qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: "Michael S. Tsirkin" <mst@redhat.com>
To: Isaku Yamahata <yamahata@valinux.co.jp>
Cc: qemu-devel@nongnu.org
Subject: [Qemu-devel] Re: [PATCH v3 3/3] piix_pci: optimize set irq path
Date: Mon, 21 Mar 2011 13:37:07 +0200	[thread overview]
Message-ID: <20110321113707.GC27445@redhat.com> (raw)
In-Reply-To: <533593a679f7c243ddce87c9c3a7c31c6f67acd7.1300540833.git.yamahata@valinux.co.jp>

On Sat, Mar 19, 2011 at 10:24:52PM +0900, Isaku Yamahata wrote:
> optimize irq routing in piix_pic.c which has been a TODO.
> So far piix3 tracks each pirq level and checks whether a given pic pins is
> asserted by seeing if each pirq is mapped into the pic pin.
> This is independent on irq routing, but data path is on slow path.
> 
> Given that irq routing is rarely changed and asserting pic pins is on
> data path, the path that asserts pic pins should be optimized and
> chainging irq routing should be on slow path.
> The new behavior with this patch series is to use bitmap which is addressed
> by pirq and pic pins with a given irq routing.
> When pirq is asserted, the bitmap is set and see if the pic pins is
> asserted by checking the bitmaps.
> When irq routing is changed, rebuild the bitmap and re-assert pic pins.
> 
> Cc: Michael S. Tsirkin <mst@redhat.com>
> Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
> ---
> Changes v1 -> v2:
> - some minor clean ups
> - commit log message
> ---
>  hw/piix_pci.c |   94 +++++++++++++++++++++++++++++++++++++++++++++++++--------
>  1 files changed, 81 insertions(+), 13 deletions(-)
> 
> diff --git a/hw/piix_pci.c b/hw/piix_pci.c
> index e88df43..f07e19d 100644
> --- a/hw/piix_pci.c
> +++ b/hw/piix_pci.c
> @@ -37,8 +37,27 @@
>  
>  typedef PCIHostState I440FXState;
>  
> +#define PIIX_NUM_PIC_IRQS       16      /* i8259 * 2 */
> +#define PIIX_NUM_PIRQS          4ULL    /* PIRQ[A-D] */
> +#define PIIX_PIRQC              0x60
> +
>  typedef struct PIIX3State {
>      PCIDevice dev;
> +
> +    /*
> +     * bitmap to track pic levels.
> +     * The pic level is the logical OR of all the PCI irqs mapped to it
> +     * So one PIC level is tracked by PIIX_NUM_PIRQS bits.
> +     *
> +     * PIRQ is mapped to PIC pins, we track it by
> +     * PIIX_NUM_PIRQS * PIIX_NUM_PIC_IRQS = 64 bits with
> +     * pic_irq * PIIX_NUM_PIRQS + pirq
> +     */
> +#if PIIX_NUM_PIC_IRQS * PIIX_NUM_PIRQS > 64
> +#error "unable to encode pic state in 64bit in pic_levels."
> +#endif
> +    uint64_t pic_levels;
> +
>      qemu_irq *pic;
>  
>      /* This member isn't used. Just for save/load compatibility */
> @@ -254,25 +273,63 @@ PCIBus *i440fx_init(PCII440FXState **pi440fx_state, int *piix3_devfn, qemu_irq *
>  }
>  
>  /* PIIX3 PCI to ISA bridge */
> +static void piix3_set_irq_pic(PIIX3State *piix3, int pic_irq)
> +{
> +    qemu_set_irq(piix3->pic[pic_irq],
> +                 !!(piix3->pic_levels &
> +                    ((PIIX_NUM_PIRQS - 1) << (pic_irq * PIIX_NUM_PIRQS))));
> +}
> +
> +static void piix3_set_irq_level(PIIX3State *piix3, int irq_num, int level,
> +                                bool propagate)
> +{
> +    int pic_irq;
> +    uint64_t mask;
> +
> +    pic_irq = piix3->dev.config[PIIX_PIRQC + irq_num];
> +    if (pic_irq >= PIIX_NUM_PIC_IRQS) {
> +        return;
> +    }
> +
> +    mask = 1ULL << ((pic_irq * PIIX_NUM_PIRQS) + irq_num);
> +    piix3->pic_levels &= ~mask;
> +    piix3->pic_levels |= mask * !!level;
> +
> +    if (propagate) {
> +        piix3_set_irq_pic(piix3, pic_irq);
> +    }
> +}
>  
>  static void piix3_set_irq(void *opaque, int irq_num, int level)
>  {
> -    int i, pic_irq, pic_level;
>      PIIX3State *piix3 = opaque;
> +    piix3_set_irq_level(piix3, irq_num, level, true);
> +}
>  
> -    /* now we change the pic irq level according to the piix irq mappings */
> -    /* XXX: optimize */
> -    pic_irq = piix3->dev.config[0x60 + irq_num];
> -    if (pic_irq < 16) {
> -        /* The pic level is the logical OR of all the PCI irqs mapped
> -           to it */
> -        pic_level = 0;
> -        for (i = 0; i < 4; i++) {
> -            if (pic_irq == piix3->dev.config[0x60 + i]) {
> -                pic_level |= pci_bus_get_irq_level(piix3->dev.bus, i);
> -            }
> +/* irq routing is changed. so rebuild bitmap */
> +static void piix3_update_irq_levels(PIIX3State *piix3)
> +{
> +    int pirq;
> +
> +    piix3->pic_levels = 0;
> +    for (pirq = 0; pirq < PIIX_NUM_PIRQS; pirq++) {
> +        piix3_set_irq_level(piix3, pirq,
> +                            pci_bus_get_irq_level(piix3->dev.bus, pirq),
> +                            false);
> +    }
> +}
> +
> +static void piix3_write_config(PCIDevice *dev,
> +                               uint32_t address, uint32_t val, int len)
> +{
> +    pci_default_write_config(dev, address, val, len);
> +    if (ranges_overlap(address, len, PIIX_PIRQC, 4)) {
> +        PIIX3State *piix3 = DO_UPCAST(PIIX3State, dev, dev);
> +        int pic_irq;
> +        piix3_update_irq_levels(piix3);
> +        for (pic_irq = 0; pic_irq < PIIX_NUM_PIC_IRQS; pic_irq++) {
> +            piix3_set_irq_pic(piix3, pic_irq);
>          }

