Skip to contentSkip to conversation

fix resource leak on register_filesystem failure

BKPATCH Will Dyson 1 participant 1 message Sep 24, 2003

Subsystems (2)
lore
Share & tools
lore

Will Dyson

author
#0 lore raw
Quoted 3 lines · Marco Cova noticed a bug in befs's init code. He

It seems there's a (very) rare-case bug in linuxvfs.c:init_befs_fs In the current implementation, if register_filesystem() fails, the memory allocated through befs_init_inodecache() is not deallocated.

He also provided the trivial patch below (modulo my renaming of the goto labels).

# This is a BitKeeper generated patch for the following project:
# Project Name: Linux kernel tree
# This patch format is intended for GNU patch command version 2.5 or higher.
# This patch includes the following deltas:
# ChangeSet 1.1228 -> 1.1230
# fs/befs/linuxvfs.c 1.8 -> 1.10
#
# The following is the BitKeeper ChangeSet Log
# --------------------------------------------
# 03/09/22 will@thalience.(none) 1.1229
# Fix slab memory leak when register_filesystem fails.
# Caught by Marco Cova
# --------------------------------------------
# 03/09/22 will@thalience.(none) 1.1230
# Fixup for previous cset
# --------------------------------------------
#

fs/befs/linuxvfs.c +12 -2
diff -Nru a/fs/befs/linuxvfs.c b/fs/befs/linuxvfs.c
@@ -939,9 +941,19 @@
 
 	err = befs_init_inodecache();
 	if (err)
-		return err;
+		goto unaquire_none;
+
+	err = register_filesystem(&befs_fs_type);
+	if (err)
+		goto unaquire_inodecache;
+
+	return 0;
+
+unaquire_inodecache:
+	befs_destroy_inodecache();
 
-	return register_filesystem(&befs_fs_type);
+unaquire_none:
+	return err;
 }
 
 static void __exit
Signature · 2 lines
Will Dyson
"Back off man, I'm a scientist!" -Dr. Peter Venkman