Re: [PATCH] block/blkio: Make driver nvme-io_uring take a "path" instead of a "filename"

[Stefan Hajnoczi <stefanha@redhat.com>]

[-- Attachment #1: Type: text/plain, Size: 1238 bytes --]

On Sat, Oct 29, 2022 at 10:50:40AM +0100, Alberto Faria wrote:
> On Sat, Oct 29, 2022 at 7:05 AM Markus Armbruster <armbru@redhat.com> wrote:
> > Alberto Faria <afaria@redhat.com> writes:
> >
> > > The nvme-io_uring driver expects a character special file such as
> > > /dev/ng0n1. Follow the convention of having a "filename" option when a
> > > regular file is expected, and a "path" option otherwise.
> >
> > I suspect this is by accident, not by design.  Is it desirable?
> 
> I'm not sure. Naturally I'd be happier if either "filename" or "path"
> was used everywhere. Maybe we should settle on a single one for all
> the libblkio drivers? Or maybe we should just leave things as is?

It wasn't an accident but maybe it was still a bad idea on my part.

My thinking was that io_uring takes regular files and therefore the
"filename" option name is appropriate. UNIX domain sockets and special
devices usually have the "path" option name (e.g. --chardev socket,path=
for AF_UNIX).

I agree with changing the option to "path" for nvme-io_uring.

The overall naming strategy is debatable. I think we can keep it, but if
the majority wants everything to be "filename" or "path" I'd be okay
with that.

Stefan

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]