid qerror_report() outside QMP command handlers
QemuOpts: Wean off qerror_report_err()
qdev-monitor: Propagate errors through qdev_device_add()
qdev-monitor: Propagate errors through set_property()
qdev-monitor: Convert qbus_find() to Error
qdev-monitor: Fix check for full bus
qdev-monitor: Stop error avalanche in qbus_find_recursive()
disas: Remove uses of CPU env
monitor: Split mon_get_cpu fn to remove ENV_GET_CPU
monitor: Fix failure path for "S" argument
monitor: Point to "help" command on syntax error
...
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Gerd Hoffmann [Thu, 18 Jun 2015 15:45:47 +0000 (17:45 +0200)]
virtio-input: move properties, use virtio_instance_init_common
Move properties from virtio-*-pci to virtio-*-device.
Also make better use of QOM and attach common properties
to the abstract parent classes (virtio-input-device and
virtio-input-pci-device).
Switch the hid device instance init functions over to use
virtio_instance_init_common, so we get the properties of the
virtio device aliased properly to the virtio pci proxy.
int qmp_FOO(Monitor *mon, const QDict *params, QObject **ret_data);
doesn't provide for returning an Error object. Instead, the handler
is expected to stash it in the monitor with qerror_report().
When we rebased QMP on top of QAPI, we didn't change this interface.
Instead, commit 776574d introduced "middle mode" as a temporary aid
for converting existing QMP commands to QAPI one by one. More than
three years later, we're still using it.
it generates input marshallers conforming to the traditional QMP
command handler interface.
* It suppresses generation of code to register them with
qmp_register_command()
This permits giving them internal linkage.
As long as we need qmp-commands.hx, we can't use the registry behind
qmp_register_command(), so the latter has to stay for now.
The former has to go to get rid of qerror_report(). Changing all QMP
commands to fit the QAPI mold in one go was impractical back when we
started, but by now there are just a few stragglers left:
do_qmp_capabilities(), qmp_qom_set(), qmp_qom_get(), qmp_object_add(),
qmp_netdev_add(), do_device_add().
Switch middle mode to generate native input marshallers, and adapt the
stragglers. Simplifies both the monitor code and the stragglers.
Rename do_qmp_capabilities() to qmp_capabilities(), and
do_device_add() to qmp_device_add, because that's how QMP command
handlers are named today.
Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Luiz Capitulino <lcapitulino@redhat.com>
qerror_report() is a transitional interface to help with converting
existing monitor commands to QMP. It should not be used elsewhere.
Replace by error_report().
Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Luiz Capitulino <lcapitulino@redhat.com>
Error classes other than ERROR_CLASS_GENERIC_ERROR should not be used
in new code. Hiding them in QERR_ macros makes new uses hard to spot.
Fortunately, there's just one such macro left. Eliminate it with this
coccinelle semantic patch:
qerror_report() is a transitional interface to help with converting
existing monitor commands to QMP. It should not be used elsewhere.
Replace by error_report() in initial startup helpers parse_sandbox()
and parse_add_fd().
Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Luiz Capitulino <lcapitulino@redhat.com>
qerror_report_err() is a transitional interface to help with
converting existing monitor commands to QMP. It should not be used
elsewhere.
The only remaining user in qemu-option.c is qemu_opts_parse(). Is it
used in QMP context? If not, we can simply replace
qerror_report_err() by error_report_err().
The uses in qemu-img.c, qemu-io.c, qemu-nbd.c and under tests/ are
clearly not in QMP context.
The uses in vl.c aren't either, because the only QMP command handlers
there are qmp_query_status() and qmp_query_machines(), and they don't
call it.
Remaining uses:
* drive_def(): Command line -drive and such, HMP drive_add and pci_add
* hmp_chardev_add(): HMP chardev-add
* monitor_parse_command(): HMP core
* tmp_config_parse(): Command line -tpmdev
* net_host_device_add(): HMP host_net_add
* net_client_parse(): Command line -net and -netdev
* usb_net_init(): Command line -usbdevice, HMP usb_add
Propagate errors through qemu_opts_parse(). Create a convenience
function qemu_opts_parse_noisily() that passes errors to
error_report_err(). Switch all non-QMP users outside tests to it.
That leaves vnc_parse_func(). Propagate errors through it. Since I'm
touching it anyway, rename it to vnc_parse().
Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Luiz Capitulino <lcapitulino@redhat.com>
As usual, the conversion breaks printing explanatory messages after
the error: actual printing of the error gets delayed, so the
explanations precede rather than follow it.
Pity. Disable them for now. See also commit 7216ae3.
While there, eliminate QERR_BUS_NOT_FOUND, and clean up unusual
spelling in the error message.
Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com>
Property bus has always been too screwed up to be really usable for
values other than plain bus IDs. This just fixes a bug that crept in
in commit 1395af6 "qdev: add a maximum device allowed field for the
bus."
