lad: Add Timeout waiting for Response FIFO

Add a timeout to the logic which opens the client response FIFO
in the LAD daemon during LAD_CONNECT. There are some corner cases
that might result in this hanging forever waiting for the response
FIFO, in which case, all other LAD communication will be blocked
forever.

Such cases might be:
- Client application crashes after sending LAD_connect request
- Communication with LAD through the LADCMDS fifo takes too long
  and the Client times out waiting for the response FIFO (5 seconds).
  If this happens, Client will never try to open the FIFO again due
  to failing when doing the stat() on the FIFO.

With this patch, if the FIFO connection is not established within the
timeout, the FIFO will be unlinked and the original state restored so
that the connection can be attempted again.

The timeout is chosen as 1 second so as not to block all other LAD
communication for a long period of time. 1 second should (hopefully) be
sufficient time for the Client and LAD to both open the FIFO.

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>

2 files changed, 30 insertions, 0 deletions

diff --git a/linux/include/_lad.h b/linux/include/_lad.h
index 4bdfd8e..0124608 100644
--- a/linux/include/_lad.h
+++ b/linux/include/_lad.h
@@ -147,6 +147,7 @@ extern struct timeval start_tv;
 #define LAD_MAXNUMCLIENTS  32      /* max simultaneous clients */
 #define LAD_CONNECTTIMEOUT 5.0  /* LAD connect response timeout (sec) */
 #define LAD_DISCONNECTTIMEOUT   5.0  /* LAD disconnect timeout (sec) */
+#define LAD_RESPONSEFIFOTIMEOUT 1.0  /* LAD timeout for opening the response fifo */
 #define LAD_MAXLENGTHFIFONAME   128  /* max length client FIFO name */
 #define LAD_MAXLENGTHCOMMAND    512  /* size limit for LAD command string */
 #define LAD_MAXLENGTHRESPONSE   512  /* size limit for LAD response string */
diff --git a/linux/src/daemon/lad.c b/linux/src/daemon/lad.c
index 758e098..4160400 100644
--- a/linux/src/daemon/lad.c
+++ b/linux/src/daemon/lad.c
@@ -47,6 +47,7 @@
 #include <signal.h>
 #include <unistd.h>
 #include <dirent.h>
+#include <time.h>
 
 #include <ti/ipc/MessageQ.h>
 #include <_MessageQ.h>
@@ -1065,6 +1066,11 @@ static Int connectToLAD(String clientName, Int pid, String clientProto, Int *cli
     Int clientId = -1;
     Bool connectDenied = FALSE;
     Int status = LAD_SUCCESS;
+    time_t currentTime;
+    time_t startTime;
+    struct stat statBuf;
+    double delta = 0;
+    int filePtrFd = 0;
     FILE * filePtr;
     Int statusIO;
 
@@ -1133,7 +1139,30 @@ openResponseFIFO:
     /* set FIFO permissions to read/write */
     chmod(clientName, 0666);
 
+    startTime = time ((time_t *) 0);
+    while (delta <= LAD_RESPONSEFIFOTIMEOUT) {
+        /* open a file for writing to FIFO, non-blocking so we can timeout */
+        filePtrFd = open(clientName, O_WRONLY | O_TRUNC | O_NONBLOCK);
+        if (filePtrFd != -1) {
+            break;
+        }
+        LOG1("\nWARN: Client %s has not yet opened, will retry\n", clientName)
+        usleep(100);
+        currentTime = time ((time_t *) 0);
+        delta = difftime(currentTime, startTime);
+    }
+
+    if (delta > LAD_RESPONSEFIFOTIMEOUT) {
+        LOG1("\nERROR: timed out waiting for Client to open response FIFO %s\n", clientName);
+        unlink(clientName);
+        status = LAD_IOFAILURE;
+        goto doneconnect;
+    }
+
     filePtr = fopen(clientName, "w");
+
+    close(filePtrFd);
+
     if (filePtr == NULL) {
         LOG1("\nERROR: unable to open response FIFO %s\n", clientName)