Quoting Paolo Bonzini (2014-02-18 07:53:33)
> Il 18/02/2014 13:44, Andreas Färber ha scritto:
> >> > There isn't really a standard criterion. It's up to each maintainer to
> >> > be stricter or looser on what goes to stable.
> > The criteria is pretty simple: Was the breakage in the last release
>
On 2014-02-18 16:28, Peter Maydell wrote:
> The ethernet device in the musicpal only has two tx queues,
> but we modelled it with four CTDP registers, presumably a
> cut and paste from the rx queue registers. Since the tx_queue[]
> array is only 2 entries long this allowed a guest to overrun
> this
Quoting mrhi...@linux.vnet.ibm.com (2014-02-17 20:34:06)
> From: "Michael R. Hines"
>
> 1. Fix small memory leak in parsing inet address from command line in
> data_init()
> 2. Fix ibv_post_send() return value check and pass error code back up
> correctly.
> 3. Fix rdma_destroy_qp() segfault af
From: Yang Zhang
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
Cc:
From: Yang Zhang
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
From: Yang Zhang
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]
From: Yang Zhang
basic gfx passthrough support:
- add a vga type for gfx passthrough
- retrieve VGA bios from host 0xC, then load it to guest 0xC
- register/unregister legacy VGA I/O ports and MMIOs for passthroughed gfx
The original patch is from Weidong Han
Signed-off-by: Yang Zhang
From: Yang Zhang
The following patches are ported from Xen Qemu-traditional branch which are
adding Intel IGD passthrough supporting to Qemu upstream.
To pass through IGD to guest, user need to add following lines in Xen config
file:
gfx_passthru=1
pci=['00:02.0@2']
Besides, since Xen + Qemu up
From: Yang Zhang
Some VBIOSs and drivers assume the IGD BDF (bus:device:function) is
always 00:02.0, so this patch reserves 00:02.0 for assigned IGD in
guest.
The original patch is from Weidong Han
Signed-off-by: Yang Zhang
Cc: Weidong Han
---
hw/pci/pci.c |6 ++
1 files changed, 6
Il 27/01/2014 15:18, Alexander Graf ha scritto:
Commit 94ccff13 introduced a more verbose failure message and retry
operations on KVM VM creation. However, it ended up using a variable
for its failure message that hasn't been initialized yet.
Fix it to use the value it meant to set.
Signed-off-
Il 21/02/2014 05:34, Matt Lupfer ha scritto:
This doesn't appear to be a solution, because with the timer rewrite, QEMU
moves its periodic (1 ms) qemu_notify_event() call to break out of
the main event loop from a SIGALRM handler to the rearm of a QEMU timer.
Presumably QEMU is counting on these
Quoting Alexander Graf (2014-01-27 08:18:09)
> Commit 94ccff13 introduced a more verbose failure message and retry
> operations on KVM VM creation. However, it ended up using a variable
> for its failure message that hasn't been initialized yet.
>
> Fix it to use the value it meant to set.
>
> Si
On 02/10/2014 05:52 PM, Alexey Kardashevskiy wrote:
> The changelog is:
> > version: update to 20140204
> > virtio-9p: disable unused structure
> > Make "boot net:dhcp" boot from IPv4 only
> > Fix virtio device shutdown
> > Change shutdown method name for virtio-scsi
> > Add support for
Quoting Richard Henderson (2014-01-10 14:39:56)
> Parity should be set for a zero result.
>
> Signed-off-by: Richard Henderson
ping for 1.7.1
> ---
> target-i386/cc_helper.c | 2 +-
> target-i386/translate.c | 2 +-
> 2 files changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/target-i3
Currently, the default msix vectors for virtio-net-pci is 3 which is
obvious not suitable for multiqueue guest, so we depends on the user
or management tools to pass a correct vectors parameter. In fact, we
can simplifying this by calculate the number of vectors on init.
