[AMD Public Use] > -----Original Message----- > From: Marcos Scriven <marcos@xxxxxxxxxxx> > Sent: Friday, August 14, 2020 4:53 AM > To: Deucher, Alexander <Alexander.Deucher@xxxxxxx> > Cc: Bjorn Helgaas <helgaas@xxxxxxxxxx>; Shah, Nehal-bakulchandra <Nehal- > bakulchandra.Shah@xxxxxxx>; Kevin Buettner <kevinb@xxxxxxxxxx>; > linux-pci@xxxxxxxxxxxxxxx; Bjorn Helgaas <bhelgaas@xxxxxxxxxx>; Alex > Williamson <alex.williamson@xxxxxxxxxx>; Koenig, Christian > <Christian.Koenig@xxxxxxx> > Subject: Re: [PATCH] PCI: Avoid FLR for AMD Starship USB 3.0 > > On Mon, 13 Jul 2020 at 23:48, Deucher, Alexander > <Alexander.Deucher@xxxxxxx> wrote: > > > > [AMD Public Use] > > > > > -----Original Message----- > > > From: Bjorn Helgaas <helgaas@xxxxxxxxxx> > > > Sent: Monday, July 13, 2020 6:15 PM > > > To: Marcos Scriven <marcos@xxxxxxxxxxx> > > > Cc: Shah, Nehal-bakulchandra <Nehal-bakulchandra.Shah@xxxxxxx>; > > > Deucher, Alexander <Alexander.Deucher@xxxxxxx>; Kevin Buettner > > > <kevinb@xxxxxxxxxx>; linux-pci@xxxxxxxxxxxxxxx; Bjorn Helgaas > > > <bhelgaas@xxxxxxxxxx>; Alex Williamson > <alex.williamson@xxxxxxxxxx>; > > > Koenig, Christian <Christian.Koenig@xxxxxxx> > > > Subject: Re: [PATCH] PCI: Avoid FLR for AMD Starship USB 3.0 > > > > > > On Mon, Jul 13, 2020 at 01:44:44PM +0100, Marcos Scriven wrote: > > > > On Thu, 25 Jun 2020 at 11:22, Marcos Scriven <marcos@xxxxxxxxxxx> > wrote: > > > > > On Tue, 9 Jun 2020 at 12:47, Shah, Nehal-bakulchandra > > > > > <nehal-bakulchandra.shah@xxxxxxx> wrote: > > > > > > On 6/8/2020 11:17 PM, Marcos Scriven wrote: > > > > > > > On Thu, 28 May 2020 at 09:12, Marcos Scriven > > > > > > > <marcos@xxxxxxxxxxx> > > > > > wrote: > > > > > > >> On Wed, 27 May 2020 at 22:42, Deucher, Alexander > > > > > > >> <Alexander.Deucher@xxxxxxx> wrote: > > > > > > >>>> -----Original Message----- > > > > > > >>>> From: Bjorn Helgaas <helgaas@xxxxxxxxxx> > > > > > > >>>> > > > > > > >>>> [+cc Alex D, Christian -- do you guys have any contacts > > > > > > >>>> or insight > > > > > into why we > > > > > > >>>> suddenly have three new AMD devices that advertise FLR > > > > > > >>>> support but > > > > > it > > > > > > >>>> doesn't work? Are we doing something wrong in Linux, or > > > > > > >>>> are these > > > > > devices > > > > > > >>>> defective? > > > > > > >>> +Nehal who handles our USB drivers. > > > > > > >>> > > > > > > >>> Nehal any ideas about FLR or whether it should be advertised? > > > > > > >>> > > > > > > Sorry for the delay. We are looking into this with BIOS team. > > > > > > I shall > > > > > revert soon on this. > > > > > > > > > > Sorry to keep pestering about this, but wondering if there's any > > > > > movement on this? > > > > > > > > > > Is it something that's likely to be fixed and actually rolled > > > > > out by motherboard manufacturers? > > > > > > > > > > There's been some grumblings in the community about adding > > > > > workarounds rather than fixing, so it would be good to pass on > > > expectations here. > > > > > > > > Any word on this please? Would be keen to know if the BIOS can be > > > > fixed, and this workaround can eventually be dropped. > > > > > > Just to clarify what the possible outcomes are: > > > > > > 1) If these AMD devices are defective, but future ones are fixed, we > > > keep the quirk. > > > > > > 2) If these AMD devices are defective *and* future ones are also > > > defective, we keep the quirk and keep adding device IDs to it. > > > > > > 