On Thu, Nov 17, 2016 at 08:52:36AM +0200, Noa Osherovich wrote: > On 11/17/2016 12:41 AM, Bjorn Helgaas wrote: > > > On Tue, Nov 15, 2016 at 09:59:57AM +0200, Noa Osherovich wrote: > >> The following 3 patches unmark some Mellanox devices as having broken > >> INTx masking. > >> > >> The first patch moves all broken INTx masking to FINAL instead of > >> HEADER so that the fixup can execute on all archs. > >> The second fixup changes the Mellanox marking to be according to a > >> list of devices rather than a general fixup by vendor. > >> The third patch excludes some Mellanox devices from the INTx broken > >> marking according to device ID / firmware version. > >> > >> v0 -> v1 > >> Fixes according to Gavin Shan's comments: > >> - Patch numbering was wrong > >> - Fixed logic of mellanox_check_broken_intx_masking > >> > >> Noa Osherovich (3): > >> PCI: Use FINAL fixup to mark broken INTx masking > >> PCI: Convert Mellanox quirks to be for listed devices only > >> PCI: Add checks to mellanox_check_broken_intx_masking > >> > >> drivers/pci/quirks.c | 175 ++++++++++++++++++++++++++++++++++++++++----------- > >> 1 file changed, 139 insertions(+), 36 deletions(-) > > I applied these (with the tweaks Gavin & I talked about) to > > pci/virtualization for v4.10. Please take a look and make sure it still > > works for you. > > > > I also removed the device ID #defines in the interest of brevity and moved > > the Mellanox code out of the middle of the other INTx quirks. > > > > Bjorn > > Hi Bjorn, > > This may have broken the kbuild robot - it sent an error saying those defines were undeclared. > How do you want to proceed? I broke it; I'll fix it :) Bjorn -- To unsubscribe from this list: send the line "unsubscribe linux-pci" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
