Re: [PATCH 1/1] New driver: rtl8xxxu (mac80211)

[Jes Sorensen <Jes.Sorensen@redhat.com>]

Larry Finger <Larry.Finger@lwfinger.net> writes:
> On 08/31/2015 11:54 PM, Jes Sorensen wrote:
>> As I mentioned in a direct email - calling writeN from slow_writeN is
>> not a solution. I found some devices fail if you try to use the fast
>> writeN routine. Maybe rtl8xxxu_raw_write32() - without the byteswap
>> would be a better approach, or simply putting the relevant code into the
>> slow_writeN function.
>>
>> I am flying back to New York tomorrow, and I'll have some time on the
>> plane, so I'll try to have a look at that.
>
> I also found that trying to write all N bytes at once failed, but
> writing 64 at a time is OK. I will add the USB calls to slow_writeN.

I see, it's possible we need to add it as a parameter to writeN or add
the size to the fops struct for the different device types. In fact the
latter might be a cleaner approach, especially if it works for 8192/8188
devices with 64 byte writes. Getting rid of slow_writeN altogether would
be nice.

Jes