atal. This preserves the existing
behaviour. Then to use the API with your optional semantic a null
parameter can be conveyed.
Signed-off-by: Eric Auger
---
RFC -> v1:
- get rid of qemu_fdt_getprop_optional and implement Peter's suggestion
that consists in using the error API
Signed-
This new helper routine returns the node path of a device
referred to by its node name and compat string.
Signed-off-by: Eric Auger
---
RFC -> v1:
- improve error handling according to Alex' comments
---
device_tree.c| 45
odes (dma and ptp) also are created. It is checked those clocks
are fixed on host side.
Signed-off-by: Eric Auger
---
RFC -> v1:
- use qemu_fdt_getprop with Error **
- free substrings in sysfs_to_dt_name
- add some comments related to endianess in add_amd_xgbe_fdt_node
- reword commit mes
Hi Alex,
On 12/17/2015 04:13 PM, Alex Bennée wrote:
>
> Peter Maydell writes:
>
>> On 17 December 2015 at 13:28, Alex Bennée wrote:
>>> Usually I would expect to see a pre-declaration of a function at the
>>> head of the file and only if it is used before the actual definition of
>>> the functi
Hi Peter,
On 12/18/2015 03:10 PM, Peter Maydell wrote:
> On 17 December 2015 at 12:29, Eric Auger wrote:
>> This function returns the host device tree blob from sysfs
>> (/sys/firmware/devicetree/base). It uses a recursive function
>> inspired from dtc read_fstree.
>>
&
Hi Peter,
On 12/18/2015 03:36 PM, Peter Maydell wrote:
> On 17 December 2015 at 12:29, Eric Auger wrote:
>> Current qemu_fdt_getprop exits if the property is not found. It is
>> sometimes needed to read an optional property, in which case we do
>> not wish to exit but simply
Hi Peter,
On 12/18/2015 03:23 PM, Peter Maydell wrote:
> On 17 December 2015 at 12:29, Eric Auger wrote:
>> This new helper routine returns the node path of a device
>> referred to by its node name and compat string.
>>
>> Signed-off-by: Eric Auger
>>
>>
On 12/18/2015 03:54 PM, Peter Maydell wrote:
> On 17 December 2015 at 12:29, Eric Auger wrote:
>> Some passthrough'ed devices depend on clock nodes. Those need to be
>> generated in the guest device tree. This patch introduces some helpers
>> to build a clock node fro
On 01/05/2016 06:55 PM, Peter Maydell wrote:
> On 5 January 2016 at 16:20, Eric Auger wrote:
>> Hi Peter,
>> On 12/18/2015 03:23 PM, Peter Maydell wrote:
>>> On 17 December 2015 at 12:29, Eric Auger wrote:
>>>> This new helper routine returns the node path
This new helper routine returns the node path of a device
referred to by its node name and compat string.
Signed-off-by: Eric Auger
---
v1 -> v2:
- move doc comment in header file
- do not use a fixed size buffer
- break on errors in while loop
- use strcmp instead of strncmp
RFC -&
ore on dtc binary: load_device_tree_from_sysfs is
self-contained and implements something similar to dtc read_fstree.
- take into account Alex' comments
- remove qemu_fdt_getprop_optional and use error API instead
Best Regards
Eric
Eric Auger (7):
hw/vfio/platform: amd-xgbe device
This patch introduces the amd-xgbe VFIO platform device. It
allows the guest to do passthrough on a device exposing an
"amd,xgbe-seattle-v1a" compat string.
Signed-off-by: Eric Auger
Reviewed-by: Alex Benné
---
RFC -> v1:
- add Alex' R-b
---
hw/vfio/Makefile.objs |
This function returns the host device tree blob from sysfs
(/proc/device-tree). It uses a recursive function inspired
from dtc read_fstree.
Signed-off-by: Eric Auger
---
v1 -> v2:
- do not implement/expose read_fstree and load_device_tree_from_sysfs
if CONFIG_LINUX is not defined (lstat
atal. This preserves the existing
behaviour. Then to use the API with your optional semantic a null
parameter can be conveyed.
