| Message ID | 1309477662-18680-4-git-send-email-linuxram@us.ibm.com (mailing list archive) |
|---|---|
| State | New, archived |
| Delegated to: | Bjorn Helgaas |
| Headers | show |
On 01.07.2011 01:47, Ram Pai wrote: > From: Yinghai Lu <yinghai@kernel.org> > > From: Yinghai Lu <yinghai@kernel.org> > > Allocate resources to SRIOV BARs only after all other required > resource-requests are satisfied. Dont retry if resource allocation for SRIOV > BARs fail. > > Signed-off-by: Ram Pai <linuxram@us.ibm.com> > Signed-off-by: Yinghai Lu <yinghai@kernel.org> Tested-by: Oliver Hartkopp <socketcan@hartkopp.net> I applied the whole patchset and it looks very similar to the dmsg output before the problematic commit that caused the regression. Tnx! Oliver > --- > drivers/pci/setup-bus.c | 10 ++++++++++ > 1 files changed, 10 insertions(+), 0 deletions(-) > > diff --git a/drivers/pci/setup-bus.c b/drivers/pci/setup-bus.c > index 9f2f15f..a0555cb 100644 > --- a/drivers/pci/setup-bus.c > +++ b/drivers/pci/setup-bus.c > @@ -669,6 +669,16 @@ static int pbus_size_mem(struct pci_bus *bus, unsigned long mask, > if (r->parent || (r->flags & mask) != type) > continue; > r_size = resource_size(r); > +#ifdef CONFIG_PCI_IOV > + /* put SRIOV requested res to the optional list */ > + if (add_head && i >= PCI_IOV_RESOURCES && > + i <= PCI_IOV_RESOURCE_END) { > + r->end = r->start - 1; > + add_to_list(add_head, dev, r, r_size, 1); > + children_add_size += r_size; > + continue; > + } > +#endif > /* For bridges size != alignment */ > align = pci_resource_alignment(dev, r); > order = __ffs(align) - 20; -- To unsubscribe from this list: send the line "unsubscribe linux-pci" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Fri, 01 Jul 2011 08:01:33 +0200 Oliver Hartkopp <socketcan@hartkopp.net> wrote: > On 01.07.2011 01:47, Ram Pai wrote: > > From: Yinghai Lu <yinghai@kernel.org> > > > > From: Yinghai Lu <yinghai@kernel.org> > > > > Allocate resources to SRIOV BARs only after all other required > > resource-requests are satisfied. Dont retry if resource allocation for SRIOV > > BARs fail. > > > > Signed-off-by: Ram Pai <linuxram@us.ibm.com> > > Signed-off-by: Yinghai Lu <yinghai@kernel.org> > > Tested-by: Oliver Hartkopp <socketcan@hartkopp.net> > > I applied the whole patchset and it looks very similar to the dmsg output > before the problematic commit that caused the regression. Ok Oliver, can I ask you to test one more tree? We'll shoot for including Ram's patches in -next, but I've just included his other patch to disable the new realloc code by default for the current kernel. Can you give my for-linus tree a try and make sure it gives you back the old behavior? Thanks a lot for your diligent testing so far. git://git.kernel.org/pub/scm/linux/kernel/git/jbarnes/pci-2.6 is the tree, then check out the "for-linus" branch. Thanks,
diff --git a/drivers/pci/setup-bus.c b/drivers/pci/setup-bus.c index 9f2f15f..a0555cb 100644 --- a/drivers/pci/setup-bus.c +++ b/drivers/pci/setup-bus.c @@ -669,6 +669,16 @@ static int pbus_size_mem(struct pci_bus *bus, unsigned long mask, if (r->parent || (r->flags & mask) != type) continue; r_size = resource_size(r); +#ifdef CONFIG_PCI_IOV + /* put SRIOV requested res to the optional list */ + if (add_head && i >= PCI_IOV_RESOURCES && + i <= PCI_IOV_RESOURCE_END) { + r->end = r->start - 1; + add_to_list(add_head, dev, r, r_size, 1); + children_add_size += r_size; + continue; + } +#endif /* For bridges size != alignment */ align = pci_resource_alignment(dev, r); order = __ffs(align) - 20;