Consider we have N queues,
Am 20.02.2014 23:18, schrieb Peter Maydell:
> On 20 February 2014 21:18, Stefan Weil wrote:
>> MinGW-w64's gcc has cpuid.h, so my 32 and 64 bit cross builds work
>> without problems. We can use that code for MinGW, too, but we could also
>> stop supporting MinGW (which has several other deficits).
Am 20.02.2014 23:22, schrieb Peter Maydell:
> On 20 February 2014 21:26, Stefan Weil wrote:
>> There was a suggestion to remove some dependencies on windows.h (which
>> causes the trouble here). I recently started doing this, and that
>> approach fixes the warning, too. Maybe I can send a patch ne
Sorry, for some reason I thought this was the CloudStack mailing list...
On Thu, Feb 20, 2014 at 10:31 PM, Marcus wrote:
> Nested virtualization will work, if the hypervisor you are using
> supports passing the vmx flag in the guest. As far as I'm aware,
> VMware fusion does, but qemu on windows
Nested virtualization will work, if the hypervisor you are using
supports passing the vmx flag in the guest. As far as I'm aware,
VMware fusion does, but qemu on windows probably doesn't.
I once made a few code changes to allow KVM guests to work without the
need for vmx (for devcloud), but the r
On 02/10/2014 05:32 PM, Alexey Kardashevskiy wrote:
> At the moment if the user asked for huge pages and there is no more huge
> pages, QEMU prints warning and falls back to the anonymous memory
> allocator which is quite easy not to notice. QEMU also does so even
> if the user specified -mem-preal
On 02/21/2014 12:32 AM, Dr. David Alan Gilbert wrote:
I'm happy to use more memory to get FT, all I'm trying to do is see
if it's possible to put a lower bound than 2x on it while still maintaining
full FT, at the expense of performance in the case where it uses
a lot of memory.
The bottom lin
Hello,
After upgrading to QEMU 1.7.0, CentOS 5.x guests often fail to boot
with the following kernel apic=debug output:
> ACPI: Core revision 20060707
> enabled ExtINT on CPU#0
> ENABLING IO-APIC IRQs
> ..TIMER: vector=0x31 apic1=0 pin1=2 apic2=-1 pin2=-1
> ..MP-BIOS bug: 8254 timer not connecte
Hi,
I wanted to know if I have a Windows Machine, and I run Ubuntu Guest on my
Windows machine using QEMU, will I be able to enable kvm inside my Ubuntu
VM? (Basically is nested virtualization support avaiable for Windows -
Ubuntu - Another Guest?)
Thankyou
Shehbaz.
On 02/14/2014 12:27 PM, Alexey Kardashevskiy wrote:
> When the guests adds buffers to receive queue, the network device
> should flush its queue of pending packets. This is done with
> qemu_flush_queued_packets.
>
> This adds a call to qemu_flush_queued_packets() which wakes up the main
> loop and
On 02/21/2014 12:50 AM, Alexey Kardashevskiy wrote:
> From: Paolo Bonzini
>
> This is a first step towards QOMifying /machine.
>
> Signed-off-by: Paolo Bonzini
I got interesting conversation about "sob" in my team so here it is:
Signed-off-by: Alexey Kardashevskiy
Is that enough or I better
On Wed, Feb 19, 2014 at 5:26 PM, Igor Mammedov wrote:
> On Wed, 19 Feb 2014 15:53:54 +0800
> Hu Tao wrote:
>
> > From: Wanlong Gao
> >
> > Add the numa_info structure to contain the numa nodes memory,
> > VCPUs information and the future added numa nodes host memory
> > policies.
> this is old
Alexander Graf writes:
> On 02/18/2014 05:17 PM, Cornelia Huck wrote:
>> Hm. So whatever_le for 1.0 devices, and virtio_whatever (checking the
>> byteswap value) for legacy devices? The device implementation will be
>> aware of the virtio version anyway.
