qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: "Richard W.M. Jones" <rjones@redhat.com>
To: Kevin Wolf <kwolf@redhat.com>
Cc: "Philippe Mathieu-Daudé" <philmd@linaro.org>,
	qemu-devel@nongnu.org, "Paolo Bonzini" <pbonzini@redhat.com>,
	"Hanna Reitz" <hreitz@redhat.com>, "Fam Zheng" <fam@euphon.net>,
	qemu-arm@nongnu.org, "Michael S. Tsirkin" <mst@redhat.com>,
	qemu-block@nongnu.org, "Jason Wang" <jasowang@redhat.com>,
	"Peter Maydell" <peter.maydell@linaro.org>,
	"Raphael Norwitz" <raphael.norwitz@nutanix.com>,
	"Edgar E. Iglesias" <edgar.iglesias@gmail.com>,
	"Alistair Francis" <alistair@alistair23.me>,
	"Gerd Hoffmann" <kraxel@redhat.com>
Subject: Re: [PATCH 4/7] hw/scsi/virtio-scsi: Use VIRTIO_SCSI_COMMON() macro
Date: Tue, 31 Oct 2023 13:48:56 +0000	[thread overview]
Message-ID: <20231031134856.GL7636@redhat.com> (raw)
In-Reply-To: <ZUD+hMTWf9DidzDb@redhat.com>

On Tue, Oct 31, 2023 at 02:17:56PM +0100, Kevin Wolf wrote:
> Am 17.10.2023 um 16:01 hat Philippe Mathieu-Daudé geschrieben:
> > Access QOM parent with the proper QOM VIRTIO_SCSI_COMMON() macro.
> > 
> > Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
> > ---
> >  hw/scsi/virtio-scsi.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > diff --git a/hw/scsi/virtio-scsi.c b/hw/scsi/virtio-scsi.c
> > index 45b95ea070..fa53f0902c 100644
> > --- a/hw/scsi/virtio-scsi.c
> > +++ b/hw/scsi/virtio-scsi.c
> > @@ -761,7 +761,7 @@ static void virtio_scsi_fail_cmd_req(VirtIOSCSIReq *req)
> >  
> >  static int virtio_scsi_handle_cmd_req_prepare(VirtIOSCSI *s, VirtIOSCSIReq *req)
> >  {
> > -    VirtIOSCSICommon *vs = &s->parent_obj;
> > +    VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(s);
> >      SCSIDevice *d;
> >      int rc;
> 
> Why is a dynamic cast more "proper" than a static type-safe access, even
> more so in a hot I/O path?
> 
> Rich Jones posted a flamegraph the other day that surprised me because
> object_class_dynamic_class_assert() and object_dynamic_cast_assert()
> were shown to be a big part of scsi_req_new(). In the overall
> performance, it's probably dwarved by other issues, but unnecessary
> little things can add up, too.

I think Kevin is referring to one of these flamegraphs:

  http://oirase.annexia.org/tmp/2023-kvm-build-on-device.svg
  http://oirase.annexia.org/tmp/2023-kvm-build.svg

Here's a zoom showing scsi_req_new (hopefully this URL is stable ...):

  http://oirase.annexia.org/tmp/2023-kvm-build-on-device.svg?s=scsi_req_new&x=512.9&y=501

Note that qemu has been compiled with QOM cast debug.  This is the
default for Fedora (not RHEL) because we'd like to get early detection
of bugs from Fedora users.

There was another patch recently where a simple change saved about 5%
of total runtime in RISC-V TCG guests (admittedly a much more hot path
than this one).

  https://lists.nongnu.org/archive/html/qemu-devel/2023-10/msg02388.html

Rich.

-- 
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones
Read my programming and virtualization blog: http://rwmj.wordpress.com
Fedora Windows cross-compiler. Compile Windows programs, test, and
build Windows installers. Over 100 libraries supported.
http://fedoraproject.org/wiki/MinGW



  parent reply	other threads:[~2023-10-31 13:49 UTC|newest]

Thread overview: 26+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-10-17 14:01 [PATCH 0/7] hw: Few more QOM/QDev cleanups Philippe Mathieu-Daudé
2023-10-17 14:01 ` [PATCH 1/7] hw/virtio/virtio-pmem: Replace impossible check by assertion Philippe Mathieu-Daudé
2023-10-17 17:20   ` Manos Pitsidianakis
2023-10-17 14:01 ` [PATCH 2/7] hw/block/vhost-user-blk: Use DEVICE() / VIRTIO_DEVICE() macros Philippe Mathieu-Daudé
2023-10-17 17:15   ` Manos Pitsidianakis
2023-10-17 20:09   ` Mark Cave-Ayland
2023-10-17 14:01 ` [PATCH 3/7] hw/display/virtio-gpu: Use VIRTIO_DEVICE() macro Philippe Mathieu-Daudé
2023-10-17 17:19   ` Manos Pitsidianakis
2023-10-17 20:10   ` Mark Cave-Ayland
2023-10-17 14:01 ` [PATCH 4/7] hw/scsi/virtio-scsi: Use VIRTIO_SCSI_COMMON() macro Philippe Mathieu-Daudé
2023-10-17 17:19   ` Manos Pitsidianakis
2023-10-17 20:13   ` Mark Cave-Ayland
2023-10-31 13:17   ` Kevin Wolf
2023-10-31 13:21     ` Peter Maydell
2023-10-31 13:48     ` Richard W.M. Jones [this message]
2023-10-31 16:35       ` Kevin Wolf
2023-10-31 16:42       ` Kevin Wolf
2023-10-31 16:48         ` Richard W.M. Jones
2023-10-17 14:01 ` [PATCH 5/7] hw/dma: Declare link using static DEFINE_PROP_LINK() macro Philippe Mathieu-Daudé
2023-10-17 20:16   ` Mark Cave-Ayland
2023-10-17 14:01 ` [PATCH 6/7] hw/net: " Philippe Mathieu-Daudé
2023-10-17 20:17   ` Mark Cave-Ayland
2023-10-17 14:01 ` [PATCH 7/7] hw/usb: " Philippe Mathieu-Daudé
2023-10-17 20:19   ` Mark Cave-Ayland
2023-10-17 14:09 ` [PATCH 0/7] hw: Few more QOM/QDev cleanups Michael S. Tsirkin
2023-10-19 21:41 ` Philippe Mathieu-Daudé

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=20231031134856.GL7636@redhat.com \
    --to=rjones@redhat.com \
    --cc=alistair@alistair23.me \
    --cc=edgar.iglesias@gmail.com \
    --cc=fam@euphon.net \
    --cc=hreitz@redhat.com \
    --cc=jasowang@redhat.com \
    --cc=kraxel@redhat.com \
    --cc=kwolf@redhat.com \
    --cc=mst@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=peter.maydell@linaro.org \
    --cc=philmd@linaro.org \
    --cc=qemu-arm@nongnu.org \
    --cc=qemu-block@nongnu.org \
    --cc=qemu-devel@nongnu.org \
    --cc=raphael.norwitz@nutanix.com \
    /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).