pci_irq_handler documents that it must be called with 0 <= irq_num <=
3 and level either 0 or 1.  Add assertions that the caller has passed
us in valid arguments.

In particular, if a device model fails to set the PCI_INTERRUPT_PIN
field in its config space correctly to indicate that it has an
interrupt, and then tries to raise an interrupt (either by calling
pci_set_irq(), or by getting a qemu_irq from pci_allocate_irq() and
then calling qemu_set_irq() on that) we will now diagnose this device
model bug with an assertion rather than engaging in the undefined
behaviour of shifting by a negative number.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
 hw/pci/pci.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/hw/pci/pci.c b/hw/pci/pci.c
index 0eadcdbc9ec..38aefe22ab3 100644
--- a/hw/pci/pci.c
+++ b/hw/pci/pci.c
@@ -1449,6 +1449,9 @@ static void pci_irq_handler(void *opaque, int irq_num, int level)
     PCIDevice *pci_dev = opaque;
     int change;
 
+    assert(irq_num >= 0 && irq_num < PCI_NUM_PINS);
+    assert(level == 0 || level == 1);
+
     change = level - pci_irq_state(pci_dev, irq_num);
     if (!change)
         return;
-- 
2.20.1

On 3/23/21 5:46 PM, Peter Maydell wrote:
> pci_irq_handler documents that it must be called with 0 <= irq_num <=
> 3 and level either 0 or 1.  Add assertions that the caller has passed
> us in valid arguments.
> 
> In particular, if a device model fails to set the PCI_INTERRUPT_PIN
> field in its config space correctly to indicate that it has an
> interrupt, and then tries to raise an interrupt (either by calling
> pci_set_irq(), or by getting a qemu_irq from pci_allocate_irq() and
> then calling qemu_set_irq() on that) we will now diagnose this device
> model bug with an assertion rather than engaging in the undefined
> behaviour of shifting by a negative number.
> 
> Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
> ---
>  hw/pci/pci.c | 3 +++
>  1 file changed, 3 insertions(+)

Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>

I included (mostly) same patch into my patch series just for patch
completeness.
Please choose whichever you like.

Reviewed-by: Isaku Yamahata <isaku.yamahata@intel.com>

On Tue, Mar 23, 2021 at 04:46:01PM +0000,
Peter Maydell <peter.maydell@linaro.org> wrote:

> pci_irq_handler documents that it must be called with 0 <= irq_num <=
> 3 and level either 0 or 1.  Add assertions that the caller has passed
> us in valid arguments.
> 
> In particular, if a device model fails to set the PCI_INTERRUPT_PIN
> field in its config space correctly to indicate that it has an
> interrupt, and then tries to raise an interrupt (either by calling
> pci_set_irq(), or by getting a qemu_irq from pci_allocate_irq() and
> then calling qemu_set_irq() on that) we will now diagnose this device
> model bug with an assertion rather than engaging in the undefined
> behaviour of shifting by a negative number.
> 
> Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
> ---
>  hw/pci/pci.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/hw/pci/pci.c b/hw/pci/pci.c
> index 0eadcdbc9ec..38aefe22ab3 100644
> --- a/hw/pci/pci.c
> +++ b/hw/pci/pci.c
> @@ -1449,6 +1449,9 @@ static void pci_irq_handler(void *opaque, int irq_num, int level)
>      PCIDevice *pci_dev = opaque;
>      int change;
>  
> +    assert(irq_num >= 0 && irq_num < PCI_NUM_PINS);
> +    assert(level == 0 || level == 1);
> +
>      change = level - pci_irq_state(pci_dev, irq_num);
>      if (!change)
>          return;
> -- 
> 2.20.1
> 
> 

-- 
Isaku Yamahata <isaku.yamahata@gmail.com>