On 2014/2/28 17:51, Dr. David Alan Gilbert wrote:
> * Gonglei (Arei) (arei.gong...@huawei.com) wrote:
>> Avoid the hot pages cache replacing by others to remarkable decrease cache
>> missing. The counter of updating dirty bitmap is used to indicate the cached
>> page age.
>
> It seems worth notin
On 2014/2/28 17:19, Dr. David Alan Gilbert wrote:
> * Gonglei (Arei) (arei.gong...@huawei.com) wrote:
>
> Hi,
>
>> a. Optimization the xbzrle remarkable decrease the cache misses.
>> The efficiency of compress increases more than fifty times.
>> Before the patch set, the cache almost tot
Il 28/02/2014 06:27, Stefan Weil ha scritto:
>> -$(SUBDIR_RULES): libqemuutil.a libqemustub.a $(common-obj-y)
>> +$(SUBDIR_RULES): libqemuutil.a libqemustub.a $(common-obj-y) $(block-obj-y)
Using SUBDIR_RULES here instead of SOFTMMU_SUBDIR_RULES (as in my patch)
includes more dependencies: $(blo
Signed-off-by: BALATON Zoltan
---
v2: Sorry, I was too fast to send it. Found two more places to change.
hw/pci-host/q35.c | 10 +-
include/hw/i386/ich9.h| 2 +-
include/hw/pci-host/q35.h | 24
3 files changed, 18 insertions(+), 18 deletions(-)
dif
Il 28/02/2014 11:41, Joaquim Barrera ha scritto:
Thanks for the answer. Something is still not clear to me. Are we in
front of a bug (that means, something that could be fixed) or is this
behaviour somehow expected for some reason? More and more tests I am
doing, I get allways the same through
On 28 February 2014 01:37, Rob Herring wrote:
> From: Rob Herring
>
> Offset 4 is UARTRSR/UARTECR, not the UARTCR. As framing and parity errors
> will never occur, we can ignore writes to this register.
It's true that we'll never get framing or parity errors,
but we can get break (and we ought t
On Wed, Feb 26, 2014 at 10:53:34AM +0100, Igor Mammedov wrote:
> On Fri, 21 Feb 2014 15:46:56 +0100
> Stefan Hajnoczi wrote:
>
> > get_pointer() assumes the string has unspecified lifetime (at least as
> > long as the object is alive). In some cases we can only produce a
> > temporary string tha
Signed-off-by: BALATON Zoltan
---
hw/pci-host/q35.c | 6 +++---
include/hw/i386/ich9.h| 2 +-
include/hw/pci-host/q35.h | 24
3 files changed, 16 insertions(+), 16 deletions(-)
diff --git a/hw/pci-host/q35.c b/hw/pci-host/q35.c
index 4bc2e01..eda64c5 100644
* Gonglei (Arei) (arei.gong...@huawei.com) wrote:
> Reducing data copy can reduce cpu overheah.
(Note a few typos in subject/title)
> Signed-off-by: ChenLiang
> Signed-off-by: Gonglei
> ---
> arch_init.c | 8 +++-
> 1 file changed, 3 insertions(+), 5 deletions(-)
>
> diff --git a/arch_ini
On 02/28/2014 09:27 PM, Stefan Hajnoczi wrote:
> On Fri, Feb 28, 2014 at 03:05:12PM +1100, Alexey Kardashevskiy wrote:
>> However once socket_scm_helper talked to "mon" (./qemu.monitor), I cannot
>> get any response from it via "socat UNIX-CONNECT:./qemu.monitor STDIN".
>
> Have you tried with QMP
On Thu, Feb 27, 2014 at 11:15:22AM +0100, Peter Lieven wrote:
> Signed-off-by: Peter Lieven
> ---
> block/vmdk.c |2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
Thanks, applied to my block tree:
https://github.com/stefanha/qemu/commits/block
Stefan
On Fri, Feb 28, 2014 at 03:05:12PM +1100, Alexey Kardashevskiy wrote:
> However once socket_scm_helper talked to "mon" (./qemu.monitor), I cannot
> get any response from it via "socat UNIX-CONNECT:./qemu.monitor STDIN".
