On 09/16/2014 10:51 PM, Alex Williamson wrote:
> On Tue, 2014-09-16 at 00:01 +0200, Eric Auger wrote:
>> On 09/12/2014 01:05 AM, Christoffer Dall wrote:
>>> On Thu, Sep 11, 2014 at 04:51:14PM -0600, Alex Williamson wrote:
>>>> On Thu, 2014-09-11 at 15:23 -0700, Christ
qemu, add
-trace events=events.txt
Complete details can be found in docs/tracing.txt
Best Regards
Eric
On 09/03/2014 10:45 AM, Eric Auger wrote:
> This patch removes all DPRINTF and replace them by trace points.
> A few DPRINTF used in error cases were transformed into error_
Hi,
I am currently doing some benchmarks using virtio-net-device using
virtio-mmio (non PCI) with qemu_system_arm with KVM. Purpose is to
compare with device passthrough performance.
I heard about the availability of a multi-queue option that greatly
improves the performance but I currently fail
On 09/16/2014 11:23 PM, Alex Williamson wrote:
> On Tue, 2014-09-16 at 14:51 -0600, Alex Williamson wrote:
>> On Tue, 2014-09-16 at 00:01 +0200, Eric Auger wrote:
>>> On 09/12/2014 01:05 AM, Christoffer Dall wrote:
>>>> On Thu, Sep 11, 2014 at 04:51:14PM -0600, Alex
On 09/19/2014 03:42 AM, Gonglei (Arei) wrote:
>> Subject: [Qemu-devel] [question] virtio-net-device and multi-queue option
>>
>> Hi,
>>
>>
>> I am currently doing some benchmarks using virtio-net-device using
>> virtio-mmio (non PCI) with qemu_system_arm with KVM. Purpose is to
>> compare with devi
uot;PRIx64
- "%s write(%04x:%02x:%02x.%x:BAR%d+0x%"PRIx64", 0x%"PRIx64", %d)"
replaced by
"%s write(%04x:%02x:%02x.%x:BAR%d+0x%"PRIx64", 0x%"PRIx64", %d"
Best Regards
Eric
On 09/19/2014 11:03 AM, Stefan Hajnoczi wrote:
> On
On 09/19/2014 04:30 AM, Gonglei (Arei) wrote:
>> From: Eric Auger [mailto:eric.au...@linaro.org]
>> Sent: Friday, September 19, 2014 10:23 AM
>> To: Gonglei (Arei); qemu list; Michael S. Tsirkin
>> Subject: Re: [Qemu-devel] [question] virtio-net-device and multi-queue optio
__init__.py", line 225, in build
>> return Event(name, props, fmt, args, arg_fmts)
>> File "scripts/tracetool/__init__.py", line 185, in __init__
>> % ", ".join(unknown_props))
>> ValueError: Unknown properties: foo(int, rc)
>
>> Cc: Lluí
t; res.append(Event.build(line))
> File "scripts/tracetool/__init__.py", line 225, in build
> return Event(name, props, fmt, args, arg_fmts)
> File "scripts/tracetool/__init__.py", line 185, in __init__
> % ", ".join(unknown_prop
This patch removes all DPRINTF and replace them by trace points.
A few DPRINTF used in error cases were transformed into error_report.
Signed-off-by: Eric Auger
---
- __func__ is removed since trace point name does the same job
- HWADDR_PRIx were replaced by PRIx64
- this transformation just
On 06/04/2014 02:28 PM, Alexander Graf wrote:
> This patch adds a new device class called "platform device". This is an
> abstract class for consumption of actual classes that implement devices.
>
> The new thing about platform devices is that they have awareness of all
> memory regions and IRQs t
On 06/04/2014 02:28 PM, Alexander Graf wrote:
> For e500 our approach to supporting platform devices is to create a simple
> bus from the guest's point of view within which we map platform devices
> dynamically.
>
> We allocate memory regions always within the "platform" hole in address
> space an
too.
vfio_mmap_bar becomes vfio_map_region
Signed-off-by: Eric Auger
---
v4->v5:
- remove fd field from VFIORegion
- change error_report format string in vfio_region_write/read
- remove #ifdef DEBUG_VFIO in the same function
- correct missing initialization of bar region's vbased
ement in that derived device yet. May be re-introduced later
on if needed typically for reset/migration.