Signed-off-by: Eric Auger
---
v1 -> v2:
- add a doc comment in the header file
RFC -> v1:
- get rid of qemu_fdt_getprop_optional and implement Peter's su
ifdef CONFIG_LINUX.
Signed-off-by: Eric Auger
---
v1 -> v2:
- add CONFIG_LINUX protection
- improves robustness in sysfs_to_dt_name
- output messages to end-user on misc failures and self-exits:
error management becomes more violent than before assuming if
the end-user wants passthroug
qemu_fdt_setprop self-exists in case of error hence no need to check
the returned value.
Signed-off-by: Eric Auger
---
hw/arm/sysbus-fdt.c | 15 +++
1 file changed, 3 insertions(+), 12 deletions(-)
diff --git a/hw/arm/sysbus-fdt.c b/hw/arm/sysbus-fdt.c
index 66fa766..68d7e53 100644
guest device tree node
- fdt_build_clock_node builds a guest clock node and checks the host
fellow clock is a fixed one.
fdt_build_clock_node will become static as soon as it gets used. A
dummy pre-declaration is needed for compilation of this patch.
Signed-off-by: Eric Auger
---
v1 -> v2:
-
Hi Peter,
On 01/07/2016 01:20 AM, Peter Crosthwaite wrote:
> On Wed, Jan 6, 2016 at 7:13 AM, Eric Auger wrote:
>> Current qemu_fdt_getprop exits if the property is not found. It is
>> sometimes needed to read an optional property, in which case we do
>> not wish to exit but
guest device tree node
- fdt_build_clock_node builds a guest clock node and checks the host
fellow clock is a fixed one.
fdt_build_clock_node will become static as soon as it gets used. A
dummy pre-declaration is needed for compilation of this patch.
Signed-off-by: Eric Auger
---
v1 -> v2:
-
atal. This preserves the existing
behaviour. Then to use the API with your optional semantic a null
parameter can be conveyed.
Signed-off-by: Eric Auger
---
v1 -> v2:
- add a doc comment in the header file
RFC -> v1:
- get rid of qemu_fdt_getprop_optional and implement Peter's su
ifdef CONFIG_LINUX.
Signed-off-by: Eric Auger
---
v1 -> v2:
- add CONFIG_LINUX protection
- improves robustness in sysfs_to_dt_name
- output messages to end-user on misc failures and self-exits:
error management becomes more violent than before assuming if
the end-user wants passthroug
qemu_fdt_setprop self-exists in case of error hence no need to check
the returned value.
Signed-off-by: Eric Auger
---
hw/arm/sysbus-fdt.c | 15 +++
1 file changed, 3 insertions(+), 12 deletions(-)
diff --git a/hw/arm/sysbus-fdt.c b/hw/arm/sysbus-fdt.c
index 66fa766..68d7e53 100644
o account Alex' comments
- remove qemu_fdt_getprop_optional and use error API instead
Best Regards
Eric
Eric Auger (8):
hw/vfio/platform: amd-xgbe device
device_tree: introduce load_device_tree_from_sysfs
device_tree: introduce qemu_fdt_node_path
device_tree: qemu_fdt_getprop conver
This function returns the host device tree blob from sysfs
(/proc/device-tree). It uses a recursive function inspired
from dtc read_fstree.
Signed-off-by: Eric Auger
---
v1 -> v2:
- do not implement/expose read_fstree and load_device_tree_from_sysfs
if CONFIG_LINUX is not defined (lstat
This patch introduces the amd-xgbe VFIO platform device. It
allows the guest to do passthrough on a device exposing an
"amd,xgbe-seattle-v1a" compat string.
Signed-off-by: Eric Auger
Reviewed-by: Alex Bennée
---
RFC -> v1:
- add Alex' R-b
---
hw/vfio/Makefile.objs
This new helper routine returns the node path of a device
referred to by its node name and compat string.
Signed-off-by: Eric Auger
---
v1 -> v2:
- move doc comment in header file
- do not use a fixed size buffer
- break on errors in while loop
- use strcmp instead of strncmp
RFC -&
atal. This preserves the existing
behaviour. Then to use the API with your optional semantic a null
parameter can be conveyed.
Signed-off-by: Eric Auger
---
v1 -> v2:
- add a doc comment in the header file
RFC -> v1:
- get rid of qemu_fdt_getprop_optional and implement Peter's su
use the new lenp parameter to interpret
the error if any.
Signed-off-by: Eric Auger
---
v3 : creation
---
device_tree.c| 21 ++---
hw/arm/boot.c| 6 --
hw/arm/vexpress.c| 6 --
include/sysemu/device_tree.h
guest device tree node
- fdt_build_clock_node builds a guest clock node and checks the host
fellow clock is a fixed one.
fdt_build_clock_node will become static as soon as it gets used. A
dummy pre-declaration is needed for compilation of this patch.
Signed-off-by: Eric Auger
---
v1 -> v2:
-
qemu_fdt_setprop self-asserts in case of error hence no need to check
the returned value.
Signed-off-by: Eric Auger
---
v3 -> v4: fix returned value
---
hw/arm/sysbus-fdt.c | 19 +--
1 file changed, 5 insertions(+), 14 deletions(-)
diff --git a/hw/arm/sysbus-fdt.c b/hw/
ifdef CONFIG_LINUX.
Signed-off-by: Eric Auger
---
v3 -> v4:
- add_amd_xgbe_fdt_node explicitly returns 0. Reword comment to emphasize
the fact the function self-asserts in case of error
v1 -> v2:
- add CONFIG_LINUX protection
- improves robustness in sysfs_to_dt_name
- output messages to e
This patch introduces the amd-xgbe VFIO platform device. It
allows the guest to do passthrough on a device exposing an
"amd,xgbe-seattle-v1a" compat string.
Signed-off-by: Eric Auger
Reviewed-by: Alex Bennée
---
RFC -> v1:
- add Alex' R-b
---
hw/vfio/Makefile.objs
ve qemu_fdt_setprop returned value check
- see individual commits for full details
RFC -> v1:
- no dependency anymore on dtc binary: load_device_tree_from_sysfs is
self-contained and implements something similar to dtc read_fstree.
- take into account Alex' comments
- remove qemu_fdt_get
This function returns the host device tree blob from sysfs
(/proc/device-tree). It uses a recursive function inspired
from dtc read_fstree.
Signed-off-by: Eric Auger
Reviewed-by: Peter Maydell
---
v6 -> v7:
- rebase on d38ea87ac54af64ef611de434d07c12dc0399216:
"all: Clean up include
atal. This preserves the existing
behaviour. Then to use the API with your optional semantic a null
parameter can be conveyed.
Signed-off-by: Eric Auger
Reviewed-by: Peter Crosthwaite
---
v4 -> v5:
- add Peter's R-b
- remove comment about error_fatal
v1 -> v2:
- add a doc comment in
qemu_fdt_setprop asserts in case of error hence no need to check
the returned value.
Signed-off-by: Eric Auger
Reviewed-by: Peter Maydell
---
v6 -> v7:
- add Peter's R-b
v3 -> v4: fix returned value
---
hw/arm/sysbus-fdt.c | 19 +--
1 file changed, 5 insert
This new helper routine returns a NULL terminated array of
node paths matching a node name and a compat string.
Signed-off-by: Eric Auger
Reviewed-by: Peter Maydell
---
v6 -> v7:
- add Peter's R-b
v5 -> v6:
- in case of error, free the resources and return NULL
- update the doc
to a guest device tree node
- fdt_build_clock_node builds a guest clock node and checks the host
fellow clock is a fixed one.
fdt_build_clock_node will become static as soon as it gets used. A
dummy pre-declaration is needed for compilation of this patch.
Signed-off-by: Eric Auger
Reviewed-by:
use the new lenp parameter to interpret
the error if any.
Signed-off-by: Eric Auger
Reviewed-by: Peter Crosthwaite
---
v4 -> v5:
- Add Peter's R-b
- remove comment about error_fatal
v3 : creation
---
device_tree.c| 21 ++---
hw/arm/boot.c
ifdef CONFIG_LINUX.
Signed-off-by: Eric Auger
Reviewed-by: Peter Maydell
---
v6 -> v7:
- add Peter's R-b
v5 -> v6:
- in sysfs_to_dt_name check substrings[0] before substrings[1]
- on qemu_fdt_node_path, check the returned value is not null
(new dt parsing error case)
Hi,
On 02/17/2016 08:59 PM, Alex Williamson wrote:
> On Tue, 16 Feb 2016 18:40:47 +
> Peter Maydell wrote:
>
>> On 1 February 2016 at 13:51, Eric Auger wrote:
>>> This series allows to set up AMD XGBE passthrough. This was tested on AMD
>>> Seattle.
&g
Hi Peter,
On 02/16/2016 07:14 PM, Peter Maydell wrote:
> On 29 January 2016 at 16:53, Eric Auger wrote:
>> This patch stores information about assigned legacy interrupt numbers in
>> GPEX host structure.
>> This is used during GPEX INTx number determination from a pin d
Hi Peter,
On 02/16/2016 07:21 PM, Peter Maydell wrote:
> On 29 January 2016 at 16:53, Eric Auger wrote:
>> Registers a 16x64kB reserved iova region. Currently this iova
>> region is used by the host kernel to map host MSI controller frames
>> (GICv2m, GITS_TRANSLATER). The h
346a68ade46905a17fefe9d:
>>
>> hw/arm/sysbus-fdt: remove qemu_fdt_setprop returned value check
>> (2016-02-18 12:43:13 -0700)
>>
>> VFIO updates 2016-02-18
>>
>> - AER pre-enable and misc fixes (Cao jin and Chen Fan)
>> - PCI_CAP_LIST_NEXT clea
fstree.
- take into account Alex' comments
- remove qemu_fdt_getprop_optional and use error API instead
Best Regards
Eric
Eric Auger (8):
hw/vfio/platform: amd-xgbe device
device_tree: introduce load_device_tree_from_sysfs
device_tree: introduce qemu_fdt_node_path
device_tree: qemu_fd
atal. This preserves the existing
behaviour. Then to use the API with your optional semantic a null
parameter can be conveyed.
Signed-off-by: Eric Auger
Reviewed-by: Peter Crosthwaite
---
v4 -> v5:
- add Peter's R-b
- remove comment about error_fatal
v1 -> v2:
- add a doc comment in
This patch introduces the amd-xgbe VFIO platform device. It
allows the guest to do passthrough on a device exposing an
"amd,xgbe-seattle-v1a" compat string.
Signed-off-by: Eric Auger
Reviewed-by: Alex Bennée
---
RFC -> v1:
- add Alex' R-b
---
hw/vfio/Makefile.objs
This new helper routine returns a NULL terminated array of
node paths matching a node name and a compat string.
Signed-off-by: Eric Auger
Reviewed-by: Peter Maydell
---
v7 -> v8:
- don't use g_slist_free_full anymore (glib 2.28)
v6 -> v7:
- add Peter's R-b
v5 -> v6:
- in
This function returns the host device tree blob from sysfs
(/proc/device-tree). It uses a recursive function inspired
from dtc read_fstree.
Signed-off-by: Eric Auger
Reviewed-by: Peter Maydell
---
v6 -> v7:
- rebase on d38ea87ac54af64ef611de434d07c12dc0399216:
"all: Clean up include
to a guest device tree node
- fdt_build_clock_node builds a guest clock node and checks the host
fellow clock is a fixed one.
fdt_build_clock_node will become static as soon as it gets used. A
dummy pre-declaration is needed for compilation of this patch.
Signed-off-by: Eric Auger
Reviewed-by:
qemu_fdt_setprop asserts in case of error hence no need to check
the returned value.
