Re: [PATCH v2 04/16] mm/slab: introduce slab_alloc_context

[Vlastimil Babka (SUSE)]

On 6/12/26 05:10, Hao Li wrote:
> On Wed, Jun 10, 2026 at 05:40:06PM +0200, Vlastimil Babka (SUSE) wrote:
>> @@ -5389,13 +5401,18 @@ void *_kmalloc_nolock_noprof(DECL_TOKEN_PARAMS(size, token), gfp_t gfp_flags, in
>>  	if (ret)
>>  		goto success;
>>  
>> +	struct slab_alloc_context ac = {
>> +		.caller_addr = _RET_IP_,
>> +		.orig_size = orig_size,
>> +	};
> 
> It might be better to move this to the beginning of the function, to avoid
> patch09 jump to `success` before ac is initialized.

Hm right, didn't compilers actually complain about goto skipping over
declarations? But neither gcc nor clang do for me, hm. Will move, thanks.

>> +
>>  	/*
>>  	 * Do not call slab_alloc_node(), since trylock mode isn't
>>  	 * compatible with slab_pre_alloc_hook/should_failslab and
>>  	 * kfence_alloc. Hence call __slab_alloc_node() (at most twice)
>>  	 * and slab_post_alloc_hook() directly.
>>  	 */
>> -	ret = __slab_alloc_node(s, alloc_gfp, node, _RET_IP_, orig_size);
>> +	ret = __slab_alloc_node(s, alloc_gfp, node, &ac);
>>  
>>  	/*
>>  	 * It's possible we failed due to trylock as we preempted someone with
>> @@ -7237,10 +7254,13 @@ static bool __kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags,
>>  	int i;
>>  
>>  	if (IS_ENABLED(CONFIG_SLUB_TINY) || kmem_cache_debug(s)) {
>> +		struct slab_alloc_context ac = {
>> +			.caller_addr = _RET_IP_,
>> +			.orig_size = s->object_size,
>> +		};
>>  		for (i = 0; i < size; i++) {
>>  
>> -			p[i] = ___slab_alloc(s, flags, NUMA_NO_NODE, _RET_IP_,
>> -					     s->object_size);
>> +			p[i] = ___slab_alloc(s, flags, NUMA_NO_NODE, &ac);
>>  			if (unlikely(!p[i]))
>>  				goto error;
>>  
>> 
>> -- 
>> 2.54.0
>>