Igor Mammedov writes:
> On Thu, 29 May 2014 14:25:31 +0200
> Andreas Färber wrote:
>
>> Am 29.05.2014 14:21, schrieb Marcel Apfelbaum:
>> > On Thu, 2014-05-29 at 12:47 +0200, Andreas Färber wrote:
>> >> Am 29.05.2014 11:47, schrieb Igor Mammedov:
>> >>> ... fixes freeing constant from vl.c by ma
From: "Edgar E. Iglesias"
Hi,
This is a second round of AArch64 EL2/3 patches working on the exception
model. Among other things adding HVC/SMC, interrupt routing to EL2/3 and
Virtual IRQs/FIQs. The VIRQ/VFIQ support only adds the external signal
delivery method.
Patch 3 is a bug fix.
Patch 14
From: "Edgar E. Iglesias"
Break out code to save/restore AArch64 SP into functions.
Signed-off-by: Edgar E. Iglesias
---
target-arm/internals.h | 29 -
target-arm/kvm64.c | 13 +++--
target-arm/op_helper.c | 6 +-
3 files changed, 24 insertions(+),
From: "Edgar E. Iglesias"
Signed-off-by: Edgar E. Iglesias
---
target-arm/op_helper.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/target-arm/op_helper.c b/target-arm/op_helper.c
index 2e2429a..581dc09 100644
--- a/target-arm/op_helper.c
+++ b/target-arm/op_helper.c
@@ -4
From: "Edgar E. Iglesias"
Signed-off-by: Edgar E. Iglesias
---
target-arm/helper-a64.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/target-arm/helper-a64.c b/target-arm/helper-a64.c
index cccda74..bc153cb 100644
--- a/target-arm/helper-a64.c
+++ b/target-arm/helper-a6
From: "Edgar E. Iglesias"
Signed-off-by: Edgar E. Iglesias
---
target-arm/cpu.h| 2 +-
target-arm/helper.c | 8
2 files changed, 9 insertions(+), 1 deletion(-)
diff --git a/target-arm/cpu.h b/target-arm/cpu.h
index 172a631..f8ca1da 100644
--- a/target-arm/cpu.h
+++ b/target-arm/cp
From: "Edgar E. Iglesias"
No functional change.
Prepares for future additions of the EL2 and 3 versions of this reg.
Signed-off-by: Edgar E. Iglesias
---
target-arm/cpu.c| 2 +-
target-arm/cpu.h| 2 +-
target-arm/helper-a64.c | 4 ++--
target-arm/helper.c | 12 ++
From: "Edgar E. Iglesias"
Signed-off-by: Edgar E. Iglesias
---
target-arm/cpu.h| 35 +++
target-arm/helper.c | 27 +++
2 files changed, 62 insertions(+)
diff --git a/target-arm/cpu.h b/target-arm/cpu.h
index ef6a95d..b446478 100644
--
From: "Edgar E. Iglesias"
Signed-off-by: Edgar E. Iglesias
---
target-arm/cpu.h| 2 +-
target-arm/helper.c | 6 ++
2 files changed, 7 insertions(+), 1 deletion(-)
diff --git a/target-arm/cpu.h b/target-arm/cpu.h
index f8ca1da..ef6a95d 100644
--- a/target-arm/cpu.h
+++ b/target-arm/cpu.
From: "Edgar E. Iglesias"
Signed-off-by: Edgar E. Iglesias
---
target-arm/cpu.h| 15 +++
target-arm/helper.c | 20
2 files changed, 35 insertions(+)
diff --git a/target-arm/cpu.h b/target-arm/cpu.h
index b446478..28521d4 100644
--- a/target-arm/cpu.h
+++ b/
From: "Edgar E. Iglesias"
Introduce new_el and new_mode in preparation for future patches
that add support for taking exceptions to and from EL2 and 3.
No functional change.
