This regression is still present in 4.15-rc9. It's been ongoing since 3.16. Could someone take a look so it can finally be fixed? I've been trying to adapt the three patches I originally identified from Yinghailu's for-pci-v4.1-rc8 branch onto the current mainline, and linux-next, but he did a forced rebase and the branch commits changed with some being removed. https://bugzilla.kernel.org/show_bug.cgi?id=81431 On 17/06/15 19:11, TJ wrote: > I've finally had the opportunity to bisect this regession down to: > > PCI: Restrict 64-bit prefetchable bridge windows to 64-bit resources > > commit 5b28541552ef5eeffc41d6936105f38c2508e566 > Author: Yinghai Lu <yinghai@xxxxxxxxxx> > Date: Mon May 19 17:01:55 2014 -0600 > > The patch is quite involved so could someone with familiarity with the > allocation strategies take a look please? > > https://bugzilla.kernel.org/show_bug.cgi?id=81431 > -- > 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