- no GSI routing table anymore
v3->v4 changes (Eric Auger, Alvise Rigo)
- rebase on last VFIO PCI code (v2.1.0-rc0)
- full git history rework to ease PCI code change review
- mv include files
This prepares for the introduction of VFIOPlatformDevice
Signed-off-by: Eric Auger
---
hw/vfio/pci.c | 209 +-
1 file changed, 105 insertions(+), 104 deletions(-)
diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c
index 188fdd2..c2cdd73 100644
From: Kim Phillips
This is done in preparation for the addition of VFIO platform
device support.
Signed-off-by: Kim Phillips
---
LICENSE | 2 +-
MAINTAINERS | 2 +-
hw/Makefile.objs | 1 +
hw/misc/Makefile.objs| 1 -
hw/
_intx, vfio_disable_irqindex which now
apply to VFIODevice. They are renamed as *_irqindex.
The index is passed as parameter to anticipate their usage for
platform IRQs
Signed-off-by: Eric Auger
---
v4->v5:
- fix style issues
- in vfio_initfn, rework allocation of vdev->vbasedev.name an
ame deallocation done in vfio_put_device instead of
vfio_put_base_device
- vfio_put_base_device done at vfio_get_device level.
Signed-off-by: Eric Auger
---
hw/vfio/pci.c | 181 ++
1 file changed, 120 insertions(+), 61 deletions(-)
diff --
A new common module is created. It implements all functions
that have no device specificity (PCI, Platform).
This patch only consists in move (no functional changes)
Signed-off-by: Kim Phillips
Signed-off-by: Eric Auger
---
v4 -> v5:
- integrate "sPAPR/IOMMU: Fix TCE entry pe
Generates the device node of VFIO devices, if any are invoked in
-device option. In case VFIO devices require more complex node
generation, they can be handled before.
Signed-off-by: Eric Auger
---
hw/arm/dyn_sysbus_devtree.c | 138
1 file changed
Minimal VFIO platform implementation supporting
- register space user mapping,
- IRQ assignment based on eventfds handled on qemu side.
irqfd kernel acceleration comes in a subsequent patch.
Signed-off-by: Kim Phillips
Signed-off-by: Eric Auger
---
v4 -> v5:
- vfio-plaform.h included fi
set kvm_irqfds_allowed
Signed-off-by: Eric Auger
---
hw/intc/arm_gic_kvm.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/hw/intc/arm_gic_kvm.c b/hw/intc/arm_gic_kvm.c
index 5038885..08b7bf9 100644
--- a/hw/intc/arm_gic_kvm.c
+++ b/hw/intc/arm_gic_kvm.c
@@ -576,6 +576,8 @@ static void
1st access to any region after IRQ hit.
This removes the need for fast/slow path swap.
Overall this brings significant performance improvements.
It depends on host kernel KVM irqfd/GSI routing capability.
Signed-off-by: Alvise Rigo
Signed-off-by: Eric Auger
---
v4 -> v5:
- addition of sys
Dear all,
I would like to test VFIO-PCI after rework done for [PATCH v5 00/10] KVM
platform device passthrough.
Do you have advises about best test environment (was thinking about TCG
system platform suitable for VFIO PCI assignment?).
All inputs welcome
Thank you in advance
Best Regards
Eric
On 08/11/2014 11:40 AM, Alexander Graf wrote:
>
> On 09.08.14 16:25, Eric Auger wrote:
>> Generates the device node of VFIO devices, if any are invoked in
>> -device option. In case VFIO devices require more complex node
>> generation, they can be handled before.