Signed-off-by: Edgar E. Iglesias
---
target-arm/cpu.h| 11 +++
target-arm/helper-a64.c | 24
From: "Edgar E. Iglesias"
Signed-off-by: Edgar E. Iglesias
---
cpu-exec.c | 5 ++---
target-arm/cpu.h | 16
2 files changed, 18 insertions(+), 3 deletions(-)
diff --git a/cpu-exec.c b/cpu-exec.c
index 38e5f02..a579ffc 100644
--- a/cpu-exec.c
+++ b/cpu-exec.c
@@ -478,7 +
From: "Edgar E. Iglesias"
Signed-off-by: Edgar E. Iglesias
---
target-arm/cpu.h | 7 +++
1 file changed, 7 insertions(+)
diff --git a/target-arm/cpu.h b/target-arm/cpu.h
index 9eddcc1..66c58bd 100644
--- a/target-arm/cpu.h
+++ b/target-arm/cpu.h
@@ -1133,6 +1133,13 @@ bool write_cpustate_t
From: "Edgar E. Iglesias"
Not all exception types update both FAR and ESR.
Signed-off-by: Edgar E. Iglesias
---
target-arm/helper-a64.c | 6 ++
1 file changed, 2 insertions(+), 4 deletions(-)
diff --git a/target-arm/helper-a64.c b/target-arm/helper-a64.c
index 7d94a74..c91005f 100644
---
From: "Edgar E. Iglesias"
Signed-off-by: Edgar E. Iglesias
---
target-arm/cpu.h | 7 ++-
target-arm/helper-a64.c| 1 +
target-arm/helper.c| 39 +++
target-arm/helper.h| 1 +
target-arm/internals.h | 6 ++
target-a
From: "Edgar E. Iglesias"
Signed-off-by: Edgar E. Iglesias
---
target-arm/cpu.h | 1 +
target-arm/helper-a64.c| 1 +
target-arm/helper.c| 6 ++
target-arm/helper.h| 1 +
target-arm/internals.h | 6 ++
target-arm/op_helper.c | 11 +++
ta
From: "Edgar E. Iglesias"
Signed-off-by: Edgar E. Iglesias
---
target-arm/cpu.h| 10 ++
target-arm/helper.c | 16
2 files changed, 26 insertions(+)
diff --git a/target-arm/cpu.h b/target-arm/cpu.h
index b3631f2..d15e8d2 100644
--- a/target-arm/cpu.h
+++ b/target-ar
From: "Edgar E. Iglesias"
Signed-off-by: Edgar E. Iglesias
---
cpu-exec.c | 12
target-arm/cpu.c| 20 ++--
target-arm/cpu.h| 24 ++--
target-arm/helper-a64.c | 2 ++
target-arm/helper.c | 3 +++
target-arm/inter
On Wed, 14 May 2014 16:18:42 -0300
Eduardo Habkost wrote:
> If a given machine have max_cpus set, not just smp_cpus needs to be
> limited, but the total number of CPUs (considering CPU hotplug) for the
> machine.
>
> We also had yet another max_cpus limit check at smp_parse(), ensuring
> that ma
On 30.05.14 07:58, Alexey Kardashevskiy wrote:
On 05/28/2014 09:35 PM, Alexander Graf wrote:
On 28.05.14 03:18, Alexey Kardashevskiy wrote:
On 05/28/2014 10:41 AM, Alexander Graf wrote:
On 28.05.14 02:34, Alexey Kardashevskiy wrote:
On 05/28/2014 09:55 AM, Alexander Graf wrote:
...
How do
On Thu, May 29, 2014 at 07:36:53PM +0200, Petar Jovanovic wrote:
> From: Petar Jovanovic
>
> From MIPS documentation (Volume III):
>
> UserLocal Register (CP0 Register 4, Select 2)
> Compliance Level: Recommended.
>
> The UserLocal register is a read-write register that is not interpreted by
>
On Thu, 29 May 2014 17:28:34 +0200
Paolo Bonzini wrote:
> Il 29/05/2014 16:22, Peter Crosthwaite ha scritto:
> >> >
> >> > +bool memory_region_is_mapped(MemoryRegion *mr)
> >> > +{
> > Is it not enough to just return mr->parent? Memory mapping assertion
> > will happen if you try and map the same
On Fri, 30 May 2014 00:05:51 +1000
Peter Crosthwaite wrote:
> On Tue, May 27, 2014 at 11:01 PM, Igor Mammedov wrote:
> > Provides framework for splitting host RAM allocation/
> > policies into a separate backend that could be used
> > by devices.
