Hi Carl, Il giorno mer 9 set 2020 alle ore 13:09 Carl Yin(殷张成) <carl....@quectel.com> ha scritto: > > Hi Deniele: > > I have an idea, by now in order to use QMAP, > must execute shell command 'echo mux_id > > /sys/class/net/<iface>/add_mux' in user space, > maybe we can expand usage of sys attribute 'add_mux', like next: > 'echo mux_id mux_size mux_version > /sys/class/net/<iface>/add_mux'. > Users can set correct 'mux_size and mux_version' according to the > response of 'QMI_WDA_SET_DATA_FORMAT '. > If 'mux_size and mux_version' miss, qmi_wwan can use default values. >
not sure this is something acceptable, let's wait for Bjørn to comment. > If fixed set as 32KB, but MDM9x07 chips only support 4KB, or uses do > not enable QMAP, > Maybe cannot reach max throughput for no enough rx urbs. > I did not test for performance, but you could be right since for example, if I'm not wrong, rx_qlen for an high-speed device would be 2 with the changed rx_urb_size. So, we'll probably need to modify also usbnet_update_max_qlen, but not sure about the direction (maybe fallback to a default value to guarantee a minimum number if rx_qlen is < than a threshold?). And this is also a change affecting all the drivers using usbnet, so it requires additional care. Let's wait for the maintainers' advice also on this. Regards, Daniele > > > -----邮件原件----- > > 发件人: Daniele Palmas [mailto:dnl...@gmail.com] > > 发送时间: Wednesday, September 09, 2020 5:13 PM > > 收件人: Bjørn Mork <bj...@mork.no> > > 抄送: Kristian Evensen <kristian.even...@gmail.com>; Paul Gildea > > <paul.gil...@gmail.com>; Carl Yin(殷张成) <carl....@quectel.com>; David S . > > Miller <da...@davemloft.net>; Jakub Kicinski <k...@kernel.org>; > > netdev@vger.kernel.org; linux-...@vger.kernel.org; Daniele Palmas > > <dnl...@gmail.com> > > 主题: [PATCH net-next 1/1] net: usb: qmi_wwan: add default rx_urb_size > > > > Add default rx_urb_size to support QMAP download data aggregation without > > needing additional setup steps in userspace. > > > > The value chosen is the current highest one seen in available modems. > > > > The patch has the side-effect of fixing a babble issue in raw-ip mode > > reported by > > multiple users. > > > > Signed-off-by: Daniele Palmas <dnl...@gmail.com> > > --- > > Resending with mailing lists added: sorry for the noise. > > > > Hi Bjørn and all, > > > > this patch tries to address the issue reported in the following threads > > > > https://www.spinics.net/lists/netdev/msg635944.html > > https://www.spinics.net/lists/linux-usb/msg198846.html > > https://www.spinics.net/lists/linux-usb/msg198025.html > > > > so I'm adding the people involved, maybe you can give it a try to double > > check if > > this is good for you. > > > > On my side, I performed tests with different QC chipsets without > > experiencing > > problems. > > > > Thanks, > > Daniele > > --- > > drivers/net/usb/qmi_wwan.c | 4 ++++ > > 1 file changed, 4 insertions(+) > > > > diff --git a/drivers/net/usb/qmi_wwan.c b/drivers/net/usb/qmi_wwan.c index > > 07c42c0719f5..92d568f982b6 100644 > > --- a/drivers/net/usb/qmi_wwan.c > > +++ b/drivers/net/usb/qmi_wwan.c > > @@ -815,6 +815,10 @@ static int qmi_wwan_bind(struct usbnet *dev, struct > > usb_interface *intf) > > } > > dev->net->netdev_ops = &qmi_wwan_netdev_ops; > > dev->net->sysfs_groups[0] = &qmi_wwan_sysfs_attr_group; > > + > > + /* Set rx_urb_size to allow QMAP rx data aggregation */ > > + dev->rx_urb_size = 32768; > > + > > err: > > return status; > > } > > -- > > 2.17.1 >
