[v2,2/4] mempool: add non-IO flag
Checks
Commit Message
From: Dmitry Kozlyuk <dkozlyuk@oss.nvidia.com>
Mempool is a generic allocator that is not necessarily used for device
IO operations and its memory for DMA. Add MEMPOOL_F_NON_IO flag to mark
such mempools.
Signed-off-by: Dmitry Kozlyuk <dkozlyuk@oss.nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
---
doc/guides/rel_notes/release_21_11.rst | 3 +++
lib/mempool/rte_mempool.h | 4 ++++
2 files changed, 7 insertions(+)
Comments
29/09/2021 16:52, dkozlyuk@oss.nvidia.com:
> From: Dmitry Kozlyuk <dkozlyuk@oss.nvidia.com>
>
> Mempool is a generic allocator that is not necessarily used for device
> IO operations and its memory for DMA. Add MEMPOOL_F_NON_IO flag to mark
> such mempools.
>
> Signed-off-by: Dmitry Kozlyuk <dkozlyuk@oss.nvidia.com>
> Acked-by: Matan Azrad <matan@nvidia.com>
> ---
> doc/guides/rel_notes/release_21_11.rst | 3 +++
> lib/mempool/rte_mempool.h | 4 ++++
> 2 files changed, 7 insertions(+)
>
> diff --git a/doc/guides/rel_notes/release_21_11.rst b/doc/guides/rel_notes/release_21_11.rst
> index f85dc99c8b..873beda633 100644
> --- a/doc/guides/rel_notes/release_21_11.rst
> +++ b/doc/guides/rel_notes/release_21_11.rst
> @@ -155,6 +155,9 @@ API Changes
> +* mempool: Added ``MEMPOOL_F_NON_IO`` flag to give a hint to DPDK components
> + that objects from this pool will not be used for device IO (e.g. DMA).
This is not a breaking change, but I am OK to add this note.
Any other opinion?
> +#define MEMPOOL_F_NON_IO 0x0040 /**< Not used for device IO (DMA). */
> + * - MEMPOOL_F_NO_IO: If set, the mempool is considered to be
> + * never used for device IO, i.e. DMA operations,
> + * which may affect some PMD behavior.
Not limited to PMD, it may affect some libs.
I would reword the last line like this:
"No impact on mempool behaviour, but it is a hint for other components."
On 10/5/21 7:39 PM, Thomas Monjalon wrote:
> 29/09/2021 16:52, dkozlyuk@oss.nvidia.com:
>> From: Dmitry Kozlyuk <dkozlyuk@oss.nvidia.com>
>>
>> Mempool is a generic allocator that is not necessarily used for device
>> IO operations and its memory for DMA. Add MEMPOOL_F_NON_IO flag to mark
>> such mempools.
>>
>> Signed-off-by: Dmitry Kozlyuk <dkozlyuk@oss.nvidia.com>
>> Acked-by: Matan Azrad <matan@nvidia.com>
>> ---
>> doc/guides/rel_notes/release_21_11.rst | 3 +++
>> lib/mempool/rte_mempool.h | 4 ++++
>> 2 files changed, 7 insertions(+)
>>
>> diff --git a/doc/guides/rel_notes/release_21_11.rst b/doc/guides/rel_notes/release_21_11.rst
>> index f85dc99c8b..873beda633 100644
>> --- a/doc/guides/rel_notes/release_21_11.rst
>> +++ b/doc/guides/rel_notes/release_21_11.rst
>> @@ -155,6 +155,9 @@ API Changes
>> +* mempool: Added ``MEMPOOL_F_NON_IO`` flag to give a hint to DPDK components
>> + that objects from this pool will not be used for device IO (e.g. DMA).
>
> This is not a breaking change, but I am OK to add this note.
> Any other opinion?
Why is it in API changes section? Shouldn't it be in
"New Features"?
@@ -155,6 +155,9 @@ API Changes
the crypto/security operation. This field will be used to communicate
events such as soft expiry with IPsec in lookaside mode.
+* mempool: Added ``MEMPOOL_F_NON_IO`` flag to give a hint to DPDK components
+ that objects from this pool will not be used for device IO (e.g. DMA).
+
ABI Changes
-----------
@@ -263,6 +263,7 @@ struct rte_mempool {
#define MEMPOOL_F_SC_GET 0x0008 /**< Default get is "single-consumer".*/
#define MEMPOOL_F_POOL_CREATED 0x0010 /**< Internal: pool is created. */
#define MEMPOOL_F_NO_IOVA_CONTIG 0x0020 /**< Don't need IOVA contiguous objs. */
+#define MEMPOOL_F_NON_IO 0x0040 /**< Not used for device IO (DMA). */
/**
* @internal When debug is enabled, store some statistics.
@@ -992,6 +993,9 @@ typedef void (rte_mempool_ctor_t)(struct rte_mempool *, void *);
* "single-consumer". Otherwise, it is "multi-consumers".
* - MEMPOOL_F_NO_IOVA_CONTIG: If set, allocated objects won't
* necessarily be contiguous in IO memory.
+ * - MEMPOOL_F_NO_IO: If set, the mempool is considered to be
+ * never used for device IO, i.e. DMA operations,
+ * which may affect some PMD behavior.
* @return
* The pointer to the new allocated mempool, on success. NULL on error
* with rte_errno set appropriately. Possible rte_errno values include: