From: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
To: Igor Mammedov <imammedo@redhat.com>
Cc: qemu-devel@nongnu.org, pbonzini@redhat.com,
maxime.coquelin@redhat.com, mst@redhat.com, groug@kaod.org
Subject: Re: [Qemu-devel] [PATCH v5 4/7] vhost: Merge sections added to temporary list
Date: Mon, 8 Jan 2018 18:07:30 +0000 [thread overview]
Message-ID: <20180108180729.GI2462@work-vm> (raw)
In-Reply-To: <20171227133403.42d43cf1@igors-macbook-pro.local>
* Igor Mammedov (imammedo@redhat.com) wrote:
> On Mon, 18 Dec 2017 20:13:37 +0000
> "Dr. David Alan Gilbert (git)" <dgilbert@redhat.com> wrote:
>
> > From: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
> >
> > As sections are reported by the listener to the _nop and _add
> > methods, add them to the temporary section list but now merge them
> > with the previous section if the new one abuts and the backend allows.
> >
> > Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
> beside small nit patch looks good to me, so with it fixed
>
> Reviewed-by: Igor Mammedov <imammedo@redhat.com>
>
> > ---
> > hw/virtio/trace-events | 2 ++
> > hw/virtio/vhost.c | 70 +++++++++++++++++++++++++++++++++++++++++++-------
> > 2 files changed, 63 insertions(+), 9 deletions(-)
> >
> > diff --git a/hw/virtio/trace-events b/hw/virtio/trace-events
> > index 4a493bcd46..0e63c8739d 100644
> > --- a/hw/virtio/trace-events
> > +++ b/hw/virtio/trace-events
> > @@ -1,6 +1,8 @@
> > # See docs/devel/tracing.txt for syntax documentation.
> >
> > # hw/virtio/vhost.c
> > +vhost_region_add_section(const char *name, uint64_t gpa, uint64_t size, uint64_t host) "%s: 0x%"PRIx64"+0x%"PRIx64" @ 0x%"PRIx64
> > +vhost_region_add_section_abut(const char *name, uint64_t new_size) "%s: 0x%"PRIx64
> > vhost_section(const char *name, int r) "%s:%d"
> >
> > # hw/virtio/virtio.c
> > diff --git a/hw/virtio/vhost.c b/hw/virtio/vhost.c
> > index 18611f0d40..57d15acd2b 100644
> > --- a/hw/virtio/vhost.c
> > +++ b/hw/virtio/vhost.c
> > @@ -709,14 +709,65 @@ out:
> > return;
> > }
> >
> > -static void vhost_add_section(struct vhost_dev *dev,
> > - MemoryRegionSection *section)
> > +/* Adds the section data to the tmp_section structure.
> > + * It relies on the listener calling us in memory address order
> > + * and for each region (via the _add and _nop methods) to
> > + * join neighbours.
> > + */
> > +static void vhost_region_add_section(struct vhost_dev *dev,
> > + MemoryRegionSection *section)
> wrong alignment, should be on '('
Fixed, thanks.
Dave
> > {
> > - ++dev->n_tmp_sections;
> > - dev->tmp_sections = g_renew(MemoryRegionSection, dev->tmp_sections,
> > - dev->n_tmp_sections);
> > - dev->tmp_sections[dev->n_tmp_sections - 1] = *section;
> > - memory_region_ref(section->mr);
> > + bool need_add = true;
> > + uint64_t mrs_size = int128_get64(section->size);
> > + uint64_t mrs_gpa = section->offset_within_address_space;
> > + uintptr_t mrs_host = (uintptr_t)memory_region_get_ram_ptr(section->mr) +
> > + section->offset_within_region;
> > +
> > + trace_vhost_region_add_section(section->mr->name, mrs_gpa, mrs_size,
> > + mrs_host);
> > +
> > + if (dev->n_tmp_sections) {
> > + /* Since we already have at least one section, lets see if
> > + * this extends it; since we're scanning in order, we only
> > + * have to look at the last one, and the FlatView that calls
> > + * us shouldn't have overlaps.