>>
On 08/11/2014 11:37 AM, Alexander Graf wrote:
>
> On 09.08.14 16:25, Eric Auger wrote:
>> set kvm_irqfds_allowed
>>
>> Signed-off-by: Eric Auger
>> ---
>> hw/intc/arm_gic_kvm.c | 2 ++
>> 1 file changed, 2 insertions(+)
>>
>> diff --git a/
On 08/11/2014 02:05 PM, Alexander Graf wrote:
>
> On 11.08.14 14:04, Eric Auger wrote:
>> On 08/11/2014 11:37 AM, Alexander Graf wrote:
>>> On 09.08.14 16:25, Eric Auger wrote:
>>>> set kvm_irqfds_allowed
>>>>
>>>> Signed-off-by: Eric Auger
On 08/11/2014 03:08 PM, Alexander Graf wrote:
>
> On 08.08.14 17:03, Eric Auger wrote:
>> This new module implements routines which help in dynamic device
>> binding (mmio regions, irq). They are supposed to be used by machine
>> files that support dynamic sysbus instantia
On 08/11/2014 03:16 PM, Alexander Graf wrote:
>
> On 08.08.14 17:03, Eric Auger wrote:
>> This module will be used by ARM machine files to generate
>> device tree nodes of dynamically instantiated sysbus devices (ie.
>> those instantiated with -device option).
>>
&g
On 08/11/2014 03:10 PM, Alexander Graf wrote:
>
> On 08.08.14 17:03, Eric Auger wrote:
>> load_dtb will be used by machvirt for dynamic instantiation of
>> platform devices
>>
>> Signed-off-by: Eric Auger
>> ---
>> hw/arm/boot.c| 2 +-
&g
On 08/11/2014 03:50 PM, Will Deacon wrote:
> I'm playing with PCI device assignment with kvmtool, so I could do some
> basic testing if you like. Can you put the patches on a git tree somewhere
> please?
Hi Will,
Thanks for your proposal,
the kernel part is at:
git://git.linaro.org/people/eric.a
On 08/11/2014 10:13 PM, Alex Williamson wrote:
> On Sat, 2014-08-09 at 15:25 +0100, Eric Auger wrote:
>> Minimal VFIO platform implementation supporting
>> - register space user mapping,
>> - IRQ assignment based on eventfds handled on qemu side.
>>
>> irq
On 08/11/2014 09:20 PM, Alex Williamson wrote:
> On Sat, 2014-08-09 at 15:25 +0100, Eric Auger wrote:
>> A new common module is created. It implements all functions
>> that have no device specificity (PCI, Platform).
>>
>> This patch only consists in move (no functiona
On 08/11/2014 09:25 PM, Alex Williamson wrote:
> On Sat, 2014-08-09 at 15:25 +0100, Eric Auger wrote:
>> diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h
>> new file mode 100644
>> index 000..4684ee5
>> --- /dev/null
>> +++ b/include
On 08/11/2014 07:57 PM, Will Deacon wrote:
> On Mon, Aug 11, 2014 at 03:28:40PM +0100, Eric Auger wrote:
>> On 08/11/2014 03:50 PM, Will Deacon wrote:
>>> I'm playing with PCI device assignment with kvmtool, so I could do some
>>> basic testing if you like. Can y
On 08/12/2014 04:41 AM, David Gibson wrote:
> On Sat, Aug 09, 2014 at 03:25:44PM +0100, Eric Auger wrote:
>> vfio_get_device now takes a VFIODevice as argument. The function is split
>> into 4 functional parts: dev_info query, device check, region populate
>> and interrupt pop
On 08/12/2014 04:41 AM, Li Liu wrote:
> Hi all,
>
> Is anyone there can tell the current status of vhost-net on kvm-arm?