>
> Yeah, but I would hope we want to share
Greg Kurz writes:
> On Tue, 18 Feb 2014 20:25:15 +0100
> Andreas Färber wrote:
>> Am 18.02.2014 13:38, schrieb Greg Kurz:
>> > diff --git a/include/hw/virtio/virtio-access.h
>> > b/include/hw/virtio/virtio-access.h new file mode 100644
>> > index 000..2e22a47
>> > --- /dev/null
>> > +++ b/inc
Resizing the xbzrle cache during migration causes qemu-crash,
because the main-thread and migration-thread modify the xbzrle
cache size concurrently without lock-protection.
Signed-off-by: ChenLiang
Signed-off-by: Gonglei
---
Changes against the previous version:
*Remove function cache_max_num_
On Wed, 2014-02-19 at 15:25 -0800, Nicholas A. Bellinger wrote:
> On Wed, 2014-02-19 at 19:03 +1100, Vadim Rozenfeld wrote:
> > On Tue, 2014-02-18 at 13:00 -0800, Nicholas A. Bellinger wrote:
> > > On Mon, 2014-02-10 at 11:05 -0800, Nicholas A. Bellinger wrote:
> > >
> > >
> > >
> > > > > > > Hi
On 02/21/2014 02:58 AM, Thomas Huth wrote:
> On Thu, 20 Feb 2014 15:10:16 +
> Peter Maydell wrote:
>
>> Check that the C++ compiler works with the C compiler; if it
>> does not, then don't pass CXX to the build process. This
>> fixes a regression where QEMU was no longer building if the
>> bu
This patch implements the ARM PMCCNTR register including
the disable and reset components of the PMCR register.
Signed-off-by: Alistair Francis
---
V8: Only implement the register in system mode to make sure it doesn't
break compilation of linux-user targets.
V7: Fixed a bug that caused the cycle
On Fri, Feb 21, 2014 at 10:07 AM, Peter Maydell
wrote:
> On 20 February 2014 23:58, Alistair Francis
> wrote:
>> On Fri, Feb 21, 2014 at 1:44 AM, Peter Maydell
>> wrote:
>>> This breaks compilation of the arm-linux-user target
>>> (the qemu_clock_get_* functions are only available in
>>> syste
On 20/02/14 7:09 PM, Peter Maydell wrote:
On 20 February 2014 23:55, Brad Smith wrote:
On 20/02/14 12:50 PM, Peter Maydell wrote:
Win32 doesn't have a cpuid.h, and MacOSX may have one but without
the __cpuid() function we use, which means that commit 9d2eec20
broke the build for those platfor
于 2014/2/21 0:38, Markus Armbruster 写道:
> Wenchao Xia writes:
>
>> By default, any union will automatically generate a enum type as
>> "[UnionName]Kind" in C code, and it is duplicated when the discriminator
>> is specified as a pre-defined enum type in schema. After this patch,
>> the pre-define
Hello, is this patch OK to be merged?
于 2014/2/20 20:22, Markus Armbruster 写道:
> Wenchao Xia writes:
>
>> Before this patch, 'QAPISchemaError' scans whole input until 'pos'
>> to get error line number. After this patch, the scan is avoided since
>> line number is remembered in schema parsing. This patch also benefits
>> other error r
On 20 February 2014 23:55, Brad Smith wrote:
> On 20/02/14 12:50 PM, Peter Maydell wrote:
>>
>> Win32 doesn't have a cpuid.h, and MacOSX may have one but without
>> the __cpuid() function we use, which means that commit 9d2eec20
>> broke the build for those platforms. Fix this by tightening up
>>
On 20 February 2014 23:58, Alistair Francis wrote:
> On Fri, Feb 21, 2014 at 1:44 AM, Peter Maydell
> wrote:
>> This breaks compilation of the arm-linux-user target
>> (the qemu_clock_get_* functions are only available in
>> system mode). You probably want to do something similar
>> to how we ha
On Fri, Feb 21, 2014 at 1:44 AM, Peter Maydell wrote:
> On 19 February 2014 23:57, Alistair Francis
> wrote:
>> This patch implements the ARM PMCCNTR register including
>> the disable and reset components of the PMCR register.
