Hi Ferruh,
On 09/21/2016 06:15 PM, Ferruh Yigit wrote:
> Hi Olivier,
>
> On 9/21/2016 4:12 PM, Olivier Matz wrote:
>> Hi Ferruh,
>>
>> On 09/20/2016 06:17 PM, Ferruh Yigit wrote:
>>> Fixes: ce94a51ff05c ("mempool: add flag for removing phys contiguous
>>> constraint")
>>>
>>> Signed-off-by: Ferruh Yigit <ferruh.yigit at intel.com>
>>> ---
>>> lib/librte_mempool/rte_mempool.c | 4 +++-
>>> lib/librte_mempool/rte_mempool.h | 3 +++
>>> 2 files changed, 6 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/lib/librte_mempool/rte_mempool.c
>>> b/lib/librte_mempool/rte_mempool.c
>>> index e96d14f..d767f39 100644
>>> --- a/lib/librte_mempool/rte_mempool.c
>>> +++ b/lib/librte_mempool/rte_mempool.c
>>> @@ -340,7 +340,9 @@ rte_mempool_free_memchunks(struct rte_mempool *mp)
>>> }
>>>
>>> /* Add objects in the pool, using a physically contiguous memory
>>> - * zone. Return the number of objects added, or a negative value
>>> + * zone if MEMPOOL_F_NO_PHYS_CONTIG flag is not set.
>>> + *
>>> + * Return the number of objects added, or a negative value
>>> * on error.
>>> */
>>> int
>>> diff --git a/lib/librte_mempool/rte_mempool.h
>>> b/lib/librte_mempool/rte_mempool.h
>>> index 6fc331a..291c168 100644
>>> --- a/lib/librte_mempool/rte_mempool.h
>>> +++ b/lib/librte_mempool/rte_mempool.h
>>> @@ -798,6 +798,9 @@ rte_mempool_free(struct rte_mempool *mp);
>>> * Add a virtually and physically contiguous memory chunk in the pool
>>> * where objects can be instanciated.
>>> *
>>> + * mempool flag MEMPOOL_F_NO_PHYS_CONTIG changes behavior of the function
>>> + * and removes physical contiguous constraint.
>>> + *
>>> * @param mp
>>> * A pointer to the mempool structure.
>>> * @param vaddr
>>>
>>
>> Not sure I'm getting why you add these comments, as I don't see any
>> usage of MEMPOOL_F_NO_PHYS_CONTIG in rte_mempool_populate_phys().
>>
>> Could you describe what makes you think the API comments are wrong here?
>
> rte_mempool_populate_phys() now can get RTE_BAD_PHYS_ADDR as "paddr"
> variable, and when paddr == RTE_BAD_PHYS_ADDR, function no more
> quarantines that populated items are physically continuous.
>
> Like how this is used in rte_mempool_populate_phys_tab(), if
> MEMPOOL_F_NO_PHYS_CONTIG is set rte_mempool_populate_phys() called for
> whole virtually continuous memory (len: pg_num * pg_size), so resulting
> items may not be in physically continuous addresses, so function comment
> becomes wrong.
>
> Although MEMPOOL_F_NO_PHYS_CONTIG is not used within the
> rte_mempool_populate_phys(), as far as I can see flag has a side effect
> on the functionality.
>
> And please help on wording if it is not accurate.
OK, so I think we should just talk about paddr in the API comment of
rte_mempool_populate_phys(), not about the flag which is already
documented in mempool_create*().
I suggest the following:
--- a/lib/librte_mempool/rte_mempool.h
+++ b/lib/librte_mempool/rte_mempool.h
@@ -800,6 +800,10 @@ rte_mempool_free(struct rte_mempool *mp);
* Add a virtually and physically contiguous memory chunk in the pool
* where objects can be instanciated.
*
+ * If the given physical address is unknown (paddr = RTE_BAD_PHYS_ADDR),
+ * the chunk doesn't need to be physically contiguous (only virtually),
+ * and allocated objects may span two pages.
+ *
* @param mp
* A pointer to the mempool structure.
* @param vaddr
What do you think?
Regards
Olivier
