Re: [PATCH 5/9] usb: xhci: move ring initialization

["Neronin, Niklas" <niklas.neronin@linux.intel.com>]

On 30/03/2026 11.42, Michal Pecio wrote:
> On Fri, 27 Mar 2026 13:34:36 +0100, Niklas Neronin wrote:
>> Move ring initialization from xhci_ring_alloc() to xhci_ring_init().
>> Call xhci_ring_init() after xhci_ring_alloc();
> 
> This adds more code and more opportunities for bugs.
> Can't ring_alloc() just call ring_init() itself?

Sure, but the naming would not be accurate.
ring_create() or ring_alloc_and_init() would be accurate.

> 
>> in the future it can also be used to re-initialize the ring during
>> resume.
> 
> Yes, but it seems we don't have the opposite problem: there is no
> need to allocate rings but never initialize them.

Technically there is, during ring linking. The "new" ring does not
need to be initialized, as it will be initialized according to the
"dst" rings values.

This is not implemented in this patch series, but I planned to do it
later.

> 
>>
>> Additionally, remove xhci_dbg_trace() from xhci_mem_init(). The
>> command ring's first segment DMA address is now printed during the
>> trace call in xhci_ring_init().
>>
>> This refactoring lays also the groundwork for eventually replacing:
>> * xhci_dbc_ring_init()
>> * xhci_clear_command_ring()
> 
> Or xhci_clear_command_ring() could just call memset() and ring_init(),
> instead of duplicating this sequence in every place which needs it.

xhci_clear_command_ring() clears the command ring specifically.

The idea is to have (naming not set in stone):
ring_alloc() <- allocate necessary ring memory
ring_free()  <- free all ring memory
ring_init()  <- initialize necessary ring memory
ring_reset() <- resets the ring

The reader can then easily understand what each function does.

>  
>>
>> Signed-off-by: Niklas Neronin <niklas.neronin@linux.intel.com>