> >
> > Initially only legacy RAM backend is prov
On Thu, 29 May 2014 10:41:32 -0600
Eric Blake wrote:
> On 05/27/2014 07:01 AM, Igor Mammedov wrote:
> > Provides framework for splitting host RAM allocation/
> > policies into a separate backend that could be used
> > by devices.
> >
> > Initially only legacy RAM backend is provided, which
> > u
On Fri, 30 May 2014 00:06:23 +1000
Peter Crosthwaite wrote:
> On Tue, May 27, 2014 at 11:01 PM, Igor Mammedov wrote:
> > Add following parameters:
> > "slots" - total number of hotplug memory slots
> > "maxmem" - maximum possible memory
> >
> > "slots" and "maxmem" should go in pair and "max
On 29.05.14 16:12, Tom Musta wrote:
This is a follow up to the patch series initiated by Doug Kwan
http://lists.nongnu.org/archive/html/qemu-devel/2014-05/msg01936.html
The resubmission of V4 cleans up some attribution issues (corrects Doug's
email address in the Signed-off-by and also cites Do
On 29.05.14 13:05, Peter Maydell wrote:
Fix a typo in the ppce500_pci vmstate definition which meant that
we were migrating the struct pci_inbound using the vmstate for
pci_outbound. Fortunately the two structures have exactly the same
format at the moment (four uint32_ts) so this was harmless,
Igor Mammedov writes:
> On Thu, 29 May 2014 10:41:32 -0600
> Eric Blake wrote:
>
>> On 05/27/2014 07:01 AM, Igor Mammedov wrote:
[...]
>> > +goto out;
>> > +}
>> > +if (!value) {
>> > +error_setg(&local_err, "Property '%s.%s' doesn't take value '%"
>> > +
On 12.05.14 14:15, Jens Freimann wrote:
From: David Hildenbrand
When restoring the previously saved instruction in
kvm_arch_remove_sw_breakpoint(), we only restored one byte. Let's use
the sizeof() operator to make sure we restore the entire instruction.
While we are at it, let's remove the d
On 30 May 2014 07:46, Paolo Bonzini wrote:
> Il 29/05/2014 22:38, Peter Maydell ha scritto:
>
>>> > +#ifdef TARGET_WORDS_BIGENDIAN
>>> > +cpsr |= CPSR_E;
>>
>> This is wrong for BE32, where CPSR_E doesn't exist and both code
>> and data accesses are big-endian.
>>
>
> Is it okay for simplicity
On 12.05.14 14:15, Jens Freimann wrote:
From: David Hildenbrand
This patch makes use of the hw debugging support in kvm (provided by the guest's
PER facility) on s390. It enables the following features, available using the
gdbserver:
- single-stepping
- hw breakpoints
- hw watchpoints
Signed-
On 30 May 2014 09:19, Alexander Graf wrote:
> Ouch. I'm fairly sure we don't have migration users on e500,
> so we're safe. Thanks a lot for catching this!
Thank clang 3.4, which noticed that the vmstate_pci_inbound
variable was never used...
-- PMM
On 30/05/14 10:21, Alexander Graf wrote:
>
> On 12.05.14 14:15, Jens Freimann wrote:
>> From: David Hildenbrand
>>
>> When restoring the previously saved instruction in
>> kvm_arch_remove_sw_breakpoint(), we only restored one byte. Let's use
>> the sizeof() operator to make sure we restore the en
On Fri, 30 May 2014 10:21:04 +0200
Markus Armbruster wrote:
> Igor Mammedov writes:
>
> > On Thu, 29 May 2014 10:41:32 -0600
> > Eric Blake wrote:
> >
> >> On 05/27/2014 07:01 AM, Igor Mammedov wrote:
> [...]
> >> > +goto out;
> >> > +}
> >> > +if (!value) {
> >> > +err
On Thu, May 29, 2014 at 07:11:27PM +1000, Peter Crosthwaite wrote:
> On Fri, May 23, 2014 at 1:40 AM, Stefan Hajnoczi wrote:
> > object_initialize() leaves the object with a refcount of 1.
> > object_property_add_child() adds its own reference which is dropped
> > again when the property is delete
On 30/05/14 10:32, Alexander Graf wrote:
>> +case KVM_HW_BP:
>> +if (find_hw_breakpoint(arch_info->addr, -1, arch_info->type)) {
>> +ret = EXCP_DEBUG;
>> +}
>> +break;
>> +case KVM_SINGLESTEP:
>> +if (cs->singlestep_enabled) {
>> +ret
v4:
* Fix some typos in the patch head description.