>
> Half a year has passed from Isa Ansharullah asked this question:
> http://www.spinics.net/lists/kvm-arm/msg08152.html
>
> I have found two patches which have provided the
On 08/12/2014 09:59 AM, bharat.bhus...@freescale.com wrote:
>
>
>> -Original Message-
>> From: Alexander Graf [mailto:ag...@suse.de]
>> Sent: Monday, August 11, 2014 3:06 PM
>> To: Eric Auger; eric.au...@st.com; christoffer.d...@linaro.org; qemu-
>> d
On 08/18/2014 11:54 PM, Joel Schopp wrote:
>
> +static void vfio_fdt_add_device_node(SysBusDevice *sbdev, void *opaque)
> +{
> +PlatformDevtreeData *data = opaque;
> +void *fdt = data->fdt;
> +const char *parent_node = data->node;
> +int compat_str_len;
> +char *nodename;
> +
On 08/19/2014 12:11 AM, Peter Maydell wrote:
> On 18 August 2014 22:54, Joel Schopp wrote:
>>
>> +static void vfio_fdt_add_device_node(SysBusDevice *sbdev, void *opaque)
>> +{
>> +PlatformDevtreeData *data = opaque;
>> +void *fdt = data->fdt;
>> +const char *parent_node = data->node;
>
On 08/19/2014 12:26 AM, Joel Schopp wrote:
>
> On 08/18/2014 05:11 PM, Peter Maydell wrote:
>> On 18 August 2014 22:54, Joel Schopp wrote:
>>> +static void vfio_fdt_add_device_node(SysBusDevice *sbdev, void *opaque)
>>> +{
>>> +PlatformDevtreeData *data = opaque;
>>> +void *fdt = data->fd
On 08/13/2014 05:32 AM, David Gibson wrote:
> On Tue, Aug 12, 2014 at 08:54:34AM +0200, Eric Auger wrote:
>> On 08/12/2014 04:41 AM, David Gibson wrote:
>>> On Sat, Aug 09, 2014 at 03:25:44PM +0100, Eric Auger wrote:
>>>> vfio_get_device now takes a VFIODevice as a
On 05/11/2014 07:13 PM, Alvise Rigo wrote:
> The flag is mandatory for the ARM SMMU so we always add it if the MMIO
> handles it.
Hi Alvise,
Refering to the root problem explanation found in
https://lkml.org/lkml/2014/2/8/176, I understand the problem is specific
to devices that fetch instruction
On 09/09/2014 01:11 PM, Paolo Bonzini wrote:
> Il 09/09/2014 09:54, Eric Auger ha scritto:
>> Allows sysbus devices to be instantiated from command line by
>> using -device option
>>
>> Signed-off-by: Alexander Graf
>> Signed-off-by: Eric Auger
>>
Dear all,
The goal of this mail is to summarize how dynamic sysbus device tree
nodes were created on ARM with "machvirt dynamic sysbus device
instantiation",
https://lists.gnu.org/archive/html/qemu-devel/2014-09/msg01626.html
and request some advises after commit "hw/arm/boot: load DTB as a ROM
im
Hi,
Thanks everyone for entering the thread & reading my long email.
Alex, I indeed can register the notifier in the machine file after the
platform bus instantiation. This indeed guarantees the notifiers are
called in the right order ...
Thanks
Best Regards
Eric
On 10/23/2014 01:26 PM, Alexa
On 10/23/2014 01:41 PM, Eric Auger wrote:
> Hi,
>
> Thanks everyone for entering the thread & reading my long email.
>
> Alex, I indeed can register the notifier in the machine file after
s/after/before
Eric
the
> platform bus instantiation. This indeed guarantees the not
vfio-calxeda-xgmac now can be instantiated using the -device option.
The node creation function generates a very basic dt node composed
of the compat, reg and interrupts properties
Signed-off-by: Eric Auger
---
v6 -> v7:
- compat string re-formatting removed since compat string is not expo
This patch removes all DPRINTF and replace them by trace points.
A few DPRINTF used in error cases were transformed into error_report.
Signed-off-by: Eric Auger
---
- __func__ is removed since trace point name does the same job
- HWADDR_PRIx were replaced by PRIx64
- this transformation just
egions, vfio_populate_interrupts and
vfio_get_device.
- correct misuse of errno
- vfio_populate_regions always returns 0
- VFIODevice .name deallocation done in vfio_put_device instead of
vfio_put_base_device
- vfio_put_base_device done at vfio_get_device level.
Signed-off-by: Eric Auger
---
h
Minimal VFIO platform implementation supporting
- register space user mapping,
- IRQ assignment based on eventfds handled on qemu side.
irqfd kernel acceleration comes in a subsequent patch.
