On Wed, Jul 01, 2020 at 11:18:29AM +0900, Kunihiko Hayashi wrote: [...] > > > > > -static void uniphier_pcie_irq_handler(struct irq_desc *desc) > > > > > +static void uniphier_pcie_misc_isr(struct pcie_port *pp, bool is_msi) > > > > > { > > > > > - struct pcie_port *pp = irq_desc_get_handler_data(desc); > > > > > struct dw_pcie *pci = to_dw_pcie_from_pp(pp); > > > > > struct uniphier_pcie_priv *priv = to_uniphier_pcie(pci); > > > > > - struct irq_chip *chip = irq_desc_get_chip(desc); > > > > > - unsigned long reg; > > > > > - u32 val, bit, virq; > > > > > + u32 val, virq; > > > > > - /* INT for debug */ > > > > > val = readl(priv->base + PCL_RCV_INT); > > > > > if (val & PCL_CFG_BW_MGT_STATUS) > > > > > dev_dbg(pci->dev, "Link Bandwidth Management Event\n"); > > > > > + > > > > > if (val & PCL_CFG_LINK_AUTO_BW_STATUS) > > > > > dev_dbg(pci->dev, "Link Autonomous Bandwidth Event\n"); > > > > > - if (val & PCL_CFG_AER_RC_ERR_MSI_STATUS) > > > > > - dev_dbg(pci->dev, "Root Error\n"); > > > > > - if (val & PCL_CFG_PME_MSI_STATUS) > > > > > - dev_dbg(pci->dev, "PME Interrupt\n"); > > > > > + > > > > > + if (is_msi) { > > > > > + if (val & PCL_CFG_AER_RC_ERR_MSI_STATUS) > > > > > + dev_dbg(pci->dev, "Root Error Status\n"); > > > > > + > > > > > + if (val & PCL_CFG_PME_MSI_STATUS) > > > > > + dev_dbg(pci->dev, "PME Interrupt\n"); > > > > > + > > > > > + if (val & (PCL_CFG_AER_RC_ERR_MSI_STATUS | > > > > > + PCL_CFG_PME_MSI_STATUS)) { > > > > > + virq = irq_linear_revmap(pp->irq_domain, 0); > > > > > + generic_handle_irq(virq); > > > > > + } > > > > > + } > > > > > > > > Please have two handlers: one for interrupts that are from the RC, > > > > another for interrupts coming from the endpoints. > > > I assume that this handler treats interrupts from the RC only and > > > this is set on the member ".msi_host_isr" added in the patch 1/6. > > > I think that the handler for interrupts coming from endpoints should be > > > treated as a normal case (after calling .msi_host_isr in > > > dw_handle_msi_irq()). > > > > It looks pretty odd that you end-up dealing with both from the > > same "parent" interrupt. I guess this is in keeping with the > > rest of the DW PCIe hacks... :-/ > > It might be odd, however, in case of UniPhier SoC, > both MSI interrupts from endpoints and PME/AER interrupts from RC are > asserted by same "parent" interrupt. In other words, PME/AER interrupts > are notified using the parent interrupt for MSI. > > MSI interrupts are treated as child interrupts with reference to > the status register in DW core. This is handled in a for-loop in > dw_handle_msi_irq(). > > PME/AER interrupts are treated with reference to the status register > in UniPhier glue layer, however, this couldn't be handled in the same way > directly. > > So I'm trying to add .msi_host_isr function to handle this > with reference to the SoC-specific registers. > > This exported function asserts MSI-0 as a shared child interrupt. > As a result, PME/AER are registered like the followings in dmesg: > > pcieport 0000:00:00.0: PME: Signaling with IRQ 25 > pcieport 0000:00:00.0: AER: enabled with IRQ 25 > > And these interrupts are shared as MSI-0: > > # cat /proc/interrupts | grep 25: > 25: 0 0 0 0 PCI-MSI 0 Edge PCIe PME, aerdrv > > This might be a special case, though, I think that this is needed to handle > interrupts from RC sharing MSI parent. Can you please send me (with this series *applied* of course and if possible with an endpoint MSI/MSI-X capable enabled): - full dmesg log - lspci -vv output - cat /proc/interrupts I need to understand how this system HW works before commenting any further. > > It is for Lorenzo to make up his mind about this anyway. > > I'd like to Lorenzo's opinion, too. I am trying to understand how the HW is wired up (and that's what Marc asked as well) so first things first, please send the logs. Lorenzo