On 07/16/2015 03:11 PM, David Gibson wrote:
On Tue, Jul 14, 2015 at 10:21:54PM +1000, Alexey Kardashevskiy wrote:
This makes use of the new "memory registering" feature. The idea is
to provide the userspace ability to notify the host kernel about pages
which are going to be used for DMA. Having
On Thu, 16 Jul 2015 19:22:21 +0200
Paolo Bonzini wrote:
>
>
> On 16/07/2015 14:47, Michael S. Tsirkin wrote:
> > I think for 2.4 it's a good idea to avoid enabling modern interface
> > by default. Therefore, for 2.4 we can keep scsi enabled unless modern
> > is requested by user.
>
> I agree.
Commit 032a74a1c0fcdd5fd1c69e56126b4c857ee36611
("virtio-net: byteswap virtio-net header") breaks any layout by
requiring out_sg[0].iov_len >= n->guest_hdr_len. Fixing this by
copying header to temporary buffer if swap is needed, and then use
this buffer as part of out_sg.
Fixes 032a74a1c0fcdd5fd1
On 07/16/2015 11:39 PM, Eduardo Habkost wrote:
This fixes the following crash, introduced by commit
49d2e648e8087d154d8bf8b91f27c8e05e79d5a6:
$ gdb --args qemu-system-x86_64 -machine pc,mem-merge=off -object
memory-backend-ram,id=ram-node0,size=1024
[...]
Program received signal SIGABR
Signed-off-by: Jason Wang
---
Changes from V2:
- Remove duplicated code of gtest_start()
Changes from V1:
- replace the magic value 12 with a macro
---
tests/Makefile | 2 +-
tests/virtio-net-test.c | 185 ++--
2 files changed, 179 insertions
On 17 July 2015 at 07:53, Thomas Huth wrote:
> Ok, assuming that my "Network traffic dumping for -netdev devices" patch
> series is going to solve the dumping-for-netdev problem, how do we
> tackle the remaining problems that we have to solve before we can
> deprecate -net? Does anybody have a sur
Rx should be recovered after cont.
Signed-off-by: Jason Wang
---
Changes from V2:
- really query status before "cont"
Changes from V1:
- query status before sending cont, this makes sure that the packet
were queued.
---
tests/virtio-net-test.c | 52 +
On 07/16/2015 03:24 PM, Igor Mammedov wrote:
On Thu, 16 Jul 2015 16:00:07 +0530
Bharata B Rao wrote:
On Thu, Jul 16, 2015 at 10:25:09AM +0530, Bharata B Rao wrote:
Hi,
When I use -machine pc,mem-merge=off|on -m 1G,slots=4,maxmem=2G, adding
memory-backend-ram object fails like below. Same fai
Unfortunately I seem to be unable to get this bug triggered with the
reproducer. It could be a detail of the guest setup I am missing. Since I do
not have access to RHEL I used CentOS 6.3 in a 8core guest with 2 virtio disks.
Host was 14.04. Left the script running for quite a bit but no crash h
On 07/17/2015 09:25 AM, Peter Maydell wrote:
> On 17 July 2015 at 07:53, Thomas Huth wrote:
>> Ok, assuming that my "Network traffic dumping for -netdev devices" patch
>> series is going to solve the dumping-for-netdev problem, how do we
>> tackle the remaining problems that we have to solve befor
On Fri, Jul 17, 2015 at 10:26:16AM +0300, Marcel Apfelbaum wrote:
> On 07/16/2015 03:24 PM, Igor Mammedov wrote:
> >On Thu, 16 Jul 2015 16:00:07 +0530
> >Bharata B Rao wrote:
> >
> >>On Thu, Jul 16, 2015 at 10:25:09AM +0530, Bharata B Rao wrote:
> >>>Hi,
> >>>
> >>>When I use -machine pc,mem-merge
On Fri, Jul 17, 2015 at 08:53:08AM +0200, Thomas Huth wrote:
> On 05/26/2015 04:29 PM, Markus Armbruster wrote:
> > Stefan Hajnoczi writes:
> [...]