>>
>> Signed-off-by: Alistair Francis
>
> This breaks compilation of
On 20/02/14 12:50 PM, Peter Maydell wrote:
Win32 doesn't have a cpuid.h, and MacOSX may have one but without
the __cpuid() function we use, which means that commit 9d2eec20
broke the build for those platforms. Fix this by tightening up
our configure cpuid.h check to test that the functions we nee
On 20 February 2014 23:26, Edgar E. Iglesias wrote:
> On Thu, Feb 20, 2014 at 03:58:15PM +, Peter Maydell wrote:
>> However, doesn't this also apply to interrupt_request ?
>> If we have a pending asserted interrupt on the CPU
>> (ie the IRQ line into the chip is being held high)
>> this should
On Thu, Feb 20, 2014 at 03:58:15PM +, Peter Maydell wrote:
> On 16 February 2014 02:07, Edgar E. Iglesias wrote:
> > On Sat, Feb 15, 2014 at 03:42:56PM +, Peter Maydell wrote:
> >> On 13 February 2014 05:07, wrote:
> >> > From: "Edgar E. Iglesias"
> >> >
> >> > cpu->exit_request is part
On 02/20/2014 02:47 PM, Peter Maydell wrote:
> On 20 February 2014 22:36, Mario Smarduch wrote:
>> On 02/20/2014 02:10 PM, Peter Maydell wrote:
>>> On 20 February 2014 21:59, Mario Smarduch wrote:
On 02/20/2014 11:35 AM, Peter Maydell wrote:
> On 20 February 2014 19:09, Mario Smarduch w
On 10/15/2013 04:18 PM, Xiexiangyou wrote:
> Thanks for your reply :-)
> The QEMU version is 1.5.1,and the KVM version is 3.6
>
> QEMU command:
> /usr/bin/qemu-kvm -name win2008_dc_5 -S -machine
> pc-i440fx-1.5,accel=kvm,usb=off -m 2048 -realtime mlock=off -smp
> 4,maxcpus=64,sockets=16,cores=4,
On 20 February 2014 22:36, Mario Smarduch wrote:
> On 02/20/2014 02:10 PM, Peter Maydell wrote:
>> On 20 February 2014 21:59, Mario Smarduch wrote:
>>> On 02/20/2014 11:35 AM, Peter Maydell wrote:
On 20 February 2014 19:09, Mario Smarduch wrote:
> host features since you don't know what
On 02/20/2014 02:10 PM, Peter Maydell wrote:
> On 20 February 2014 21:59, Mario Smarduch wrote:
>> On 02/20/2014 11:35 AM, Peter Maydell wrote:
>>> On 20 February 2014 19:09, Mario Smarduch wrote:
host features since you don't know what virtio device will be plugged
in later.
>>>
>>> I
On 20 February 2014 21:26, Stefan Weil wrote:
> There was a suggestion to remove some dependencies on windows.h (which
> causes the trouble here). I recently started doing this, and that
> approach fixes the warning, too. Maybe I can send a patch next weekend.
That would probably be the nicest ap
On 20 February 2014 21:18, Stefan Weil wrote:
> MinGW-w64's gcc has cpuid.h, so my 32 and 64 bit cross builds work
> without problems. We can use that code for MinGW, too, but we could also
> stop supporting MinGW (which has several other deficits).
We need the conditionals for MacOSX builds anyw
On 20 February 2014 21:33, Stefan Weil wrote:
> (I personally prefer ULL instead of ull because it avoids constants
> ending with 'full', and if we ignore the disas code, ULL is also more
> common in QEMU)
I agree, as it happens, but I opted to stick with the existing case here.
thanks
-- PMM
On 20 February 2014 21:59, Mario Smarduch wrote:
> On 02/20/2014 11:35 AM, Peter Maydell wrote:
>> On 20 February 2014 19:09, Mario Smarduch wrote:
>>> host features since you don't know what virtio device will be plugged
>>> in later.
>>
>> I think this function is the right place to set these p
On 02/20/2014 11:35 AM, Peter Maydell wrote:
> On 20 February 2014 19:09, Mario Smarduch wrote:
>> host features since you don't know what virtio device will be plugged
>> in later.
>
> I think this function is the right place to set these properties,
> yes. What I'm saying is that I don't see wh
Am 20.02.2014 20:44, schrieb Peter Maydell:
> The Win32 build warns about trace/control-internal.h:
>
> warning: 'trace_event_count' declared inline after being called
>
> Fix this by simply reordering trace_event_id() and
> trace_event_count().
>
> Signed-off-by: Peter Maydell
> ---
> trace/c
Am 20.02.2014 20:47, schrieb Peter Maydell:
> 64 bit constants need the "ULL" suffix, not just "UL", because
> on 32 bit platforms 'long' is not large enough and this will
> cause a compiler warning.
>
> Signed-off-by: Peter Maydell
> ---
> I think plain "UL" as a suffix is pretty much never righ
Am 20.02.2014 20:45, schrieb Peter Maydell:
> The Windows headers provided by MinGW define MOD_SHIFT. Avoid
> it by using SPITZ_MOD_* for our constants here.
>
> Signed-off-by: Peter Maydell
> ---
> The other approach would be just to #undef MOD_SHIFT, (and
> looking back through the archives I s
Am 20.02.2014 18:50, schrieb Peter Maydell:
> Win32 doesn't have a cpuid.h, and MacOSX may have one but without
> the __cpuid() function we use, which means that commit 9d2eec20
> broke the build for those platforms. Fix this by tightening up
> our configure cpuid.h check to test that the functions
On 02/20/2014 01:42 PM, Peter Maydell wrote:
> Win32 doesn't have a cpuid.h, and MacOSX may have one but without
> the __cpuid() function we use, which means that commit 9d2eec20
> broke the build for those platforms. Fix this by tightening up
> our configure cpuid.h check to test that the function
On Thu, Feb 20, 2014 at 03:57:18PM +0100, Kevin Wolf wrote:
> This new function checks if any help option ('help' or '?') occurs
> anywhere in an option string, so that things like 'cluster_size=4k,help'
> are recognised.
>
> Signed-off-by: Kevin Wolf
Reviewed-by: Jeff Cody
> ---
> include/qe
On Thu, Feb 20, 2014 at 03:57:22PM +0100, Kevin Wolf wrote:
> This patch allows using 'qemu-img $subcmd -o help' for the create,
> convert and amend subcommands, without specifying the previously
> required filename arguments.
>
> Note that it's still allowed and meaningful to specify a filename:
spapr_pci_msi_init() does this:
memory_region_init_io(&spapr->msiwindow, NULL, &spapr_msi_ops, spapr,
"msi", getpagesize());
That means this device's memory region size will depend on
the host OS CPU and configuration, which seems like a bad idea,
especially if this
64 bit constants need the "ULL" suffix, not just "UL", because
on 32 bit platforms 'long' is not large enough and this will
cause a compiler warning.
Signed-off-by: Peter Maydell
---
I think plain "UL" as a suffix is pretty much never right;
it should either be "U" or "ULL".
target-ppc/translat
The Windows headers provided by MinGW define MOD_SHIFT. Avoid
it by using SPITZ_MOD_* for our constants here.
Signed-off-by: Peter Maydell
---
The other approach would be just to #undef MOD_SHIFT, (and
looking back through the archives I see Stefan posted a patch
to do just that last year) but I
The Win32 build warns about trace/control-internal.h:
warning: 'trace_event_count' declared inline after being called
Fix this by simply reordering trace_event_id() and
trace_event_count().
Signed-off-by: Peter Maydell
---
trace/control-internal.h | 10 +-
1 file changed, 5 insertions(
Win32 doesn't have a cpuid.h, and MacOSX may have one but without
the __cpuid() function we use, which means that commit 9d2eec20
broke the build for those platforms. Fix this by tightening up
our configure cpuid.h check to test that the functions we need
are present, and adding some missing #ifdef
On Thu, Feb 20, 2014 at 03:57:21PM +0100, Kevin Wolf wrote:
> Instead of ignoring all option values but the last one, multiple -o
> options now have the same meaning as having a single option with all
> settings in the order of their respective -o options.
>
> Signed-off-by: Kevin Wolf
Reviewed-
On 02/20/2014 11:50 AM, Peter Maydell wrote:
> +unsigned a, b, c, d;
> +int max = __get_cpuid_max(0, 0);
> +
> +if (max >= 1) {
> +__cpuid(1, a, b, c, d);
> +}
> +return 0;
> }
You might as well check for __cpuid_count too while you're at it:
if (max >= 7)
__cpu
On 02/20/2014 12:33 PM, Jeff Cody wrote:
> On Thu, Feb 20, 2014 at 03:57:20PM +0100, Kevin Wolf wrote:
>> Instead of ignoring all option values but the last one, multiple -o
>> options now have the same meaning as having a single option with all
>> settings in the order of their respective -o optio
On 20 February 2014 19:09, Mario Smarduch wrote:
> Questionable in this patch - is cutting through several layers to set
> proxy properties. They should be set from "device" instance init
> before it's realized. The problem though is that unlike PCI
> that sets proxy and virtio-net properties via
On Thu, Feb 20, 2014 at 03:57:20PM +0100, Kevin Wolf wrote:
> Instead of ignoring all option values but the last one, multiple -o
> options now have the same meaning as having a single option with all
> settings in the order of their respective -o options.
>
> Signed-off-by: Kevin Wolf
> ---
> q
On Thu, Feb 20, 2014 at 03:57:19PM +0100, Kevin Wolf wrote:
> If you specified multiple -o options for qemu-img create, it would
> silently ignore all but the last one. This patch fixes the problem.
>
> Now multiple -o options has the same meaning as having a single option
> with all settings in t
Signed-off-by: Stefan Weil
---
tcg/tcg.c |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tcg/tcg.c b/tcg/tcg.c
index 7ff1a8f..44ec42e 100644
--- a/tcg/tcg.c
+++ b/tcg/tcg.c
@@ -41,7 +41,7 @@
#include "qemu/host-utils.h"
#include "qemu/timer.h"
-/* Note: the long term pl
Peter thanks.
Questionable in this patch - is cutting through several layers to set
proxy properties. They should be set from "device" instance init
before it's realized. The problem though is that unlike PCI
that sets proxy and virtio-net properties via its virtio_net_properites[],
the virtio-mmi
It does not need qemu-common.h. Including exec/gdbstub.h fixes a warning
from static code analyzers and avoids mismatching declarations for
xml_builtin.
Signed-off-by: Stefan Weil
---
stubs/gdbstub.c |4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/stubs/gdbstub.c b/stub
On Thu, Feb 20, 2014 at 03:57:23PM +0100, Kevin Wolf wrote:
> Signed-off-by: Kevin Wolf
> ---
> tests/qemu-iotests/082 | 187 +++
> tests/qemu-iotests/082.out | 436
> +
> tests/qemu-iotests/group | 1 +
> 3 files changed, 624 i
On 02/20/2014 11:38 AM, Amos Kong wrote:
> Stefan Fritsch just fixed a virtio-net driver bug [1], virtio-net won't
> filter out VLAN-tagged packets if VIRTIO_NET_F_CTRL_VLAN isn't negotiated.
>
> This patch added a new field to @RxFilterInfo to indicate if management
> uses the vlan table.
>
> [1
On 20 February 2014 18:12, Mario Smarduch wrote:
>
> Hello,
>
> any feedback on this patch, after a brief email exchange Anthony deferred to
> Peter.
>
> Lack of improper host features handling lowers 1g & 10g performance
> substantially on arm-kvm compared to xeon.
