mbox series

[V4,00/10] Add RCEC handling to PCI/AER

Message ID 20200917162548.2079894-1-sean.v.kelley@intel.com (mailing list archive)
Headers show
Series Add RCEC handling to PCI/AER | expand

Message

Kelley, Sean V Sept. 17, 2020, 4:25 p.m. UTC
Changes since v3 [1]:

- Merge usage patch and API (pcie_walk_rcec).
(Sathyanarayanan Kuppuswamy)

- Considering the possible ways to call pcie_do_recovery(), make
the flow more understandable through assignments of the incoming
devices in terms of 'bridges'.
- reset_link() may be misnamed. The point is really to reset any devices
below 'dev', which is a bridge. Using reset_subordinate_devices() makes it
more clear that we pass a bridge and reset the devices *below* it.
- In pcie_walk_rcec() an RCEC bus number does not indicate association
in the Assocated Bus Numbers register so skip it (7.9.10.3)
- Potential for a lot of churning to call pci_get_slot() in
pcie_walk_rciep_Devfn(). Attempt to reduce the number of calls by identifying
the associations through bus walks.
- Change pr_dbg of link RCiEP to "PME & error events reported via.."
(Bjorn Helgaas)

- In some cases (AER via APEI) there may not exist an RCEC. In these cases
the bridge is NULL. Account for that in the bridge walk.
(Jonathan Cameron)


[1] https://lore.kernel.org/linux-pci/20200812164659.1118946-1-sean.v.kelley@intel.com/


Jonathan Cameron (1):
  PCI/AER: Extend AER error handling to RCECs

Qiuxu Zhuo (5):
  PCI/RCEC: Add RCEC class code and extended capability
  PCI/RCEC: Bind RCEC devices to the Root Port driver
  PCI/AER: Apply function level reset to RCiEP on fatal error
  PCI/RCEC: Add RCiEP's linked RCEC to AER/ERR
  PCI/AER: Add RCEC AER error injection support

Sean V Kelley (4):
  PCI/RCEC: Cache RCEC capabilities in pci_init_capabilities()
  PCI/RCEC: Add pcie_link_rcec() to associate RCiEPs
  PCI/AER: Add pcie_walk_rcec() to RCEC AER handling
  PCI/PME: Add pcie_walk_rcec() to RCEC PME handling

 drivers/pci/pci.h               |  26 ++++-
 drivers/pci/pcie/Makefile       |   2 +-
 drivers/pci/pcie/aer.c          |  36 +++++--
 drivers/pci/pcie/aer_inject.c   |   5 +-
 drivers/pci/pcie/err.c          | 102 ++++++++++++++----
 drivers/pci/pcie/pme.c          |  15 ++-
 drivers/pci/pcie/portdrv_core.c |   8 +-
 drivers/pci/pcie/portdrv_pci.c  |   8 +-
 drivers/pci/pcie/rcec.c         | 185 ++++++++++++++++++++++++++++++++
 drivers/pci/probe.c             |   3 +-
 include/linux/pci.h             |   5 +
 include/linux/pci_ids.h         |   1 +
 include/uapi/linux/pci_regs.h   |   7 ++
 13 files changed, 359 insertions(+), 44 deletions(-)
 create mode 100644 drivers/pci/pcie/rcec.c

--
2.28.0

Comments

Bjorn Helgaas Sept. 17, 2020, 5:36 p.m. UTC | #1
On Thu, Sep 17, 2020 at 09:25:38AM -0700, Sean V Kelley wrote:
> Changes since v3 [1]:

This series claims "V4 00/10", i.e., there should be this cover letter
plus 10 patches, but I only got 3 patches.  I don't know if some got
lost, or if only those 3 patches were updated, or what?  If it's the
latter, it's too hard for me to collect the right versions of
everything into a single series.

Either way, can you resend the entire series as a V5?

Bjorn
Kelley, Sean V Sept. 17, 2020, 5:55 p.m. UTC | #2
On 17 Sep 2020, at 10:36, Bjorn Helgaas wrote:

> On Thu, Sep 17, 2020 at 09:25:38AM -0700, Sean V Kelley wrote:
>> Changes since v3 [1]:
>
> This series claims "V4 00/10", i.e., there should be this cover letter
> plus 10 patches, but I only got 3 patches.  I don't know if some got
> lost, or if only those 3 patches were updated, or what?  If it's the
> latter, it's too hard for me to collect the right versions of
> everything into a single series.
>
> Either way, can you resend the entire series as a V5?
>
> Bjorn

That's weird.  I can see all 10 got sent. There's something awry with 
the mailer as I got the copies.  You are right.  Lore only shows 3.  I 
will see if something happened with the smtp access.  Will resend as V5.

Thanks,

Sean
Jonathan Cameron Sept. 18, 2020, 8:44 a.m. UTC | #3
On Thu, 17 Sep 2020 10:55:16 -0700
Sean V Kelley <sean.v.kelley@intel.com> wrote:

> On 17 Sep 2020, at 10:36, Bjorn Helgaas wrote:
> 
> > On Thu, Sep 17, 2020 at 09:25:38AM -0700, Sean V Kelley wrote:  
> >> Changes since v3 [1]:  
> >
> > This series claims "V4 00/10", i.e., there should be this cover letter
> > plus 10 patches, but I only got 3 patches.  I don't know if some got
> > lost, or if only those 3 patches were updated, or what?  If it's the
> > latter, it's too hard for me to collect the right versions of
> > everything into a single series.
> >
> > Either way, can you resend the entire series as a V5?
> >
> > Bjorn  
> 
> That's weird.  I can see all 10 got sent. There's something awry with 
> the mailer as I got the copies.  You are right.  Lore only shows 3.  I 
> will see if something happened with the smtp access.  Will resend as V5.

I got all 10.  

May be coincidence but I saw this with a couple of series I was trying
to pick up with b4 (so from lore.kernel.org). I assumed it was a problem
at the sender, but that wasn't intel..

Perhaps something more general going on at vger?

Thanks,

Jonathan

> 
> Thanks,
> 
> Sean