> >> We thought the QEMU "vlan" concept would be dropped completely in the
> >> future, so it was never added to -netdev. No patches to do that have
On Fri, Jul 17, 2015 at 06:17:47AM +0200, Paolo Bonzini wrote:
>
>
> On 16/07/2015 11:56, Paolo Bonzini wrote:
> > @@ -286,13 +283,15 @@ bool aio_poll(AioContext *ctx, bool blocking)
> > npfd = 0;
> > ctx->walking_handlers--;
> >
> > +if (blocking) {
> > +atomic_sub(&ctx->
On 17/07/2015 06:44, Paolo Bonzini wrote:
>
>
> On 16/07/2015 21:05, Richard W.M. Jones wrote:
>>
>> Sorry to spoil things, but I'm still seeing this bug, although it is
>> now a lot less frequent with your patch. I would estimate it happens
>> more often than 1 in 5 runs with qemu.git, and pr
On Fri, 17 Jul 2015, Nils Carlson wrote:
> Hi,
>
> The commit 812c1057f, Handle G_IO_HUP in tcp_chr_read for tcp chardev, broke
> CloudStack. CloudStack was relying on fire-and-forget style messaging across a
> unix socket to the VM. Because the host "fires" the message and then closes
> the sock
Alvise Rigo writes:
> Create a new pair of instructions that implement a LoadLink/StoreConditional
> mechanism.
>
> It has not been possible to completely include the two new opcodes
> in the plain variants, since the StoreConditional will always require
> one more argument to store the success
16.07.2015 20:52, Paolo Bonzini пишет:
On 16/07/2015 16:41, Ефимов Василий wrote:
The main problem is rendering memory tree to FlatView.
I don't believe it's necessary to render a memory tree to the FlatView.
You can use existing AddressSpaces.
+/* Read from RAM and write to PCI */
+
QEMU silently ignores aio=native if libaio is unavailable. It is
confusing when aio=native performance is identical to aio=threads
because the binary was accidentally built without libaio.
Use error_report() because failing would break backward compatibility.
There are probably users using aio=na
You are right. I will fix this case as in gen_ldst_i32().
Thanks,
alvise
On Fri, Jul 17, 2015 at 11:49 AM, Alex Bennée wrote:
>
> Alvise Rigo writes:
>
>> Create a new pair of instructions that implement a LoadLink/StoreConditional
>> mechanism.
>>
>> It has not been possible to completely incl
On 17/07/2015 11:50, Ефимов Василий wrote:
> I will try solution based on address spaces as stated above.
If it works, that would be great.
Paolo
On 2015-07-16 22:25, Richard Henderson wrote:
> This saves 2 insns and 10 bytes from the implementation of
> each memory operation.
>
> Signed-off-by: Richard Henderson
> ---
> tcg/i386/tcg-target.c | 120
> +++---
> 1 file changed, 56 insertions(+),
On 2015-07-16 22:25, Richard Henderson wrote:
> Removing the ??? comment explaining why it (mostly) worked.
>
> Signed-off-by: Richard Henderson
> ---
> tcg/i386/tcg-target.c | 105
> +++---
> 1 file changed, 65 insertions(+), 40 deletions(-)
>
> dif
On 2015-07-16 22:29, Richard Henderson wrote:
> On 07/15/2015 09:54 PM, Aurelien Jarno wrote:
> >While I understand why we need the new trunc_shr_i32 opcode for MIPS64
> >(the 32-bit values must be kept sign-extended), I currently fail to
> >see why it is needed for SPARC.
>
> As far as I recall,
On 2015-07-17 07:38, Richard Henderson wrote:
> On 07/15/2015 12:03 PM, Aurelien Jarno wrote:
> >Now that we have real size changing ops, we don't need to marked high
> >bits of the destination as garbage. The goal of the optimizer is to
> >predict the value of the temps (and not of the registers)
Hi
On Fri, Jul 17, 2015 at 4:25 AM, Paolo Bonzini wrote:
> But LOG_BASE makes little sense across processes, and LOG_FD is unused
> in QEMU, isn't it? So this patch is not enough to add support of live
> migration.