* Improve some comments.
* Given that xen_pt_register_vga_regions()/xen_pt_unregister_vga_regions()
are called unconditionally, so we just return 0 there.
* Remove one spurious change.
* Remove some unnecessary "return" in void foo().
* Given t
basic gfx passthrough support:
- add a vga type for gfx passthrough
- retrieve VGA bios from sysfs, then load it to guest at 0xC
- register/unregister legacy VGA I/O ports and MMIOs for passthrough GFX
The original patch is from Weidong Han
Signed-off-by: Yang Zhang
Signed-off-by: Tiejun Ch
Some registers of Intel IGD are mapped in host bridge, so it needs to
passthrough these registers of physical host bridge to guest because
emulated host bridge in guest doesn't have these mappings.
The original patch is from Weidong Han < weidong.han @ intel.com >
Signed-off-by: Yang Zhang
Signe
ISA bridge is needed since Intel gfx drive will probe it instead
of Dev31:Fun0 to make graphics device passthrough work easy for VMM, that
only need to expose ISA bridge to let driver know the real hardware underneath.
The original patch is from Allen Kay [allen.m@intel.com]
Signed-off-by: Ya
On Thu, May 29, 2014 at 07:03:42PM +1000, Peter Crosthwaite wrote:
> On Fri, May 23, 2014 at 1:40 AM, Stefan Hajnoczi wrote:
> > virtio-blk-pci, virtio-blk-s390, and virtio-blk-ccw all duplicate the
> > qdev properties of their VirtIOBlock child. This approach does not work
> > well with string o
Implement that pci host bridge to specific to passthrough. Actually
this just inherit the standard one.
Signed-off-by: Tiejun Chen
---
v4:
* Fix one typo in the patch head description.
* Use (xen_enabled() && xen_has_gfx_passthru) to make sure we only work
in this scenario.
v3:
* Just fix th
On 30.05.14 10:57, Christian Borntraeger wrote:
On 30/05/14 10:32, Alexander Graf wrote:
+case KVM_HW_BP:
+if (find_hw_breakpoint(arch_info->addr, -1, arch_info->type)) {
+ret = EXCP_DEBUG;
+}
+break;
+case KVM_SINGLESTEP:
+if (cs->singlestep
On 05/30/2014 06:00 PM, Alexander Graf wrote:
>
> On 30.05.14 07:58, Alexey Kardashevskiy wrote:
>> On 05/28/2014 09:35 PM, Alexander Graf wrote:
>>> On 28.05.14 03:18, Alexey Kardashevskiy wrote:
On 05/28/2014 10:41 AM, Alexander Graf wrote:
> On 28.05.14 02:34, Alexey Kardashevskiy wrot
Il 30/05/2014 11:04, Peter Lieven ha scritto:
Am 30.05.2014 07:50, schrieb Paolo Bonzini:
Il 29/05/2014 22:25, Peter Lieven ha scritto:
Am 29.05.2014 21:17, schrieb Paolo Bonzini:
Il 29/05/2014 20:41, Peter Lieven ha scritto:
static inline unsigned geo_rand_range(double m, double n) {
ret
Am 30.05.2014 07:50, schrieb Paolo Bonzini:
> Il 29/05/2014 22:25, Peter Lieven ha scritto:
>> Am 29.05.2014 21:17, schrieb Paolo Bonzini:
>>> Il 29/05/2014 20:41, Peter Lieven ha scritto:
static inline unsigned geo_rand_range(double m, double n) {
return exp((log(m) + (double)rand()
On Fri, 30 May 2014 00:21:27 +1000
Peter Crosthwaite wrote:
> On Tue, May 27, 2014 at 11:01 PM, Igor Mammedov wrote:
> > From: Vasilis Liaskovitis
> >
> > Each hotplug-able memory slot is a DimmDevice.
