Always assume SKIP_GONE_DEVS behaviour and kill the flag

...i.e. GET_DEVS == (GET_DEVS|SKIP_GONE_DEVS)

A null pointer dereference in Incremental.c can be triggered by
replugging a disk while the old name is in use.  When mdadm -I is called
on the new disk we fail the call to sysfs_read().  I audited all the
locations that use GET_DEVS and it appears they can tolerate missing a
drive.  So just make SKIP_GONE_DEVS the default behaviour.

Also fix up remaining unchecked usages of the sysfs_read() return value.

Reported-by: Dave Jiang <dave.jiang@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>

1 files changed, 9 insertions, 11 deletions

diff --git a/sysfs.c b/sysfs.c
index ebf9d8a6..17f25673 100644
--- a/sysfs.c
+++ b/sysfs.c
@@ -273,22 +273,20 @@ struct mdinfo *sysfs_read(int fd, int devnum, unsigned long options)
 
 		strcpy(dbase, "block/dev");
 		if (load_sys(fname, buf)) {
+			/* assume this is a stale reference to a hot
+			 * removed device
+			 */
 			free(dev);
-			if (options & SKIP_GONE_DEVS)
-				continue;
-			else
-				goto abort;
+			continue;
 		}
 		sscanf(buf, "%d:%d", &dev->disk.major, &dev->disk.minor);
 
 		/* special case check for block devices that can go 'offline' */
-		if (options & SKIP_GONE_DEVS) {
-			strcpy(dbase, "block/device/state");
-			if (load_sys(fname, buf) == 0 &&
-			    strncmp(buf, "offline", 7) == 0) {
-				free(dev);
-				continue;
-			}
+		strcpy(dbase, "block/device/state");
+		if (load_sys(fname, buf) == 0 &&
+		    strncmp(buf, "offline", 7) == 0) {
+			free(dev);
+			continue;
 		}
 
 		/* finally add this disk to the array */