[Qemu-devel] [PATCH 0/2] Generic PCIe host bridge legacy interrupts routing support

[Qemu-devel] [PATCH 0/2] Generic PCIe host bridge legacy interrupts routing support

[Pranavkumar Sawargaonkar]

This patch adds:
- Function to determine an irq number from qemu_irq.
- Function to determine and route legacy interrupts to a VM.


Pranavkumar Sawargaonkar (2):
  hw: irq: Add API to get irq number from qemu_irq
  pci: GPEX: Add support to route legacy interrupts

 hw/core/irq.c      |  9 +++++++++
 hw/pci-host/gpex.c | 12 ++++++++++++
 include/hw/irq.h   |  1 +
 3 files changed, 22 insertions(+)

-- 
1.9.1

[Qemu-devel] [PATCH 1/2] hw: irq: Add API to get irq number from qemu_irq

[Pranavkumar Sawargaonkar]

This patch adds a new function to get irq number from qemu irq.

Signed-off-by: Pranavkumar Sawargaonkar <pranavkumar@linaro.org>
Signed-off-by: Tushar Jagad <tushar.jagad@linaro.org>
---
 hw/core/irq.c    | 9 +++++++++
 include/hw/irq.h | 1 +
 2 files changed, 10 insertions(+)

diff --git a/hw/core/irq.c b/hw/core/irq.c
index 8a62a36..0eeffdf 100644
--- a/hw/core/irq.c
+++ b/hw/core/irq.c
@@ -35,6 +35,15 @@ struct IRQState {
     int n;
 };
 
+int qemu_get_irq_num(qemu_irq irq)
+{
+    if (!irq) {
+        return -1;
+    }
+
+    return irq->n;
+}
+
 void qemu_set_irq(qemu_irq irq, int level)
 {
     if (!irq)
diff --git a/include/hw/irq.h b/include/hw/irq.h
index 4c4c2ea..dd5220c 100644
--- a/include/hw/irq.h
+++ b/include/hw/irq.h
@@ -9,6 +9,7 @@ typedef struct IRQState *qemu_irq;
 
 typedef void (*qemu_irq_handler)(void *opaque, int n, int level);
 
+int qemu_get_irq_num(qemu_irq irq);
 void qemu_set_irq(qemu_irq irq, int level);
 
 static inline void qemu_irq_raise(qemu_irq irq)
-- 
1.9.1

[Qemu-devel] [PATCH 2/2] pci: GPEX: Add support to route legacy interrupts

[Pranavkumar Sawargaonkar]

This patch adds routine to route legacy interrupts to VM.

Signed-off-by: Pranavkumar Sawargaonkar <pranavkumar@linaro.org>
Signed-off-by: Tushar Jagad <tushar.jagad@linaro.org>
---
 hw/pci-host/gpex.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/hw/pci-host/gpex.c b/hw/pci-host/gpex.c
index 9d8fb5a..35842e4 100644
--- a/hw/pci-host/gpex.c
+++ b/hw/pci-host/gpex.c
@@ -42,6 +42,17 @@ static void gpex_set_irq(void *opaque, int irq_num, int level)
     qemu_set_irq(s->irq[irq_num], level);
 }
 
+static PCIINTxRoute gpex_route_intx_pin_to_irq(void *opaque, int pin)
+{
+    PCIINTxRoute route;
+    GPEXHost *s = opaque;
+
+    route.mode = PCI_INTX_ENABLED;
+    route.irq = qemu_get_irq_num(s->irq[pin]);
+
+    return route;
+}
+
 static void gpex_host_realize(DeviceState *dev, Error **errp)
 {
     PCIHostState *pci = PCI_HOST_BRIDGE(dev);
@@ -66,6 +77,7 @@ static void gpex_host_realize(DeviceState *dev, Error **errp)
                                 &s->io_ioport, 0, 4, TYPE_PCIE_BUS);
 
     qdev_set_parent_bus(DEVICE(&s->gpex_root), BUS(pci->bus));
+    pci_bus_set_route_irq_fn(pci->bus, gpex_route_intx_pin_to_irq);
     qdev_init_nofail(DEVICE(&s->gpex_root));
 }
 
