On Sun, Feb 20, 2022 at 08:33:34PM +0100, Marek Behún wrote: > --- a/drivers/pci/hotplug/pciehp_hpc.c > +++ b/drivers/pci/hotplug/pciehp_hpc.c > @@ -788,6 +788,7 @@ static int pciehp_poll(void *data) > @@ -800,12 +801,17 @@ static void pcie_enable_notification(struct controller *ctrl) > * next power fault detected interrupt was notified again. > */ > > + pcie_capability_read_dword(ctrl_dev(ctrl), PCI_EXP_LNKCAP, &link_cap); > /* > - * Always enable link events: thus link-up and link-down shall > - * always be treated as hotplug and unplug respectively. Enable > - * presence detect only if Attention Button is not present. > + * Enable link events if their support is indicated in Link Capability > + * register: thus link-up and link-down shall always be treated as > + * hotplug and unplug respectively. Enable presence detect only if > + * Attention Button is not present. > */ > - cmd = PCI_EXP_SLTCTL_DLLSCE; > + cmd = 0; > + if (link_cap & PCI_EXP_LNKCAP_DLLLARC) > + cmd |= PCI_EXP_SLTCTL_DLLSCE; The Data Link Layer Link Active Reporting Capable bit is cached in ctrl_dev(ctrl)->link_active_reporting. Please use that instead of re-reading it from the register. According to PCIe r6.0, sec. 7.5.3.6, "For a hot-plug capable Downstream Port [...], this bit must be hardwired to 1b." That has been part of the spec since PCIe r1.1, sec. 7.8.6. PCIe r1.0 did not contain the sentence because it did not support DLLLARC (it defined bit 20 as RsvdP). In other words, what you're doing here is add support for PCIe r1.0. I'm not opposed to that, but I'd assume that aardvark supports a more recent spec version. More likely it doesn't comply with the spec? What is the user-visible issue that you're experiencing without this commit? Does aardvark somehow misbehave if the DLLLARC bit is set to 1? Since the bit is RsvdP, setting it shouldn't have any negative side effects. > --- a/drivers/pci/hotplug/pnv_php.c > +++ b/drivers/pci/hotplug/pnv_php.c > @@ -840,6 +840,7 @@ static void pnv_php_init_irq(struct pnv_php_slot *php_slot, int irq) > struct pci_dev *pdev = php_slot->pdev; > u32 broken_pdc = 0; > u16 sts, ctrl; > + u32 link_cap; > int ret; > > /* Allocate workqueue */ pnv_php.c is a driver for PowerNV, yet this patch is for a series targeting an ARM PCIe controller. That doesn't make sense, changes to pnv_php.c seem wrong here. Thanks, Lukas
