Re: [PATCH v4] vfio/cdx: add support for CDX bus

From: Alex Williamson
Date: Tue Apr 18 2023 - 16:41:36 EST


On Tue, 18 Apr 2023 12:50:13 +0000
"Gupta, Nipun" <Nipun.Gupta@xxxxxxx> wrote:

> > -----Original Message-----
> > From: Jason Gunthorpe <jgg@xxxxxxxx>
> > Sent: Tuesday, April 18, 2023 5:40 PM
> > To: Gupta, Nipun <Nipun.Gupta@xxxxxxx>
> > Cc: alex.williamson@xxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx;
> > kvm@xxxxxxxxxxxxxxx; masahiroy@xxxxxxxxxx; nathan@xxxxxxxxxx;
> > ndesaulniers@xxxxxxxxxx; nicolas@xxxxxxxxx; git (AMD-Xilinx) <git@xxxxxxx>;
> > Anand, Harpreet <harpreet.anand@xxxxxxx>; Jansen Van Vuuren, Pieter
> > <pieter.jansen-van-vuuren@xxxxxxx>; Agarwal, Nikhil
> > <nikhil.agarwal@xxxxxxx>; Simek, Michal <michal.simek@xxxxxxx>
> > Subject: Re: [PATCH v4] vfio/cdx: add support for CDX bus
> >
> > Caution: This message originated from an External Source. Use proper caution
> > when opening attachments, clicking links, or responding.
> >
> >
> > On Tue, Apr 18, 2023 at 05:06:55PM +0530, Nipun Gupta wrote:
> >
> > > diff --git a/drivers/vfio/Kconfig b/drivers/vfio/Kconfig
> > > index 89e06c981e43..aba36f5be4ec 100644
> > > --- a/drivers/vfio/Kconfig
> > > +++ b/drivers/vfio/Kconfig
> > > @@ -57,6 +57,7 @@ source "drivers/vfio/pci/Kconfig"
> > > source "drivers/vfio/platform/Kconfig"
> > > source "drivers/vfio/mdev/Kconfig"
> > > source "drivers/vfio/fsl-mc/Kconfig"
> > > +source "drivers/vfio/cdx/Kconfig"
> >
> > keep sorted
>
> Since it is not sorted as of now, should a separate patch to be created for
> sorting, before adding vfio-cdx?

These are essentially in chronological order rather than alphabetical,
so I don't really understand this request from Jason. Perhaps if it
was already alphabetical the request would be justified, but I don't
see any obligation here.

> > > endif
> > >
> > > source "virt/lib/Kconfig"
> > > diff --git a/drivers/vfio/Makefile b/drivers/vfio/Makefile
> > > index 70e7dcb302ef..1a27b2516612 100644
> > > --- a/drivers/vfio/Makefile
> > > +++ b/drivers/vfio/Makefile
> > > @@ -14,3 +14,4 @@ obj-$(CONFIG_VFIO_PCI) += pci/
> > > obj-$(CONFIG_VFIO_PLATFORM) += platform/
> > > obj-$(CONFIG_VFIO_MDEV) += mdev/
> > > obj-$(CONFIG_VFIO_FSL_MC) += fsl-mc/
> > > +obj-$(CONFIG_VFIO_CDX) += cdx/
> >
> > keep sorted
>
> Is there Linux guideline here on how objects and folders in Makefile are sorted,
> as there are mix and match of files and folders in "drivers/vfio/Makefile".
> I could not find any reference for sorting in other Makefiles as well.

Same here, and I also don't know of a best practices reference that
suggests an alphabetical ordering policy.

> > > diff --git a/drivers/vfio/cdx/Makefile b/drivers/vfio/cdx/Makefile
> > > new file mode 100644
> > > index 000000000000..82e4ef412c0f
> > > --- /dev/null
> > > +++ b/drivers/vfio/cdx/Makefile
> > > @@ -0,0 +1,8 @@
> > > +# SPDX-License-Identifier: GPL-2.0
> > > +#
> > > +# Copyright (C) 2022-2023, Advanced Micro Devices, Inc.
> > > +#
> > > +
> > > +obj-$(CONFIG_VFIO_CDX) += vfio-cdx.o
> > > +
> > > +vfio-cdx-objs := vfio_cdx.o
> >
> > Linus has asked about "suttering" filenames before, suggest to call
> > this
> >
> > "vfio/cdx/main.c"
>
> Okay, I do not any strong affiliation towards the name.
> Alex, your thoughts on this please?

I think Jason means "stuttering" file paths, or essentially self
redundancy, ex. vfio/cdx/vfio_cdx.c. Yes, Linus has scolded us for
vfio/pci/vfio_pci_* naming in the past, so in the spirit of not being
further scolded I'd agree with the suggestion here. Thanks,

Alex