diff mbox series

PCI/DPC: Yet another quirk for PIO log size on Intel Raptor Lake-P

Message ID 20250102164315.7562-1-tiwai@suse.de (mailing list archive)
State New
Delegated to: Bjorn Helgaas
Headers show
Series PCI/DPC: Yet another quirk for PIO log size on Intel Raptor Lake-P | expand

Commit Message

Takashi Iwai Jan. 2, 2025, 4:43 p.m. UTC
There is yet another PCI entry for Intel Raptor Lake-P that shows the
  error "DPC: RP PIO log size 0 is invalid":
  0000:00:07.0 PCI bridge [0604]: Intel Corporation Raptor Lake-P Thunderbolt 4 PCI Express Root Port #0 [8086:a76e]
  0000:00:07.2 PCI bridge [0604]: Intel Corporation Raptor Lake-P Thunderbolt 4 PCI Express Root Port #2 [8086:a72f]

Add the corresponding quirk entry for 8086:a72f.

Note that the one for 8086:a76e has been already added by the commit
627c6db20703 ("PCI/DPC: Quirk PIO log size for Intel Raptor Lake Root
Ports").

Link: https://bugzilla.suse.com/show_bug.cgi?id=1234623
Signed-off-by: Takashi Iwai <tiwai@suse.de>
---
 drivers/pci/quirks.c | 1 +
 1 file changed, 1 insertion(+)

Comments

Bjorn Helgaas Jan. 3, 2025, 10:53 p.m. UTC | #1
On Thu, Jan 02, 2025 at 05:43:13PM +0100, Takashi Iwai wrote:
> There is yet another PCI entry for Intel Raptor Lake-P that shows the
>   error "DPC: RP PIO log size 0 is invalid":
>   0000:00:07.0 PCI bridge [0604]: Intel Corporation Raptor Lake-P Thunderbolt 4 PCI Express Root Port #0 [8086:a76e]
>   0000:00:07.2 PCI bridge [0604]: Intel Corporation Raptor Lake-P Thunderbolt 4 PCI Express Root Port #2 [8086:a72f]
> 
> Add the corresponding quirk entry for 8086:a72f.
> 
> Note that the one for 8086:a76e has been already added by the commit
> 627c6db20703 ("PCI/DPC: Quirk PIO log size for Intel Raptor Lake Root
> Ports").

Intel folks, what's the long-term resolution of this?  I'm kind of
tired of adding quirks like this.  So far we have these (not including
the current patch), dating back to Aug 2022:

  627c6db20703 ("PCI/DPC: Quirk PIO log size for Intel Raptor Lake Root Ports")
  3b8803494a06 ("PCI/DPC: Quirk PIO log size for Intel Ice Lake Root Ports")
  5459c0b70467 ("PCI/DPC: Quirk PIO log size for certain Intel Root Ports")

I *thought* this problem was caused by BIOS defects that were supposed
to be fixed, but nothing seems to be happening.

If Linux can't figure out the correct value by itself (and I assume it
can't because we wouldn't need the register if we could), maybe we
can just silently ignore it and live without the RP logs?

> Link: https://bugzilla.suse.com/show_bug.cgi?id=1234623
> Signed-off-by: Takashi Iwai <tiwai@suse.de>
> ---
>  drivers/pci/quirks.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
> index 76f4df75b08a..4ed3704ce92e 100644
> --- a/drivers/pci/quirks.c
> +++ b/drivers/pci/quirks.c
> @@ -6253,6 +6253,7 @@ DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x9a2b, dpc_log_size);
>  DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x9a2d, dpc_log_size);
>  DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x9a2f, dpc_log_size);
>  DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x9a31, dpc_log_size);
> +DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0xa72f, dpc_log_size);
>  DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0xa73f, dpc_log_size);
>  DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0xa76e, dpc_log_size);
>  #endif
> -- 
> 2.43.0
>
Mika Westerberg Jan. 5, 2025, 8:25 a.m. UTC | #2
Hi,

On Fri, Jan 03, 2025 at 04:53:15PM -0600, Bjorn Helgaas wrote:
> On Thu, Jan 02, 2025 at 05:43:13PM +0100, Takashi Iwai wrote:
> > There is yet another PCI entry for Intel Raptor Lake-P that shows the
> >   error "DPC: RP PIO log size 0 is invalid":
> >   0000:00:07.0 PCI bridge [0604]: Intel Corporation Raptor Lake-P Thunderbolt 4 PCI Express Root Port #0 [8086:a76e]
> >   0000:00:07.2 PCI bridge [0604]: Intel Corporation Raptor Lake-P Thunderbolt 4 PCI Express Root Port #2 [8086:a72f]
> > 
> > Add the corresponding quirk entry for 8086:a72f.
> > 
> > Note that the one for 8086:a76e has been already added by the commit
> > 627c6db20703 ("PCI/DPC: Quirk PIO log size for Intel Raptor Lake Root
> > Ports").
> 
> Intel folks, what's the long-term resolution of this?  I'm kind of
> tired of adding quirks like this.  So far we have these (not including
> the current patch), dating back to Aug 2022:
> 
>   627c6db20703 ("PCI/DPC: Quirk PIO log size for Intel Raptor Lake Root Ports")
>   3b8803494a06 ("PCI/DPC: Quirk PIO log size for Intel Ice Lake Root Ports")
>   5459c0b70467 ("PCI/DPC: Quirk PIO log size for certain Intel Root Ports")
> 
> I *thought* this problem was caused by BIOS defects that were supposed
> to be fixed, but nothing seems to be happening.

As far as I know it should be fixed already. I just checked my MTLP and PTL
systems (both with integrated TBT PCIe root ports) and I don't see the
message anymore. I don't have RPL system though. This is on reference
hardware and BIOS so it is possible that the fix has not been taken into
the OEM BIOS.
diff mbox series

Patch

diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
index 76f4df75b08a..4ed3704ce92e 100644
--- a/drivers/pci/quirks.c
+++ b/drivers/pci/quirks.c
@@ -6253,6 +6253,7 @@  DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x9a2b, dpc_log_size);
 DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x9a2d, dpc_log_size);
 DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x9a2f, dpc_log_size);
 DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x9a31, dpc_log_size);
+DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0xa72f, dpc_log_size);
 DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0xa73f, dpc_log_size);
 DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0xa76e, dpc_log_size);
 #endif