[Xen-changelog] [xen-unstable] tapaio check return value from read()

[Xen patchbot-unstable <patchbot-unstable@xxxxxxxxxxxxxxxxxxx>]

# HG changeset patch
# User Keir Fraser <keir.fraser@xxxxxxxxxx>
# Date 1198758538 0
# Node ID 003542d9ab7713ffcecb7a91a2639083b7da4af5
# Parent  9fe92a88912b45c0c44184d7e3b16983bf5aed47
tapaio check return value from read()

In tools/blktap/drivers/tapaio.c there is a call to read(2) whose
return value is not checked.  The attached patch attempts to do
something vaguely sensible in cases of error.

Fully comprehensive error handling in this area would be quite tough
to introduce now but at least with this change when things go wrong
you stand a chance of getting some information about what happened.

Signed-off-by: Ian Jackson <ian.jackson@xxxxxxxxxxxxx>
---
 tools/blktap/drivers/tapaio.c |   20 ++++++++++++++++++--
 1 files changed, 18 insertions(+), 2 deletions(-)

diff -r 9fe92a88912b -r 003542d9ab77 tools/blktap/drivers/tapaio.c
--- a/tools/blktap/drivers/tapaio.c     Thu Dec 27 12:27:34 2007 +0000
+++ b/tools/blktap/drivers/tapaio.c     Thu Dec 27 12:28:58 2007 +0000
@@ -30,6 +30,8 @@
 #include "tapaio.h"
 #include "tapdisk.h"
 #include <unistd.h>
+#include <errno.h>
+#include <string.h>
 
 /**
  * We used a kernel patch to return an fd associated with the AIO context
@@ -149,8 +151,22 @@ tap_aio_get_events(tap_aio_context_t *ct
         if (!ctx->poll_in_thread)
                 nr_events = io_getevents(ctx->aio_ctx, 1,
                                          ctx->max_aio_events, ctx->aio_events, 
NULL);
-        else
-                read(ctx->completion_fd[0], &nr_events, sizeof(nr_events));
+        else {
+               int r;
+               r = read(ctx->completion_fd[0], &nr_events, sizeof(nr_events));
+               if (r < 0) {
+                       if (errno == EAGAIN || errno == EINTR)
+                               return 0;
+                       /* This is pretty bad, we'll probably spin */
+                       DPRINTF("Aargh, read completion_fd failed: %s",
+                               strerror(errno));
+               } else if (r != sizeof(nr_events)) {
+                       /* Should never happen because sizeof(nr_events)
+                        * fits in the guaranteed atomic pipe write size.
+                        * Blundering on is slightly nicer than asserting */
+                       DPRINTF("Aargh, read completion_fd short read %d", r);
+               }
+       }
 
         return nr_events;
 }

_______________________________________________
Xen-changelog mailing list
Xen-changelog@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-changelog