>
> We would like to have this
On Thu, Feb 20, Stefano Stabellini wrote:
> On Thu, 6 Feb 2014, Olaf Hering wrote:
> > @@ -253,6 +254,8 @@ static int ioreq_parse(struct ioreq *ioreq)
> > case BLKIF_OP_WRITE:
> > ioreq->prot = PROT_READ; /* from memory */
> > break;
> > +case BLKIF_OP_DISCARD:
> > +
On Thu, 20 Feb 2014, Olaf Hering wrote:
> On Thu, Feb 20, Stefano Stabellini wrote:
>
> > On Thu, 6 Feb 2014, Olaf Hering wrote:
> > > @@ -253,6 +254,8 @@ static int ioreq_parse(struct ioreq *ioreq)
> > > case BLKIF_OP_WRITE:
> > > ioreq->prot = PROT_READ; /* from memory */
> > >
On Thu, Feb 20, Stefano Stabellini wrote:
> On Thu, 6 Feb 2014, Olaf Hering wrote:
> > @@ -253,6 +254,8 @@ static int ioreq_parse(struct ioreq *ioreq)
> > case BLKIF_OP_WRITE:
> > ioreq->prot = PROT_READ; /* from memory */
> > break;
> > +case BLKIF_OP_DISCARD:
> > +
From: Olaf Hering
bdrv_acct_done was called unconditional. But in case the ioreq has no
segments there is no matching bdrv_acct_start call. This could lead to
bogus accounting values.
Found by code inspection.
Signed-off-by: Olaf Hering
Signed-off-by: Stefano Stabellini
---
hw/block/xen_disk
Signed-off-by: Stefano Stabellini
Acked-by: Paolo Bonzini
---
hw/ide/piix.c |3 +--
hw/xen/xen_platform.c |3 ++-
include/hw/ide.h |1 +
3 files changed, 4 insertions(+), 3 deletions(-)
diff --git a/hw/ide/piix.c b/hw/ide/piix.c
index 0eda301..40757eb 100644
--- a/hw/id
The following changes since commit 2ca92bb993991d6dcb8f68751aca9fc2ec2b8867:
Merge remote-tracking branch 'remotes/kraxel/tags/pull-usb-3' into staging
(2014-02-20 15:25:05 +)
are available in the git repository at:
git://xenbits.xen.org/people/sstabellini/qemu-dm.git xen-140220
for y
Hello,
any feedback on this patch, after a brief email exchange Anthony deferred to
Peter.
Lack of improper host features handling lowers 1g & 10g performance
substantially on arm-kvm compared to xeon.
We would like to have this fixed so we don't have to patch every new release
of qemu, espe
On 02/20/14 16:38, Gabriel L. Somlo wrote:
> On Thu, Feb 20, 2014 at 04:27:27PM +0100, Gerd Hoffmann wrote:
>> On Mi, 2014-02-19 at 15:40 -0500, Gabriel L. Somlo wrote:
>>> So I gave up on that relatively quickly, as there's no easy and
>>> convenient way to "harvest" a binary of just one table typ
On Thu, 6 Feb 2014, Olaf Hering wrote:
> @@ -253,6 +254,8 @@ static int ioreq_parse(struct ioreq *ioreq)
> case BLKIF_OP_WRITE:
> ioreq->prot = PROT_READ; /* from memory */
> break;
> +case BLKIF_OP_DISCARD:
> +return 0;
> default:
> xen_be_printf(&b
On 02/19/2014 10:54 PM, Wenchao Xia wrote:
> After this patch, hidden enum type BlockdevOptionsKind will not
> be generated, and other API can use enum BlockdevDriver.
>
> Signed-off-by: Wenchao Xia
> ---
> qapi-schema.json | 14 +-
> 1 files changed, 13 insertions(+), 1 deletions(
Paolo Bonzini writes:
> Il 20/02/2014 17:02, Andreas Färber ha scritto:
>> Not unexpected, it's the older; the - convention was introduced possibly
>> with QOM around start of 2012. Or at least there it's been enforced, and
QMP preferred '-' from the start, but insufficient decoupling from
comma
On 02/20/2014 09:54 AM, Markus Armbruster wrote:
>> +# When c is upper and no "_" appears before, do more checks
>> +if c.isupper() and (i > 0) and c_fun_str[i - 1] != "_":
>
> c_fun_str[i - 1]... what if i == 0?