We wouldn't need this loop if piix3_update_irq_levels
called piix3_set_irq_level with propagate enabled.

> -        qemu_set_irq(piix3->pic[pic_irq], pic_level);
>      }
>  }
>  
> @@ -312,6 +369,15 @@ static void piix3_reset(void *opaque)
>      pci_conf[0xab] = 0x00;
>      pci_conf[0xac] = 0x00;
>      pci_conf[0xae] = 0x00;
> +
> +    d->pic_levels = 0;
> +}
> +
> +static int piix3_post_load(void *opaque, int version_id)
> +{
> +    PIIX3State *piix3 = opaque;
> +    piix3_update_irq_levels(piix3);

Couldn't figure out why would we not want to
propagate the interrupts here.
Could you explain please?
What happens if we do propagate them?
Nothing bad, right?

> +    return 0;
>  }
>  
>  static void piix3_pre_save(void *opaque)
> @@ -330,6 +396,7 @@ static const VMStateDescription vmstate_piix3 = {
>      .version_id = 3,
>      .minimum_version_id = 2,
>      .minimum_version_id_old = 2,
> +    .post_load = piix3_post_load,
>      .pre_save = piix3_pre_save,
>      .fields      = (VMStateField []) {
>          VMSTATE_PCI_DEVICE(dev, PIIX3State),
> @@ -372,6 +439,7 @@ static PCIDeviceInfo i440fx_info[] = {
>          .qdev.no_user = 1,
>          .no_hotplug   = 1,
>          .init         = piix3_initfn,
> +        .config_write = piix3_write_config,
>      },{
>          /* end of list */
>      }
> -- 
> 1.7.1.1

  reply	other threads:[~2011-03-21 11:37 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-03-19 13:24 [Qemu-devel] [PATCH v3 0/3] piix_pci: optimize irq data path Isaku Yamahata
2011-03-19 13:24 ` [Qemu-devel] [PATCH v3 1/3] pci: add accessor function to get irq levels Isaku Yamahata
2011-03-19 13:24 ` [Qemu-devel] [PATCH v3 2/3] piix_pci: eliminate PIIX3State::pci_irq_levels Isaku Yamahata
2011-03-19 13:24 ` [Qemu-devel] [PATCH v3 3/3] piix_pci: optimize set irq path Isaku Yamahata
2011-03-21 11:37   ` Michael S. Tsirkin [this message]
2011-03-21 12:10     ` [Qemu-devel] " Isaku Yamahata
2011-03-21 12:31       ` Michael S. Tsirkin
2011-03-21 12:56         ` Isaku Yamahata
2011-03-21 13:01           ` Michael S. Tsirkin
2011-03-21 12:26     ` Isaku Yamahata
2011-03-21 14:10   ` Michael S. Tsirkin
2011-03-22  0:50     ` Isaku Yamahata
2011-03-22 13:40       ` Michael S. Tsirkin
2011-03-22 14:07         ` Isaku Yamahata
2011-03-22 16:24           ` Michael S. Tsirkin

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=20110321113707.GC27445@redhat.com \
    --to=mst@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=yamahata@valinux.co.jp \
    /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).