Re: [PATCH 1/2] bcachefs: Using vmalloc_array() for array space allocation

[Liao Yuanhong <liaoyuanhong@vivo.com>]

On 8/11/2025 8:50 PM, Alan Huang wrote:
> [You don't often get email from mmpgouride@gmail.com. Learn why this is important at https://aka.ms/LearnAboutSenderIdentification ]
>
> On Aug 11, 2025, at 20:09, Liao Yuanhong <liaoyuanhong@vivo.com> wrote:
>> Replace kmalloc(size * sizeof) with kmalloc_array() for safer memory
>> allocation and overflow prevention.
>>
>> Signed-off-by: Liao Yuanhong <liaoyuanhong@vivo.com>
>> ---
>> fs/bcachefs/btree_key_cache.c    | 2 +-
>> fs/bcachefs/btree_trans_commit.c | 2 +-
>> 2 files changed, 2 insertions(+), 2 deletions(-)
>>
>> diff --git a/fs/bcachefs/btree_key_cache.c b/fs/bcachefs/btree_key_cache.c
>> index 4890cbc88e7c..8dd70024e513 100644
>> --- a/fs/bcachefs/btree_key_cache.c
>> +++ b/fs/bcachefs/btree_key_cache.c
>> @@ -136,7 +136,7 @@ static struct bkey_cached *__bkey_cached_alloc(unsigned key_u64s, gfp_t gfp)
>> struct bkey_cached *ck = kmem_cache_zalloc(bch2_key_cache, gfp);
>> if (unlikely(!ck))
>> return NULL;
>> - ck->k = kmalloc(key_u64s * sizeof(u64), gfp);
>> + ck->k = kmalloc_array(key_u64s, sizeof(u64), gfp);
>> if (unlikely(!ck->k)) {
>> kmem_cache_free(bch2_key_cache, ck);
>> return NULL;
>> diff --git a/fs/bcachefs/btree_trans_commit.c b/fs/bcachefs/btree_trans_commit.c
>> index 4d58bdb233e9..4102a3cb2410 100644
>> --- a/fs/bcachefs/btree_trans_commit.c
>> +++ b/fs/bcachefs/btree_trans_commit.c
>> @@ -396,7 +396,7 @@ btree_key_can_insert_cached_slowpath(struct btree_trans *trans, unsigned flags,
>> bch2_trans_unlock_updates_write(trans);
>> bch2_trans_unlock(trans);
>>
>> - new_k = kmalloc(new_u64s * sizeof(u64), GFP_KERNEL);
>> + new_k = kmalloc_array(new_u64s, sizeof(u64), GFP_KERNEL);
> No, it’s not an array.

Yes, there is a problem with this patch. Please ignore it.

Thanks,

Liao

>> if (!new_k) {
>> struct bch_fs *c = trans->c;
>> bch_err(c, "error allocating memory for key cache key, btree %s u64s %u",
>> --
>> 2.34.1
>>
>>