Signed-off-by: Kim Phillips
Signed-off-by: Eric Auger
---
v6 -> v7:
- compat is not exposed anymore
too.
vfio_mmap_bar becomes vfio_map_region
Signed-off-by: Eric Auger
---
v4->v5:
- remove fd field from VFIORegion
- change error_report format string in vfio_region_write/read
- remove #ifdef DEBUG_VFIO in the same function
- correct missing initialization of bar region's vbased
notifier in machvirt.
Signed-off-by: Eric Auger
---
The registration of the IRQ start notifier could also happen in
the platform bus.
---
hw/arm/virt.c | 9 +
1 file changed, 9 insertions(+)
diff --git a/hw/arm/virt.c b/hw/arm/virt.c
index 3a09d58..911dbfc 100644
--- a/hw/arm/virt.c
+++ b
for fast/slow path swap.
Overall this brings significant performance improvements.
it depends on host kernel KVM irqfd.
Signed-off-by: Alvise Rigo
Signed-off-by: Eric Auger
---
v5 -> v6
- rely on kvm_irqfds_enabled() and kvm_resamplefds_enabled()
- guard KVM code with #ifdef CONFIG_KVM
ht come latter)
- platform_bus_base removed from PlatformDevtreeData
Eric Auger (6):
hw/arm/boot: load_dtb becomes non static arm_load_dtb
hw/arm/boot: dtb start and limit moved in arm_boot_info
hw/arm/boot: do not free VirtBoardInfo fdt in arm_load_dtb
hw/arm: add a new modify_dtb_opaque field in arm_boot
virtual
IRQ, the corresponding physical IRQ is completed and the same
physical IRQ can hit again.
A new x-forward property enables to force forwarding off although
enabled by the kernel.
Signed-off-by: Eric Auger
---
hw/vfio/platform.c | 52
Syncup KVM related linux headers from linux-next tree using
scripts/update-linux-headers.sh.
Integrate updated KVM-VFIO API related to forwarded IRQ
Signed-off-by: Eric Auger
---
linux-headers/linux/kvm.h | 9 +
1 file changed, 9 insertions(+)
diff --git a/linux-headers/linux/kvm.h b
A new common module is created. It implements all functions
that have no device specificity (PCI, Platform).
This patch only consists in move (no functional changes)
Signed-off-by: Kim Phillips
Signed-off-by: Eric Auger
---
v6 -> v7:
- integrate Revert "vfio: Make BARs native endian&q
be called by
ARM machine files that support platform bus and their dynamic
sysbus. Addition of dynamic sysbus nodes is done only if the
user did not provide any dtb.
Signed-off-by: Alexander Graf
Signed-off-by: Eric Auger
---
v3 -> v4:
- dyn_sysbus_devtree.c renamed into sysbus-fdt.c
- use
load_dtb is renamed into arm_load_dtb and becomes non static.
it will be used by machvirt for dynamic instantiation of
platform devices
Signed-off-by: Eric Auger
---
v2 -> v3:
load_dtb renamed into arm_load_dtb
Conflicts:
hw/arm/boot.c
---
hw/arm/boot.c|
the device is now used in platform for forwarded IRQ setup
Signed-off-by: Eric Auger
---
hw/vfio/common.c | 3 ++-
include/hw/vfio/vfio-common.h | 5 +
2 files changed, 7 insertions(+), 1 deletion(-)
diff --git a/hw/vfio/common.c b/hw/vfio/common.c
index fbd9e7f..99ff89a
The platform device class has become abstract. The device can be be
instantiated on command line using such option.
-device vfio-calxeda-xgmac,host="fff51000.ethernet"
Signed-off-by: Eric Auger
---
v5 -> v6
- back again following Alex Graf advises
- fix a bug related to compat
better fit in the rest of the namespace
Signed-off-by: Eric Auger
---
hw/vfio/pci.c | 22 +++---
1 file changed, 11 insertions(+), 11 deletions(-)
diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c
index 0ee6f7f..2216bd4 100644
--- a/hw/vfio/pci.c
+++ b/hw/vfio/pci.c
@@ -283,7 +283,7
instantiation, Eric Auger
[3] [PATCH v3 0/2] actual checks of KVM_CAP_IRQFD and KVM_CAP_IRQFD_RESAMPLE,
Eric Auger
http://lists.nongnu.org/archive/html/qemu-devel/2014-09/msg00589.html
[4] [PATCH v2] vfio: migration to trace points, Eric Auger
https://patchwork.ozlabs.org/patch/394785
This patch removes all DPRINTF and replace them by trace points.