> > A hot-add operation for a DIMM:
> > - creates a new DimmDevice and makes hotplug controll
On 30/05/14 11:01, Alexander Graf wrote:
>
> On 30.05.14 10:57, Christian Borntraeger wrote:
>> On 30/05/14 10:32, Alexander Graf wrote:
>>
+case KVM_HW_BP:
+if (find_hw_breakpoint(arch_info->addr, -1, arch_info->type)) {
+ret = EXCP_DEBUG;
+}
>>
Am 29.05.2014 um 13:14 hat Chrysostomos Nanakos geschrieben:
> Hello team,
>
> this is a patch implementing support for a new storage layer,
> Archipelago [1][2].
>
> We've been using Archipelago in our IaaS public cloud production
> environment for over a year now, along with Google Ganeti [3]
>
The OpRegion shouldn't be mapped 1:1 because the address in the host
can't be used in the guest directly.
This patch traps read and write access to the opregion of the Intel
GPU config space (offset 0xfc).
The original patch is from Jean Guyader
Signed-off-by: Yang Zhang
Signed-off-by: Tiejun
On Fri, 30 May 2014, Paolo Bonzini wrote:
> Il 30/05/2014 00:33, BALATON Zoltan ha scritto:
> Can you get the output of "info mtree" at the point where the unassigned
> access happens?
memory
- (prio 0, RW): system
-07ff (prio 0, RW):
Kevin Wolf writes:
> Some code in the block layer makes potentially huge allocations. Failure
> is not completely unexpected there, so avoid aborting qemu and handle
> out-of-memory situations gracefully.
>
> This patch addresses the allocations in the rbd block driver.
>
> Signed-off-by: Kevin W
PAPR allows having multiple interrupt sources such as PHB.
This adds a source lookup function and makes use of it.
Since at the moment QEMU only supports a single source,
no change in behaviour is expected.
Signed-off-by: Alexey Kardashevskiy
---
hw/intc/xics.c | 30 +--
There are few helpers already to support array migration. However they all
require the destination side to preallocate arrays before migration which
is not always possible due to unknown array size as it might be some
sort of dynamic state. One of the examples is an array of MSIX-enabled
devices in
This removes @next_irq from sPAPREnvironment which was used in old
IRQ allocator as XICS is now responsible for IRQs and keeps track of
allocated IRQs.
Signed-off-by: Alexey Kardashevskiy
---
hw/ppc/spapr.c | 3 +--
include/hw/ppc/spapr.h | 1 -
2 files changed, 1 insertion(+), 3 deletio
This implements interrupt release function so IRQs can be returned back
to the pool for reuse in cases such as PCI hot plug.
Signed-off-by: Alexey Kardashevskiy
---
Changes:
v3:
* xics_free() now accepts "irq" instead of server:srcno and does sanity
checks
---
hw/intc/xics.c| 27
Currently SPAPR PHB keeps track of all allocated MSI (here and below
MSI stands for both MSI and MSIX) interrupt because
XICS used to be unable to reuse interrupts. This is a problem for
dynamic MSI reconfiguration which happens when guest reloads a driver
or performs PCI hotplug. Another problem i
The existing interrupt allocation scheme in SPAPR assumes that
interrupts are allocated at the start time, continously and the config
will not change. However, there are cases when this is not going to work
such as:
1. migration - we will have to have an ability to choose interrupt
numbers for dev
1. Debug bdrv_drain_all() and find out whether there are any I/O
requests remaining.
I believe that's what happens:
Context 1:
- commit_one_iteration makes write request (req A)
- request A is handled to io thread, qemu_coroutine_yield() is called
Context 2:
- VM makes write request (req B
The current allocator returns IRQ numbers from a pool and does not
support IRQs reuse in any form as it did not keep track of what it
previously returned, it only keeps the last returned IRQ. Some use
cases such as PCI hot(un)plug may require IRQ release and reallocation.
This moves an allocator f
This moves interrupts allocation business from SPAPR to XICS
and makes use of it.
Changes:
v3:
* replaces static array of descriptors in SPAPR PHB with GHashTable
* implements migration of hash table via temporary array
v2:
* s/server/source/
* fixed typos, code style, added an assert
* added pat
On Fri, May 30, 2014 at 7:01 PM, Igor Mammedov wrote:
> On Fri, 30 May 2014 00:21:27 +1000
> Peter Crosthwaite wrote:
>
>> On Tue, May 27, 2014 at 11:01 PM, Igor Mammedov wrote:
>> > From: Vasilis Liaskovitis
>> >
>> > Each hotplug-able memory slot is a DimmDevice.