> > + */
> > + MemoryRegionSection *prev_sec = dev->tmp_sections +
> > + (dev->n_tmp_sections - 1);
> > + uint64_t prev_gpa_start = prev_sec->offset_within_address_space;
> > + uint64_t prev_size = int128_get64(prev_sec->size);
> > + uint64_t prev_gpa_end = range_get_last(prev_gpa_start, prev_size);
> > + uint64_t prev_host_start =
> > + (uintptr_t)memory_region_get_ram_ptr(prev_sec->mr) +
> > + prev_sec->offset_within_region;
> > + uint64_t prev_host_end = range_get_last(prev_host_start, prev_size);
> > +
> > + if (prev_gpa_end + 1 == mrs_gpa &&
> > + prev_host_end + 1 == mrs_host &&
> > + section->mr == prev_sec->mr &&
> > + (!dev->vhost_ops->vhost_backend_can_merge ||
> > + dev->vhost_ops->vhost_backend_can_merge(dev,
> > + mrs_host, mrs_size,
> > + prev_host_start, prev_size))) {
> > + /* The two sections abut */
> > + need_add = false;
> > + prev_sec->size = int128_add(prev_sec->size, section->size);
> > + trace_vhost_region_add_section_abut(section->mr->name,
> > + mrs_size + prev_size);
> > + }
> > + }
> > +
> > + if (need_add) {
> > + ++dev->n_tmp_sections;
> > + dev->tmp_sections = g_renew(MemoryRegionSection, dev->tmp_sections,
> > + dev->n_tmp_sections);
> > + dev->tmp_sections[dev->n_tmp_sections - 1] = *section;
> > + /* The flatview isn't stable and we don't use it, making it NULL
> > + * means we can memcmp the list.
> > + */
> > + dev->tmp_sections[dev->n_tmp_sections - 1].fv = NULL;
> > + memory_region_ref(section->mr);
> > + }
> > }
> >
> > static void vhost_region_add(MemoryListener *listener,
> > @@ -728,11 +779,12 @@ static void vhost_region_add(MemoryListener *listener,
> > if (!vhost_section(section)) {
> > return;
> > }
> > + vhost_region_add_section(dev, section);
> >
> > - vhost_add_section(dev, section);
> > vhost_set_memory(listener, section, true);
> > }
> >
> > +/* Called on regions that have not changed */
> > static void vhost_region_nop(MemoryListener *listener,
> > MemoryRegionSection *section)
> > {
> > @@ -743,7 +795,7 @@ static void vhost_region_nop(MemoryListener *listener,
> > return;
> > }
> >
> > - vhost_add_section(dev, section);
> > + vhost_region_add_section(dev, section);
> > }
> >
> > static void vhost_region_del(MemoryListener *listener,
>
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK
next prev parent reply other threads:[~2018-01-08 18:07 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-12-18 20:13 [Qemu-devel] [PATCH v5 0/7] Rework vhost memory region updates Dr. David Alan Gilbert (git)
2017-12-18 20:13 ` [Qemu-devel] [PATCH v5 1/7] vhost: Build temporary section list and deref after commit Dr. David Alan Gilbert (git)
2017-12-27 11:56 ` Igor Mammedov
2018-01-08 17:48 ` Dr. David Alan Gilbert
2017-12-18 20:13 ` [Qemu-devel] [PATCH v5 2/7] vhost: Move log_dirty check Dr. David Alan Gilbert (git)
2017-12-18 20:13 ` [Qemu-devel] [PATCH v5 3/7] vhost: Simplify ring verification checks Dr. David Alan Gilbert (git)
2017-12-27 12:20 ` Igor Mammedov
2017-12-27 14:06 ` Igor Mammedov
2018-01-08 17:54 ` Dr. David Alan Gilbert
2018-01-10 10:23 ` Igor Mammedov
2018-01-11 3:29 ` Jason Wang
2018-01-16 16:48 ` Dr. David Alan Gilbert
2017-12-18 20:13 ` [Qemu-devel] [PATCH v5 4/7] vhost: Merge sections added to temporary list Dr. David Alan Gilbert (git)
2017-12-27 12:34 ` Igor Mammedov
2018-01-08 18:07 ` Dr. David Alan Gilbert [this message]
2018-01-16 3:20 ` Michael S. Tsirkin
2017-12-18 20:13 ` [Qemu-devel] [PATCH v5 5/7] vhost: Regenerate region list from changed sections list Dr. David Alan Gilbert (git)
2017-12-27 13:19 ` Igor Mammedov
2017-12-28 13:03 ` Igor Mammedov
2018-01-09 16:41 ` Dr. David Alan Gilbert
2017-12-18 20:13 ` [Qemu-devel] [PATCH v5 6/7] vhost: Clean out old vhost_set_memory and friends Dr. David Alan Gilbert (git)
2017-12-22 19:14 ` Igor Mammedov
2018-01-09 15:54 ` Dr. David Alan Gilbert
2017-12-18 20:13 ` [Qemu-devel] [PATCH v5 7/7] vhost: Merge and delete unused callbacks Dr. David Alan Gilbert (git)
2017-12-27 13:27 ` Igor Mammedov
2018-01-08 18:43 ` Dr. David Alan Gilbert
2017-12-27 13:44 ` [Qemu-devel] [PATCH v5 0/7] Rework vhost memory region updates Igor Mammedov
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=20180108180729.GI2462@work-vm \
--to=dgilbert@redhat.com \
--cc=groug@kaod.org \
--cc=imammedo@redhat.com \
--cc=maxime.coquelin@redhat.com \
--cc=mst@redhat.com \
--cc=pbonzini@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).