Signed-off-by: Eric Auger
Reviewed-by: Peter Maydell
---
v6 -> v7:
- add Peter's R-b
v3 -> v4: fix returned value
---
hw/arm/sysbus-fdt.c | 19 +--
1 file changed, 5 insert
ifdef CONFIG_LINUX.
Signed-off-by: Eric Auger
Reviewed-by: Peter Maydell
---
v6 -> v7:
- add Peter's R-b
v5 -> v6:
- in sysfs_to_dt_name check substrings[0] before substrings[1]
- on qemu_fdt_node_path, check the returned value is not null
(new dt parsing error case)
use the new lenp parameter to interpret
the error if any.
Signed-off-by: Eric Auger
Reviewed-by: Peter Crosthwaite
---
v4 -> v5:
- Add Peter's R-b
- remove comment about error_fatal
v3 : creation
---
device_tree.c| 21 ++---
hw/arm/boot.c
Hi Gerd,
On 03/01/2016 03:49 PM, Gerd Hoffmann wrote:
> On Di, 2016-02-16 at 14:31 -0700, Alex Williamson wrote:
>> v3:
>
> Have a new version of this?
> Want look at this, but it doesn't apply to qemu master.
I think you have subset of this series in:
"[Qemu-devel] [PATCH 0/6] vfio: Prep for vGPU
Hi Paolo
On 06/24/2015 11:55 AM, Paolo Bonzini wrote:
>
>
> On 15/06/2015 18:33, Eric Auger wrote:
>> Now that the platform-bus is instantiated in virt, the ARM sysbus-fdt
>> infrastructure, the VFIO platform device and its first Calxeda xgmac
>> specialization al
Hi Peter,
On 06/26/2015 01:41 PM, Peter Maydell wrote:
> On 15 June 2015 at 17:33, Eric Auger wrote:
>> VFIO platform device needs to setup irqfd but it does not know the
>> gsi corresponding to the device qemu_irq. This series proposes to
>> store a hash table in kvm_state
On 06/26/2015 01:43 PM, Peter Maydell wrote:
> On 15 June 2015 at 17:33, Eric Auger wrote:
>> The arm_gic_kvm now calls kvm_irqchip_set_qemuirq_gsi to build
>> the hash table storing qemu_irq/gsi mappings. From that point on
>> irqfd can be setup directly fr
On 06/26/2015 01:57 PM, Peter Maydell wrote:
> On 15 June 2015 at 17:33, Eric Auger wrote:
>> This patch aims at optimizing IRQ handling using irqfd framework.
>>
>> Instead of handling the eventfds on user-side they are handled on
>> kernel side using
>> - the K
Anticipating for the introduction of new add/remove functions taking
a qemu_irq parameter, let's rename existing ones with a gsi suffix.
Signed-off-by: Eric Auger
Tested-by: Vikram Sethi
Acked-by: Paolo Bonzini
---
v16 -> v17:
- added Paolo's A-b
v15 -> v16:
- added Vikram
is supposed to use it.
kvm_irqchip_[add, remove]_irqfd_notifier allows to setup/tear down
irqfd directly from the qemu_irq.
Signed-off-by: Eric Auger
Tested-by: Vikram Sethi
Acked-by: Paolo Bonzini
---
v16 -> v17
- replace GPOINTER_TO_UINT by GPOINTER_TO_INT
- add Paolo's ack
v1
ings to
implement 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
-
Commit f41389ae3c54b introduced kvm_resamplefds_enabled() and
associated kvm_resamplefds_allowed boolean. This patch adds
non-KVM version for kvm_resamplefds_enabled and also declares
kvm_resamplefds_allowed in kvm-stub as it is done for fellow
kvm_irqfds_allowed.
Signed-off-by: Eric Auger
The arm_gic_kvm now calls kvm_irqchip_set_qemuirq_gsi to build
the hash table storing qemu_irq/gsi mappings. From that point on
irqfd can be setup directly from the qemu_irq using
kvm_irqchip_add_irqfd_notifier.