A few DPRINTF used in error cases were transformed into error_report.
Signed-off-by: Eric Auger
---
- __func__ is removed since trace point name does the same job
- HWADDR_PRIx were replaced by PRIx64
- this transformation just
brought in 84223598778ba08041f4297fda485df83414d57e,
"KVM: s390: irq routing for adapter interrupts". Both commits
first appear in 3.15 so there should not be any kernel
version impacted by this QEMU modification.
Signed-off-by: Eric Auger
---
v2->v3:
- changed the commit me
_intx, vfio_disable_irqindex which now
apply to VFIODevice. They are renamed as *_irqindex.
The index is passed as parameter to anticipate their usage for
platform IRQs
Signed-off-by: Eric Auger
---
v4->v5:
- fix style issues
- in vfio_initfn, rework allocation of vdev->vbasedev.name an
This prepares for the introduction of VFIOPlatformDevice
Signed-off-by: Eric Auger
---
hw/vfio/pci.c | 210 +-
1 file changed, 106 insertions(+), 104 deletions(-)
diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c
index 8514b9e..93181bf 100644
From: Kim Phillips
This is done in preparation for the addition of VFIO platform
device support.
Signed-off-by: Kim Phillips
---
LICENSE | 2 +-
MAINTAINERS | 2 +-
hw/Makefile.objs | 1 +
hw/misc/Makefile.objs| 1 -
hw/
Signed-off-by: Eric Auger
Conflicts:
trace-events
---
hw/vfio/pci.c | 213 --
trace-events | 109 --
2 files changed, 116 insertions(+), 206 deletions(-)
diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c
sysbus devices. Indeed
dynamic sysbus devices are created after machine init.
machvirt also registers a notifier that will start the VFIO
dynamic device IRQ handling.
Signed-off-by: Alexander Graf
Signed-off-by: Eric Auger
---
v3 -> v4:
- use platform bus object, instantiated in create_platform_
This field can be used by any modify_dtb() function to pass
additional arguments requested to build the modified dtb. This
is needed for creating the platform bus dynamic sysbus nodes.
Signed-off-by: Eric Auger
---
include/hw/arm/arm.h | 4
1 file changed, 4 insertions(+)
diff --git a
Use the kvm_resamplefds_enabled function
Signed-off-by: Eric Auger
---
hw/misc/vfio.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/hw/misc/vfio.c b/hw/misc/vfio.c
index b5e7981..75bfa1c 100644
--- a/hw/misc/vfio.c
+++ b/hw/misc/vfio.c
@@ -406,7 +406,7 @@ static void
destroyed, this is not possible.
Signed-off-by: Eric Auger
---
hw/arm/boot.c | 8 ++--
1 file changed, 6 insertions(+), 2 deletions(-)
diff --git a/hw/arm/boot.c b/hw/arm/boot.c
index 9f0662e..0e4b078 100644
--- a/hw/arm/boot.c
+++ b/hw/arm/boot.c
@@ -427,12 +427,16 @@ int arm_load_dtb(const
dtb is already loaded. Storing those parameters
in arm_boot_info allows to avoid computing again dtb_start and
dtb_load, as done in arm_load_kernel.
Signed-off-by: Eric Auger
---
hw/arm/boot.c| 38 +-
include/hw/arm/arm.h | 5 +++--
2 files changed, 24
user of kvm_resamplefds_enabled().
Eric Auger (2):
KVM_CAP_IRQFD and KVM_CAP_IRQFD_RESAMPLE checks
vfio: use kvm_resamplefds_enabled()
hw/intc/openpic_kvm.c | 1 -
hw/intc/xics_kvm.c| 1 -
hw/misc/vfio.c| 5 ++---
include/sysemu/kvm.h | 10 ++
kvm-all.c | 7
On 10/31/2014 03:35 PM, Alex Williamson wrote:
> On Fri, 2014-10-31 at 13:44 +0000, Eric Auger wrote:
>> This patch removes all DPRINTF and replace them by trace points.