You are right, LOG_BASE doesn't make much sense, and LOG_FD isn't
used, despite
On Thu, Jul 16, 2015 at 10:55:22AM +0100, Richard W.M. Jones wrote:
> The prototypes of the {nettle_}cbc_encrypt and cbc_decrypt functions
> are:
>
> void
> cbc_encrypt(const void *ctx, nettle_cipher_func *f,
> size_t block_size, uint8_t *iv,
> size_t length, uint8_t *dst,
On Tue, Jul 14, 2015 at 02:51:40PM +0200, Wolfgang Bumiller wrote:
> Commit 800567a61 updated the code to the generic crypto API
> and mixed up encrypt and decrypt functions in
> procotol_client_auth_vnc.
> (Used to be: deskey(key, EN0) which encrypts, and was
> changed to qcrypto_cipher_decrypt in
On Thu, Jul 16, 2015 at 06:03:59PM +0200, Radim Krčmář wrote:
> In nettle 3, cbc_encrypt() accepts 'nettle_cipher_func' instead of
> 'nettle_crypt_func' and these two differ in 'const' qualifier of the
> first argument. The build fails with:
>
> In file included from crypto/cipher.c:71:0:
> .
On Thu, Jul 16, 2015 at 06:04:00PM +0200, Radim Krčmář wrote:
> Calling a function pointer that was cast from an incompatible function
> results in undefined behavior. 'void *' isn't compatible with 'struct
> XXX *', so we can't cast to nettle_cipher_func, but have to provide a
> wrapper. (Conver
Am 17.07.2015 um 11:59 hat Stefan Hajnoczi geschrieben:
> QEMU silently ignores aio=native if libaio is unavailable. It is
> confusing when aio=native performance is identical to aio=threads
> because the binary was accidentally built without libaio.
>
> Use error_report() because failing would b
On Thu, Jul 16, 2015 at 06:04:01PM +0200, Radim Krčmář wrote:
> It's horrible both ways and I prefer this one.
>
> Signed-off-by: Radim Krčmář
As you say its ugly either way, but I think the amount of duplicated
code pattern is not too bad, and IMHO it is slightly clearer to read
without the mac
On 17/07/2015 13:00, Daniel P. Berrange wrote:
> > > On 16/07/2015 18:04, Radim Krčmář wrote:
> > > It's horrible both ways and I prefer this one.
> >
> > Let's see what Dan thinks. Since he's on vacation, for now I'm applying
> > patches 1 and 2 only.
>
> I don't have a strong opinion on patch
On Thu, Jul 16, 2015 at 06:11:12PM +0200, Paolo Bonzini wrote:
>
>
> On 16/07/2015 18:04, Radim Krčmář wrote:
> > It's horrible both ways and I prefer this one.
>
> Let's see what Dan thinks. Since he's on vacation, for now I'm applying
> patches 1 and 2 only.
I don't have a strong opinion on
Hi
Please, send any topic that you are interested in covering.
Call details:
By popular demand, a google calendar public entry with it
https://www.google.com/calendar/embed?src=dG9iMXRqcXAzN3Y4ZXZwNzRoMHE4a3BqcXNAZ3JvdXAuY2FsZW5kYXIuZ29vZ2xlLmNvbQ
(Let me know if you have any problems wi
e remote-tracking branch 'remotes/lalrae/tags/mips-20150716' into
> staging (2015-07-16 10:40:23 +0100)
>
> are available in the git repository at:
>
>
> git://git.kraxel.org/qemu tags/pull-input-20150717-1
>
> for you to fetch changes
From: Lin Ma
Signed-off-by: Lin Ma
Reviewed-by: Gonglei
Signed-off-by: Gerd Hoffmann
---
hw/usb/host-libusb.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/hw/usb/host-libusb.c b/hw/usb/host-libusb.c
index bc2944c..11429f5 100644
--- a/hw/usb/host-libusb.c
+++ b/hw/usb/host-libusb.c
This reverts commit 4e8cfbe1143d8384387595b500212d7a7f11aeae.
We should not poll via timer, and with ccid being fixed
to properly notify us about pending transfers we don't have to.
Signed-off-by: Gerd Hoffmann
---
hw/usb/hcd-xhci.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/hw/usb/hc
Properly notify the host adapter that we have
data pending, so it doesn't has to poll us.