How? We already had '(i > 0) and' prior to the use of i-1.
--
Eri
On 20 February 2014 17:50, Peter Maydell wrote:
> Win32 doesn't have a cpuid.h, and MacOSX may have one but without
> the __cpuid() function we use, which means that commit 9d2eec20
> broke the build for those platforms. Fix this by tightening up
> our configure cpuid.h check to test that the func
From: Aneesh Kumar K.V
Correctly update the htab_mask using the return value of
KVM_PPC_ALLOCATE_HTAB ioctl. Also we don't update sdr1
on GET_SREGS for HV. We check for external htab and if
found true, we don't need to update sdr1
Signed-off-by: Aneesh Kumar K.V
[ fixed pte group offset computa
From: Aneesh Kumar K.V
This support updating htab managed by the hypervisor. Currently we don't have
any user for this feature. This actually bring the store_hpte interface
in-line with the load_hpte one. We may want to use this when we want to
emulate henter hcall in qemu for HV kvm.
Signed-off
From: Aneesh Kumar K.V
For updating in kernel htab we need to provide both pte0 and pte1, hence update
the interface to take pte0 and pte1 together
Signed-off-by: Aneesh Kumar K.V
[ ldq_phys() API change, Greg Kurz ]
Signed-off-by: Greg Kurz
---
hw/ppc/spapr_hcall.c| 20 ++-
From: Aneesh Kumar K.V
With kvm enabled, we store the hash page table information in the hypervisor.
Use ioctl to read the htab contents. Without this we get the below error when
trying to read the guest address
(gdb) x/10 do_fork
0xc0098660 : Cannot access memory at address
0xc
On Mon, 17 Feb 2014 14:22:14 +0100
Greg Kurz wrote:
> Hi,
>
> This is a new tentative for the patches 2/5 to 5/5 from the "target-ppc:
> Add support for dumping guest memory using qemu gdb server" patchset:
>
> https://lists.nongnu.org/archive/html/qemu-ppc/2014-01/msg00380.html
>
> All patches
On 02/20/2014 07:57 AM, Kevin Wolf wrote:
> Signed-off-by: Kevin Wolf
> ---
> tests/qemu-iotests/082 | 187 +++
> tests/qemu-iotests/082.out | 436
> +
> tests/qemu-iotests/group | 1 +
> 3 files changed, 624 insertions(+)
> cr
Win32 doesn't have a cpuid.h, and MacOSX may have one but without
the __cpuid() function we use, which means that commit 9d2eec20
broke the build for those platforms. Fix this by tightening up
our configure cpuid.h check to test that the functions we need
are present, and adding some missing #ifdef
On 20 February 2014 16:43, Richard Henderson wrote:
> Can you try this?
[summarising an irc discussion]
That fixes W32 but not MacOSX, because there clang has a
cpuid.h but it doesn't have __get_cpuid_max() or __cpuid().
I'm just testing a patch which enhances our configure test
to check for the
Il 19/02/2014 15:06, Kevin Wolf ha scritto:
Thanks, applied to the block branch.
Conflicts with master... to help you rebasing I put my own version at
block-open-errors on git://github.com/bonzini/qemu.git.
Paolo
None of these needs QEMU_PROG, and they all take but a few seconds.
We need to point the launching script to qemu-nbd, though.
Signed-off-by: Paolo Bonzini
---
tests/qemu-iotests-quick.sh | 1 +
tests/qemu-iotests/group| 34 +-
2 files changed, 18 insertions(
Alex Williamson writes:
> On Wed, 2014-02-19 at 15:20 -0500, Bandan Das wrote:
>> Certain cards such as the Broadcom BCM57810 have rom quirks
>> that exhibit unstable system behavior duing device assignment. In
>> the particular case of 57810, rom execution hangs and if a FLR
>> follows, the devi
1 - 100 of 280 matches
Mail list logo
