[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [PATCH 2/2] acpi: x86: pcihp: add support hotplug on multifunction b
From: |
Michael S. Tsirkin |
Subject: |
Re: [PATCH 2/2] acpi: x86: pcihp: add support hotplug on multifunction bridges |
Date: |
Thu, 22 Jul 2021 13:49:34 -0400 |
On Thu, Jul 22, 2021 at 06:59:45AM -0400, Igor Mammedov wrote:
> Commit 17858a1695 (hw/acpi/ich9: Set ACPI PCI hot-plug as default on Q35)
> switched PCI hotplug from native to ACPI one by default.
>
> That however breaks ihotplug on following CLI that used to work:
s/ihotplug/hotplug/ ?
> -nodefaults -machine q35 \
> -device
> pcie-root-port,id=pcie-root-port-0,multifunction=on,bus=pcie.0,addr=0x1,chassis=1
> \
> -device
> pcie-root-port,id=pcie-root-port-1,port=0x1,addr=0x1.0x1,bus=pcie.0,chassis=2
>
> where PCI device is hotplugged to pcie-root-port-1 with error on guest side:
>
> ACPI BIOS Error (bug): Could not resolve symbol [^S0B.PCNT], AE_NOT_FOUND
> (20201113/psargs-330)
> ACPI Error: Aborting method \_SB.PCI0.PCNT due to previous error
> (AE_NOT_FOUND) (20201113/psparse-531)
> ACPI Error: Aborting method \_GPE._E01 due to previous error (AE_NOT_FOUND)
> (20201113/psparse-531)
> ACPI Error: AE_NOT_FOUND, while evaluating GPE method [_E01]
> (20201113/evgpe-515)
>
> cause is that QEMU's ACPI hotplug never supported functions other then 0
> and due to bug it was generating notification entries for not described
> functions.
>
> Technically there is no reason not to describe cold-plugged bridges
> (root ports) on functions other then 0, as they similaraly to bridge
> on function 0 are unpluggable.
>
> Fix consists of describing cold-plugged bridges[root ports] on functions
> other than 0.
I would add: since we need to describe multifunction devices
>
> Fixes: 17858a169508609ca9063c544833e5a1adeb7b52
use short hash and include subject within ("subject here") please
> Signed-off-by: Igor Mammedov <imammedo@redhat.com>
> Reported-by: Laurent Vivier <lvivier@redhat.com>
> ---
> hw/i386/acpi-build.c | 31 ++++++++++++++++++++++++-------
> 1 file changed, 24 insertions(+), 7 deletions(-)
>
> diff --git a/hw/i386/acpi-build.c b/hw/i386/acpi-build.c
> index b40e284b72..77cb7a338a 100644
> --- a/hw/i386/acpi-build.c
> +++ b/hw/i386/acpi-build.c
> @@ -364,7 +364,7 @@ static void build_append_pcihp_notify_entry(Aml *method,
> int slot)
> int32_t devfn = PCI_DEVFN(slot, 0);
>
> if_ctx = aml_if(aml_and(aml_arg(0), aml_int(0x1U << slot), NULL));
> - aml_append(if_ctx, aml_notify(aml_name("S%.02X", devfn), aml_arg(1)));
> + aml_append(if_ctx, aml_notify(aml_name("^S%.03X", devfn), aml_arg(1)));
> aml_append(method, if_ctx);
> }
>
why are you adding a parent prefix here? anything wrong with
relying on search rules?
Seems like an unrelated change.
Also there's always 8 bit in devfn. Why do we need an extra 0?
> @@ -389,18 +389,26 @@ static void build_append_pci_bus_devices(Aml
> *parent_scope, PCIBus *bus,
> PCIDeviceClass *pc;
> PCIDevice *pdev = bus->devices[devfn];
> int slot = PCI_SLOT(devfn);
> + int func = PCI_FUNC(devfn);
> + /* ACPI spec: 1.0b: Table 6-2 _ADR Object Bus Types, PCI type */
> + int adr = slot << 16 | func;
> bool hotplug_enabled_dev;
> bool bridge_in_acpi;
> bool cold_plugged_bridge;
>
> if (!pdev) {
> - if (bsel) { /* add hotplug slots for non present devices */
> + /*
> + * add hotplug slots for non present devices.
> + * hotplug is supported only for non-multifunction device
> + * so generate device description only for function 0
> + */
> + if (bsel && !PCI_FUNC(devfn)) {
replace with func here and elsewhere?
> if (pci_bus_is_express(bus) && slot > 0) {
> break;
> }
> - dev = aml_device("S%.02X", PCI_DEVFN(slot, 0));
> + dev = aml_device("S%.03X", devfn);
> aml_append(dev, aml_name_decl("_SUN", aml_int(slot)));
> - aml_append(dev, aml_name_decl("_ADR", aml_int(slot << 16)));
> + aml_append(dev, aml_name_decl("_ADR", aml_int(adr)));
> method = aml_method("_EJ0", 1, AML_NOTSERIALIZED);
> aml_append(method,
> aml_call2("PCEJ", aml_name("BSEL"), aml_name("_SUN"))
> @@ -436,9 +444,18 @@ static void build_append_pci_bus_devices(Aml
> *parent_scope, PCIBus *bus,
> continue;
> }
>
> + /*
> + * allow describing coldplugged bridges in ACPI even if they are not
> + * on function 0, as they are not unpluggale, for all other devices
unpluggable
> + * generate description only for function 0 per slot
> + */
> + if (PCI_FUNC(devfn) && !bridge_in_acpi) {
> + continue;
> + }
> +
> /* start to compose PCI slot descriptor */
> - dev = aml_device("S%.02X", PCI_DEVFN(slot, 0));
> - aml_append(dev, aml_name_decl("_ADR", aml_int(slot << 16)));
> + dev = aml_device("S%.03X", devfn);
> + aml_append(dev, aml_name_decl("_ADR", aml_int(adr)));
>
> if (bsel) {
> /*
> @@ -529,7 +546,7 @@ static void build_append_pci_bus_devices(Aml
> *parent_scope, PCIBus *bus,
> continue;
> }
>
> - aml_append(method, aml_name("^S%.02X.PCNT",
> + aml_append(method, aml_name("^S%.03X.PCNT",
> sec->parent_dev->devfn));
> }
> }
> --
> 2.27.0
Re: [PATCH 2/2] acpi: x86: pcihp: add support hotplug on multifunction bridges, Daniel P . Berrangé, 2021/07/23
[PATCH 1/2] acpi: x86: pcihp: cleanup devfn usage in build_append_pci_bus_devices(), Igor Mammedov, 2021/07/22