Re: [RFC PATCH 6/9] PCI: Add 'rcec' field to pci_dev for associated RCiEPs

From: Sean V Kelley
Date: Mon Jul 27 2020 - 11:39:45 EST

On 27 Jul 2020, at 4:23, Jonathan Cameron wrote:

On Fri, 24 Jul 2020 10:22:20 -0700
Sean V Kelley <sean.v.kelley@xxxxxxxxx> wrote:

From: Qiuxu Zhuo <qiuxu.zhuo@xxxxxxxxx>

When attempting error recovery for an RCiEP associated with an RCEC device,
there needs to be a way to update the Root Error Status, the Uncorrectable
Error Status and the Uncorrectable Error Severity of the parent RCEC.
So add the 'rcec' field to the pci_dev structure and provide a hook for the
Root Port Driver to associate RCiEPs with their respective parent RCEC.

Co-developed-by: Sean V Kelley <sean.v.kelley@xxxxxxxxx>
Signed-off-by: Qiuxu Zhuo <qiuxu.zhuo@xxxxxxxxx>
Signed-off-by: Sean V Kelley <sean.v.kelley@xxxxxxxxx>

I haven't tested yet, but I think there is one path in here that breaks
my case (no OS visible rcec / all done in firmware GHESv2 / APEI)

Jonathan

Okay, keep me up to date and if there is a way I can reproduce or better test for your use case let me know and I will do so.


---
drivers/pci/pcie/aer.c | 9 +++++----
drivers/pci/pcie/err.c | 9 +++++++++
drivers/pci/pcie/portdrv_pci.c | 15 +++++++++++++++
include/linux/pci.h | 3 +++
4 files changed, 32 insertions(+), 4 deletions(-)

diff --git a/drivers/pci/pcie/aer.c b/drivers/pci/pcie/aer.c
index 3acf56683915..f1bf06be449e 100644
--- a/drivers/pci/pcie/aer.c
+++ b/drivers/pci/pcie/aer.c
@@ -1335,17 +1335,18 @@ static int aer_probe(struct pcie_device *dev)
static pci_ers_result_t aer_root_reset(struct pci_dev *dev)
{
int aer = dev->aer_cap;
+ int rc = 0;
u32 reg32;
- int rc;
-

/* Disable Root's interrupt in response to error messages */
pci_read_config_dword(dev, aer + PCI_ERR_ROOT_COMMAND, &reg32);
reg32 &= ~ROOT_PORT_INTR_ON_MESG_MASK;
pci_write_config_dword(dev, aer + PCI_ERR_ROOT_COMMAND, reg32);

- rc = pci_bus_error_reset(dev);
- pci_info(dev, "Root Port link has been reset\n");
+ if (pci_pcie_type(dev) != PCI_EXP_TYPE_RC_EC) {
+ rc = pci_bus_error_reset(dev);
+ pci_info(dev, "Root Port link has been reset\n");
+ }

/* Clear Root Error Status */
pci_read_config_dword(dev, aer + PCI_ERR_ROOT_STATUS, &reg32);
diff --git a/drivers/pci/pcie/err.c b/drivers/pci/pcie/err.c
index 9b3ec94bdf1d..0aae7643132e 100644
--- a/drivers/pci/pcie/err.c
+++ b/drivers/pci/pcie/err.c
@@ -203,6 +203,11 @@ pci_ers_result_t pcie_do_recovery(struct pci_dev *dev,
pci_walk_dev_affected(dev, report_frozen_detected, &status);
if (pci_pcie_type(dev) == PCI_EXP_TYPE_RC_END) {
status = flr_on_rciep(dev);
+ /*
+ * The callback only clears the Root Error Status
+ * of the RCEC (see aer.c).
+ */
+ reset_link(dev->rcec);

This looks dangerous for my case where APEI / GHESv2 is used. In that case
we don't expose an RCEC at all. I don't think the reset_link callback
is safe to a null pointer here. Fix may be as simple as
if (dev->rcec)
reset_link(dev->rcec);

Good catch. Will look to NULL case as well as walk through code where dev->rcec is further referenced (below and elsewhere) to ensure nothing falls through. Glad you are pointing these out.



if (status != PCI_ERS_RESULT_RECOVERED) {
pci_warn(dev, "function level reset failed\n");
goto failed;
@@ -246,7 +251,11 @@ pci_ers_result_t pcie_do_recovery(struct pci_dev *dev,
pci_pcie_type(dev) == PCI_EXP_TYPE_RC_EC)) {
pci_aer_clear_device_status(dev);
pci_aer_clear_nonfatal_status(dev);
+ } else if (pci_pcie_type(dev) == PCI_EXP_TYPE_RC_END) {
+ pci_aer_clear_device_status(dev->rcec);
+ pci_aer_clear_nonfatal_status(dev->rcec);

These may be safe as in my both now have protections for !pcie_aer_is_native.

Iâll have a look again for the NULL case

}
+
pci_info(dev, "device recovery successful\n");
return status;

diff --git a/drivers/pci/pcie/portdrv_pci.c b/drivers/pci/pcie/portdrv_pci.c
index d5b109499b10..f9409a0110c2 100644
--- a/drivers/pci/pcie/portdrv_pci.c
+++ b/drivers/pci/pcie/portdrv_pci.c
@@ -90,6 +90,18 @@ static const struct dev_pm_ops pcie_portdrv_pm_ops = {
#define PCIE_PORTDRV_PM_OPS NULL
#endif /* !PM */

+static int pcie_hook_rcec(struct pci_dev *pdev, void *data)
+{
+ struct pci_dev *rcec = (struct pci_dev *)data;
+
+ pdev->rcec = rcec;
+ pci_info(rcec, "RCiEP(under an RCEC) %04x:%02x:%02x.%d\n",
+ pci_domain_nr(pdev->bus), pdev->bus->number,
+ PCI_SLOT(pdev->devfn), PCI_FUNC(pdev->devfn));

We may want to make this debug info at somepoint if we have a way
of discovering it from userspace. The PCI boot up is extremely
verbose already!

Iâve submitted patches to pciutils, which expose this very detail:

https://lore.kernel.org/linux-pci/20200624223940.240463-1-sean.v.kelley@xxxxxxxxxxxxxxx/

Flexible either way, could remove or make it debug info.

Thanks,

Sean


+
+ return 0;
+}
+
/*
* pcie_portdrv_probe - Probe PCI-Express port devices
* @dev: PCI-Express port device being probed
@@ -110,6 +122,9 @@ static int pcie_portdrv_probe(struct pci_dev *dev,
(pci_pcie_type(dev) != PCI_EXP_TYPE_RC_EC)))
return -ENODEV;

+ if (pci_pcie_type(dev) == PCI_EXP_TYPE_RC_EC)
+ pcie_walk_rcec(dev, pcie_hook_rcec, dev);
+
status = pcie_port_device_register(dev);
if (status)
return status;
diff --git a/include/linux/pci.h b/include/linux/pci.h
index 34c1c4f45288..e920f29df40b 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -326,6 +326,9 @@ struct pci_dev {
#ifdef CONFIG_PCIEAER
u16 aer_cap; /* AER capability offset */
struct aer_stats *aer_stats; /* AER stats for this device */
+#endif
+#ifdef CONFIG_PCIEPORTBUS
+ struct pci_dev *rcec; /* Associated RCEC device */
#endif
u8 pcie_cap; /* PCIe capability offset */
u8 msi_cap; /* MSI capability offset */