c->btree_roots_known[i].b can be NULL. In this case, a NULL pointer dereference
occurs, so you need to add code to check the variable.

Reported-by: syzbot+b468b9fef56949c3b528@syzkaller.appspotmail.com
Fixes: 7773df19c35f ("bcachefs: metadata version bucket_stripe_sectors")
Signed-off-by: Jeongjun Park <aha310510@gmail.com>
---
 fs/bcachefs/sb-downgrade.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/fs/bcachefs/sb-downgrade.c b/fs/bcachefs/sb-downgrade.c
index ae715ff658e8..8767c33c2b51 100644
--- a/fs/bcachefs/sb-downgrade.c
+++ b/fs/bcachefs/sb-downgrade.c
@@ -143,6 +143,9 @@ UPGRADE_TABLE()
 
 static int have_stripes(struct bch_fs *c)
 {
+	if (IS_ERR_OR_NULL(c->btree_roots_known[BTREE_ID_stripes].b))
+		return 0;
+
 	return !btree_node_fake(c->btree_roots_known[BTREE_ID_stripes].b);
 }
 
--

On Fri, Oct 25, 2024 at 08:56:18PM +0900, Jeongjun Park wrote:
> c->btree_roots_known[i].b can be NULL. In this case, a NULL pointer dereference
> occurs, so you need to add code to check the variable.
> 
> Reported-by: syzbot+b468b9fef56949c3b528@syzkaller.appspotmail.com
> Fixes: 7773df19c35f ("bcachefs: metadata version bucket_stripe_sectors")
> Signed-off-by: Jeongjun Park <aha310510@gmail.com>

This looks identical to the v1? It's already in my testing branch

(But it should be in my hotfix branch, doing that now)

> ---
>  fs/bcachefs/sb-downgrade.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/fs/bcachefs/sb-downgrade.c b/fs/bcachefs/sb-downgrade.c
> index ae715ff658e8..8767c33c2b51 100644
> --- a/fs/bcachefs/sb-downgrade.c
> +++ b/fs/bcachefs/sb-downgrade.c
> @@ -143,6 +143,9 @@ UPGRADE_TABLE()
>  
>  static int have_stripes(struct bch_fs *c)
>  {
> +	if (IS_ERR_OR_NULL(c->btree_roots_known[BTREE_ID_stripes].b))
> +		return 0;
> +
>  	return !btree_node_fake(c->btree_roots_known[BTREE_ID_stripes].b);
>  }
>  
> --

On Oct 26, 2024, at 00:54, Kent Overstreet <kent.overstreet@linux.dev> wrote:
> 
> On Fri, Oct 25, 2024 at 08:56:18PM +0900, Jeongjun Park wrote:
>> c->btree_roots_known[i].b can be NULL. In this case, a NULL pointer dereference
>> occurs, so you need to add code to check the variable.
>> 
>> Reported-by: syzbot+b468b9fef56949c3b528@syzkaller.appspotmail.com
>> Fixes: 7773df19c35f ("bcachefs: metadata version bucket_stripe_sectors")
>> Signed-off-by: Jeongjun Park <aha310510@gmail.com>
> 
> This looks identical to the v1? It's already in my testing branch

This version fix the “Fixes" tag, the original one is:

"Fixes: ("bcachefs: metadata version bucket_stripe_sectors”)"

> 
> (But it should be in my hotfix branch, doing that now)
> 
>> ---
>> fs/bcachefs/sb-downgrade.c | 3 +++
>> 1 file changed, 3 insertions(+)
>> 
>> diff --git a/fs/bcachefs/sb-downgrade.c b/fs/bcachefs/sb-downgrade.c
>> index ae715ff658e8..8767c33c2b51 100644
>> --- a/fs/bcachefs/sb-downgrade.c
>> +++ b/fs/bcachefs/sb-downgrade.c
>> @@ -143,6 +143,9 @@ UPGRADE_TABLE()
>> 
>> static int have_stripes(struct bch_fs *c)
>> {
>> + if (IS_ERR_OR_NULL(c->btree_roots_known[BTREE_ID_stripes].b))
>> + return 0;
>> +
>> return !btree_node_fake(c->btree_roots_known[BTREE_ID_stripes].b);
>> }
>> 
>> --
> 

On Sat, Oct 26, 2024 at 01:05:50AM +0800, Alan Huang wrote:
> On Oct 26, 2024, at 00:54, Kent Overstreet <kent.overstreet@linux.dev> wrote:
> > 
> > On Fri, Oct 25, 2024 at 08:56:18PM +0900, Jeongjun Park wrote:
> >> c->btree_roots_known[i].b can be NULL. In this case, a NULL pointer dereference
> >> occurs, so you need to add code to check the variable.
> >> 
> >> Reported-by: syzbot+b468b9fef56949c3b528@syzkaller.appspotmail.com
> >> Fixes: 7773df19c35f ("bcachefs: metadata version bucket_stripe_sectors")
> >> Signed-off-by: Jeongjun Park <aha310510@gmail.com>
> > 
> > This looks identical to the v1? It's already in my testing branch
> 
> This version fix the “Fixes" tag, the original one is:
> 
> "Fixes: ("bcachefs: metadata version bucket_stripe_sectors”)"

thanks