Signed-off-by: Gerd Hoffmann
---
hw/usb/dev-smartcard-reader.c | 5 +
1 file changed, 5 insertions(+)
diff --git a/hw/usb/dev-smartcard-reader.c b/hw/usb/dev-smartcard-reader.c
index 904b568..8952eff 1
t:
git://git.kraxel.org/qemu tags/pull-usb-20150717-1
for you to fetch changes up to 92fdfa4bef9c92addcc009dd3e0131172b4fdc78:
Revert "xhci: set timer to retry xfers" (2015-07-17 13:20:53 +0200)
usb: fixes for 2.4 (cci
QOMification dropped the parent device lookup, fix it.
Signed-off-by: Gerd Hoffmann
---
hw/usb/dev-smartcard-reader.c | 16
1 file changed, 8 insertions(+), 8 deletions(-)
diff --git a/hw/usb/dev-smartcard-reader.c b/hw/usb/dev-smartcard-reader.c
index de534ba..904b568 100644
-
Hi Naredula
On Mon, Jul 13, 2015 at 10:35 AM, Naredula Janardhana Reddy
wrote:
>
> In the QEMU 2.3.0 , The memory regions published to user space application
> is limited address space, not the complete memory regions of quest vm, due
> to this , the user level app cannot decode all the memory bl
Signed-off-by: Andrew Jones
---
hw/ppc/spapr.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/hw/ppc/spapr.c b/hw/ppc/spapr.c
index a6f19473cf278..c1cbf3387ae0c 100644
--- a/hw/ppc/spapr.c
+++ b/hw/ppc/spapr.c
@@ -30,6 +30,7 @@
#include "hw/fw-path-provider.h"
#include "elf.h"
#include
I've started playing with adding ppc support to kvm-unit-tests,
using spapr for the machine model. I wanted to link the unit test
at 0x40 to match qemu's load address, making the unit test
startup code simpler, but ended up with 0x80 instead, due to
how translate_kernel_address works. The t
I've started playing with adding ppc support to kvm-unit-tests,
using spapr for the machine model. Here are a couple patches
to spapr I've made.
Thanks,
drew
Andrew Jones (2):
spapr: add dumpdtb support
spapr: -kernel: allow linking with specified addr
hw/ppc/spapr.c | 5 +
1 file chang
pc_dimm_realize() validates the NUMA node to which memory hotplug is
being performed only in case of NUMA configuration. Include a check
to fail invalid nodes in case of non-NUMA configuration too.
Signed-off-by: Bharata B Rao
---
hw/mem/pc-dimm.c | 5 +++--
1 file changed, 3 insertions(+), 2 de
716' into
> staging (2015-07-16 10:40:23 +0100)
>
> are available in the git repository at:
>
>
> git://git.kraxel.org/qemu tags/pull-usb-20150717-1
>
> for you to fetch changes up to 92fdfa4bef9c92addcc009dd3e0131172b4fdc78:
>
>
Alvise Rigo writes:
> Implement strex and ldrex instruction relying on TCG's qemu_ldlink and
> qemu_stcond. For the time being only the 32bit instructions are supported.
>
> Suggested-by: Jani Kokkonen
> Suggested-by: Claudio Fontana
> Signed-off-by: Alvise Rigo
> ---
> target-arm/translate
Alvise Rigo writes:
> Implement strex and ldrex instruction relying on TCG's qemu_ldlink and
> qemu_stcond. For the time being only 32bit configurations are supported.
>
> Suggested-by: Jani Kokkonen
> Suggested-by: Claudio Fontana
> Signed-off-by: Alvise Rigo
> ---
> tcg/i386/tcg-target.c
On 17/07/2015 12:34, Marc-André Lureau wrote:
> On Fri, Jul 17, 2015 at 4:25 AM, Paolo Bonzini wrote:
> > But LOG_BASE makes little sense across processes, and LOG_FD is unused
> > in QEMU, isn't it? So this patch is not enough to add support of live
> > migration.
>
> You are right, LOG_BASE d
On Fri, Jul 17, 2015 at 09:30:03AM +0300, Catalin Vasile wrote:
> Do you mean vhost_net - old kernel, qemu - latest, guest - latest?
I mean whatever combination didn't work for your custom vhost module.
If vhost_net is broken too there is probably a vhost/QEMU bug that
should be fixed.
