6:30 a.m.
On Fri, Apr 21, 2017 at 11:19:36AM +0200, Michal Privoznik wrote:
On 04/20/2017 02:21 PM, Martin Kletzander wrote:
> We are currently parsing only rx_max_coalesced_frames because that's
> the only value that makes sense for us. The tun device just added
> support for this one and the others are only supported by hardware
> devices which we don't need to worry about as the only way we'd pass
> those to the domain is using <hostdev/> or <interface
type='hostdev'/>.
> And in those cases the guest can modify the settings itself.
>
> Signed-off-by: Martin Kletzander <mkletzan(a)redhat.com>
> ---
> docs/formatdomain.html.in | 24 ++++
> docs/schemas/domaincommon.rng | 131 +++++++++++++++++++++
> src/conf/domain_conf.c | 80 +++++++++++++
> src/conf/domain_conf.h | 2 +
> src/qemu/qemu_domain.c | 31 +++++
> .../qemuxml2argvdata/qemuxml2argv-net-coalesce.xml | 68 +++++++++++
> .../qemuxml2xmlout-net-coalesce.xml | 71 +++++++++++
> tests/qemuxml2xmltest.c | 1 +
> 8 files changed, 408 insertions(+)
> create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-net-coalesce.xml
> create mode 100644 tests/qemuxml2xmloutdata/qemuxml2xmlout-net-coalesce.xml
>
> diff --git a/docs/formatdomain.html.in b/docs/formatdomain.html.in
> index b1e38f00e423..ea64b7fd1193 100644
> --- a/docs/formatdomain.html.in
> +++ b/docs/formatdomain.html.in
> @@ -5405,6 +5405,30 @@ qemu-kvm -net nic,model=? /dev/null
> <span class="since">Since 3.1.0</span>
> </p>
>
> + <h5><a name="coalesce">Coalesce
settings</a></h5>
> +<pre>
> +...
> +<devices>
> + <interface type='network'>
> + <source network='default'/>
> + <target dev='vnet0'/>
> + <b><coalesce>
> +
<rx_max_coalesced_frames>5</rx_max_coalesced_frames>
> + </coalesce></b>
> + </interface>
> +</devices>
> +...</pre>
> +
> + <p>
> + This element provides means of setting coalesce settings for some
> + interface devices (currently only type <code>network</code>
> + and <code>bridge</code>. Currently there is just one sub-element
> + named <code>rx_max_coalesced_frames</code> which accepts a
non-negative
> + integer that specifies the maximum number of packets that will be received
> + before an interrupt.
> + <span class="since">Since 3.3.0</span>
> + </p>
This does not correspond to the schema or tests introduced later in the
patch.
Good catch, thanks for noticing.
> diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c
> index 705deb39a1bf..cbeebdc56880 100644
> --- a/src/conf/domain_conf.c
> +++ b/src/conf/domain_conf.c
> @@ -6772,6 +6772,77 @@ virDomainNetIPInfoParseXML(const char *source,
> return ret;
> }
>
> +
> +static virNetDevCoalescePtr
> +virDomainNetDefCoalesceParseXML(xmlNodePtr node,
> + xmlXPathContextPtr ctxt)
> +{
> + virNetDevCoalescePtr ret = NULL;
> + xmlNodePtr save = NULL;
> + char *str = NULL;
> + unsigned long long tmp = 0;
> +
> + save = ctxt->node;
> + ctxt->node = node;
> +
> + str = virXPathString("string(./rx/frames/@max)", ctxt);
> + if (!str)
> + goto cleanup;
> +
> + if (!ret && VIR_ALLOC(ret) < 0)
> + goto cleanup;
> +
> + if (virStrToLong_ullp(str, NULL, 10, &tmp) < 0) {
> + virReportError(VIR_ERR_XML_DETAIL,
> + _("cannot parse value '%s' for coalesce
parameter"),
> + str);
> + VIR_FREE(str);
> + goto error;
> + }
> + VIR_FREE(str);
> +
> + if (tmp > UINT32_MAX) {
> + virReportError(VIR_ERR_OVERFLOW,
> + _("value '%llu' is too big for coalesce "
> + "parameter, maximum is '%lu'"),
> + tmp, (unsigned long) UINT32_MAX);
> + goto error;
> + }
> + ret->rx_max_coalesced_frames = tmp;
I wonder if we can turn this into a macro (despite some of us hating
them) because potentially we'll parse all the 32 coalesce attributes (or
how many there are). But for now, this is okay as is.
I did that. Multiple times. GET_COAL(rx, frames, max) for example and
I tried various ways how to do that so it's nicely extensible. But
after a while I gave up because it would be used once and whatever.
Maybe nobody will ever want to add any other parameter...
> +
> + cleanup:
> + ctxt->node = save;
> + return ret;
> +
> + error:
> + VIR_FREE(ret);
> + goto cleanup;
> +}
Michal