Re: [PATCH v4 0/5] Add larger page size support for USB audio offload path

[Wesley Cheng <wesley.cheng@oss.qualcomm.com>]

On 10/1/2026 3:33 PM, Mathias Nyman wrote:
> On 10/1/26 01:23, Wesley Cheng wrote:
>>
>> Hi Mathias,
>>
>> Just checking to see if there was anything that we needed to account for 
>> in this series?  I think there was the discussion that came up here:
>> https://lore.kernel.org/all/ 
>> TYUPR06MB6217000B59003EDF233D7246D2B22@TYUPR06MB6217.apcprd06.prod.outlook.com/
> 
> Patch v4 1/5: Looks good, is useful even as a standalone patch. Can be 
> taken immediately.
> Patch v4 2/5: Looks good, not useful as standalone patch, can be taken with 
> rest of series.
> Patch v4 3/5: I need to revisit this and look at new dma pool lifetime vs 
> xHC accessing those DMA addresses.
> PAtch v4 Rest:  ALSA, no opinion.
> 

Great, thanks for the inputs.

>>
>> But, I think your suggestion about adding the sideband check in the 
>> xhci_setup_device() would mitigate that problem.  Esp, if we don't want 
>> xHCI core to really know about sideband ops.
> 
> That would be the bandaid in:
> https://lore.kernel.org/all/c6c9890f-46d8-42f1- 
> a7e7-8a5d551fa22c@linux.intel.com/
> 
> I can turn that into a proper patch.
> Lianqin also gave some preliminary positive feedback, but I'm not sure if 
> the code was
> ever properly tested.
> 

I can chime in a bit here.  I did work with Lianqin offline, and I was able 
to reproduce the problem (w/ this series included in the codebase), and 
your suggestion did work.  I provided some feedback to them, so hopefully 
they can convey that on the other thread.

Thanks
Wesley Cheng