Stefan
p
On Fri, Jul 17, 2015 at 2:51 PM, Alex Bennée wrote:
>
> Alvise Rigo writes:
>
>> Implement strex and ldrex instruction relying on TCG's qemu_ldlink and
>> qemu_stcond. For the time being only the 32bit instructions are supported.
>>
>> Suggested-by: Jani Kokkonen
>> Suggested-by: Claudio Fontan
On Fri, Jul 17, 2015 at 11:30:38AM +0200, Paolo Bonzini wrote:
> error: kvm run failed Function not implemented
> PC=bf671210 SP=c1f0
> X00=0a003e70 X01= X02=bf680548
> X03=0030
> X04=bbb5fc18 X05=004b7770 X06=000
On 17/07/2015 14:58, Richard W.M. Jones wrote:
> On Fri, Jul 17, 2015 at 11:30:38AM +0200, Paolo Bonzini wrote:
>> error: kvm run failed Function not implemented
>> PC=bf671210 SP=c1f0
>> X00=0a003e70 X01= X02=bf680548
>> X03=0030
Get/put routines are missing, live migration is not possible.
Signed-off-by: Pavel Fedin
---
hw/intc/Makefile.objs | 3 +
hw/intc/arm_gicv3_kvm.c | 159
2 files changed, 162 insertions(+)
create mode 100644 hw/intc/arm_gicv3_kvm.c
diff --git
This series introduces support for GICv3 by KVM. Software emulation is
currently not supported.
Differences from v4:
- Do not reintroduce several constants shared with GICv2, reuse them instead.
- Added gicv3_init_irqs_and_mmio() in base class, to be used by both software
emulation and KVM code.
Set kernel_irqchip_type according to value of the option and pass it
around where necessary. Instantiate devices and fdt nodes according
to the choice.
mac_cpus for virt machine increased to 64. GICv2 compatibility check
happens inside arm_gic_common_realize().
Signed-off-by: Pavel Fedin
---
hw
This patch introduces kernel_irqchip_type member in Machine class, which
it passed to kvm_arch_irqchip_create. It allows machine models to specify
correct GIC type during KVM capability verification. The variable is
defined as int in order to be architecture-agnostic for potential future
uses by ot
From: Shlomo Pongratz
This class is to be used by both software and KVM implementations of GICv3
Signed-off-by: Shlomo Pongratz
Signed-off-by: Pavel Fedin
---
hw/intc/Makefile.objs | 1 +
hw/intc/arm_gicv3_common.c | 260 +
hw/intc/gi
On Fri, Jul 17, 2015 at 2:56 PM, Alex Bennée wrote:
>
> Alvise Rigo writes:
>
>> Implement strex and ldrex instruction relying on TCG's qemu_ldlink and
>> qemu_stcond. For the time being only 32bit configurations are supported.
>>
>> Suggested-by: Jani Kokkonen
>> Suggested-by: Claudio Fontana
These functions are useful also for vGICv3 implementation. Make them accessible
from within other modules.
Actually kvm_dist_get() and kvm_dist_put() could also be made reusable, but
they would require two extra parameters (s->dev_fd and s->num_cpu) as well as
lots of typecasts of 's' to DeviceSta
Alvise Rigo writes:
> Suggested-by: Jani Kokkonen
> Suggested-by: Claudio Fontana
> Signed-off-by: Alvise Rigo
> ---
> include/exec/ram_addr.h | 6 +++---
> 1 file changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/include/exec/ram_addr.h b/include/exec/ram_addr.h
> index 2766541..e51
On Fri, 17 Jul 2015 10:30:38 +0100
Paolo Bonzini wrote:
>
>
> On 17/07/2015 06:44, Paolo Bonzini wrote:
> >
> >
> > On 16/07/2015 21:05, Richard W.M. Jones wrote:
> >>
> >> Sorry to spoil things, but I'm still seeing this bug, although it is
> >> now a lot less frequent with your patch. I wo
Hi
On Fri, Jul 17, 2015 at 2:57 PM, Paolo Bonzini wrote:
> LOG_FD is implemented in the kernel drivers/vhost/vhost.c. It seems to
> be an eventfd-like mechanism to save on dirty bitmap scans. However,
> it's not well documented how to implement it in a correct way.