Signed-off-by: Eric Auger
Tested-by: Vikram Sethi
Acked-by: Paolo Bonzini
---
v16
Add a new connect_irq_notifier notifier in the SysBusDeviceClass. This
notifier, if populated, is called after sysbus_connect_irq.
This mechanism is used to setup VFIO signaling once VFIO platform
devices get attached to their platform bus, on a machine init done
notifier.
Signed-off-by: Eric
for fast/slow path swap.
Overall this brings significant performance improvements.
Signed-off-by: Alvise Rigo
Signed-off-by: Eric Auger
Reviewed-by: Alex Bennée
Tested-by: Vikram Sethi
---
v16 -> v17:
- Take into account Peter's comments
- most importantly remove #ifdef CONFIG_
On 06/26/2015 06:26 PM, Paolo Bonzini wrote:
>
>
> On 26/06/2015 18:21, Alex Williamson wrote:
+#ifdef CONFIG_KVM
+if (kvm_irqfds_enabled() && kvm_resamplefds_enabled() &&
+vdev->irqfd_allowed) {
+sbc->connect_irq_notifier = vfio_start_irqfd_injection;
>> S
is supposed to use it.
kvm_irqchip_[add, remove]_irqfd_notifier allows to setup/tear down
irqfd directly from the qemu_irq.
Signed-off-by: Eric Auger
Tested-by: Vikram Sethi
Acked-by: Paolo Bonzini
---
v16 -> v17
- replace GPOINTER_TO_UINT by GPOINTER_TO_INT
- add Paolo's ack
v1
Anticipating for the introduction of new add/remove functions taking
a qemu_irq parameter, let's rename existing ones with a gsi suffix.
Signed-off-by: Eric Auger
Tested-by: Vikram Sethi
Acked-by: Paolo Bonzini
---
v16 -> v17:
- added Paolo's A-b
v15 -> v16:
- added Vikram
de generation removed from device and handled in
hw/arm/dyn_sysbus_devtree.c
- remove "hw/vfio: add an example calxeda_xgmac": with removal of
device tree node generation we do not have so many things to
implement in that derived device yet. May be re-introduced later
on if needed typ
Commit f41389ae3c54b introduced kvm_resamplefds_enabled() and
associated kvm_resamplefds_allowed boolean. This patch adds
non-KVM version for kvm_resamplefds_enabled and also declares
kvm_resamplefds_allowed in kvm-stub as it is done for fellow
kvm_irqfds_allowed.
Signed-off-by: Eric Auger
The arm_gic_kvm now calls kvm_irqchip_set_qemuirq_gsi to build
the hash table storing qemu_irq/gsi mappings. From that point on
irqfd can be setup directly from the qemu_irq using
kvm_irqchip_add_irqfd_notifier.
Signed-off-by: Eric Auger
Tested-by: Vikram Sethi
Acked-by: Paolo Bonzini
---
v16
for fast/slow path swap.
Overall this brings significant performance improvements.
Signed-off-by: Alvise Rigo
Signed-off-by: Eric Auger
Reviewed-by: Alex Bennée
Tested-by: Vikram Sethi
---
v17 -> v18:
- remove found in vfio_start_irqfd_injection
- unconditionnally set
Add a new connect_irq_notifier notifier in the SysBusDeviceClass. This
notifier, if populated, is called after sysbus_connect_irq.
This mechanism is used to setup VFIO signaling once VFIO platform
devices get attached to their platform bus, on a machine init done
notifier.
Signed-off-by: Eric
On 06/29/2015 11:47 AM, Peter Maydell wrote:
> On 26 June 2015 at 18:29, Eric Auger wrote:
>> Commit f41389ae3c54b introduced kvm_resamplefds_enabled() and
>> associated kvm_resamplefds_allowed boolean. This patch adds
>> non-KVM version for kvm_resamplefds_enab
This patch introduces the amd-xgbe VFIO platform device. It
allows the guest to do passthrough on a device exposing an
"amd,xgbe-seattle-v1a" compat string.
Signed-off-by: Eric Auger
---
hw/vfio/Makefile.objs | 1 +
hw/vfio/amd-xgbe.c
g16648.html.
- Currently host booted with ACPI is not supported.
The patches can be found at
https://git.linaro.org/people/eric.auger/qemu.git/shortlog/refs/heads/v2.4.0-xgbe-rfc
Best Regards
Eric
Eric Auger (6):
hw/vfio/platform: amd-xgbe device
device_tree: introduce load_device_tree_from
guest device tree node
- fdt_build_clock_node builds a guest clock node and checks the host
fellow clock is a fixed one.
fdt_build_clock_node will become static as soon as a they get used
Signed-off-by: Eric Auger
---
hw/arm/sysbus-fdt.c | 110
1
This new helper routine returns the node path of a device
referred to by its name and compat string.
Signed-off-by: Eric Auger
---
device_tree.c| 40
include/sysemu/device_tree.h | 3 +++
2 files changed, 43 insertions(+)
diff --git a
Current qemu_fdt_getprop exits if the property is not found. It is
sometimes needed to read an optional property, in which case we do
not wish to exit but simply returns a null value.
This is what this new qemu_fdt_getprop_optional function does.
Signed-off-by: Eric Auger
---
device_tree.c
d XGBE/PHY representation (>= 4.2 host kernel).
2 clock nodes (dma and ptp) also are created. It is mandated those clocks
are fixed on host side.
Signed-off-by: Eric Auger
---
hw/arm/sysbus-fdt.c | 171 ++--
1 file changed, 167 insertions(+), 4 delet
.
Signed-off-by: Eric Auger
---
device_tree.c| 40
include/sysemu/device_tree.h | 1 +
2 files changed, 41 insertions(+)
diff --git a/device_tree.c b/device_tree.c
index a9f5f8e..58a5329 100644
--- a/device_tree.c
+++ b/device_tree.c
KVM_SET_GSI_ROUTING overrides the kernel default irqchip table.
If no routing entry exists for the GSI, any IRQFD signaling for
this GSI will fail.
Signed-off-by: Eric Auger
---
gsi routing for ARM is functional but needs to be maturated.
The API for MSI routing still is in discussion.
See http
Hi,
On 08/12/2015 03:23 PM, Christoffer Dall wrote:
> On Wed, Aug 12, 2015 at 2:59 PM, Peter Maydell
> wrote:
>> On 12 August 2015 at 13:27, Pavel Fedin wrote:
>>> Hello!
>>>
I still think this is the wrong approach -- see my remarks
in the previous round of patch review.
>>>
>>> You
On 08/12/2015 04:24 PM, Christoffer Dall wrote:
> On Wed, Aug 12, 2015 at 4:14 PM, Eric Auger wrote:
>> Hi,
>> On 08/12/2015 03:23 PM, Christoffer Dall wrote:
>>> On Wed, Aug 12, 2015 at 2:59 PM, Peter Maydell
>>> wrote:
>>>> On 12 August 20
Re-indent in a15memmap after VIRT_PLATFORM_BUS introduction
Signed-off-by: Eric Auger
Reviewed-by: Alex Bennée
---
v11 -> v12:
- Add Alex R-b
---
hw/arm/virt.c | 28 ++--
1 file changed, 14 insertions(+), 14 deletions(-)
diff --git a/hw/arm/virt.c b/hw/arm/vir
becomes static
- load_dtb renamed into arm_load_dtb
- add copyright in hw/arm/dyn_sysbus_devtree.c
v1 -> v2:
- device node generation no more in sysbus device but in
dyn_sysbus_devtree
- VFIO region shrinked to 4MB and relocated in machvirt to avoid PCI
shrink (dynamic vfio-mmio support might come latter
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
Reviewed-by: Shannon Zhao
Reviewed-by: Alexander Graf
Reviewed-by: Alex
sysbus device nodes are integrated in the dtb.
