Re: [PATCH for-next v2 1/4] RDMA/erdma: Make the device probe process more robust

[Cheng Xu <chengyou@linux.alibaba.com>]

On 9/5/24 12:06 AM, Jason Gunthorpe wrote:
> On Mon, Sep 02, 2024 at 05:09:09PM +0800, Cheng Xu wrote:
> 
>> The hardware now requires that the former reset (issued in the
>> remove routine) must be completed before device init (issued in the
>> probe routine). Waiting the reset completed either in the remove
>> routine or in the probe routine both can meet the requirement.  This
>> patch chose to wait in the probe routine because it can speed up the
>> remove process.
> 
> But what happens if you attach VFIO or some other driver while this
> background reset is occuring? Are you OK with that?

Yes, it's OK.

To simplify the description, We have two relevant components in the
hardware: The device management engine and the RDMA engine. The PCIe
device initialization (erdma, vfio and other drivers will perform it)
is handled by device management engine without any issue. The issue described
in this patch pertains to the RDMA engine, which is only invoked by erdma
driver.

In fact, this low-probability issue is not very serious and can be resolved
by reloading the driver. After internal discussion, we have decided to eliminate
this constraint through appropriate firmware modifications.

Thanks,
Cheng Xu

> 
> Jason