svm: add unidirectional message queue

Meant for single reader/writer message exchanges. Supports multiple
message rings.

Change-Id: I925de9a6ae19226c5c39a63caff76424ed123a13
Signed-off-by: Florin Coras <fcoras@cisco.com>

4 files changed, 496 insertions, 3 deletions

diff --git a/src/svm/message_queue.c b/src/svm/message_queue.c
new file mode 100644
index 00000000000..dc0f6251120
--- /dev/null
+++ b/src/svm/message_queue.c
@@ -0,0 +1,159 @@
+/*
+ * Copyright (c) 2018 Cisco and/or its affiliates.
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at:
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#include <svm/message_queue.h>
+#include <vppinfra/mem.h>
+
+svm_msg_q_t *
+svm_msg_q_alloc (svm_msg_q_cfg_t * cfg)
+{
+  svm_msg_q_ring_t *ring;
+  svm_msg_q_t *mq;
+  int i;
+
+  if (!cfg)
+    return 0;
+
+  mq = clib_mem_alloc_aligned (sizeof (svm_msg_q_t), CLIB_CACHE_LINE_BYTES);
+  memset (mq, 0, sizeof (*mq));
+  mq->q = svm_queue_init (cfg->q_nitems, sizeof (svm_msg_q_msg_t),
+			  cfg->consumer_pid, 0);
+  vec_validate (mq->rings, cfg->n_rings - 1);
+  for (i = 0; i < cfg->n_rings; i++)
+    {
+      ring = &mq->rings[i];
+      ring->elsize = cfg->ring_cfgs[i].elsize;
+      ring->nitems = cfg->ring_cfgs[i].nitems;
+      if (cfg->ring_cfgs[i].data)
+	ring->data = cfg->ring_cfgs[i].data;
+      else
+	ring->data = clib_mem_alloc_aligned (ring->nitems * ring->elsize,
+					     CLIB_CACHE_LINE_BYTES);
+    }
+
+  return mq;
+}
+
+void
+svm_msg_q_free (svm_msg_q_t * mq)
+{
+  svm_msg_q_ring_t *ring;
+
+  vec_foreach (ring, mq->rings)
+  {
+    clib_mem_free (ring->data);
+  }
+  vec_free (mq->rings);
+  clib_mem_free (mq);
+}
+
+svm_msg_q_msg_t
+svm_msg_q_alloc_msg (svm_msg_q_t * mq, u32 nbytes)
+{
+  svm_msg_q_msg_t msg = {.as_u64 = ~0 };
+  svm_msg_q_ring_t *ring;
+
+  vec_foreach (ring, mq->rings)
+  {
+    if (ring->elsize < nbytes || ring->cursize == ring->nitems)
+      continue;
+    msg.ring_index = ring - mq->rings;
+    msg.elt_index = ring->tail;
+    ring->tail = (ring->tail + 1) % ring->nitems;
+    __sync_fetch_and_add (&ring->cursize, 1);
+    break;
+  }
+  return msg;
+}
+
+static inline svm_msg_q_ring_t *
+svm_msg_q_get_ring (svm_msg_q_t * mq, u32 ring_index)
+{
+  return vec_elt_at_index (mq->rings, ring_index);
+}
+
+static inline void *
+svm_msg_q_ring_data (svm_msg_q_ring_t * ring, u32 elt_index)
+{
+  ASSERT (elt_index < ring->nitems);
+  return (ring->data + elt_index * ring->elsize);
+}
+
+void *
+svm_msg_q_msg_data (svm_msg_q_t * mq, svm_msg_q_msg_t * msg)
+{
+  svm_msg_q_ring_t *ring = svm_msg_q_get_ring (mq, msg->ring_index);
+  return svm_msg_q_ring_data (ring, msg->elt_index);
+}
+
+void
+svm_msg_q_free_msg (svm_msg_q_t * mq, svm_msg_q_msg_t * msg)
+{
+  svm_msg_q_ring_t *ring;
+
+  if (vec_len (mq->rings) <= msg->ring_index)
+    return;
+  ring = &mq->rings[msg->ring_index];
+  if (msg->elt_index == ring->head)
+    {
+      ring->head = (ring->head + 1) % ring->nitems;
+    }
+  else
+    {
+      /* for now, expect messages to be processed in order */
+      ASSERT (0);
+    }
+  __sync_fetch_and_sub (&ring->cursize, 1);
+}
+
+static int
+svm_msq_q_msg_is_valid (svm_msg_q_t * mq, svm_msg_q_msg_t * msg)
+{
+  svm_msg_q_ring_t *ring;
+  u32 dist1, dist2;
+
+  if (vec_len (mq->rings) <= msg->ring_index)
+    return 0;
+  ring = &mq->rings[msg->ring_index];
+
