On 2/21/2024 4:15 AM, Alexander Lobakin wrote:
From: Alan Brady <[email protected]>
Date: Tue, 20 Feb 2024 16:49:40 -0800
This starts refactoring how virtchnl messages are handled by adding a
transaction manager (idpf_vc_xn_manager).
[...]
+/**
+ * struct idpf_vc_xn_params - Parameters for executing transaction
+ * @send_buf: kvec for send buffer
+ * @recv_buf: kvec for recv buffer, may be NULL, must then have zero length
+ * @timeout_ms: timeout to wait for reply
+ * @async: send message asynchronously, will not wait on completion
+ * @async_handler: If sent asynchronously, optional callback handler. The user
+ * must be careful when using async handlers as the memory for
+ * the recv_buf _cannot_ be on stack if this is async.
+ * @vc_op: virtchnl op to send
+ */
+struct idpf_vc_xn_params {
+ struct kvec send_buf;
+ struct kvec recv_buf;
+ int timeout_ms;
+ bool async;
+ async_vc_cb async_handler;
+ u32 vc_op;
+};
Sorry for not noticing this before, but this struct can be local to
idpf_virtchnl.c.
Nice catch, I can definitely move this. I'm also considering though, all
of these structs I'm adding here are better suited in idpf_virtchnl.c
all together. I think the main thing preventing that is the
idpf_vc_xn_manager field in idpf_adapter. Would it be overkill to make
the field in idpf_adapter a pointer so I can forward declare and kalloc
it? I think I can then move everything to idpf_virtchnl.c. Or do you see
a better alternative? Or is it not worth the effort? Thanks!
+
+/**
+ * struct idpf_vc_xn_manager - Manager for tracking transactions
+ * @ring: backing and lookup for transactions
+ * @free_xn_bm: bitmap for free transactions
+ * @xn_bm_lock: make bitmap access synchronous where necessary
+ * @salt: used to make cookie unique every message
+ */
[...]
Thanks,
Olek