>> > A hot-add operation for
On Wed, May 28, 2014 at 04:37:44PM +0200, Kevin Wolf wrote:
> Some code in the block layer makes potentially huge allocations. Failure
> is not completely unexpected there, so avoid aborting qemu and handle
> out-of-memory situations gracefully.
>
> This patch addresses the allocations in the qcow
On Wed, May 28, 2014 at 04:37:35PM +0200, Kevin Wolf wrote:
> Some code in the block layer makes potentially huge allocations. Failure
> is not completely unexpected there, so avoid aborting qemu and handle
> out-of-memory situations gracefully.
>
> This patch addresses bounce buffer allocations i
On Fri, 30 May 2014, BALATON Zoltan wrote:
On Fri, 30 May 2014, Paolo Bonzini wrote:
Il 30/05/2014 00:33, BALATON Zoltan ha scritto:
Can you get the output of "info mtree" at the point where the unassigned
access happens?
More exactly at the beginning of ohci_read_hcca it is the following. (No
This removes xics_set_irq_type() as it is not used anymore.
This is done by a separate patch to make the previous patch
look nicer.
Signed-off-by: Alexey Kardashevskiy
---
hw/intc/xics.c| 11 ---
include/hw/ppc/xics.h | 1 -
2 files changed, 12 deletions(-)
diff --git a/hw/int
Since islsi[] array has been merged into the ICSState struct,
we must not reset flags as they tell if the interrupt is in use.
Signed-off-by: Alexey Kardashevskiy
---
hw/intc/xics.c | 7 +++
hw/intc/xics_kvm.c | 7 +++
2 files changed, 14 insertions(+)
diff --git a/hw/intc/xics.c b/
On 30.05.14 11:34, Alexey Kardashevskiy wrote:
Currently SPAPR PHB keeps track of all allocated MSI (here and below
MSI stands for both MSI and MSIX) interrupt because
XICS used to be unable to reuse interrupts. This is a problem for
dynamic MSI reconfiguration which happens when guest reloads a
Il 30/05/2014 00:33, BALATON Zoltan ha scritto:
usb-ohci: pci-ohci: USB Operational
Unassigned mem read 07c9ae00
Unassigned mem read 07c9ae04
[...]
Unassigned mem read 07c9ae84
usb-ohci: HCCA read error at 7c9ae00
ohci_die: DMA error
The read happens through ohci_read_hc
On Fri, 30 May 2014 19:48:13 +1000
Peter Crosthwaite wrote:
> On Fri, May 30, 2014 at 7:01 PM, Igor Mammedov wrote:
> > On Fri, 30 May 2014 00:21:27 +1000
> > Peter Crosthwaite wrote:
> >
> >> On Tue, May 27, 2014 at 11:01 PM, Igor Mammedov
> >> wrote:
> >> > From: Vasilis Liaskovitis
> >> >
On Fri, May 30, 2014 at 7:48 PM, Peter Crosthwaite
wrote:
> On Fri, May 30, 2014 at 7:01 PM, Igor Mammedov wrote:
>> On Fri, 30 May 2014 00:21:27 +1000
>> Peter Crosthwaite wrote:
>>
>>> On Tue, May 27, 2014 at 11:01 PM, Igor Mammedov wrote:
>>> > From: Vasilis Liaskovitis
>>> >
>>> > Each hot
On Fri, May 30, 2014 at 8:31 PM, Igor Mammedov wrote:
> On Fri, 30 May 2014 19:48:13 +1000
> Peter Crosthwaite wrote:
>
>> On Fri, May 30, 2014 at 7:01 PM, Igor Mammedov wrote:
>> > On Fri, 30 May 2014 00:21:27 +1000
>> > Peter Crosthwaite wrote:
>> >
>> >> On Tue, May 27, 2014 at 11:01 PM, Igo
Ping. This patch was suggested to be sent separately but I can also
include it to the TZ patchset again.
http://patchwork.ozlabs.org/patch/349592/
Thanks,
Fabian
On 16 May 2014, at 14:43, Fabian Aggeler wrote:
> This patch changes some readfns/writefns to use raw_write
> and raw_read function
Kevin Wolf writes:
> A not too small part of the recent CVEs were DoS scenarios by letting
> qemu abort with too large memory allocations. We generally "fixed" these
> cases by setting some limits on values read from image files that
> influence the size of allocations.
