Re: [PATCH 0/7] rtl8xxxu: Pending patches

[Barry Day <briselec@gmail.com>]

On Mon, Nov 21, 2016 at 11:59:47AM -0500, Jes Sorensen wrote:
> Barry Day <briselec@gmail.com> writes:
> > On Sat, Nov 19, 2016 at 06:53:42PM -0500, Jes Sorensen wrote:
> >> Barry Day <briselec@gmail.com> writes:
> >> > On Fri, Nov 18, 2016 at 09:00:10PM -0500, Jes Sorensen wrote:
> >> >> Barry Day <briselec@gmail.com> writes:
> >> >> > On Fri, Nov 18, 2016 at 04:44:21PM -0500, Jes.Sorensen@redhat.com wrote:
> >> >> >> From: Jes Sorensen <Jes.Sorensen@redhat.com>
> >> >> >> 
> >> >> >> Kalle,
> >> >> >> 
> >> >> >> Please find attached a number of patches for the rtl8xxxu
> >> >> >> driver.
> >> >> >> 
> >> >> >> The issues reported with wpa_supplicant on 8723bu still needs further
> >> >> >> investigation.
> >> >> >> 
> >> >> >
> >> >> > The patch I posted that you want tested more will also fix the
> >> >> > wpa_supplicant issue.  Currently I'm looking at why the tx rate is not
> >> >> > what it should be. I feel fixing that first will be beneficial for
> >> >> > fixing any other issues.
> >> >> 
> >> >> Interesting, I was thinking that might be the case. I do want to dig
> >> >> into this further to understand it better. If we use your solution I
> >> >> will want to make sure we cover both gen1 and gen2 parts.
> >> >> 
> >> >> > The recent merge has made my local branch of rtl8xxxu-devel 14
> >> >> > commits ahead.
> >> >> > Do I need to do a reset and submit a new patch for the DWA-131 dongle?
> >> >> 
> >> >> In general you need to use 'git pull --rebase' on my tree. I rebase it
> >> >> to stay in sync with Kalle's tree.
> >> >> 
> >> >> The DWA-131 is the 8192eu? Sorry a bit behind and my mind is losing
> >> >> bits. If it's the patch you posted earlier I can dig it out and play
> >> >> with it - I am still catching up though, so please be patient.
> >> >
> >> > yes it's an 8192eu.
> >> 
> >> Gotcha - how do you do your testing to reproduce the problem btw? Most
> >> of my testing is using the 8723au as a primary device and the next
> >> device as a follow-on, so I may not see as long a run with the device
> >> active as you see.
> >> 
> >
> > Testing is simple. Connect to an AP in the usual
> > manner..disconnect..reconnect.  The 8192eu will fail to reconnect and
> > John Heenan reported the 8723bu also fails to reconnect. Even though
> > he was directly stopping and restarting wpa_supplicant, it's the same
> > thing to the driver - connect..disconnect..reconnect.
> 
> Thanks for the details - I'll have a look shortly.
> 
> > With using a usb_device pointer, each message starts with the usb bus
> > address.  Plug it into a different port and that address could
> > change. By using a pointer to the device associated with the wiphy
> > each message will begin with the driver name. Just makes it easier for
> > the average user to report what's in the log because he can just grep
> > for "rtl8xxxu".
> 
> I see - that would be problematic for me as I quite often have 3-4 of
> these things plugged in at the same time. Not knowing which port spits
> out the message would make it a lot harder to track. In fact my primary
> devel box for this (Lenovo Yoga 13) has an rtl8723au soldered on the
> motherboard, so the moment I plug in any other dongle I'll have two.
> 
> The alternative would be to add a prefer to the individual messages.
> 
> Cheers,
> Jes

I should have mentioned it also places the usb address after the driver name.