+  dist1 = ((ring->nitems + msg->ring_index) - ring->head) % ring->nitems;
+  if (ring->tail == ring->head)
+    dist2 = (ring->cursize == 0) ? 0 : ring->nitems;
+  else
+    dist2 = ((ring->nitems + ring->tail) - ring->head) % ring->nitems;
+  return (dist1 < dist2);
+}
+
+int
+svm_msg_q_add (svm_msg_q_t * mq, svm_msg_q_msg_t msg, int nowait)
+{
+  ASSERT (svm_msq_q_msg_is_valid (mq, &msg));
+  return svm_queue_add (mq->q, (u8 *) & msg, nowait);
+}
+
+int
+svm_msg_q_sub (svm_msg_q_t * mq, svm_msg_q_msg_t * msg,
+	       svm_q_conditional_wait_t cond, u32 time)
+{
+  return svm_queue_sub (mq->q, (u8 *) msg, cond, time);
+}
+
+/*
+ * fd.io coding-style-patch-verification: ON
+ *
+ * Local Variables:
+ * eval: (c-set-style "gnu")
+ * End:
+ */
diff --git a/src/svm/message_queue.h b/src/svm/message_queue.h
new file mode 100644
index 00000000000..5ec8547016e
--- /dev/null
+++ b/src/svm/message_queue.h
@@ -0,0 +1,154 @@
+/*
+ * Copyright (c) 2018 Cisco and/or its affiliates.
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at:
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+/**
+ * @file
+ * @brief Unidirectional shared-memory multi-ring message queue
+ */
+
+#ifndef SRC_SVM_MESSAGE_QUEUE_H_
+#define SRC_SVM_MESSAGE_QUEUE_H_
+
+#include <vppinfra/clib.h>
+#include <svm/queue.h>
+
+typedef struct svm_msg_q_ring_
+{
+  volatile u32 cursize;			/**< current size of the ring */
+  u32 nitems;				/**< max size of the ring */
+  u32 head;				/**< current head (for dequeue) */
+  u32 tail;				/**< current tail (for enqueue) */
+  u32 elsize;				/**< size of an element */
+  u8 *data;				/**< chunk of memory for msg data */
+} svm_msg_q_ring_t;
+
+typedef struct svm_msg_q_
+{
+  svm_queue_t *q;			/**< queue for exchanging messages */
+  svm_msg_q_ring_t *rings;		/**< rings with message data*/
+} svm_msg_q_t;
+
+typedef struct svm_msg_q_ring_cfg_
+{
+  u32 nitems;
+  u32 elsize;
+  void *data;
+} svm_msg_q_ring_cfg_t;
+
+typedef struct svm_msg_q_cfg_
+{
+  int consumer_pid;			/**< pid of msg consumer */
+  u32 q_nitems;				/**< msg queue size (not rings) */
+  u32 n_rings;				/**< number of msg rings */
+  svm_msg_q_ring_cfg_t *ring_cfgs;	/**< array of ring cfgs */
+} svm_msg_q_cfg_t;
+
+typedef union
+{
+  struct
+  {
+    u32 ring_index;			/**< ring index, could be u8 */
+    u32 elt_index;			/**< index in ring */
+  };
+  u64 as_u64;
+} svm_msg_q_msg_t;
+
+/**
+ * Allocate message queue
+ *
+ * Allocates a message queue on the heap. Based on the configuration options,
+ * apart from the message queue this also allocates (one or multiple)
+ * shared-memory rings for the messages.
+ *
+ * @param cfg 		configuration options: queue len, consumer pid,
+ * 			ring configs
+ * @return		message queue
+ */
+svm_msg_q_t *svm_msg_q_alloc (svm_msg_q_cfg_t * cfg);
+
+/**
+ * Free message queue
+ *
+ * @param mq		message queue to be freed
+ */
+void svm_msg_q_free (svm_msg_q_t * mq);
+
+/**
+ * Allocate message buffer
+ *
+ * Message is allocated on the first available ring capable of holding
+ * the requested number of bytes.
+ *
+ * @param mq		message queue
+ * @param nbytes	number of bytes needed for message
+ * @return		message structure pointing to the ring and position
+ * 			allocated
+ */
+svm_msg_q_msg_t svm_msg_q_alloc_msg (svm_msg_q_t * mq, u32 nbytes);
+
+/**
+ * Free message buffer
+ *
+ * Marks message buffer on ring as free.