Just double-checking: do
Move the x-data-plane property. Originally it was outside since not
every transport may wish to support dataplane. But that makes little
sense when we have a dedicated CONFIG_VIRTIO_BLK_DATA_PLANE ifdef
already.
This move makes it easier to switch to property aliases in the next
patch.
Signed-o
v3:
* Split qdev_alias_all_properties() into its own patch [Peter Crosthwaite]
* Do not dereference DEVICE_CLASS(class) inline [Peter Crosthwaite]
v2:
* Add qdev_alias_all_properties() instead of virtio-blk-specific function
[Paolo]
* Explain refcount handling in doc comment [Paolo]
* Fix "p
It becomes unwiedly to duplicate all virtio-blk qdev property
definitions due to an #ifdef. The C preprocessor syntax makes it a
little hard to resolve this cleanly but we can extract the #ifdef and
call a macro it defines later.
Avoiding duplication is important since it will only get worse when
Sometimes an object needs to present a property which is actually on
another object, or it needs to provide an alias name for an existing
property.
Examples:
a.foo -> b.foo
a.old_name -> a.new_name
The new object_property_add_alias() API allows objects to alias a
property on the same object o
The qdev_alias_all_properties() function creates QOM alias properties
for each qdev property on a DeviceState. This is useful for parent
objects that wish to forward property accesses to their children.
Signed-off-by: Stefan Hajnoczi
---
v3:
* Split qdev_alias_all_properties() into its own patc
There is no need to make DEFINE_VIRTIO_BLK_PROPERTIES() public. Inline
it into virtio-blk.c so it cannot be used by mistake from other source
files.
Signed-off-by: Stefan Hajnoczi
Reviewed-by: Peter Crosthwaite
---
hw/block/virtio-blk.c | 12 +++-
include/hw/virtio/virtio-blk.
This function is no longer used since parent objects now use child
aliases to set the VirtIOBlkConf directly.
Signed-off-by: Stefan Hajnoczi
Reviewed-by: Peter Crosthwaite
---
hw/block/virtio-blk.c | 6 --
include/hw/virtio/virtio-blk.h | 2 --
2 files changed, 8 deletions(-)
diff
object_initialize() leaves the object with a refcount of 1.
object_property_add_child() adds its own reference which is dropped
again when the property is deleted.
The upshot of this is that we always have a refcount >= 1. Upon hot
unplug the virtio-blk child is not finalized!
Drop our reference
On Fri, 30 May 2014 20:55:50 +1000
Peter Crosthwaite wrote:
> On Fri, May 30, 2014 at 8:31 PM, Igor Mammedov wrote:
> > On Fri, 30 May 2014 19:48:13 +1000
> > Peter Crosthwaite wrote:
> >
> >> On Fri, May 30, 2014 at 7:01 PM, Igor Mammedov wrote:
> >> > On Fri, 30 May 2014 00:21:27 +1000
> >>
When a quorum file is totally destroyed (broken NAS or SAN) the user can start a
drive-mirror job on the quorum block backend and then replace the broken
quorum file with drive-mirror-replace given it has a node-name.
Signed-off-by: Benoit Canet
---
block.c | 8 ++--
block/mi
On read operations when this parameter is set and some replicas are corrupted
while quorum can be reached quorum will proceed to rewrite the correct version
of the data to fix the corrupted replicas.
This will shine with SSD where the FTL will remap the same block at another
place on rewrite.
Sig
Tests for drive-mirror-replace whose purpose is to enable quorum file mirroring
and replacement after failure.
Signed-off-by: Benoit Canet
---
tests/qemu-iotests/041| 34 +--
tests/qemu-iotests/096| 222 ++
tests/qemu-iotests/096.out
These are the last bits required to make quorum usable in production.
v5: rebase on latest Stefan's block branch [Kevin]
v4:
update patchset to stefan's block branch
drop Max reviewed by because the series changes
Benoît Canet (3):
quorum: Add the rewrite-corrupted
Am 29.05.2014 um 13:14 hat Chrysostomos Nanakos geschrieben:
> VM Image on Archipelago volume is specified like this:
>
> file=archipelago:[/mport=[:vport=]]
>
> 'archipelago' is the protocol.
