Re: [PATCH bpf-next v2] bpf: clarify batch lookup semantics

[Yonghong Song <yonghong.song@linux.dev>]

On 2/21/24 1:38 PM, Martin Kelly wrote:
> On 2/21/24 13:18, Martin Kelly wrote:
>> The batch lookup and lookup_and_delete APIs have two parameters,
>> in_batch and out_batch, to facilitate iterative
>> lookup/lookup_and_deletion operations for supported maps. Except NULL
>> for in_batch at the start of these two batch operations, both parameters
>> need to point to memory equal or larger than the respective map key
>> size, except for various hashmaps (hash, percpu_hash, lru_hash,
>> lru_percpu_hash) where the in_batch/out_batch memory size should be
>> at least 4 bytes.
>>
>> Document these semantics to clarify the API.
>>
>> Signed-off-by: Martin Kelly <martin.kelly@crowdstrike.com>
>> ---
>>   include/uapi/linux/bpf.h       |  6 +++++-
>>   tools/include/uapi/linux/bpf.h |  6 +++++-
>>   tools/lib/bpf/bpf.h            | 17 ++++++++++++-----
>>   3 files changed, 22 insertions(+), 7 deletions(-)
>
> Yonghong, looks like I missed your comment to change from "clarify 
> batch lookup semantics" to "Clarify batch lookup/lookup_and_delete 
> semantics"; sorry about that. Feel free to change it if you merge 
> this, or I can include it in a v3 if needed.

Ok, LGTM except the subject. I guess it is up to maintainers who will either change the subject
or asking for another revision if more change is needed. From my part,

Acked-by: Yonghong Song <yonghong.song@linux.dev>

>
>>
>> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
>> index d96708380e52..d2e6c5fcec01 100644
>> --- a/include/uapi/linux/bpf.h
>> +++ b/include/uapi/linux/bpf.h
>> @@ -617,7 +617,11 @@ union bpf_iter_link_info {
>>    *        to NULL to begin the batched operation. After each 
>> subsequent
>>    *        **BPF_MAP_LOOKUP_BATCH**, the caller should pass the 
>> resultant
>>    *        *out_batch* as the *in_batch* for the next operation to
>> - *        continue iteration from the current point.
>> + *        continue iteration from the current point. Both *in_batch* 
>> and
>> + *        *out_batch* must point to memory large enough to hold a key,
>> + *        except for maps of type **BPF_MAP_TYPE_{HASH, PERCPU_HASH,
>> + *        LRU_HASH, LRU_PERCPU_HASH}**, for which batch parameters
>> + *        must be at least 4 bytes wide regardless of key size.
>>    *
>>    *        The *keys* and *values* are output parameters which must 
>> point
>>    *        to memory large enough to hold *count* items based on the 
>> key
>> diff --git a/tools/include/uapi/linux/bpf.h 
>> b/tools/include/uapi/linux/bpf.h
>> index d96708380e52..d2e6c5fcec01 100644
>> --- a/tools/include/uapi/linux/bpf.h
>> +++ b/tools/include/uapi/linux/bpf.h
>> @@ -617,7 +617,11 @@ union bpf_iter_link_info {
>>    *        to NULL to begin the batched operation. After each 
>> subsequent
>>    *        **BPF_MAP_LOOKUP_BATCH**, the caller should pass the 
>> resultant
>>    *        *out_batch* as the *in_batch* for the next operation to
>> - *        continue iteration from the current point.
>> + *        continue iteration from the current point. Both *in_batch* 
>> and
>> + *        *out_batch* must point to memory large enough to hold a key,
>> + *        except for maps of type **BPF_MAP_TYPE_{HASH, PERCPU_HASH,
>> + *        LRU_HASH, LRU_PERCPU_HASH}**, for which batch parameters
>> + *        must be at least 4 bytes wide regardless of key size.
>>    *
>>    *        The *keys* and *values* are output parameters which must 
>> point
>>    *        to memory large enough to hold *count* items based on the 
>> key
>> diff --git a/tools/lib/bpf/bpf.h b/tools/lib/bpf/bpf.h
>> index ab2570d28aec..df0db2f0cdb7 100644
>> --- a/tools/lib/bpf/bpf.h
>> +++ b/tools/lib/bpf/bpf.h
>> @@ -190,10 +190,14 @@ LIBBPF_API int bpf_map_delete_batch(int fd, 
>> const void *keys,
>>   /**
>>    * @brief **bpf_map_lookup_batch()** allows for batch lookup of BPF 
>> map elements.
>>    *
>> - * The parameter *in_batch* is the address of the first element in 
>> the batch to read.
>> - * *out_batch* is an output parameter that should be passed as 
>> *in_batch* to subsequent
>> - * calls to **bpf_map_lookup_batch()**. NULL can be passed for 
>> *in_batch* to indicate
>> - * that the batched lookup starts from the beginning of the map.
>> + * The parameter *in_batch* is the address of the first element in 
>> the batch to
>> + * read. *out_batch* is an output parameter that should be passed as 
>> *in_batch*
>> + * to subsequent calls to **bpf_map_lookup_batch()**. NULL can be 
>> passed for
>> + * *in_batch* to indicate that the batched lookup starts from the 
>> beginning of
>> + * the map. Both *in_batch* and *out_batch* must point to memory 
>> large enough to
>> + * hold a single key, except for maps of type **BPF_MAP_TYPE_{HASH, 
>> PERCPU_HASH,
>> + * LRU_HASH, LRU_PERCPU_HASH}**, for which the memory size must be at
>> + * least 4 bytes wide regardless of key size.
>>    *
>>    * The *keys* and *values* are output parameters which must point 
>> to memory large enough to
>>    * hold *count* items based on the key and value size of the map 
>> *map_fd*. The *keys*
>> @@ -226,7 +230,10 @@ LIBBPF_API int bpf_map_lookup_batch(int fd, void 
>> *in_batch, void *out_batch,
>>    *
>>    * @param fd BPF map file descriptor
>>    * @param in_batch address of the first element in batch to read, 
>> can pass NULL to
>> - * get address of the first element in *out_batch*
>> + * get address of the first element in *out_batch*. If not NULL, 
>> must be large
>> + * enough to hold a key. For **BPF_MAP_TYPE_{HASH, PERCPU_HASH, 
>> LRU_HASH,
>> + * LRU_PERCPU_HASH}**, the memory size must be at least 4 bytes wide 
>> regardless
>> + * of key size.
>>    * @param out_batch output parameter that should be passed to next 
>> call as *in_batch*
>>    * @param keys pointer to an array of *count* keys
>>    * @param values pointer to an array large enough for *count* values