and it's not used by qemu, s
as planned, the GNU ARM Eclipse QEMU branch is now fully functional for simple
LED blinking applications.
the board image is displayed in a SDL window and the LEDs are coloured
rectangles shown over the image.
a short demo video is available from:
http://gnuarmeclipse.livius.net/mw/images/f
On 07/17/15 15:28, Marc Zyngier wrote:
> On Fri, 17 Jul 2015 10:30:38 +0100
> Paolo Bonzini wrote:
>
>>
>>
>> On 17/07/2015 06:44, Paolo Bonzini wrote:
>>>
>>>
>>> On 16/07/2015 21:05, Richard W.M. Jones wrote:
Sorry to spoil things, but I'm still seeing this bug, although it is
no
On Thu, Jul 16, 2015 at 04:46:07PM -0400, Programmingkid wrote:
> @@ -2014,7 +2015,9 @@ kern_return_t GetBSDPath( io_iterator_t mediaIterator,
> char *bsdPath, CFIndex ma
> if ( bsdPathAsCFString ) {
> size_t devPathLength;
> strcpy( bsdPath, _PATH_DEV );
> -
On Fri, Jul 17, 2015 at 05:13:37PM +1000, Alexey Kardashevskiy wrote:
> On 07/16/2015 03:11 PM, David Gibson wrote:
> >On Tue, Jul 14, 2015 at 10:21:54PM +1000, Alexey Kardashevskiy wrote:
> >>This makes use of the new "memory registering" feature. The idea is
> >>to provide the userspace ability t
On 2015-07-17 12:23, Aurelien Jarno wrote:
> On 2015-07-16 22:29, Richard Henderson wrote:
> > On 07/15/2015 09:54 PM, Aurelien Jarno wrote:
> > >While I understand why we need the new trunc_shr_i32 opcode for MIPS64
> > >(the 32-bit values must be kept sign-extended), I currently fail to
> > >see
Alvise Rigo writes:
> When a vCPU is about to set a memory page as exclusive, it needs to wait
> that all the running vCPUs finish to execute the current TB and to be aware
> of the exact moment when that happens. For this, add a simple rendezvous
> mechanism that will be used in softmmu_llsc_te
On Fri, 17 Jul 2015 14:39:55 +0100
Laszlo Ersek wrote:
> On 07/17/15 15:28, Marc Zyngier wrote:
> > On Fri, 17 Jul 2015 10:30:38 +0100
> > Paolo Bonzini wrote:
> >
> >>
> >>
> >> On 17/07/2015 06:44, Paolo Bonzini wrote:
> >>>
> >>>
> >>> On 16/07/2015 21:05, Richard W.M. Jones wrote:
>
>
From: Pankaj Gupta
This patch triggers timer only when guest requests for
entropy. As soon as first request from guest for entropy
comes we set the timer. Timer bumps up the quota value
when it gets triggered.
Signed-off-by: Pankaj Gupta
Reviewed-by: Amit Shah
Message-Id: <1436962608-9961-2-gi
The following changes since commit 5b5e8cdd7da7a2214dd062afff5b866234aab228:
Merge remote-tracking branch 'remotes/kraxel/tags/pull-usb-20150717-1' into
staging (2015-07-17 12:39:12 +0100)
are available in the git repository at:
https://git.kernel.org/pub/scm/virt/qemu/amit/virt
On 17/07/2015 15:35, Marc-André Lureau wrote:
>> > LOG_FD is implemented in the kernel drivers/vhost/vhost.c. It seems to
>> > be an eventfd-like mechanism to save on dirty bitmap scans. However,
>> > it's not well documented how to implement it in a correct way.
> and it's not used by qemu, so
On Fri, Jul 17, 2015 at 02:48:40PM +0100, Marc Zyngier wrote:
> Still: there is nothing in the registers that remotely points to that
> area. X0 is the closest, but it'd take a big negative offset to get
> there.
>
> Is that a Linux kernel? or something else?
You're sure it's not this one?
https:
On (Thu) 16 Jul 2015 [02:34:34], Pankaj Gupta wrote:
> > Anyway we can look at that later, patch 1 is already a big improvement
> > so I think we should just stick with that, and think about other
> > things in a different series.
>
> Sure.
