Re: [Qemu-devel] [PATCH] CAN bus MIOe-3680 PCI (dual SJA1000 channel) emulation added.

[Deniz Eren]

Hi Pavel,

Thank you for the merge - CAN-bus support for QEMU is very important in my 
opinion. Most embedded applications utilise CAN and virtualising this interface 
gives an amazing integration test harness.



Best regards,
Deniz

Sent from my iPhone

Deniz Eren
+61 400 307 762

> On 9 Sep. 2016, at 3:10 am, Pavel Pisa <address@hidden> wrote:
> 
> Hello Deniz,
> 
> thanks much for contribution and testing.
> 
> I have applied your patches to our QEMU
> repo
> 
>  https://github.com/CTU-IIG/qemu
> 
> You find QEMU-2.6.1 based version of our merged
> patches on "merged-2.6" branch.
> 
> I have updated and shortly tested all our topic branches
> with actual QEMU-2.7.0 release. Simple test with
> Linux within Linux and SocketCAN on both sides
> works for PCI CAN-S (single SJA1000 channel) card
> well with QEMU-2.7.0.
> 
> So I have undated all topics branches "can-pci",
> "mf624" and "apohw" to this version then and created
> "merged-2.7" branch. "merged" branch is now moved
> to "merged-2.7" as well.
> 
> Your offer to help with mainlining is great as well.
> Some discussion with somebody from core QEMU team
> member is required.
> 
> I hope that actual CAN chip driver and PCI cards
> follow QOM model right way. Problem is the simple
> bus used for mesages delivery between CAN controllers.
> These buses should be converted somehow to QOM acceptable
> model. But for embedded development when you do not
> freeze emulated system state is actual solution quite
> usable as well.
> 
> It would be great to implement CAN FD (Flexible Datarate)
> controller emulation support as well.
> 
> I am trying to find some diploma student for that for
> two years already because he/she can have continuous
> time for such work and testing which I cannot find.
> Help from other instreested or project users would
> be great as well.
> 
> Best wishes,
> 
>              Pavel
> 
> 
> 
>