Have you tried with QMP? Perhaps the HMP monitor handles
connect/disconnect d
We were forgetting to set the ipv4 and ipv6 QemuOpts properties, so
socket_dgram was always trying _both_ ipv4 and ipv6. Forward the
properties correctly.
Signed-off-by: Paolo Bonzini
---
util/qemu-sockets.c | 9 -
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/util/qemu-s
Signed-off-by: Christian Borntraeger
---
pc-bios/s390-ccw.img | Bin 9336 -> 9336 bytes
1 file changed, 0 insertions(+), 0 deletions(-)
diff --git a/pc-bios/s390-ccw.img b/pc-bios/s390-ccw.img
index
6727f0ca39d6bf6d114974d1535cb7ad9e56355f..f6223e77c2aacfa86652d63b773dc05eca55570d
100644
GIT b
In some cases, "ipv4=on,ipv6=on" means "try both kinds of address";
in others, it means "try IPv6 only" just by virtue of how the code
is structured.
Fix this to make things more consistent, and adjust coding style too.
Signed-off-by: Paolo Bonzini
---
util/qemu-sockets.c | 33 +
From: Heinz Graalfs
Code restructure in order to simplify class hierarchy
- remove S390SCLPDevice abstract base class
and move function pointers into new SCLPEventFacilityClass
- implement SCLPEventFacility as SysBusDevice
- use define constants for instance creation strings
The follow
* Gonglei (Arei) (arei.gong...@huawei.com) wrote:
> It is inaccuracy and complex that using the transfer speed of
> migration thread to determine whether the convergence migration.
> The dirty page may be compressed by XBZRLE or ZERO_PAGE.The counter
> of updating dirty bitmap will be increasing co
While reviewing the L2TPv3 submission, I wanted to tell Anton that
IPv4/IPv6 is already done in qemu-sockets.c. Turns out it is not
exactly the case; these patches fixes the two issues I spotted.
Paolo Bonzini (2):
socket: treat ipv4=on,ipv6=on uniformly
socket: handle ipv4/ipv6 in socket_dgr
From: Thomas Huth
The pointer to the SCCB should not be limited to 32 bits only.
In contrast to this, the command word parameter is only 32 bits
(the upper 32 bits should be ignored).
Signed-off-by: Thomas Huth
Reviewed-by: Cornelia Huck
Signed-off-by: Christian Borntraeger
---
hw/s390x/sclp
* Gonglei (Arei) (arei.gong...@huawei.com) wrote:
Hi,
> a. Optimization the xbzrle remarkable decrease the cache misses.
> The efficiency of compress increases more than fifty times.
> Before the patch set, the cache almost totally miss when the
> number of cache item less than the d
Anthony, Peter,
The following changes since commit d5001cf787ad0514839a81d0f2e771e01e076e21:
xilinx: Delete hw/include/xilinx.h (2014-02-26 14:54:45 +1000)
are available in the git repository at:
git://github.com/borntraeger/qemu.git tags/kvm-s390-20140227
for you to fetch changes up to 7f
On Fri, Feb 21, 2014 at 09:24:47AM +0100, Alexander Graf wrote:
> Anthony / Stefan, could you please update the SLOF.git mirror on git.qemu.org?
Done. The SLOT.git mirror should now update automatically every day.
Please let me know if you notice any issues.
Stefan
This transport uses a linux specific call to get several GBit RX rate.
Might as well mention that it's recvmmsg. :)
This call can be wrapped (at some cost) in a compatibility loop using
posix compliant recvmsg instead for other systems. As I am not familiar
with the fine details on how to add
* Gonglei (Arei) (arei.gong...@huawei.com) wrote:
> Rebuild the cache_is_cached function.
>
> Signed-off-by: ChenLiang
> Signed-off-by: Gonglei
> ---
> page_cache.c | 33 +++--
> 1 file changed, 15 insertions(+), 18 deletions(-)
>
> diff --git a/page_cache.c b/page_
* Gonglei (Arei) (arei.gong...@huawei.com) wrote:
> Avoid the hot pages cache replacing by others to remarkable decrease cache
> missing. The counter of updating dirty bitmap is used to indicate the cached
> page age.
