From: "Michael S. Tsirkin" <mst@redhat.com>
To: Ani Sinha <ani@anisinha.ca>
Cc: "Eduardo Habkost" <ehabkost@redhat.com>,
jusual@redhat.com, qemu-devel@nongnu.org,
"Aleksandar Markovic" <aleksandar.qemu.devel@gmail.com>,
"Paolo Bonzini" <pbonzini@redhat.com>,
"Igor Mammedov" <imammedo@redhat.com>,
"Philippe Mathieu-Daudé" <philmd@redhat.com>,
"Aurelien Jarno" <aurelien@aurel32.net>,
"Richard Henderson" <rth@twiddle.net>
Subject: Re: [PATCH] Introduce global piix flag to disable PCI hotplug
Date: Mon, 10 Aug 2020 11:00:01 -0400 [thread overview]
Message-ID: <20200810105634-mutt-send-email-mst@kernel.org> (raw)
In-Reply-To: <CAARzgwwsuzw9rcQzu3MF7KZO8F+0PHMs3hCsiSPJCXJS-dEkhw@mail.gmail.com>
On Mon, Aug 10, 2020 at 08:24:53PM +0530, Ani Sinha wrote:
>
>
> On Mon, Aug 10, 2020 at 8:17 PM Michael S. Tsirkin <mst@redhat.com> wrote:
>
> On Mon, Aug 10, 2020 at 04:59:41PM +0530, Ani Sinha wrote:
> > We introduce a new global flag for PIIX with which we can
> > turn on or off PCI device hotplug. This flag can be used
> > to prevent all PCI devices from getting hotplugged/unplugged
> > on the PCI bus. The new options disables all hotpluh HW
> > initialization code as well as the ACPI AMLs.
> >
> > Signed-off-by: Ani Sinha <ani@anisinha.ca>
>
> Well we have a flag like this for pci bridges, right?
> So all that's left is an option to disable hotplug
> for the pci root, right?
> Wouldn't that be better than disabling it globally?
>
>
> The idea is to have just one option to disable all hotplug globally. But if you
> want to have two flags one for the bridges and one for the pci root, we can
> certainly look into it.
>
I can certainly see the attraction of a global flag.
However, with the interface we have for bridges right now, how are we
going to resolve the conflict if hotplug is disabled globally but
enabled for a given bridge?
A reasonable way would be to change the default value for bridges,
have a user-specified value override it.
This patch doesn't do it, though.
>
>
> > ---
> > hw/acpi/piix4.c   | 8 ++++++--
> >Â hw/i386/acpi-build.c | 20 ++++++++++++++------
> >Â 2 files changed, 20 insertions(+), 8 deletions(-)
> >
> > diff --git a/hw/acpi/piix4.c b/hw/acpi/piix4.c
> > index 26bac4f..8b13e86 100644
> > --- a/hw/acpi/piix4.c
> > +++ b/hw/acpi/piix4.c
> > @@ -78,6 +78,7 @@ typedef struct PIIX4PMState {
> >Â
> >Â Â Â AcpiPciHpState acpi_pci_hotplug;
> >Â Â Â bool use_acpi_hotplug_bridge;
> > +Â Â bool use_acpi_pci_hotplug;
> >Â
> >Â Â Â uint8_t disable_s3;
> >Â Â Â uint8_t disable_s4;
> > @@ -595,8 +596,9 @@ static void piix4_acpi_system_hot_add_init
> (MemoryRegion *parent,
> >Â Â Â Â Â Â Â Â Â Â Â Â Â Â "acpi-gpe0", GPE_LEN);
> >Â Â Â memory_region_add_subregion(parent, GPE_BASE, &s->io_gpe);
> >Â
> > -Â Â acpi_pcihp_init(OBJECT(s), &s->acpi_pci_hotplug, bus, parent,
> > -Â Â Â Â Â Â Â Â Â Â s->use_acpi_hotplug_bridge);
> > +Â Â if (s->use_acpi_pci_hotplug)
> > +Â Â Â Â acpi_pcihp_init(OBJECT(s), &s->acpi_pci_hotplug, bus, parent,
> > +Â Â Â Â Â Â Â Â Â Â Â Â s->use_acpi_hotplug_bridge);
> >Â
> >Â Â Â s->cpu_hotplug_legacy = true;
> >Â Â Â object_property_add_bool(OBJECT(s), "cpu-hotplug-legacy",
> > @@ -635,6 +637,8 @@ static Property piix4_pm_properties[] = {
> >Â Â Â DEFINE_PROP_UINT8(ACPI_PM_PROP_S4_VAL, PIIX4PMState, s4_val, 2),
> >Â Â Â DEFINE_PROP_BOOL("acpi-pci-hotplug-with-bridge-support",
> PIIX4PMState,
> >Â Â Â Â Â Â Â Â Â Â Â Â use_acpi_hotplug_bridge, true),
> > +Â Â DEFINE_PROP_BOOL("acpi-pci-hotplug", PIIX4PMState,
> > +Â Â Â Â Â Â Â Â Â Â Â use_acpi_pci_hotplug, true),
> >Â Â Â DEFINE_PROP_BOOL("memory-hotplug-support", PIIX4PMState,
> >Â Â Â Â Â Â Â Â Â Â Â Â acpi_memory_hotplug.is_enabled, true),
> >Â Â Â DEFINE_PROP_END_OF_LIST(),
> > diff --git a/hw/i386/acpi-build.c b/hw/i386/acpi-build.c
> > index b7bcbbb..343b9b6 100644
> > --- a/hw/i386/acpi-build.c
> > +++ b/hw/i386/acpi-build.c
> > @@ -95,6 +95,7 @@ typedef struct AcpiPmInfo {
> >Â Â Â bool s3_disabled;
> >Â Â Â bool s4_disabled;
> >Â Â Â bool pcihp_bridge_en;
> > +Â Â bool pcihp_en;
> >Â Â Â uint8_t s4_val;
> >Â Â Â AcpiFadtData fadt;
> >Â Â Â uint16_t cpu_hp_io_base;
> > @@ -245,6 +246,9 @@ static void acpi_get_pm_info(MachineState *machine,
> AcpiPmInfo *pm)
> >Â Â Â pm->pcihp_bridge_en =
> >Â Â Â Â Â object_property_get_bool(obj,
> "acpi-pci-hotplug-with-bridge-support",
> >Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â NULL);
> > +Â Â pm->pcihp_en =
> > +Â Â Â Â object_property_get_bool(obj, "acpi-pci-hotplug", NULL);
> > +
> >Â }
> >Â
> >Â static void acpi_get_misc_info(AcpiMiscInfo *info)
> > @@ -337,14 +341,16 @@ static void build_append_pcihp_notify_entry(Aml
> *method, int slot)
> >Â }
> >Â
> >Â static void build_append_pci_bus_devices(Aml *parent_scope, PCIBus *bus,
> > -Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â bool pcihp_bridge_en)
> > +Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â bool pcihp_bridge_en, bool
> pcihp_en)
> >Â {
> >Â Â Â Aml *dev, *notify_method = NULL, *method;
> > -Â Â QObject *bsel;
> > +Â Â QObject *bsel = NULL;
> >Â Â Â PCIBus *sec;
> >Â Â Â int i;
> >Â
> > -Â Â bsel = object_property_get_qobject(OBJECT(bus),
> ACPI_PCIHP_PROP_BSEL, NULL);
> > +Â Â if (pcihp_en)
> > +Â Â Â Â bsel = object_property_get_qobject(OBJECT(bus),
> > +Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â ACPI_PCIHP_PROP_BSEL, NULL);
> >Â Â Â if (bsel) {
> >Â Â Â Â Â uint64_t bsel_val = qnum_get_uint(qobject_to(QNum, bsel));
> >Â
> > @@ -439,7 +445,8 @@ static void build_append_pci_bus_devices(Aml
> *parent_scope, PCIBus *bus,
> >Â Â Â Â Â Â Â Â */
> >Â Â Â Â Â Â Â PCIBus *sec_bus = pci_bridge_get_sec_bus(PCI_BRIDGE(pdev));
> >Â
> > -Â Â Â Â Â Â build_append_pci_bus_devices(dev, sec_bus, pcihp_bridge_en);
> > +Â Â Â Â Â Â build_append_pci_bus_devices(dev, sec_bus, pcihp_bridge_en,
> > +Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â pcihp_en);
> >Â Â Â Â Â }
> >Â Â Â Â Â /* slot descriptor has been composed, add it into parent context
> */
> >Â Â Â Â Â aml_append(parent_scope, dev);
> > @@ -468,7 +475,7 @@ static void build_append_pci_bus_devices(Aml
> *parent_scope, PCIBus *bus,
> >Â Â Â }
> >Â
> >Â Â Â /* Notify about child bus events in any case */
> > -Â Â if (pcihp_bridge_en) {
> > +Â Â if (pcihp_bridge_en && pcihp_en) {
> >Â Â Â Â Â QLIST_FOREACH(sec, &bus->child, sibling) {
> >Â Â Â Â Â Â Â int32_t devfn = sec->parent_dev->devfn;
> >Â
> > @@ -1818,7 +1825,8 @@ build_dsdt(GArray *table_data, BIOSLinker *linker,
> >Â Â Â Â Â if (bus) {
> >Â Â Â Â Â Â Â Aml *scope = aml_scope("PCI0");
> >Â Â Â Â Â Â Â /* Scan all PCI buses. Generate tables to support hotplug. *
> /
> > -Â Â Â Â Â Â build_append_pci_bus_devices(scope, bus, pm->
> pcihp_bridge_en);
> > +Â Â Â Â Â Â build_append_pci_bus_devices(scope, bus, pm->
> pcihp_bridge_en,
> > +Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â pm->pcihp_en);
> >Â
> >Â Â Â Â Â Â Â if (TPM_IS_TIS_ISA(tpm)) {
> >Â Â Â Â Â Â Â Â Â if (misc->tpm_version == TPM_VERSION_2_0) {
> > --
> > 2.7.4
>
>
next prev parent reply other threads:[~2020-08-10 15:01 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-08-10 11:29 [PATCH] Introduce global piix flag to disable PCI hotplug Ani Sinha
2020-08-10 11:33 ` Ani Sinha
2020-08-10 14:46 ` Michael S. Tsirkin
2020-08-10 14:54 ` Ani Sinha
2020-08-10 15:00 ` Michael S. Tsirkin [this message]
2020-08-10 15:05 ` Ani Sinha
2020-08-10 15:19 ` Michael S. Tsirkin
2020-08-10 15:41 ` Ani Sinha
2020-08-11 13:09 ` Ani Sinha
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20200810105634-mutt-send-email-mst@kernel.org \
--to=mst@redhat.com \
--cc=aleksandar.qemu.devel@gmail.com \
--cc=ani@anisinha.ca \
--cc=aurelien@aurel32.net \
--cc=ehabkost@redhat.com \
--cc=imammedo@redhat.com \
--cc=jusual@redhat.com \
--cc=pbonzini@redhat.com \
--cc=philmd@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=rth@twiddle.net \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).