On Fri, May 08, 2026 at 06:10:29PM -0500, Bjorn Helgaas wrote: > On Fri, May 08, 2026 at 10:47:14PM +0000, Haiyang Zhang wrote: > > > -----Original Message----- > > > From: Bjorn Helgaas <[email protected]> > > > Sent: Friday, May 8, 2026 6:38 PM > > > To: Haiyang Zhang <[email protected]> > > > Cc: [email protected]; [email protected]; KY Srinivasan > > > <[email protected]>; Haiyang Zhang <[email protected]>; Wei Liu > > > <[email protected]>; Dexuan Cui <[email protected]>; Long Li > > > <[email protected]>; Andrew Lunn <[email protected]>; David S. > > > Miller <[email protected]>; Eric Dumazet <[email protected]>; Jakub > > > Kicinski <[email protected]>; Paolo Abeni <[email protected]>; Bjorn Helgaas > > > <[email protected]>; Simon Horman <[email protected]>; Shradha Gupta > > > <[email protected]>; Dipayaan Roy > > > <[email protected]>; Erni Sri Satya Vennela > > > <[email protected]>; [email protected]; linux- > > > [email protected]; Paul Rosswurm <[email protected]> > > > Subject: [EXTERNAL] Re: [PATCH net-next] net: mana: Add handler for sriov > > > configure > > > > > > On Fri, May 08, 2026 at 03:04:06PM -0700, Haiyang Zhang wrote: > > > > From: Haiyang Zhang <[email protected]> > > > > > > > > Add callback function for the pci_driver, sriov_configure. > > > > > > > > Also disable VF autoprobe when it runs as PF driver on bare metal, > > > > since the hardware side may not have the VF ready immediately. > > > > > > > > Export pci_vf_drivers_autoprobe() so the driver can toggle the VF > > > > autoprobe flag. > > > > > > Technically pci_vf_drivers_autoprobe() doesn't *toggle* the autoprobe > > > flag. That would mean setting it to the opposite of its current > > > value. > > > > > > Here I would say "so the driver can prevent autoprobing of the VFs", > > > which is the intent. > > Thanks, I will change the wording. > > > > > > > > Out of curiosity, how do the VFs eventually get probed? I guess > > > there's some other mechanism that tells you when they're ready, and > > > you manually use sysfs 'sriov_drivers_autoprobe' to enable probing, > > > then bind drivers to them via sysfs? > > We have a user program talking to the Azure backplane to get that > > information. > > @Paul Rosswurm, do you have more details? > > > > > > > The prevention of autoprobing sounds like a critical part of this > > > change; might be worth saying something in the subject, because "add > > > sriov configure" doesn't include much information. > > How about "Add handler for sriov configure with VF autoprobe off"? > > OK by me :)
Bjorn, I believe it is the wrong decision to allow toggling a user‑visible knob without the user’s awareness. In this case, they can either disable autoprobe on the PF or rely on EPROBE_DEFER. In all cases, the same functionality can be achieved without changing PCI autoprobe code. Thanks. >