>> A few DPRINTF used in error cases were transformed into error_report.
>>
>> Signed-off-by: Eric
-by: Eric Auger
---
in practice KVM_CAP_IRQFD_RESAMPLE seems to be always enabled
as soon as kernel has HAVE_KVM_IRQFD so the resamplefd check
may be unnecessary.
---
hw/intc/openpic_kvm.c | 1 -
hw/intc/xics_kvm.c| 1 -
include/sysemu/kvm.h | 10 ++
kvm-all.c | 7
On 09/01/2014 12:49 PM, Paolo Bonzini wrote:
> Il 29/08/2014 19:38, Eric Auger ha scritto:
>> Compute kvm_irqfds_allowed by checking the KVM_CAP_IRQFD extension.
>> Remove direct settings in architecture specific files.
>>
>> Add a new kvm_resamplefds_allowed variable,
On 08/13/2014 09:59 PM, Alex Williamson wrote:
> On Tue, 2014-08-12 at 08:09 +0200, Eric Auger wrote:
>> On 08/11/2014 09:25 PM, Alex Williamson wrote:
>>> On Sat, 2014-08-09 at 15:25 +0100, Eric Auger wrote:
>>>> diff --git a/include/hw/vfio/vfio-common.h b/include/
On 09/01/2014 07:41 PM, Alexander Graf wrote:
>
>
>> Am 01.09.2014 um 18:31 schrieb Eric Auger :
>>
>>> On 08/13/2014 09:59 PM, Alex Williamson wrote:
>>>> On Tue, 2014-08-12 at 08:09 +0200, Eric Auger wrote:
>>>>> On 08/11/2014 09:25 PM, Ale
This patch removes all DPRINTF and replace them by trace points.
A few DPRINTF used in error cases were transformed into error_report.
Signed-off-by: Eric Auger
---
- __func__ is removed since trace point name does the same job
- HWADDR_PRIx were replaced by PRIx64
Besides those changes
This patch serie replaces direct settings of kvm_irqfds_allowed
by actual checks of the KVM_CAP_IRQFD extension. Also A new
kvm_resamplefds_enabled() enables to check KVM_CAP_IRQFD_RESAMPLE.
in the second patch file the vfio device is the first user of
kvm_resamplefds_enabled().
Eric Auger (2
-by: Eric Auger
---
in practice KVM_CAP_IRQFD_RESAMPLE seems to be always enabled
as soon as kernel has HAVE_KVM_IRQFD so the resamplefd check
may be unnecessary.
---
hw/intc/openpic_kvm.c | 1 -
hw/intc/xics_kvm.c| 1 -
include/sysemu/kvm.h | 10 ++
kvm-all.c | 7
Use the kvm_resamplefds_enabled function
Signed-off-by: Eric Auger
---
hw/misc/vfio.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/hw/misc/vfio.c b/hw/misc/vfio.c
index 40dcaa6..24f6a3a 100644
--- a/hw/misc/vfio.c
+++ b/hw/misc/vfio.c
@@ -406,7 +406,7 @@ static void
On 09/03/2014 02:50 PM, Christian Borntraeger wrote:
> On 03/09/14 11:54, Eric Auger wrote:
> [...]
>> --- a/kvm-all.c
>> +++ b/kvm-all.c
> [...]
>> @@ -1548,6 +1549,12 @@ int kvm_init(MachineClass *mc)
>> kvm_eventfds_allowed =
>> (kvm_ch
/2014-07/msg00047.html
[2] [RFC v3] machvirt dynamic sysbus device instantiation, Eric Auger
[3] [PATCH v2 0/2] actual checks of KVM_CAP_IRQFD and KVM_CAP_IRQFD_RESAMPLE,
Eric Auger
http://lists.nongnu.org/archive/html/qemu-devel/2014-09/msg00589.html
[4] [RFC] vfio: migration to trace points
This prepares for the introduction of VFIOPlatformDevice
Signed-off-by: Eric Auger
---
hw/vfio/pci.c | 209 +-
1 file changed, 105 insertions(+), 104 deletions(-)
diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c
index 7e6a1bc..ad5da4b 100644
_intx, vfio_disable_irqindex which now
apply to VFIODevice. They are renamed as *_irqindex.