Happily plugs the virtio-rng-device into the virtio-bus provided by
virtio-serial-pci, even though its only slot is already occupied by a
virtio-serial-device.
qdev-monitor: Stop error avalanche in qbus_find_recursive()
Reproducer:
$ qemu-system-x86_64 -nodefaults -device virtio-rng-pci -device virtio-rng-pci -device virtio-rng-device,bus=virtio-bus
qemu-system-x86_64: -device virtio-rng-device,bus=virtio-bus: Bus 'virtio-bus' is full
qemu-system-x86_64: -device virtio-rng-device,bus=virtio-bus: Bus 'virtio-bus' is full
qemu-system-x86_64: -device virtio-rng-device,bus=virtio-bus: Bus 'virtio-bus' not found
qbus_find_recursive() reports the "is full" error itself, and leaves
reporting "not found" to its caller. The result is confusion. Write
it a function contract that permits leaving all error reporting to the
caller, and implement it. Update callers to detect and report "is
full".
Screwed up when commit 1395af6 added the max_dev limit and the "is
full" error condition to enforce it.
Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com>
disas does not need to access the CPU env for any reason. Change the
APIs to accept CPU pointers instead. Small change pattern needs to be
applied to all target translate.c. This brings us closer to making
disas.o a common-obj and less architecture specific in general.
Cc: Richard Henderson <rth@twiddle.net> Cc: Peter Maydell <peter.maydell@linaro.org> Cc: "Edgar E. Iglesias" <edgar.iglesias@gmail.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: Eduardo Habkost <ehabkost@redhat.com> Cc: Michael Walle <michael@walle.cc> Cc: Aurelien Jarno <aurelien@aurel32.net> Cc: Leon Alrae <leon.alrae@imgtec.com> Cc: Jia Liu <proljc@gmail.com> Cc: Alexander Graf <agraf@suse.de> Cc: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> Cc: Bastian Koppelmann <kbastian@mail.uni-paderborn.de> Cc: Guan Xuetao <gxt@mprc.pku.edu.cn> Cc: Max Filippov <jcmvbkbc@gmail.com> Reviewed-by: Richard Henderson <rth@twiddle.net> Reviewed-by: Eduardo Habkost <ehabkost@redhat.com> Signed-off-by: Peter Crosthwaite <crosthwaite.peter@gmail.com> Acked-by: Luiz Capitulino <lcapitulino@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
monitor: Split mon_get_cpu fn to remove ENV_GET_CPU
The monitor currently has one helper, mon_get_cpu() which will return
an env pointer. The target specific users of this API want an env, but
all the target agnostic users really just want the cpu pointer. These
users then need to use the target-specifically defined ENV_GET_CPU to
navigate back up to the CPU from the ENV. Split the API for the two
uses cases to remove all need for ENV_GET_CPU.
Reviewed-by: Richard Henderson <rth@twiddle.net> Reviewed-by: Andreas Färber <afaerber@suse.de> Signed-off-by: Peter Crosthwaite <crosthwaite.peter@gmail.com> Acked-by: Luiz Capitulino <lcapitulino@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
Bandan Das [Wed, 3 Jun 2015 22:38:09 +0000 (18:38 -0400)]
monitor: Point to "help" command on syntax error
When a command fails due to incorrect syntax or input, suggest using
the "help" command to get more information about the command. This
is only applicable for HMP.
Signed-off-by: Bandan Das <bsd@redhat.com> Acked-by: Luiz Capitulino <lcapitulino@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
Bandan Das [Wed, 3 Jun 2015 22:38:08 +0000 (18:38 -0400)]
monitor: cleanup parsing of cmd name and cmd arguments
There's too much going on in monitor_parse_command().
Split up the arguments parsing bits into a separate function
monitor_parse_arguments(). Let the original function check for
command validity and sub-commands if any and return data (*cmd)
that the newly introduced function can process and return a
QDict. Also, pass a pointer to the cmdline to track current
parser location.
Suggested-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Bandan Das <bsd@redhat.com> Acked-by: Luiz Capitulino <lcapitulino@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
Eric Blake [Fri, 15 May 2015 22:24:59 +0000 (16:24 -0600)]
qobject: Use 'bool' for qbool
We require a C99 compiler, so let's use 'bool' instead of 'int'
when dealing with boolean values. There are few enough clients
to fix them all in one pass.
Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Andreas Färber <afaerber@suse.de> Reviewed-by: Alberto Garcia <berto@igalia.com> Acked-by: Luiz Capitulino <lcapitulino@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
Greg Ungerer [Fri, 19 Jun 2015 13:43:26 +0000 (23:43 +1000)]
m68k: fix usp processing on interrupt entry and exception exit
The action to potentially switch sp register is not occurring at the correct
point in the interrupt entry or exception exit sequences.
For the interrupt entry case the sp on entry is used to create the stack
exception frame - but this may well be the user stack pointer, since we
haven't done the switch yet. Re-order the flow to switch the sp regs then
use the current sp to create the exception frame.
For the return from exception case the code is unwinding the sp after
switching sp registers. But it should always unwind the supervisor sp
first, then carry out any required sp switch.
Note that these problems don't effect operation unless the user sp bit is
set in the CACR register. Only a single sp is used in the default power up
state. Previously Linux only used this single sp mode. But modern versions
of Linux use the user sp mode now, so we need correct behavior for Linux
to work.
projects / osstest / qemu.git / log