ipc: add warnings about not using IPC with memory API

IPC and memory-related API's should not be mixed because memory
relies on IPC internally. Add explicit warnings to IPC API and
to the documentation about this.

Cc: stable@dpdk.org
Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>

diff --git a/doc/guides/prog_guide/env_abstraction_layer.rst b/doc/guides/prog_guide/env_abstraction_layer.rst
index c27f730..f15bcd9 100644 (file)
--- a/doc/guides/prog_guide/env_abstraction_layer.rst
+++ b/doc/guides/prog_guide/env_abstraction_layer.rst
@@ -147,6 +147,14 @@ A default validator callback is provided by EAL, which can be enabled with a
 ``--socket-limit`` command-line option, for a simple way to limit maximum amount
 of memory that can be used by DPDK application.
 
+.. warning::
+    Memory subsystem uses DPDK IPC internally, so memory allocations/callbacks
+    and IPC must not be mixed: it is not safe to allocate/free memory inside
+    memory-related or IPC callbacks, and it is not safe to use IPC inside
+    memory-related callbacks. See chapter
+    :ref:`Multi-process Support <Multi-process_Support>` for more details about
+    DPDK IPC.
+
 + Legacy memory mode
 
 This mode is enabled by specifying ``--legacy-mem`` command-line switch to the

diff --git a/doc/guides/prog_guide/multi_proc_support.rst b/doc/guides/prog_guide/multi_proc_support.rst
index 9660979..63d08b6 100644 (file)
--- a/doc/guides/prog_guide/multi_proc_support.rst
+++ b/doc/guides/prog_guide/multi_proc_support.rst
@@ -318,6 +318,11 @@ supported. However, since sending messages (not requests) does not involve an
 IPC thread, sending messages while processing another message or request is
 supported.
 
+Since the memory sybsystem uses IPC internally, memory allocations and IPC must
+not be mixed: it is not safe to use IPC inside a memory-related callback, nor is
+it safe to allocate/free memory inside IPC callbacks. Attempting to do so may
+lead to a deadlock.
+
 Asynchronous request callbacks may be triggered either from IPC thread or from
 interrupt thread, depending on whether the request has timed out. It is
 therefore suggested to avoid waiting for interrupt-based events (such as alarms)

diff --git a/lib/librte_eal/common/include/rte_eal.h b/lib/librte_eal/common/include/rte_eal.h
index 0603eaf..7db0225 100644 (file)
--- a/lib/librte_eal/common/include/rte_eal.h
+++ b/lib/librte_eal/common/include/rte_eal.h
@@ -225,6 +225,8 @@ struct rte_mp_reply {
  *
  * As we create  socket channel for primary/secondary communication, use
  * this function typedef to register action for coming messages.
+ *
+ * @note No memory allocations should take place inside the callback.
  */
 typedef int (*rte_mp_t)(const struct rte_mp_msg *msg, const void *peer);
 
@@ -234,6 +236,8 @@ typedef int (*rte_mp_t)(const struct rte_mp_msg *msg, const void *peer);
  * As we create socket channel for primary/secondary communication, use
  * this function typedef to register action for coming responses to asynchronous
  * requests.
+ *
+ * @note No memory allocations should take place inside the callback.
  */
 typedef int (*rte_mp_async_reply_t)(const struct rte_mp_msg *request,
                const struct rte_mp_reply *reply);
@@ -308,6 +312,9 @@ rte_mp_sendmsg(struct rte_mp_msg *msg);
  *
  * @note The caller is responsible to free reply->replies.
  *
+ * @note This API must not be used inside memory-related or IPC callbacks, and
+ *   no memory allocations should take place inside such callback.
+ *
  * @param req
  *   The req argument contains the customized request message.
  *