Re: [PATCH] net: cxgb: Fix skb reference leak in espibug_workaround_t204()
From: Wentao Liang
Date: Sun Nov 08 2026 - 05:35:08 EST
Hi Jakub,
I am very sorry for the late reply, and for the flood of patches that
made it necessary. You are right on both points, and I apologize for the
extra work and confusion this caused.
You are right that I sent far more patches at once than anyone could
review, and more than I could follow up on. The hard limit of 15 patches
per person, and the guidance to stay at 8 or fewer when the changes are
of low value, exist so that the feedback on one posting can be applied to
the next. I did not respect that, and that is on me.
I understand why you asked for a one-month pause on new patches from
@iscas.ac.cn for drivers/net and net/. I have also gone back to the
kernel documentation on subsystem identification, patch set organization
and testing, and on how this kind of work should be submitted, and I will
follow it.
To be clear about how these patches are produced: we use static analysis
tools to find candidate issues, and then DeepSeek and GPT-6.1 Sol to
audit the candidates and help prepare the fixes. I review every change
and sign it off myself, and I take responsibility for anything I send.
This has been an important lesson. I am going through the remaining
feedback one item at a time by hand, and replying to each one promptly.
Going forward I will not send patches in bulk, and I will keep every
posting well within what I can review and answer. I have also shared
your message with my colleagues, as you asked.
The cxgb patch is a small reference leak fix in an old T2 driver. A v2
that also covers the single-port timer callback with the same pattern is
ready; I will send it separately as one patch rather than in this reply.
If you would rather I drop it, please tell me and I will.
Sorry again, and thank you for your patience.
Wentao