Signed-off-by: Eric Auger
Reviewed-by: Shannon Zhao
Reviewed-by: Alexander Graf
Reviewed-by: Alex Bennée
---
v11 -> v12:
- Add Alex R-b
v8 -> v9:
- fix compilation with arm-linux-user
- reorder fields in ArmLoadKernelNotifier and use DO_UPCA
sysbus devices. Indeed
dynamic sysbus devices are created after machine init.
machvirt also registers a notifier that will build the device
tree nodes for the platform bus and its children dynamic sysbus
devices.
Signed-off-by: Eric Auger
Reviewed-by: Alex Bennée
---
v12 -> v13:
- Changed st
Hi Alex,
On 06/04/2015 08:11 PM, Alex Williamson wrote:
> On Wed, 2015-05-27 at 10:11 +0100, Eric Auger wrote:
>> This series aims at enabling KVM platform device passthrough.
>>
>> On kernel side, the vfio platform driver is needed, available from
>> 4.1-rc1 onwards.
t;> >
>> >
>> > git://github.com/awilliam/qemu-vfio.git tags/vfio-update-20150608.0
>> >
>> > for you to fetch changes up to 88be7b6bf4c86d89e70585881b6b3326f9bac822:
>> >
>> > hw/vfio/platform: calxeda xgmac device (2015-06-08 09:25:
The platform device class has become abstract. This patch introduces
a calxeda xgmac device that derives from it.
Signed-off-by: Eric Auger
Reviewed-by: Alex Bennée
---
v16 -> v17:
- properly attach vfio_platform_calxeda_xgmac_vmstate
v15 -> v16:
- added Vikram's T-b
v10 -> v
Gonglei,
Thanks for the report and please apologize for the inconvenience. I am
going to send a fix reverting to simple g_malloc0 then.
Best Regards
Eric
On 06/11/2015 04:21 AM, Gonglei wrote:
> On 2015/6/9 23:55, Alex Williamson wrote:
>> +static int vfio_populate_device(VFIODevice *vbasedev)
g_malloc0_n() is introduced since glib-2.24 while QEMU currently
requires glib-2.22. This may cause a link error on some distributions.
Signed-off-by: Eric Auger
---
hw/vfio/platform.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/hw/vfio/platform.c b/hw/vfio/platform.c
g_malloc0_n() is introduced since glib-2.24 while QEMU currently
requires glib-2.22. This may cause a link error on some distributions.
Signed-off-by: Eric Auger
---
v1 -> v2:
- replace g_malloc0 by g_new0
---
hw/vfio/platform.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
d
Hi Gonglei,
On 06/11/2015 09:55 AM, Gonglei wrote:
> On 2015/6/11 15:14, Eric Auger wrote:
>> g_malloc0_n() is introduced since glib-2.24 while QEMU currently
>> requires glib-2.22. This may cause a link error on some distributions.
>>
>> Signed-off-by: Eric Auger
&g
Replace g_malloc0 occurences by g_new0 where that makes obvious sense.
This looks to be the best practice bringing extra compile-time checking
and integer overflow protection.
Signed-off-by: Eric Auger
---
hw/vfio/common.c | 8
hw/vfio/pci.c | 24
hw
Hi Peter,
On 06/12/2015 04:54 AM, Peter Crosthwaite wrote:
> On Tue, Jun 2, 2015 at 9:33 AM, Peter Maydell
> wrote:
>> From: Eric Auger
>>
>> Device tree nodes for the platform bus and its children dynamic sysbus
>> devices are added in a machine init done
On 06/12/2015 10:25 AM, Eric Auger wrote:
> Hi Peter,
> On 06/12/2015 04:54 AM, Peter Crosthwaite wrote:
>> On Tue, Jun 2, 2015 at 9:33 AM, Peter Maydell
>> wrote:
>>> From: Eric Auger
>>>
>>> Device tree nodes for the platform bus and its child
commit ac9d32e39664e060cd1b538ff190980d57ad69e4 had the consequence to
register the do_cpu_reset after the rom_reset one. Hence they get
executed in the wrong order. This commit restores the registration of
do_cpu_reset in arm_load_kernel.
Signed-off-by: Eric Auger
---
hw/arm/boot.c | 20
501 - 600 of 4501 matches
Mail list logo
