| Message ID | 20170510092903.23649-1-marcel@redhat.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On 10/05/2017 11:29, Marcel Apfelbaum wrote: > +++ b/hw/pci-bridge/pci_bridge_dev.c > @@ -163,7 +163,7 @@ static Property pci_bridge_dev_properties[] = { > DEFINE_PROP_ON_OFF_AUTO(PCI_BRIDGE_DEV_PROP_MSI, PCIBridgeDev, msi, > ON_OFF_AUTO_AUTO), > DEFINE_PROP_BIT(PCI_BRIDGE_DEV_PROP_SHPC, PCIBridgeDev, flags, > - PCI_BRIDGE_DEV_F_SHPC_REQ, false), > + PCI_BRIDGE_DEV_F_SHPC_REQ, true), > DEFINE_PROP_END_OF_LIST(), > }; > > diff --git a/include/hw/compat.h b/include/hw/compat.h > index 846b90e..eff021f 100644 > --- a/include/hw/compat.h > +++ b/include/hw/compat.h > @@ -18,10 +18,6 @@ > .property = "old-multiple-chip-handling",\ > .value = "on",\ > },{\ > - .driver = "pci-bridge",\ > - .property = "shpc",\ > - .value = "on",\ > - },{\ Unless we make a conscious decision to break PCI bridge backwards compatibility, you now have to disable it for 2.9 and leave this HW_COMPAT_2_8 hunk intact. Peter Xu has a patch to add the 2.10 machine types. Paolo > .driver = TYPE_PCI_DEVICE,\ > .property = "x-pcie-extcap-init",\ > .value = "off",\ > -- 2.9.3
----- Original Message ----- > From: "Paolo Bonzini" <pbonzini@redhat.com> > To: "Marcel Apfelbaum" <marcel@redhat.com>, qemu-devel@nongnu.org > Cc: qemu-stable@nongnu.org, mst@redhat.com > Sent: Wednesday, May 10, 2017 6:38:45 PM > Subject: Re: [Qemu-devel] [PATCH] Revert "hw/pci: disable pci-bridge's shpc by default" > > > > On 10/05/2017 11:29, Marcel Apfelbaum wrote: > > +++ b/hw/pci-bridge/pci_bridge_dev.c > > @@ -163,7 +163,7 @@ static Property pci_bridge_dev_properties[] = { > > DEFINE_PROP_ON_OFF_AUTO(PCI_BRIDGE_DEV_PROP_MSI, PCIBridgeDev, msi, > > ON_OFF_AUTO_AUTO), > > DEFINE_PROP_BIT(PCI_BRIDGE_DEV_PROP_SHPC, PCIBridgeDev, flags, > > - PCI_BRIDGE_DEV_F_SHPC_REQ, false), > > + PCI_BRIDGE_DEV_F_SHPC_REQ, true), > > DEFINE_PROP_END_OF_LIST(), > > }; > > > > diff --git a/include/hw/compat.h b/include/hw/compat.h > > index 846b90e..eff021f 100644 > > --- a/include/hw/compat.h > > +++ b/include/hw/compat.h > > @@ -18,10 +18,6 @@ > > .property = "old-multiple-chip-handling",\ > > .value = "on",\ > > },{\ > > - .driver = "pci-bridge",\ > > - .property = "shpc",\ > > - .value = "on",\ > > - },{\ > Hi Paolo, > Unless we make a conscious decision to break PCI bridge backwards > compatibility, you now have to disable it for 2.9 and leave this > HW_COMPAT_2_8 hunk intact. Peter Xu has a patch to add the 2.10 machine > types. > I thought is not an interesting enough scenario to keep it "broken" for 2.9, but I'll do that, sure, I'll add it on top of Peter's patch. Thanks, Marcel > Paolo > > > .driver = TYPE_PCI_DEVICE,\ > > .property = "x-pcie-extcap-init",\ > > .value = "off",\ > > -- 2.9.3 >
On Wed, May 10, 2017 at 05:38:45PM +0200, Paolo Bonzini wrote: > > > On 10/05/2017 11:29, Marcel Apfelbaum wrote: > > +++ b/hw/pci-bridge/pci_bridge_dev.c > > @@ -163,7 +163,7 @@ static Property pci_bridge_dev_properties[] = { > > DEFINE_PROP_ON_OFF_AUTO(PCI_BRIDGE_DEV_PROP_MSI, PCIBridgeDev, msi, > > ON_OFF_AUTO_AUTO), > > DEFINE_PROP_BIT(PCI_BRIDGE_DEV_PROP_SHPC, PCIBridgeDev, flags, > > - PCI_BRIDGE_DEV_F_SHPC_REQ, false), > > + PCI_BRIDGE_DEV_F_SHPC_REQ, true), > > DEFINE_PROP_END_OF_LIST(), > > }; > > > > diff --git a/include/hw/compat.h b/include/hw/compat.h > > index 846b90e..eff021f 100644 > > --- a/include/hw/compat.h > > +++ b/include/hw/compat.h > > @@ -18,10 +18,6 @@ > > .property = "old-multiple-chip-handling",\ > > .value = "on",\ > > },{\ > > - .driver = "pci-bridge",\ > > - .property = "shpc",\ > > - .value = "on",\ > > - },{\ > > Unless we make a conscious decision to break PCI bridge backwards > compatibility, you now have to disable it for 2.9 and leave this > HW_COMPAT_2_8 hunk intact. Peter Xu has a patch to add the 2.10 machine > types. > > Paolo > > > .driver = TYPE_PCI_DEVICE,\ > > .property = "x-pcie-extcap-init",\ > > .value = "off",\ > > -- 2.9.3 It's going upstream now, I just need to rebase.
diff --git a/hw/pci-bridge/pci_bridge_dev.c b/hw/pci-bridge/pci_bridge_dev.c index 647ad80..5dbd933 100644 --- a/hw/pci-bridge/pci_bridge_dev.c +++ b/hw/pci-bridge/pci_bridge_dev.c @@ -163,7 +163,7 @@ static Property pci_bridge_dev_properties[] = { DEFINE_PROP_ON_OFF_AUTO(PCI_BRIDGE_DEV_PROP_MSI, PCIBridgeDev, msi, ON_OFF_AUTO_AUTO), DEFINE_PROP_BIT(PCI_BRIDGE_DEV_PROP_SHPC, PCIBridgeDev, flags, - PCI_BRIDGE_DEV_F_SHPC_REQ, false), + PCI_BRIDGE_DEV_F_SHPC_REQ, true), DEFINE_PROP_END_OF_LIST(), }; diff --git a/include/hw/compat.h b/include/hw/compat.h index 846b90e..eff021f 100644 --- a/include/hw/compat.h +++ b/include/hw/compat.h @@ -18,10 +18,6 @@ .property = "old-multiple-chip-handling",\ .value = "on",\ },{\ - .driver = "pci-bridge",\ - .property = "shpc",\ - .value = "on",\ - },{\ .driver = TYPE_PCI_DEVICE,\ .property = "x-pcie-extcap-init",\ .value = "off",\
This reverts commit dc0ae767700c156894e36fab89a745a2dc4173de. Disabling the shpc controller has an undesired side effect. The PCI bridge remains with no attached devices at boot time, and the guest operating systems do not all allocate any resources for it leaving the bridge unusable. Note that the behaviour is dictated by the pci bridge specification. Revert the commit and leave the shpc controller even if is not actually used by any architecture. Slot 0 remains unusable at boot time. Signed-off-by: Marcel Apfelbaum <marcel@redhat.com> --- hw/pci-bridge/pci_bridge_dev.c | 2 +- include/hw/compat.h | 4 ---- 2 files changed, 1 insertion(+), 5 deletions(-)