I think we can apply patch 1 for 2.4, since it reduces t
On Fri, Jul 17, 2015 at 3:45 PM, Alex Bennée wrote:
>
> Alvise Rigo writes:
>
>> When a vCPU is about to set a memory page as exclusive, it needs to wait
>> that all the running vCPUs finish to execute the current TB and to be aware
>> of the exact moment when that happens. For this, add a simple
On 07/17/15 15:48, Marc Zyngier wrote:
> On Fri, 17 Jul 2015 14:39:55 +0100
> Laszlo Ersek wrote:
>
>> On 07/17/15 15:28, Marc Zyngier wrote:
>>> On Fri, 17 Jul 2015 10:30:38 +0100
>>> Paolo Bonzini wrote:
>>>
On 17/07/2015 06:44, Paolo Bonzini wrote:
>
>
> On 16/07/20
On Fri, Jul 17, 2015 at 12:56:15PM +0200, Kevin Wolf wrote:
> Am 17.07.2015 um 11:59 hat Stefan Hajnoczi geschrieben:
> > QEMU silently ignores aio=native if libaio is unavailable. It is
> > confusing when aio=native performance is identical to aio=threads
> > because the binary was accidentally b
On Fri, 17 Jul 2015 14:53:06 +0100
"Richard W.M. Jones" wrote:
> On Fri, Jul 17, 2015 at 02:48:40PM +0100, Marc Zyngier wrote:
> > Still: there is nothing in the registers that remotely points to that
> > area. X0 is the closest, but it'd take a big negative offset to get
> > there.
> >
> > Is t
On 17/07/2015 15:28, Marc Zyngier wrote:
> >
> > Marc, does it ring any bell?
> Well, this is an example of a guest accessing non-memory using an
> instruction that we cannot safely emulate - not an IO accessor (load
> multiple, for example).
>
> In this case, we kill the guest (we could as wel
This patchset sets the stage for extending the vhost user
protocol, with full backwards compatibility.
The approach is to negotiate feature bits queried and
acknowledged during device setup.
For now, there's no new functionality: two new messages
are added that will allow negotiating new messages
This reverts commit 830d70db692e374b5f4407f96a1ceefdcc97.
The interface as merged isn't fully backwards-compatible with existing
clients. Revert it. Let's redo this after 2.4, based on protocol
extensions in follow-up patches.
Signed-off-by: Michael S. Tsirkin
---
qapi-schema.json
Support a separate bitmask for vhost-user protocol features,
and messages to get/set protocol features.
Invoke them at init.
No features are defined yet.
Signed-off-by: Michael S. Tsirkin
---
include/hw/virtio/vhost.h | 1 +
hw/net/vhost_net.c| 2 ++
hw/virtio/vhost-user.c| 52 ++
translate at the outset, have rest of code use vhost-user features
exclusively.
Signed-off-by: Michael S. Tsirkin
---
hw/virtio/vhost-user.c | 87 ++
1 file changed, 46 insertions(+), 41 deletions(-)
diff --git a/hw/virtio/vhost-user.c b/hw/virtio
Data is empty for now, but do make sure master
sets the new feature bit flag.
Signed-off-by: Michael S. Tsirkin
---
tests/vhost-user-test.c | 19 +++
1 file changed, 19 insertions(+)
diff --git a/tests/vhost-user-test.c b/tests/vhost-user-test.c
index 75fedf0..228acb6 100644
---
Just speficying ops = NULL in some cases can be more convenient than having
two functions. See vGICv3 v5 patch set for at least one such case.
GICv2 code can be refactored in a way similar to what is done in the mentioned
patchset, killing
some code duplication.
Signed-off-by: Pavel Fedin
---
On 07/17/2015 08:09 AM, Michael S. Tsirkin wrote:
> This reverts commit 830d70db692e374b5f4407f96a1ceefdcc97.
>
> The interface as merged isn't fully backwards-compatible with existing
> clients. Revert it. Let's redo this after 2.4, based on protocol
> extensions in follow-up patches.
>
> S
On Fri, 17 Jul 2015 15:04:27 +0100
Paolo Bonzini wrote:
>
>
> On 17/07/2015 15:28, Marc Zyngier wrote:
> > >
> > > Marc, does it ring any bell?
