Skip to contentSkip to conversation

xfs_io: don't leak fd in open -Tr failure case

PATCH In review Eric Sandeen 3 participants 4 messages Feb 27, 2014

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 xfs_io: don't leak fd in open -Tr failure case

Coverity spotted this.

It complained that we didn't close the fd before returning in this case of incompatible options, but it seems like we should just test for the incompatible flags before even trying to open the file, no?

(The open would have failed in any case, but with a somewhat cryptic "Invalid argument" - so it's probably better to state it plainly and bail immediately.)

---

io/open.c +5 -5
index 6bb0d46..c106fa7 100644
@@ -342,6 +342,11 @@ open_f(
 	if (optind != argc - 1)
 		return command_usage(&open_cmd);
 
+	if ((flags & (IO_READONLY|IO_TMPFILE)) == (IO_READONLY|IO_TMPFILE)) {
+		fprintf(stderr, _("-T and -r options are incompatible\n"));
+		return -1;
+	}
+
 	fd = openfile(argv[optind], &geometry, flags, mode);
 	if (fd < 0)
 		return 0;
@@ -349,11 +354,6 @@ open_f(
 	if (!platform_test_xfs_fd(fd))
 		flags |= IO_FOREIGN;
 
-	if ((flags & (IO_READONLY|IO_TMPFILE)) == (IO_READONLY|IO_TMPFILE)) {
-		fprintf(stderr, _("-T and -r options are incompatible\n"));
-		return -1;
-	}
-
 	addfile(argv[optind], fd, &geometry, flags);
 	return 0;
 }

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

#1 lore raw
Quoted 10 lines · Eric Sandeen

Coverity spotted this.

It complained that we didn't close the fd before returning in this case of incompatible options, but it seems like we should just test for the incompatible flags before even trying to open the file, no?

(The open would have failed in any case, but with a somewhat cryptic "Invalid argument" - so it's probably better to state it plainly and bail immediately.)

So actually, we wouldn't leak, because the open would fail. So I guess it's not the best subject & description...

-Eric

Quoted 32 lines

---

io/open.c +5 -5
index 6bb0d46..c106fa7 100644
@@ -342,6 +342,11 @@ open_f(
 	if (optind != argc - 1)
 		return command_usage(&open_cmd);
 
+	if ((flags & (IO_READONLY|IO_TMPFILE)) == (IO_READONLY|IO_TMPFILE)) {
+		fprintf(stderr, _("-T and -r options are incompatible\n"));
+		return -1;
+	}
+
 	fd = openfile(argv[optind], &geometry, flags, mode);
 	if (fd < 0)
 		return 0;
@@ -349,11 +354,6 @@ open_f(
 	if (!platform_test_xfs_fd(fd))
 		flags |= IO_FOREIGN;
 
-	if ((flags & (IO_READONLY|IO_TMPFILE)) == (IO_READONLY|IO_TMPFILE)) {
-		fprintf(stderr, _("-T and -r options are incompatible\n"));
-		return -1;
-	}
-
 	addfile(argv[optind], fd, &geometry, flags);
 	return 0;
 }
Quoted 5 lines

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

Reviewed-by
#2 lore raw
Quoted 8 lines · Eric Sandeen
Quoted 7 lines · Eric Sandeen

Coverity spotted this.

It complained that we didn't close the fd before returning in this case of incompatible options, but it seems like we should just test for the incompatible flags before even trying to open the file, no?

Quoted 6 lines
Quoted 3 lines

(The open would have failed in any case, but with a somewhat cryptic "Invalid argument" - so it's probably better to state it plainly and bail immediately.)

So actually, we wouldn't leak, because the open would fail. So I guess it's not the best subject & description...

It's not, but doing the check earlier seems sensible.

(with a better description)

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

#3 lore raw

PATCH v2 1/1 xfs_io: test for invalid -Tr flag combination before open

In reply to Eric Sandeen

Coverity spotted this.

It complained that we didn't close the fd before returning in the error case of incompatible options, but in reality, we wouldn't have gotten that far because open(O_RDONLY|O_TMPFILE) would be rejected with EINVAL.

So the error handling test would never actually be true.

Fix this by moving the error checking prior to the open so the user gets a more useful error message than "Invalid Argument."

---

V2: New summary & commit message

io/open.c +5 -5
index 6bb0d46..c106fa7 100644
@@ -342,6 +342,11 @@ open_f(
 	if (optind != argc - 1)
 		return command_usage(&open_cmd);
 
+	if ((flags & (IO_READONLY|IO_TMPFILE)) == (IO_READONLY|IO_TMPFILE)) {
+		fprintf(stderr, _("-T and -r options are incompatible\n"));
+		return -1;
+	}
+
 	fd = openfile(argv[optind], &geometry, flags, mode);
 	if (fd < 0)
 		return 0;
@@ -349,11 +354,6 @@ open_f(
 	if (!platform_test_xfs_fd(fd))
 		flags |= IO_FOREIGN;
 
-	if ((flags & (IO_READONLY|IO_TMPFILE)) == (IO_READONLY|IO_TMPFILE)) {
-		fprintf(stderr, _("-T and -r options are incompatible\n"));
-		return -1;
-	}
-
 	addfile(argv[optind], fd, &geometry, flags);
 	return 0;
 }

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs