From: "Michael S. Tsirkin" <mst@redhat.com>
To: Igor Mammedov <imammedo@redhat.com>
Cc: qemu-devel@nongnu.org
Subject: Re: [Qemu-devel] [PATCH] pcihp: reduce number of device check events
Date: Mon, 27 Jan 2014 15:27:54 +0200 [thread overview]
Message-ID: <20140127132754.GB11895@redhat.com> (raw)
In-Reply-To: <20140127134844.23dd840f@nial.usersys.redhat.com>
On Mon, Jan 27, 2014 at 01:48:44PM +0100, Igor Mammedov wrote:
> On Sun, 26 Jan 2014 12:37:09 +0200
> "Michael S. Tsirkin" <mst@redhat.com> wrote:
>
> > PIIX created a made-up value for the UP register since it was read by
> > guest 32 times for each interrupt.
> > There's no reason to do this for the new PCIHP: register is only read
> > once for each interrupt, so clean up code by making read act as an
> > interrupt acknowledgement: the new UP register clear on read.
> >
> > In this way we cut down the number of bus rescans
> > by a factor of 32, and drop a bunch of code that's
> > now unused.
> >
> > Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
> > ---
> > hw/acpi/pcihp.c | 21 +++++----------------
> > include/hw/acpi/pcihp.h | 2 +-
> > 2 files changed, 6 insertions(+), 17 deletions(-)
> >
> > diff --git a/hw/acpi/pcihp.c b/hw/acpi/pcihp.c
> > index 3fa3d7c..4345f5d 100644
> > --- a/hw/acpi/pcihp.c
> > +++ b/hw/acpi/pcihp.c
> > @@ -116,7 +116,6 @@ static void acpi_pcihp_eject_slot(AcpiPciHpState *s, unsigned bsel, unsigned slo
> > {
> > BusChild *kid, *next;
> > int slot = ffs(slots) - 1;
> > - bool slot_free = true;
> > PCIBus *bus = acpi_pcihp_find_hotplug_bus(s, bsel);
> >
> > if (!bus) {
> > @@ -125,21 +124,17 @@ static void acpi_pcihp_eject_slot(AcpiPciHpState *s, unsigned bsel, unsigned slo
> >
> > /* Mark request as complete */
> > s->acpi_pcihp_pci_status[bsel].down &= ~(1U << slot);
> > + s->acpi_pcihp_pci_status[bsel].up &= ~(1U << slot);
> >
> > QTAILQ_FOREACH_SAFE(kid, &bus->qbus.children, sibling, next) {
> > DeviceState *qdev = kid->child;
> > PCIDevice *dev = PCI_DEVICE(qdev);
> > if (PCI_SLOT(dev->devfn) == slot) {
> > - if (acpi_pcihp_pc_no_hotplug(s, dev)) {
> > - slot_free = false;
> > - } else {
> > + if (!acpi_pcihp_pc_no_hotplug(s, dev)) {
> > object_unparent(OBJECT(qdev));
> > }
> > }
> > }
> > - if (slot_free) {
> > - s->acpi_pcihp_pci_status[bsel].device_present &= ~(1U << slot);
> > - }
> > }
> >
> > static void acpi_pcihp_update_hotplug_bus(AcpiPciHpState *s, int bsel)
> > @@ -153,7 +148,6 @@ static void acpi_pcihp_update_hotplug_bus(AcpiPciHpState *s, int bsel)
> > }
> >
> > s->acpi_pcihp_pci_status[bsel].hotplug_enable = ~0;
> > - s->acpi_pcihp_pci_status[bsel].device_present = 0;
> >
> > if (!bus) {
> > return;
> > @@ -166,8 +160,6 @@ static void acpi_pcihp_update_hotplug_bus(AcpiPciHpState *s, int bsel)
> > if (acpi_pcihp_pc_no_hotplug(s, pdev)) {
> > s->acpi_pcihp_pci_status[bsel].hotplug_enable &= ~(1U << slot);
> > }
> > -
> > - s->acpi_pcihp_pci_status[bsel].device_present |= (1U << slot);
> > }
> > }
> >
> > @@ -187,7 +179,7 @@ void acpi_pcihp_reset(AcpiPciHpState *s)
> >
> > static void enable_device(AcpiPciHpState *s, unsigned bsel, int slot)
> > {
> > - s->acpi_pcihp_pci_status[bsel].device_present |= (1U << slot);
> > + s->acpi_pcihp_pci_status[bsel].up |= (1U << slot);
> > }
> >
> > static void disable_device(AcpiPciHpState *s, unsigned bsel, int slot)
> > @@ -208,7 +200,6 @@ int acpi_pcihp_device_hotplug(AcpiPciHpState *s, PCIDevice *dev,
> > * it is present on boot, no hotplug event is necessary. We do send an
> > * event when the device is disabled later. */
> > if (state == PCI_COLDPLUG_ENABLED) {
> > - s->acpi_pcihp_pci_status[bsel].device_present |= (1U << slot);
> > return 0;
> > }
> >
> > @@ -233,10 +224,8 @@ static uint64_t pci_read(void *opaque, hwaddr addr, unsigned int size)
> >
> > switch (addr) {
> > case PCI_UP_BASE - PCI_HOTPLUG_ADDR:
> > - /* Manufacture an "up" value to cause a device check on any hotplug
> > - * slot with a device. Extra device checks are harmless. */
> > - val = s->acpi_pcihp_pci_status[bsel].device_present &
> > - s->acpi_pcihp_pci_status[bsel].hotplug_enable;
> > + val = s->acpi_pcihp_pci_status[bsel].up;
> > + s->acpi_pcihp_pci_status[bsel].up = 0;
> > ACPI_PCIHP_DPRINTF("pci_up_read %" PRIu32 "\n", val);
> > break;
> > case PCI_DOWN_BASE - PCI_HOTPLUG_ADDR:
> > diff --git a/include/hw/acpi/pcihp.h b/include/hw/acpi/pcihp.h
> > index 6230e60..aa297c2 100644
> > --- a/include/hw/acpi/pcihp.h
> > +++ b/include/hw/acpi/pcihp.h
> > @@ -32,7 +32,7 @@
> > #include "hw/pci/pci.h" /* for PCIHotplugState */
> >
> > typedef struct AcpiPciHpPciStatus {
> > - uint32_t up; /* deprecated, maintained for migration compatibility */
> > + uint32_t up;
> > uint32_t down;
> > uint32_t hotplug_enable;
> > uint32_t device_present;
>
> After patch field AcpiPciHpPciStatus.device_present seem to be unused,
> should it be deleted here as well?
I agree, will do in a follow-up patch unless there's need to respin this
one.
prev parent reply other threads:[~2014-01-27 13:23 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-01-26 10:37 [Qemu-devel] [PATCH] pcihp: reduce number of device check events Michael S. Tsirkin
2014-01-27 12:48 ` Igor Mammedov
2014-01-27 13:27 ` Michael S. Tsirkin [this message]
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=20140127132754.GB11895@redhat.com \
--to=mst@redhat.com \
--cc=imammedo@redhat.com \
--cc=qemu-devel@nongnu.org \
/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).