Marvell's own product brief implies the 92xx series are a closely related family, and sure enough it turns out that 9235 seems to need the same quirk as the other three, although possibly only when certain ports are used. CC: stable@xxxxxxxxxxxxxxx Reported-by: Jason Adriaanse <jason_a69@xxxxxxxxxxx> Link: https://lore.kernel.org/linux-iommu/2a699a99-545c-1324-e052-7d2f41fed1ae@xxxxxxxxxxx/ Signed-off-by: Robin Murphy <robin.murphy@xxxxxxx> --- Note that the actual regression which started the thread is a different matter, wherein a particular combination of parameters which used to put intel-iommu into passthrough mode now enables full translation instead. Take #2, hopefully not royally screwing up my email alises this time. Sorry about that... drivers/pci/quirks.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c index f4e2a88729fd..3186f2c84eab 100644 --- a/drivers/pci/quirks.c +++ b/drivers/pci/quirks.c @@ -4174,6 +4174,8 @@ DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_MARVELL_EXT, 0x9220, /* https://bugzilla.kernel.org/show_bug.cgi?id=42679#c49 */ DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_MARVELL_EXT, 0x9230, quirk_dma_func1_alias); +DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_MARVELL_EXT, 0x9235, + quirk_dma_func1_alias); DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_TTI, 0x0642, quirk_dma_func1_alias); DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_TTI, 0x0645, -- 2.39.2.101.g768bb238c484.dirty
