Re: xhci_hcd / ASMedia ASM4242: Bulk-OUT -EPROTO with Logitec 0789:0308 during DVD+RW recording

On Thu, 10 Sep 2026 20:38:43 +0900, wakasio wrote:

Sorry, I do not know why the attachment did not go through. Pasting
the full log below instead.

Thanks. Here's the repeating error event:

TRB 00000000fffbf8e0 status 'USB Transaction Error' len 23552 slot 1 ep 4 type 'Transfer Event' flags e:c

In three runs so far, it always shows 23KB remaining out of 32KB and
that's the only consistent pattern. It happened quite far from any
segment boundary this time, practically in the middle of a segment.

My patch logged TRB addresses wrong and added a zero in the middle:

fffbf08b0: Buffer 00000000fffbd000 length 31 TD size 0 intr 0 type 'Normal' flags b:i:I:c:s:i:e:C
fffbf08c0: Buffer 00000000ff620000 length 32768 TD size 0 intr 0 type 'Normal' flags b:i:I:c:s:i:e:C
fffbf08d0: Buffer 00000000fffbd000 length 31 TD size 0 intr 0 type 'Normal' flags b:i:I:c:s:i:e:C
fffbf08e0: Buffer 00000000ff628000 length 32768 TD size 0 intr 0 type 'Normal' flags b:i:I:c:s:i:e:C

Nothing special is happening here, it's just a single 32KB buffer,
aligned to 32KB, not crossing any 64KB boundary. TD size is 0 as it
should be for the last (and only) TRB of a TD, chain bit is clear,
only IOC and cycle are set. Next TRB's cycle still isn't set:

fffbf08f0: Buffer 00000000ff670000 length 32768 TD size 0 intr 0 type 'Normal' flags b:i:I:c:s:i:e:c

So I don't know what's wrong with it, I think it's a HW bug happening
for no obvious reason. Retries aren't helping for 100ms, so the HW
seems to be stuck for good in some invalid state.

And it only happens with one combination of host and device, so we
can't even be sure which side is wrong. Maybe the most sensible
solution is to simply stop using this device with this host and keep
watching for firmware upgrades.

Regards,
Michal