>
> 'mport' is the port number on which mapperd is listening. This is optional
> and if not specified,
The Friday 30 May 2014 à 13:18:42 (+0200), Benoît Canet wrote :
Oh sorry I made again commits lines ending with a dot.
Will fix this in the next iteration.
Best regards
Benoît
> On read operations when this parameter is set and some replicas are corrupted
> while quorum can be reached quorum wi
Am 30.05.2014 um 13:14 hat Markus Armbruster geschrieben:
> Kevin Wolf writes:
>
> > A not too small part of the recent CVEs were DoS scenarios by letting
> > qemu abort with too large memory allocations. We generally "fixed" these
> > cases by setting some limits on values read from image files
On Fri, 30 May 2014, Paolo Bonzini wrote:
Il 30/05/2014 00:33, BALATON Zoltan ha scritto:
usb-ohci: pci-ohci: USB Operational
Unassigned mem read 07c9ae00
Unassigned mem read 07c9ae04
[...]
Unassigned mem read 07c9ae84
usb-ohci: HCCA read error at 7c9ae00
ohci_die: DMA e
Il 30/05/2014 13:32, BALATON Zoltan ha scritto:
On Fri, 30 May 2014, Paolo Bonzini wrote:
Il 30/05/2014 00:33, BALATON Zoltan ha scritto:
usb-ohci: pci-ohci: USB Operational
Unassigned mem read 07c9ae00
Unassigned mem read 07c9ae04
[...]
Unassigned mem read 07c9ae84
usb
virtio-blk-pci, virtio-blk-s390, and virtio-blk-ccw all duplicate the
qdev properties of their VirtIOBlock child. This approach does not work
well with string or pointer properties since we must be careful about
leaking or double-freeing them.
Use the QOM alias property to forward property access
Am 29.05.2014 um 00:19 hat Max Reitz geschrieben:
> If bs->drv is NULL, iov.iov_base should not be leaked.
>
> Signed-off-by: Max Reitz
Funny how such things go unnoticed for months and then three people fix
it independently in the same week. Anyway, I won: commit bd604369 ;-)
Kevin
The Wednesday 28 May 2014 à 09:30:11 (-0600), Eric Blake wrote :
> On 05/28/2014 08:57 AM, Igor Mammedov wrote:
> > ... allowing to get state of present memory devices.
> > Currently implemented only for DimmDevice.
> >
> > Signed-off-by: Igor Mammedov
> > ---
> > hw/mem/dimm.c|
On 05/29/14 22:57, Eric Blake wrote:
> On 05/29/2014 02:36 PM, Laszlo Ersek wrote:
>
>> Emitting the events unconditionally shouldn't hurt anything, I think. I
>> added the property for two reasons:
>>
>> - I was sure someone would ask for it. :)
>
> Someone still might :) But it may be a pre-ma
On Fri, 30 May 2014 13:43:52 +0200
Benoît Canet wrote:
> The Wednesday 28 May 2014 à 09:30:11 (-0600), Eric Blake wrote :
> > On 05/28/2014 08:57 AM, Igor Mammedov wrote:
> > > ... allowing to get state of present memory devices.
> > > Currently implemented only for DimmDevice.
> > >
> > > Signe
The Wednesday 28 May 2014 à 16:37:34 (+0200), Kevin Wolf wrote :
> This function returns NULL instead of aborting when an allocation fails.
>
> Signed-off-by: Kevin Wolf
> Reviewed-by: Stefan Hajnoczi
> ---
> block.c | 5 +
> include/block/block.h | 1 +
> include/qemu/osdep
The Wednesday 28 May 2014 à 16:37:35 (+0200), Kevin Wolf wrote :
> Some code in the block layer makes potentially huge allocations. Failure
> is not completely unexpected there, so avoid aborting qemu and handle
> out-of-memory situations gracefully.
>
> This patch addresses bounce buffer allocati
The Wednesday 28 May 2014 à 16:37:36 (+0200), Kevin Wolf wrote :
> Some code in the block layer makes potentially huge allocations. Failure
> is not completely unexpected there, so avoid aborting qemu and handle
> out-of-memory situations gracefully.
>
> This patch addresses the allocations in the
1 - 100 of 313 matches
Mail list logo