It seems worth noting that you've changed 'it_age's use - that's ok, but
perhaps
On 26/02/14 17:25, Peter Maydell wrote:
On 26 February 2014 16:59, Fabien Chouteau wrote:
On 02/26/2014 08:56 AM, Sebastian Huber wrote:
Hello,
exists there someone who is able to commit this?
I'm sorry Sebastian, as you probably understood the SPARC maintainer is
missing which makes comm
* Gonglei (Arei) (arei.gong...@huawei.com) wrote:
> Add counters to log the times of updating the dirty bitmap.
Reviewed-by: Dr. David Alan Gilbert
> Signed-off-by: ChenLiang
> Signed-off-by: Gonglei
> ---
> arch_init.c | 20
> 1 file changed, 20 insertions(+)
>
> diff -
From: Thomas Huth
Some of the state in the kernel can not be reset from QEMU yet.
For this we've got to use the KVM_S390_INITIAL_RESET ioctl to make
sure that the state in the kernel is set to the right values during
initial CPU reset, too.
Signed-off-by: Thomas Huth
Acked-by: Cornelia Huck
Si
From: Thomas Huth
When loading S390 kernels, the current code expects an ELF file with the
start address 0x1. Other ELF files cause a segmentation fault. To avoid
these crashes, we should get the start address from the ELF file instead
of always using a hard-coded address.
Signed-off-by: Tho
From: Thomas Huth
The SET-ARCHITECTURE handler in QEMU caused a program interruption.
This is wrong according to the "Principles of Operations" specification
(since SIGP should never cause a program interrupt) and was likely only
introduced for debugging purposes. Since we handle SET-ARCHITECTURE
From: Heinz Graalfs
init/exit functionality of abstract SCLPEvent class is now exploiting
realize/unrealize.
Signed-off-by: Heinz Graalfs
Reviewed-by: Cornelia Huck
Signed-off-by: Christian Borntraeger
---
hw/s390x/event-facility.c | 27 ++-
1 file changed, 18 inserti
From: Jens Freimann
This patch implements a floating-interrupt controller device (flic)
which interacts with the s390 flic kvm_device.
Signed-off-by: Jens Freimann
Reviewed-by: Cornelia Huck
Signed-off-by: Christian Borntraeger
---
default-configs/s390x-softmmu.mak | 1 +
hw/intc/Makefile.
From: Thomas Huth
The s390_cpu_initial_reset() function had two deficiencies: First, it
used an ioctl for the destination CPU, and this ioctl could block
nearly forever, as long as the destination CPU was running in the SIE
loop. Second, it also cleared the general purpose registers - something
i
From: Heinz Graalfs
Add support for live migration using VMStateDescription.
Signed-off-by: Heinz Graalfs
Reviewed-by: Cornelia Huck
Signed-off-by: Christian Borntraeger
---
hw/s390x/event-facility.c | 12
1 file changed, 12 insertions(+)
diff --git a/hw/s390x/event-facility.c
From: Frank Blaschka
The current implementation uses the second byte of the instruction
to identify the instruction handler. This is not sufficient to
support instructions not starting with 0xb2. This patch
adds separate handlers for 0xb2, 0xb9 and 0xeb to be able to
support the full instruction
From: Thomas Huth
The SIGP order CPU RESET was still missing in the list of our
supported handler. This patch now adds a simple implementation,
by using the cpu_reset() function that is already available in
target-s390x/cpu.c.
Signed-off-by: Thomas Huth
Reviewed-by: Christian Borntraeger
Revie
We have to set the cssid to 0, otherwise the stsch code will
return an operand exception without the m bit. In the same way
we should set m=0.
This case was triggered in some cases during reboot, if for some
reason the location of blk_schid.cssid contains 1 and m was 0.
Turns out that the qemu elf
As a followup to commit 5f04c14a10fa7f259bc0808f35a0beda49f7821e
(s390-sclp: Define New SCLP Codes) we should mask the sclp command
not only in base sclp, but also in the event facility.
Based on an initial patch from Ralf Hoppe.
Signed-off-by: Christian Borntraeger
---
hw/s390x/event-facility.
From: Heinz Graalfs
Do some renaming to shorten some identifiers and to emphasize sclp.
