On Mon, Sep 6, 2021 at 7:49 PM Niklas Schnelle <schnelle@xxxxxxxxxxxxx> wrote: > > Patch 3 I already sent separately resulting in the discussion below but without > a final conclusion. > > https://lore.kernel.org/lkml/20210720150145.640727-1-schnelle@xxxxxxxxxxxxx/ > > I believe even though there were some doubts about the use of > pci_dev_is_added() by arch code the existing uses as well as the use in the > final patch of this series warrant this export. The use of pci_dev_is_added() in arch/powerpc was because in the past pci_bus_add_device() could be called before pci_device_add(). That was fixed a while ago so It should be safe to remove those calls now. > Patch 4 "PCI: Export pci_dev_lock()" is basically an extension to commit > e3a9b1212b9d ("PCI: Export pci_dev_trylock() and pci_dev_unlock()") which > already exported pci_dev_trylock(). In the final patch we make use of > pci_dev_lock() to wait for any other exclusive uses of the pdev to be finished > before starting recovery. Hmm, I noticed the EEH (arch/powerpc/kernel/eeh_driver.c:eeh_pe_report_edev()) and the generic PCIe error recovery code (see drivers/pci/pcie/err.c:report_error_detected()) only call device_lock() before entering the driver's error handling callbacks. I wonder if they should be using pci_dev_lock() instead. The only real difference is that pci_dev_lock() will also block user space from accessing the device's config space while error recovery is in progress which seems sensible enough.