> > Well, this is an example of a guest accessing non-memory using an
> > instruction that we cannot safely emulate - not an IO accessor (load
> > mul
CONFIG_LINUX_AIO is an implementation detail of raw-posix.c. Don't
mention CONFIG_LINUX_AIO in blockdev.c. Let raw-posix.c decide what to
do with BDRV_O_NATIVE_AIO if CONFIG_LINUX_AIO is not defined.
Signed-off-by: Stefan Hajnoczi
---
blockdev.c | 2 --
1 file changed, 2 deletions(-)
diff --g
v2:
* Banish CONFIG_LINUX_AIO from blockdev.c, that is raw-posix.c's business
[Kevin]
* Print the warning in the same way as the aio=native,cache.direct=off
deprecation warning [Kevin]
Open question: what about the Windows case? We now pass the
FILE_FLAG_OVERLAPPED flag which we didn't do
raw-posix.c silently ignores BDRV_O_NATIVE_AIO if libaio is unavailable.
It is confusing when aio=native performance is identical to aio=threads
because the binary was accidentally built without libaio.
Print a deprecation warning if -drive aio=native is used with a binary
that does not support li
On 16 July 2015 at 12:34, Christoffer Dall wrote:
> Some registers like the CNTVCT register should only be written to the
> kernel as part of machine initialization or on vmload operations, but
> never during runtime, as this can potentially make time go backwards or
> create inconsistent time obs
We had build problems due to the git version checking in the ipxe build
system in the past. Don't remember the details, but the problem seems
to be gone now, so lets remove the workaround.
Signed-off-by: Gerd Hoffmann
[ most likely ipxe commit 6153c09c41034250408f3596555fcaae715da46c:
[build]
ipxe build now generates empty local header files in case they are
not present. Let's remove them on cleanup to make sure we store a
fresh copy on the next build.
Signed-off-by: Gerd Hoffmann
Acked-by: Laszlo Ersek
---
roms/Makefile | 1 +
1 file changed, 1 insertion(+)
diff --git a/roms/Make
Activate the opt-out added by one not-upstream ipxe patch:
[efi] make load file protocol optional
Signed-off-by: Gerd Hoffmann
Reviewed-by: Laszlo Ersek
---
roms/config.ipxe.general.h | 1 +
1 file changed, 1 insertion(+)
diff --git a/roms/config.ipxe.general.h b/roms/config.ipxe.general.
Signed-off-by: Gerd Hoffmann
Reviewed-by: Laszlo Ersek
---
pc-bios/efi-e1000.rom| Bin 197120 -> 192512 bytes
pc-bios/efi-eepro100.rom | Bin 197632 -> 192512 bytes
pc-bios/efi-ne2k_pci.rom | Bin 195584 -> 190976 bytes
pc-bios/efi-pcnet.rom| Bin 195584 -> 190976 bytes
pc-bios/efi-rtl81
Apply qemu-project.org branding, so the official builds can easily be
identified in the banner.
Signed-off-by: Gerd Hoffmann
Acked-by: Laszlo Ersek
---
roms/Makefile | 4 ++--
roms/config.ipxe.branding.h | 2 ++
2 files changed, 4 insertions(+), 2 deletions(-)
create mode 100644
0100)
are available in the git repository at:
git://git.kraxel.org/qemu tags/pull-ipxe-20150717-1
for you to fetch changes up to 4f0c601b71e53a7d225a1913b784242400788991:
ipxe: update binaries (2015-07-16 17:39:12 +0200)
update ipxe
Add two patches we've been struggling to get upstream.
They are available from "git://git.qemu.org/ipxe.git qemu"
git shortlog
Gerd Hoffmann (1):
[efi] make load file protocol optional
Laszlo Ersek (1):
efi_snp: improve compliance with the EFI_SIMPLE_NETWORK_PROTOCOL spe
git shortlog
Alex Williamson (1):
[dhcp] Extract timing parameters out to config/dhcp.h
Bernd Wiebelt (1):
[tg3] Add support for BCM57766
Christian Hesse (3):
[intel] Add PCI device IDs for Intel I218-LM and I218-V
[build] Add missing "const" qualifiers
1 - 100 of 158 matches
Mail list logo