3) If the BIOS is defective, we keep the quirk. If anybody cares > > > about FLR enough, they can make the quirk smart enough to identify > > > fixed BIOS versions and enable FLR. > > > > > > 4) If Linux is defective, we can fix Linux and drop the quirk. > > > > > > The ideal outcome would be 4), but we don't have any indication that > > > Linux is doing something wrong. > > > > > > What we're really trying to avoid is 2) because that means new > > > devices will break Linux until somebody figures out the problem > > > again, updates the quirk, and gets the update into distro kernels. > > > > > > In case 3), we don't drop the quirk because that forces people to > > > upgrade their BIOS, and most people will not. We can't drop the > > > quirk, reintroduce the problem on old BIOSes, and hide behind the > > > excuse of "you need to upgrade the BIOS." That wastes the user's time > and our time. > > > > > > > Understood. Just trying to find the right people internally to understand > what has been validated and productized with respect to FLR on various > peripherals. > > > > Alex > > > > Hi Alex > > Sorry to keep bugging - wondering if you'd had any success finding the > people internally to look at this? Sorry, I have not. > > My main personal concern is that I faced some criticism from users in > submitting the quirk, as people felt that took the pressure off AMD to fix. There is hardware out there that apparently needs the quirk so even if it were a bios issue or something like that, that doesn't help the hardware that is already out in the wild. If there is ultimately some programming fix, we can always revert the patch once that is available. If FLR is actually broken, then there is nothing to fix, the quirk is correct. Alex > > Thanks > > Marcos > > > > > > > >>>>> diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c > > > > > > >>>>> index 43a0c2ce635e..b1db58d00d2b 100644 > > > > > > >>>>> --- a/drivers/pci/quirks.c > > > > > > >>>>> +++ b/drivers/pci/quirks.c > > > > > > >>>>> @@ -5133,6 +5133,7 @@ > > > > > > >>>> DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_INTEL, 0x443, > > > > > > >>>> quirk_intel_qat_vf_cap); > > > > > > >>>>> * FLR may cause the following to devices to hang: > > > > > > >>>>> * > > > > > > >>>>> * AMD Starship/Matisse HD Audio Controller 0x1487 > > > > > > >>>>> + * AMD Starship USB 3.0 Host Controller 0x148c > > > > > > >>>>> * AMD Matisse USB 3.0 Host Controller 0x149c > > > > > > >>>>> * Intel 82579LM Gigabit Ethernet Controller 0x1502 > > > > > > >>>>> * Intel 82579V Gigabit Ethernet Controller 0x1503 @@ > > > > > > >>>>> -5143,6 > > > > > +5144,7 > > > > > > >>>>> @@ static void quirk_no_flr(struct pci_dev *dev) > > > > > > >>>>> dev->dev_flags |= PCI_DEV_FLAGS_NO_FLR_RESET; } > > > > > > >>>>> DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_AMD, 0x1487, > > > > > > >>>>> quirk_no_flr); > > > > > > >>>>> +DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_AMD, > 0x148c, > > > > > > >>>> quirk_no_flr); > > > > > > >>>>> DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_AMD, > 0x149c, > > > > > > >>>> quirk_no_flr); > > > > > > >>>>> DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_INTEL, > 0x1502, > > > > > > >>>> quirk_no_flr); > > > > > > >>>>> DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_INTEL, > 0x1503, > > > > > > >>>> quirk_no_flr); > > > > > > > > > > > > Regard > > > > > > > > > > > > Nehal Shah > > > > > > > > > > >
