Re: [PATCH v2] exfat: fix memory leak in exfat_parse_param()

Re: [PATCH v2] exfat: fix memory leak in exfat_parse_param()

[Markus Elfring]

> Additionally, there's no point copying param->string in
> exfat_parse_param() - just steal it, leaving NULL in param->string.
> That's independent from the leak or fix thereof - it's simply
> avoiding an extra copy.

I find it clearer to provide such a source code adjustment
by a separate update step.
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/submitting-patches.rst?id=d6f9469a03d832dcd17041ed67774ffb5f3e73b3#n138

Please move it into another patch.


…
> +++ b/fs/exfat/super.c
…
> @@ -686,7 +685,12 @@  static int exfat_get_tree(struct fs_context *fc)
>
>  static void exfat_free(struct fs_context *fc)
>  {
> -	kfree(fc->s_fs_info);
> +	struct exfat_sb_info *sbi = fc->s_fs_info;
> +
> +	if (sbi) {
> +		exfat_free_iocharset(sbi);
> +		kfree(sbi);
> +	}
>  }
…

Can it be helpful to annotate the added check according to branch prediction?
Are valid pointers likely at this place?

Regards,
Markus

[PATCH v2] exfat: fix memory leak in exfat_parse_param()

[Namjae Jeon]

From: Al Viro <viro@zeniv.linux.org.uk>

butt3rflyh4ck reported memory leak found by syzkaller.

A param->string held by exfat_mount_options.

BUG: memory leak

unreferenced object 0xffff88801972e090 (size 8):
  comm "syz-executor.2", pid 16298, jiffies 4295172466 (age 14.060s)
  hex dump (first 8 bytes):
    6b 6f 69 38 2d 75 00 00                          koi8-u..
  backtrace:
    [<000000005bfe35d6>] kstrdup+0x36/0x70 mm/util.c:60
    [<0000000018ed3277>] exfat_parse_param+0x160/0x5e0
fs/exfat/super.c:276
    [<000000007680462b>] vfs_parse_fs_param+0x2b4/0x610
fs/fs_context.c:147
    [<0000000097c027f2>] vfs_parse_fs_string+0xe6/0x150
fs/fs_context.c:191
    [<00000000371bf78f>] generic_parse_monolithic+0x16f/0x1f0
fs/fs_context.c:231
    [<000000005ce5eb1b>] do_new_mount fs/namespace.c:2812 [inline]
    [<000000005ce5eb1b>] do_mount+0x12bb/0x1b30 fs/namespace.c:3141
    [<00000000b642040c>] __do_sys_mount fs/namespace.c:3350 [inline]
    [<00000000b642040c>] __se_sys_mount fs/namespace.c:3327 [inline]
    [<00000000b642040c>] __x64_sys_mount+0x18f/0x230 fs/namespace.c:3327
    [<000000003b024e98>] do_syscall_64+0xf6/0x7d0
arch/x86/entry/common.c:295
    [<00000000ce2b698c>] entry_SYSCALL_64_after_hwframe+0x49/0xb3

exfat_free() should call exfat_free_iocharset(), to prevent a leak
in case we fail after parsing iocharset= but before calling
get_tree_bdev().

Additionally, there's no point copying param->string in
exfat_parse_param() - just steal it, leaving NULL in param->string.
That's independent from the leak or fix thereof - it's simply
avoiding an extra copy.

Fixes: 719c1e182916 ("exfat: add super block operations")
Cc: stable@vger.kernel.org # v5.7
Reported-by: butt3rflyh4ck <butterflyhuangxx@gmail.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 v2:
   - update patch description in more detail.

 fs/exfat/super.c | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)

diff --git a/fs/exfat/super.c b/fs/exfat/super.c
index 405717e4e3ea..e650e65536f8 100644
--- a/fs/exfat/super.c
+++ b/fs/exfat/super.c
@@ -273,9 +273,8 @@ static int exfat_parse_param(struct fs_context *fc, struct fs_parameter *param)
 		break;
 	case Opt_charset:
 		exfat_free_iocharset(sbi);
-		opts->iocharset = kstrdup(param->string, GFP_KERNEL);
-		if (!opts->iocharset)
-			return -ENOMEM;
+		opts->iocharset = param->string;
+		param->string = NULL;
 		break;
 	case Opt_errors:
 		opts->errors = result.uint_32;
@@ -686,7 +685,12 @@ static int exfat_get_tree(struct fs_context *fc)
 
 static void exfat_free(struct fs_context *fc)
 {
-	kfree(fc->s_fs_info);
+	struct exfat_sb_info *sbi = fc->s_fs_info;
+
+	if (sbi) {
+		exfat_free_iocharset(sbi);
+		kfree(sbi);
+	}
 }
 
 static const struct fs_context_operations exfat_context_ops = {
-- 
2.17.1