+ *
+ * @param mq		message queue
+ * @param msg		message to be freed
+ */
+void svm_msg_q_free_msg (svm_msg_q_t * mq, svm_msg_q_msg_t * msg);
+/**
+ * Producer enqueue one message to queue
+ *
+ * Prior to calling this, the producer should've obtained a message buffer
+ * from one of the rings by calling @ref svm_msg_q_alloc_msg.
+ *
+ * @param mq		message queue
+ * @param msg		message (pointer to ring position) to be enqueued
+ * @param nowait	flag to indicate if request is blocking or not
+ * @return		success status
+ */
+int svm_msg_q_add (svm_msg_q_t * mq, svm_msg_q_msg_t msg, int nowait);
+
+/**
+ * Consumer dequeue one message from queue
+ *
+ * This returns the message pointing to the data in the message rings.
+ * The consumer is expected to call @ref svm_msg_q_free_msg once it
+ * finishes processing/copies the message data.
+ *
+ * @param mq		message queue
+ * @param msg		pointer to structure where message is to be received
+ * @param cond		flag that indicates if request should block or not
+ * @return		success status
+ */
+int svm_msg_q_sub (svm_msg_q_t * mq, svm_msg_q_msg_t * msg,
+		   svm_q_conditional_wait_t cond, u32 time);
+
+/**
+ * Get data for message in queu
+ *
+ * @param mq		message queue
+ * @param msg		message for which the data is requested
+ * @return		pointer to data
+ */
+void *svm_msg_q_msg_data (svm_msg_q_t * mq, svm_msg_q_msg_t * msg);
+
+#endif /* SRC_SVM_MESSAGE_QUEUE_H_ */
+
+/*
+ * fd.io coding-style-patch-verification: ON
+ *
+ * Local Variables:
+ * eval: (c-set-style "gnu")
+ * End:
+ */
diff --git a/src/svm/ssvm.c b/src/svm/ssvm.c
index 04e0efa3d62..dfb68c039ad 100644
--- a/src/svm/ssvm.c
+++ b/src/svm/ssvm.c
@@ -29,7 +29,6 @@ int
 ssvm_master_init_shm (ssvm_private_t * ssvm)
 {
   int ssvm_fd, mh_flags = MHEAP_FLAG_DISABLE_VM | MHEAP_FLAG_THREAD_SAFE;
-  svm_main_region_t *smr = svm_get_root_rp ()->data_base;
   clib_mem_vm_map_t mapa = { 0 };
   u8 junk = 0, *ssvm_filename;
   ssvm_shared_header_t *sh;
@@ -56,8 +55,13 @@ ssvm_master_init_shm (ssvm_private_t * ssvm)
 
   if (fchmod (ssvm_fd, S_IRUSR | S_IWUSR | S_IRGRP | S_IWGRP) < 0)
     clib_unix_warning ("ssvm segment chmod");
-  if (fchown (ssvm_fd, smr->uid, smr->gid) < 0)
-    clib_unix_warning ("ssvm segment chown");
+  if (svm_get_root_rp ())
+    {
+      /* TODO: is this really needed? */
+      svm_main_region_t *smr = svm_get_root_rp ()->data_base;
+      if (fchown (ssvm_fd, smr->uid, smr->gid) < 0)
+	clib_unix_warning ("ssvm segment chown");
+    }
 
   if (lseek (ssvm_fd, ssvm->ssvm_size, SEEK_SET) < 0)
     {
diff --git a/src/svm/test_svm_message_queue.c b/src/svm/test_svm_message_queue.c
new file mode 100644
index 00000000000..69ffd131ac2
--- /dev/null
+++ b/src/svm/test_svm_message_queue.c
@@ -0,0 +1,176 @@
+/*
+ * Copyright (c) 2018 Cisco and/or its affiliates.