-- 
1.9.1

Re: [Qemu-devel] [PATCH 2/2] pci: GPEX: Add support to route legacy interrupts

[Alexander Graf]

On 04/17/2015 10:55 AM, Pranavkumar Sawargaonkar wrote:
> This patch adds routine to route legacy interrupts to VM.
>
> Signed-off-by: Pranavkumar Sawargaonkar <pranavkumar@linaro.org>
> Signed-off-by: Tushar Jagad <tushar.jagad@linaro.org>

I don't think the device model should have awareness of any global irq 
number spaces. Can you move this into the board model instead? Then you 
also don't need the first patch.


Alex

> ---
>   hw/pci-host/gpex.c | 12 ++++++++++++
>   1 file changed, 12 insertions(+)
>
> diff --git a/hw/pci-host/gpex.c b/hw/pci-host/gpex.c
> index 9d8fb5a..35842e4 100644
> --- a/hw/pci-host/gpex.c
> +++ b/hw/pci-host/gpex.c
> @@ -42,6 +42,17 @@ static void gpex_set_irq(void *opaque, int irq_num, int level)
>       qemu_set_irq(s->irq[irq_num], level);
>   }
>   
> +static PCIINTxRoute gpex_route_intx_pin_to_irq(void *opaque, int pin)
> +{
> +    PCIINTxRoute route;
> +    GPEXHost *s = opaque;
> +
> +    route.mode = PCI_INTX_ENABLED;
> +    route.irq = qemu_get_irq_num(s->irq[pin]);
> +
> +    return route;
> +}
> +
>   static void gpex_host_realize(DeviceState *dev, Error **errp)
>   {
>       PCIHostState *pci = PCI_HOST_BRIDGE(dev);
> @@ -66,6 +77,7 @@ static void gpex_host_realize(DeviceState *dev, Error **errp)
>                                   &s->io_ioport, 0, 4, TYPE_PCIE_BUS);
>   
>       qdev_set_parent_bus(DEVICE(&s->gpex_root), BUS(pci->bus));
> +    pci_bus_set_route_irq_fn(pci->bus, gpex_route_intx_pin_to_irq);
>       qdev_init_nofail(DEVICE(&s->gpex_root));
>   }
>   

Re: [Qemu-devel] [PATCH 0/2] Generic PCIe host bridge legacy interrupts routing support

[Peter Maydell]

On 17 April 2015 at 09:55, Pranavkumar Sawargaonkar
<pranavkumar@linaro.org> wrote:
> This patch adds:
> - Function to determine an irq number from qemu_irq.
> - Function to determine and route legacy interrupts to a VM.

Can you describe in more detail what this patch series
is supposed to do, please? Interrupt routing already
works fine, via the interrupt controller...

thanks
-- PMM

Re: [Qemu-devel] [PATCH 0/2] Generic PCIe host bridge legacy interrupts routing support

[Pranavkumar Sawargaonkar]

Hi PMM,

On 17 April 2015 at 15:23, Peter Maydell <peter.maydell@linaro.org> wrote:
> On 17 April 2015 at 09:55, Pranavkumar Sawargaonkar
> <pranavkumar@linaro.org> wrote:
>> This patch adds:
>> - Function to determine an irq number from qemu_irq.
>> - Function to determine and route legacy interrupts to a VM.
>
> Can you describe in more detail what this patch series
> is supposed to do, please? Interrupt routing already
> works fine, via the interrupt controller...

GPEX driver is not registering a callback to route INTx and
pci_device_route_intx_to_irq() throws a following warning
"PCI: Bug - unimplemented PCI INTx routing".
It also returns -1 as an interrupt number as well as marking it as a
disabled one.
Hence to complete the GPEX driver in terms of INTx routing I have
added this patch but at the same time routing works without this
patchset :) .

>
> thanks
> -- PMM

Thanks,
Pranav