[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-devel] [PULL 3/3] vfio/common: Introduce vfio_set_irq_signalin
From: |
Auger Eric |
Subject: |
Re: [Qemu-devel] [PULL 3/3] vfio/common: Introduce vfio_set_irq_signaling helper |
Date: |
Tue, 2 Jul 2019 18:33:19 +0200 |
User-agent: |
Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101 Thunderbird/60.4.0 |
Hi,
On 7/2/19 6:30 PM, Alex Williamson wrote:
> On Tue, 2 Jul 2019 16:58:02 +0100
> Peter Maydell <address@hidden> wrote:
>
>> On Tue, 2 Jul 2019 at 16:56, Alex Williamson <address@hidden> wrote:
>>> When we're emulating writes to the MSI-X vector table we have no
>>> failure path up to the guest. Real hardware cannot fail to enable a
>>> vector that's available in hardware, thus we can either log the issue,
>>> ignore the issue, or fault. I guess Coverity is simply noting that
>>> other cases are tested while this is not, therefore we should either
>>> explicitly ignore the return value with a cast to void or take this as
>>> an opportunity to log the fault, which might be useful in debugging a
>>> device that isn't working properly. Thanks,
>>
>> Yeah, Coverity's check here is purely a heuristic ("did we seem
>> to check returns from this function in other places?") so it's
>> wrong sometimes. If you want me to mark this as a false positive
>> in the coverity UI I can do that.
>
> TBH, it seems like a good nag to log it properly. Eric, do you mind
> posting a fix to do that? Thanks,
Sure I will send a fix.
Thanks
Eric
>
> Alex
>