[dpdk-dev] [PATCH] mempool: fix mempool virt populate with small chunks

Burakov, Anatoly anatoly.burakov at intel.com
Thu Jan 9 15:29:08 CET 2020


On 09-Jan-20 2:23 PM, Olivier Matz wrote:
> On Thu, Jan 09, 2020 at 01:52:41PM +0000, Burakov, Anatoly wrote:
>> On 09-Jan-20 1:27 PM, Olivier Matz wrote:
>>> To populate a mempool with a virtual area, the mempool code calls
>>> rte_mempool_populate_iova() for each iova-contiguous area. It happens
>>> (rarely) that this area is too small to store one object. In this case,
>>> rte_mempool_populate_iova() returns an error, which is forwarded by
>>> rte_mempool_populate_virt().
>>>
>>> This case should not throw an error in
>>> rte_mempool_populate_virt(). Instead, the area that is too small should
>>> just be ignored.
>>>
>>> To fix this issue, change the return value of
>>> rte_mempool_populate_iova() to -ENOBUFS when no object can be populated,
>>> so it can be ignored by the caller. As this would be an API change, add
>>> a compat wrapper to keep the current API unchanged. The wrapper will be
>>> removed for 20.11.
>>>
>>> Fixes: 354788b60cfd ("mempool: allow populating with unaligned virtual area")
>>> Cc: stable at dpdk.org
>>>
>>> Signed-off-by: Olivier Matz <olivier.matz at 6wind.com>
>>> ---
>>>
>>
>> The approach fixes the issue on my end, so
>>
>> Tested-by: Anatoly Burakov <anatoly.burakov at intel.com>
>>
>>> Is there a simple way to ensure that we won't forget to remove the
>>> wrapper for 20.11? Anatoly suggested me to use versioned symbols, but
>>> it's not clear to me how.
>>>
>>
>> Yes, i'd like to do better than "ah shur we won't forget pinky swear".
>>
>> Can't we do this with ABI versioning? E.g.
>>
>> rte_populate_iova_v20() ... returns EINVAL
>>
>> rte_populate_iova_v21() ... returns ENOBUFS
>>
>> I'm pretty sure, even if it doesn't break, it will still be more likely to
>> not be forgotten because there's almost a guarantee that someone will grep
>> for symbol versioning macros across the codebase around 20.11 timeframe.
> 
> Without using symbol versionning, would this be ok too?
> 
>    int
>    rte_mempool_populate_iova(struct rte_mempool *mp, char *vaddr,
>           rte_iova_t iova, size_t len, rte_mempool_memchunk_free_cb_t *free_cb,
>           void *opaque)
>    {
>           int ret;
> 
>           ret = __rte_mempool_populate_iova(mp, vaddr, iova, len, free_cb, opaque);
> 
>    #if RTE_VERSION < RTE_VERSION_NUM(20, 11, 0, 0)
>           if (ret == -ENOBUFS)
>                   ret = -EINVAL;
>    #endif
> 
>           return ret;
>    }
> 
> 

Well it would certainly work :) it just makes it more likely that this 
will be missed.

How about, we leave your patch as is, and then you submit another patch 
marked for [20.11] and mark it as Deferred straight away? This is 
probably the best method to not forget that i can think of, if you're so 
averse to symbol versioning :D

> 
>> -- 
>> Thanks,
>> Anatoly


-- 
Thanks,
Anatoly


More information about the dev mailing list