Re: [Qemu-devel] [PATCH] hw/pcie-pci-bridge: restrict to X86 and ARM

[Marcel Apfelbaum <marcel@redhat.com>]

On 08/11/2017 18:52, Thomas Huth wrote:
> On 08.11.2017 16:20, Marcel Apfelbaum wrote:
>> The PCIE-PCI bridge is specific to "pure" PCI systems
>

Hi Thomas,

> s/PCI/PCIe/ ?
> 

In this context maybe it doesn't really matter,
but if I'll send a v2 I'll make the change, sure.

>> (on QEMU we have X86 and ARM), it does not make sense to
>> have if in other archs.
>>
>> Reported-by: Thomas Huth <thuth@redhat.com>
>> Signed-off-by: Marcel Apfelbaum <marcel@redhat.com>
>> ---
>>   default-configs/arm-softmmu.mak    | 1 +
>>   default-configs/i386-softmmu.mak   | 1 +
>>   default-configs/x86_64-softmmu.mak | 1 +
>>   hw/pci-bridge/Makefile.objs        | 3 ++-
>>   4 files changed, 5 insertions(+), 1 deletion(-)
>>
>> diff --git a/default-configs/arm-softmmu.mak b/default-configs/arm-softmmu.mak
>> index 5059d134c8..33934b1dc4 100644
>> --- a/default-configs/arm-softmmu.mak
>> +++ b/default-configs/arm-softmmu.mak
>> @@ -122,6 +122,7 @@ CONFIG_FSL_IMX25=y
>>   CONFIG_IMX_I2C=y
>>   
>>   CONFIG_PCIE_PORT=y
>> +CONFIG_PCIE_PCI_BRIDGE=y
> 
> Actually, why not simply use CONFIG_PCIE_PORT instead to determine
> whether the PCIe-to-PCI bridge should be included? The device only makes
> sense if there is a PCIe root port available on the system...
> 

I thought about it, but then we should also change all the devices
that makes sense only if connected to PCIe Root Ports, like Intel
Upstream Ports (Switches) and maybe more. (otherwise people will
start asking why it is different.)

We could use CONFIG_PCIE_PCI_BRIDGE=CONFIG_PCIE_PORT or something
but it looks weird.

This is why I've chosen the "independent" config option.
But, to be fair, I have nothing against your suggestion,
let's see what others think.

Thanks for the review,
Marcel

>   Thomas
>