Re: [PATCH resend v3] bcachefs: Fix NULL pointer dereference in bch2_opt_to_text

From: Hongbo Li
Date: Mon Oct 07 2024 - 22:23:20 EST




On 2024/10/5 21:02, Mohammed Anees wrote:
This patch adds a bounds check to the bch2_opt_to_text function to prevent
NULL pointer dereferences when accessing the opt->choices array. This
ensures that the index used is within valid bounds before dereferencing.
The new version enhances the readability.

Reported-and-tested-by: syzbot+37186860aa7812b331d5@xxxxxxxxxxxxxxxxxxxxxxxxx
Closes: https://syzkaller.appspot.com/bug?extid=37186860aa7812b331d5
Signed-off-by: Mohammed Anees <pvmohammedanees2003@xxxxxxxxx>
---
v3:
- Moved bounds check above as default case.
- Removed the nesterd if clauses.
---
fs/bcachefs/opts.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/fs/bcachefs/opts.c b/fs/bcachefs/opts.c
index 232be8a44051..84097235eea9 100644
--- a/fs/bcachefs/opts.c
+++ b/fs/bcachefs/opts.c
@@ -427,7 +427,9 @@ void bch2_opt_to_text(struct printbuf *out,
prt_printf(out, "%lli", v);
break;
case BCH_OPT_STR:
- if (flags & OPT_SHOW_FULL_LIST)
+ if (v < opt->min || v >= opt->max - 1)
As definition, max is the right bound for value. opt->max - 1 is valid. May be you should remove the equals sign.

It should be no problem to just solve this null pointer issue() (How is this stack triggered, may be I lost something.), but this should be a general boundary check condition, strictly speaking, it's a check for the validity of the options' value.

Thanks,
Hongbo

+ prt_printf(out, "(invalid option %lli)", v);
+ else if (flags & OPT_SHOW_FULL_LIST)
prt_string_option(out, opt->choices, v);
else
prt_str(out, opt->choices[v]);