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at:
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#include <svm/ssvm.h>
+#include <svm/message_queue.h>
+
+#define test1_error(_fmt, _args...)			\
+{							\
+    ssvm_pop_heap (oldheap);				\
+    error = clib_error_return (0, _fmt, ##_args);	\
+    goto done;						\
+}
+
+clib_error_t *
+test1 (int verbose)
+{
+  ssvm_private_t _ssvm, *ssvm = &_ssvm;
+  svm_msg_q_cfg_t _cfg, *cfg = &_cfg;
+  svm_msg_q_msg_t msg1, msg2, msg[12];
+  ssvm_shared_header_t *sh;
+  clib_error_t *error = 0;
+  svm_msg_q_t *mq;
+  void *oldheap;
+  int i;
+
+  memset (ssvm, 0, sizeof (*ssvm));
+
+  ssvm->ssvm_size = 1 << 20;
+  ssvm->i_am_master = 1;
+  ssvm->my_pid = getpid ();
+  ssvm->name = format (0, "%s%c", "test", 0);
+  ssvm->requested_va = 0;
+
+  if (ssvm_master_init (ssvm, SSVM_SEGMENT_SHM))
+    return clib_error_return (0, "failed: segment allocation");
+  sh = ssvm->sh;
+
+  svm_msg_q_ring_cfg_t rc[2]= {{8, 8, 0}, {8, 16, 0}};
+  cfg->consumer_pid = ~0;
+  cfg->n_rings = 2;
+  cfg->q_nitems = 16;
+  cfg->ring_cfgs = rc;
+
+  oldheap = ssvm_push_heap (sh);
+  mq = svm_msg_q_alloc (cfg);
+  if (!mq)
+    test1_error ("failed: alloc");
+
+  if (vec_len (mq->rings) != 2)
+      test1_error ("failed: ring allocation");
+
+  msg1 = svm_msg_q_alloc_msg (mq, 8);
+  if (mq->rings[0].cursize != 1
+      || msg1.ring_index != 0
+      || msg1.elt_index != 0)
+    test1_error ("failed: msg alloc1");
+
+  msg2 = svm_msg_q_alloc_msg (mq, 15);
+  if (mq->rings[1].cursize != 1
+      || msg2.ring_index != 1
+      || msg2.elt_index != 0)
+      test1_error ("failed: msg alloc2");
+
+  svm_msg_q_free_msg (mq, &msg1);
+  if (mq->rings[0].cursize != 0)
+    test1_error("failed: free msg");
+
+  for (i = 0; i < 12; i++)
+    {
+      msg[i] = svm_msg_q_alloc_msg (mq, 7);
+      *(u32 *)svm_msg_q_msg_data (mq, &msg[i]) = i;
+    }
+
+  if (mq->rings[0].cursize != 8
+      || mq->rings[1].cursize != 5)
+      test1_error ("failed: msg alloc3");
+
+  *(u32 *)svm_msg_q_msg_data (mq, &msg2) = 123;
+  svm_msg_q_add (mq, msg2, SVM_Q_NOWAIT);
+  for (i = 0; i < 12; i++)
+    svm_msg_q_add (mq, msg[i], SVM_Q_NOWAIT);
+
+  if (svm_msg_q_sub (mq, &msg2, SVM_Q_NOWAIT, 0))
+    test1_error ("failed: dequeue1");
+
+  if (msg2.ring_index != 1 || msg2.elt_index != 0)
+    test1_error ("failed: dequeue1 result");
+  if (*(u32 *)svm_msg_q_msg_data (mq, &msg2) != 123)
+    test1_error ("failed: dequeue1 wrong data");
+
+  svm_msg_q_free_msg (mq, &msg2);
+
+  for (i = 0; i < 12; i++)
+    {
+      if (svm_msg_q_sub (mq, &msg[i], SVM_Q_NOWAIT, 0))
+	test1_error ("failed: dequeue2");
+      if (i < 8)
+	{
+	  if (msg[i].ring_index != 0 || msg[i].elt_index != (i + 1) % 8)
+	    test1_error ("failed: dequeue2 result2");
+	}
+      else
+	{
+	  if (msg[i].ring_index != 1 || msg[i].elt_index != (i - 8) + 1)
+	    test1_error ("failed: dequeue2 result3");
+	}
+      if (*(u32 *)svm_msg_q_msg_data (mq, &msg[i]) != i)
+        test1_error ("failed: dequeue2 wrong data");
+      svm_msg_q_free_msg (mq, &msg[i]);
+    }
+  if (mq->rings[0].cursize != 0 || mq->rings[1].cursize != 0)
+    test1_error ("failed: post dequeue");
+
+  ssvm_pop_heap (oldheap);
+
+done:
+  ssvm_delete (ssvm);
+  return error;
+}
+
+int
+test_svm_message_queue (unformat_input_t * input)
+{
+  clib_error_t *error = 0;
+  int verbose = 0;
+  int test_id = 0;
+
+  while (unformat_check_input (input) != UNFORMAT_END_OF_INPUT)
+    {
+      if (unformat (input, "test1"))
+      	test_id = 1;
+      else
+	{
+	  error = clib_error_create ("unknown input `%U'\n",
+				     format_unformat_error, input);
+	  goto out;
+	}
+    }
+
+  switch (test_id)
+    {
+    case 1:
+      error = test1 (verbose);
+    }
+out:
+  if (error)
+    clib_error_report (error);
+  else
+    clib_warning ("success");
+
+  return 0;
+}
+
+int
+main (int argc, char *argv[])
+{
+  unformat_input_t i;
+  int r;
+
+  unformat_init_command_line (&i, argv);
+  r = test_svm_message_queue (&i);
+  unformat_free (&i);
+  return r;
+}