The index is passed as parameter to anticipate their usage for
platform IRQs
Signed-off-by: Eric Auger
---
v4->v5:
- fix style issues
- in vfio_initfn, rework allocation of vdev->vbasedev.name an
From: Kim Phillips
This is done in preparation for the addition of VFIO platform
device support.
Signed-off-by: Kim Phillips
---
LICENSE | 2 +-
MAINTAINERS | 2 +-
hw/Makefile.objs | 1 +
hw/misc/Makefile.objs| 1 -
hw/
too.
vfio_mmap_bar becomes vfio_map_region
Signed-off-by: Eric Auger
---
v4->v5:
- remove fd field from VFIORegion
- change error_report format string in vfio_region_write/read
- remove #ifdef DEBUG_VFIO in the same function
- correct missing initialization of bar region's vbased
The platform device class has become abstract. The device can be be
instantiated on command line using such option.
-device vfio-calxeda-xgmac,host="fff51000.ethernet"
compat string is hardcoded in the code except if user overrides it
Signed-off-by: Eric Auger
---
v5 -> v6
Minimal VFIO platform implementation supporting
- register space user mapping,
- IRQ assignment based on eventfds handled on qemu side.
irqfd kernel acceleration comes in a subsequent patch.
Signed-off-by: Kim Phillips
Signed-off-by: Eric Auger
---
v5 -> v6:
- vfio_device property rena
egions, vfio_populate_interrupts and
vfio_get_device.
- correct misuse of errno
- vfio_populate_regions always returns 0
- VFIODevice .name deallocation done in vfio_put_device instead of
vfio_put_base_device
- vfio_put_base_device done at vfio_get_device level.
Signed-off-by: Eric Auger
---
h
vfio-calxeda-xgmac now can be instantiated using the -device option
Signed-off-by: Eric Auger
---
v2 -> v3:
- correct bug of reg_attr[2*i] in vfio_fdt_add_device_node
- fix a bug related to compat_str_len computed on original compat
instead of corrected compat
- wrap_vfio_fdt_add_node tak
for fast/slow path swap.
Overall this brings significant performance improvements.
it depends on host kernel KVM irqfd.
Signed-off-by: Alvise Rigo
Signed-off-by: Eric Auger
---
v5 -> v6
- rely on kvm_irqfds_enabled() and kvm_resamplefds_enabled()
- guard KVM code with #ifdef CONFIG_KVM
Signed-off-by: Eric Auger
---
hw/vfio/pci.c | 213 --
trace-events | 105 ++---
2 files changed, 111 insertions(+), 207 deletions(-)
diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c
index 5623539..c617b79 100644
--- a/hw
he IRQ is completed
on another timer timeout to emulate eoi on write/read access.
for instance, following options
x-fake-irq[0]=1,x-fake-period[0]=10,x-fake-duration[0]=50,
x-fake-irq[1]=2,x-fake-period[i]=20,x-fake-duration[1]=100
set vfio platform IRQ indexed #1 and #2 as fake IRQ
Signed-off-by
Syncup KVM related linux headers from linux-next tree using
scripts/update-linux-headers.sh.
Integrate updated KVM-VFIO API related to forwarded IRQ
Signed-off-by: Eric Auger
---
linux-headers/linux/kvm.h | 9 +
1 file changed, 9 insertions(+)
diff --git a/linux-headers/linux/kvm.h b
the device is now used in platform for forwarded IRQ setup
Signed-off-by: Eric Auger
---
hw/vfio/common.c | 3 ++-
include/hw/vfio/vfio-common.h | 5 +
2 files changed, 7 insertions(+), 1 deletion(-)
diff --git a/hw/vfio/common.c b/hw/vfio/common.c
index 252c0b8..466b0e8
1 - 100 of 3719 matches
Mail list logo
