On Tue, Apr 22, 2025 at 11:19 AM Pavel Begunkov asml.silence@gmail.com wrote:
On 4/22/25 14:56, Mina Almasry wrote:
On Tue, Apr 22, 2025 at 1:43 AM Pavel Begunkov asml.silence@gmail.com wrote:
On 4/18/25 00:15, Mina Almasry wrote:
Currently net_iovs support only pp ref counts, and do not support a page ref equivalent.
Makes me wonder why it's needed. In theory, nobody should ever be taking page references without going through struct ubuf_info handling first, all in kernel users of these pages should always be paired with ubuf_info, as it's user memory, it's not stable, and without ubuf_info the user is allowed to overwrite it.
The concern about the stability of the from-userspace data is already called out in the MSG_ZEROCOPY documentation that we're piggybacking devmem TX onto:
Sure, I didn't object that. There is no problem as long as the ubuf_info semantics is followed, which by extension mean that any ref manipulation should already be gated on ubuf_info, and there should be no need in changing generic paths.
I'm sorry I'm not following. skb_frag_ref is how the net stack obtains references on an skb_frag, regardless on whether the frag is a MSG_ZEROCOPY one with ubuf info, or a regular tx frag without a ubuf_info, or even an io_uring frag which I think have the msg->ubuf_info like we discussed previously. I don't see the net stack in the current code special casing how it obtains refs on frags, and I don't see the need to add special casing. Can you elaborate in more detail what is the gating you expect, and why? Are you asking that I check the skb has a ubuf_info before allowing to grab the reference on the dmabuf binding? Or something else?