On Fri, Mar 07, 2025 at 06:14:26PM +0800, Bo Sun wrote:
[...]
So, I propose the following solution as a workaround to handle these edge cases.
diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c index 82b21e34c545..af8efebc7e7d 100644 --- a/drivers/pci/quirks.c +++ b/drivers/pci/quirks.c @@ -6181,6 +6181,13 @@ DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_INTEL, 0x1536, rom_bar_overlap_defect); DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_INTEL, 0x1537, rom_bar_overlap_defect); DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_INTEL, 0x1538, rom_bar_overlap_defect);
+static void quirk_marvell_cn96xx_cn10xxx_reassign_all_busnr(struct pci_dev *dev) +{
if (!pci_has_flag(PCI_PROBE_ONLY))pci_add_flags(PCI_REASSIGN_ALL_BUS);+} +DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_CAVIUM, 0xa002, quirk_marvell_cn96xx_cn10xxx_reassign_all_busnr);
LGTM. Please add a comment about this quirk too.
OK, I'll add the comment. Should I send the v2 patch?
Yes, ofc. You should send v2 with the quirk change.
- Mani