From: Yury Murashka <[email protected]>

pcie_do_recovery() only clears the error status bits of an error-reporting
device if recovery succeeds, but not if it fails.  (E.g. because reset
failed, drivers are missing pci_error_handlers or those handlers failed.)

One undesirable consequence is that the AER driver may subsequently
identify a wrong device when searching for error-reporting devices in
is_error_source(), due to stale bits on an unaffected device.

Avoid by clearing error status bits on recovery failure.

If recovery succeeds, Fatal Errors are generally cleared by a call to
pci_restore_state() from drivers' ->slot_reset() callbacks.  If recovery
fails, those callbacks may not have been invoked or they may not be
implemented.  Hence clear Fatal Errors as well on recovery failure.

Link: 
https://lore.kernel.org/r/CAPzpGcRCTCZtaX1EVaJNZ103THZKsoszZduY7=gwfydcrmo...@mail.gmail.com/
Signed-off-by: Yury Murashka <[email protected]>
[lukas: drop cmdline param, rewrite commit msg, clear fatal errors]
Signed-off-by: Lukas Wunner <[email protected]>
---
Changes v1 -> v2:
  * Clear Fatal Errors as well on recovery failure (Sathya, sashiko).
 
 drivers/pci/pcie/err.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/drivers/pci/pcie/err.c b/drivers/pci/pcie/err.c
index 24a8a8263c2e..a450146fb285 100644
--- a/drivers/pci/pcie/err.c
+++ b/drivers/pci/pcie/err.c
@@ -289,6 +289,13 @@ pci_ers_result_t pcie_do_recovery(struct pci_dev *dev,
        return status;
 
 failed:
+       if (host->native_aer) {
+               pcie_capability_write_word(dev, PCI_EXP_DEVSTA,
+                                          PCI_EXP_DEVSTA_UNC_MASK);
+               pci_aer_clear_nonfatal_status(dev);
+               pci_aer_clear_fatal_status(dev);
+       }
+
        pci_walk_bridge(bridge, pci_pm_runtime_put, NULL);
 
        pci_walk_bridge(bridge, report_perm_failure_detected, NULL);
-- 
2.53.0


Reply via email to