Signed-off-by: Heinz Graalfs
Acked-by: Cornelia Huck
Signed-off-by: Christian Borntraeger
---
hw/s390x/event-facility.c | 18 +-
1 file changed, 9 insertions(+), 9 deletions(-)
diff --git
This updates the kvm headers to
commit d3714010c307d26df251c45be9cd12ab6d41f0c4
KVM: x86: emulator_cmpxchg_emulated should mark_page_dirty
in kvm/next.
Signed-off-by: Christian Borntraeger
---
linux-headers/asm-arm/kvm.h | 28 +
linux-headers/asm-arm64/kvm.h | 30 ++
The current code does not initialize next_idx in the virtio ring.
As the ccw bios will always use guest memory at a fixed location,
this queue might != 0 after a reboot.
Lets make the initialization explicit.
Signed-off-by: Christian Borntraeger
Reviewed-by: Cornelia Huck
---
pc-bios/s390-ccw/v
From: Thomas Huth
If the 51 most significant bits of the SCCB address are zero or equal to
the prefix, we should throw an specification exception, too.
Also moved the check for privileged mode to sclp_service_call() to have
all program checks in one place now.
Signed-off-by: Thomas Huth
Reviewe
From: Thomas Huth
In the SCLP handler function, the condition code register must
only be set if no exception occured.
Signed-off-by: Thomas Huth
Reviewed-by: Cornelia Huck
Signed-off-by: Christian Borntraeger
---
target-s390x/kvm.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
dif
From: Dominik Dingel
S390 can also use async page faults, to enhance guest scheduling.
In case of live migration we want to disable the feature and let
all pending request finish.
Signed-off-by: Dominik Dingel
Signed-off-by: Jens Freimann
Signed-off-by: Christian Borntraeger
---
hw/intc/s390
From: Thomas Huth
The handler for diag 500 did not check whether the requested function
was in the supported range, so illegal values could crash QEMU in the
worst case.
Signed-off-by: Thomas Huth
Reviewed-by: Cornelia Huck
Signed-off-by: Christian Borntraeger
CC: qemu-sta...@nongnu.org
---
From: Thomas Huth
So far, the DIAG 500 hypervisor call was only setting -EINVAL in
R2 when a guest tried to call this function with an illegal subcode.
This patch now changes the behavior so that a specification exception
is thrown instead, since this is the common behavior of other DIAG
function
* Gonglei (Arei) (arei.gong...@huawei.com) wrote:
> The page may not be inserted into cache after executing save_xbzrle_page.
> In case of failure to insert, the original page should be sent rather
> than the page in the cache.
>
> Signed-off-by: ChenLiang
> Signed-off-by: Gonglei
Nice catch.
ping?
Richard, do you prefer to spin your polishing from last year instead?
We can start with the pure refactoring patch as you suggested, and in
this case, you could look here for some things I'd look for,
in particular that we move gradually, that we group instructions that
are handled in a sim
Hi all,
On behalf of Cisco Systems I am authorized to contribute a new transport
to the network subsystem in qemu.
Specifically, we would like to contribute a new transport:
L2TPv3 static tunnel network transport.
Earlier versions of this (we have patchsets going back to qemu-1.0) have
been use
Il 28/02/2014 05:55, Michael S. Tsirkin ha scritto:
Why?
We already have different ACPI tables for TCG vs KVM.
Specifically apic interrupt override flag in MADT is set
for KVM but not TCG.
It used to be this way, but
bool kvm_allows_irq0_override(void)
{
return !kvm_irqchip_in_kerne
Il 28/02/2014 05:55, Michael S. Tsirkin ha scritto:
Why?
We already have different ACPI tables for TCG vs KVM.
Specifically apic interrupt override flag in MADT is set
for KVM but not TCG.
It used to be this way, but
bool kvm_allows_irq0_override(void)
{
return !kvm_irqchip_in_kerne
On 27 February 2014 15:49, Michael S. Tsirkin wrote:
> > Michael: Luke has asked to increase the virtio-net virtqueue size.
> > Thoughts?
> >
> > Stefan
>
> Heh you want to increase the bufferbloat?
>
I'm sensitive to this. (I have actually built a commercial anti-bufferbloat
network device for
201 - 252 of 252 matches
Mail list logo
