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

From: Jes Sorensen <Jes.Sorensen@redhat.com>
To: Barry Day <briselec@gmail.com>
Cc: linux-wireless@vger.kernel.org, kvalo@codeaurora.org,
	Larry.Finger@lwfinger.net
Subject: Re: [PATCH 0/7] rtl8xxxu: Pending patches
Date: Sat, 19 Nov 2016 18:53:42 -0500	[thread overview]
Message-ID: <wrfj37inauwp.fsf@redhat.com> (raw)
In-Reply-To: <20161119023830.GA7324@testbox> (Barry Day's message of "Sat, 19 Nov 2016 12:38:31 +1000")

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.

> Would you accept a patch that adds a struct device pointer to rtl8xxxu_priv
> and used as the device pointer in the logging functions? Then all the messages
> will start with the driver name making them easier to find.

How do you mean? Right now I have a struct usb_device pointer and
dereference that for ->dev to use with dev_info() messages etc. Do you
want to see more than that?

Cheers,
Jes