Re: [PATCH net-next v3 2/3] net: macb: add safeguards for jumbo frame larger than 10240

[Paolo Abeni]

On 3/16/26 7:26 PM, Charles Perry wrote:
> On Mon, Mar 16, 2026 at 06:21:38PM +0100, Théo Lebrun wrote:
>> Hello Charles,
>>
>> On Fri Mar 13, 2026 at 3:06 PM CET, Charles Perry wrote:
>>> The RX buffers for GEM can have a maximum size of 16320 bytes
>>> (0xff in the RXBS field of the DMACFG register means 255*64 =
>>> 16320 bytes).
>>>
>>> The GEM IP has configurable maximum jumbo frame length that can go up to
>>> 16383. The actual value for this limit can be found in the
>>>        "jumbo_max_length" field (bits 0..13) of the DCFG2 register.
>>> Currently, the macb driver doesn't use the DCFG2 register when
>>> determining the max MTU, instead an hardcoded value (jumbo_max_len in
>>> struct macb_config) is used for each platform. Right now the maximum
>>> value for jumbo_max_len is 10240 (0x2800).
>>
>> If DCFG2 contains the value then we can runtime detect it. With that, we
>> could make the macb_config->jumbo_max_len attribute optional. Then
>> start dropping it from platforms where we know we can trust the DCFG2
>> value.
>>
> 
> Hello Théo,
> 
> That would be a good idea. We could use "jumbo_max_len == 0" as a way to
> signal that the DCFG2 register should be used for determining the max MTU.
> 
> However, that's a new feature and it doesn't belong in this patch. All I
> want to do in this patchset is put the real value of jumbo_max_length in
> the PIC64-HPSC macb_config and make sure the driver doesn't overflow when
> that's used.

FWIW, I agree that is better suited for a follow-up than for the initial
bring-up.

/P