On Mon, Sep 4, 2017 at 4:32 PM, De Lara Guarch, Pablo
<[email protected]> wrote:
> Hi Jan,
>
>> -----Original Message-----
>> From: dev [mailto:[email protected]] On Behalf Of De Lara Guarch,
>> Pablo
>> Sent: Saturday, July 15, 2017 12:05 PM
>> To: Jan Blunck <[email protected]>; [email protected]
>> Cc: Doherty, Declan <[email protected]>
>> Subject: Re: [dpdk-dev] [PATCH v2 3/4] cryptodev: rework PMD init to not
>> require rte_vdev.h
>>
>> Hi
>>
>> > -----Original Message-----
>> > From: Jan Blunck [mailto:[email protected]] On Behalf Of Jan Blunck
>> > Sent: Wednesday, July 12, 2017 8:59 PM
>> > To: [email protected]
>> > Cc: Doherty, Declan <[email protected]>; De Lara Guarch, Pablo
>> > <[email protected]>
>> > Subject: [PATCH v2 3/4] cryptodev: rework PMD init to not require
>> > rte_vdev.h
>> >
>> > The rte_cryptodev_vdev_pmd_init() is a helper for vdev-based drivers.
>> > By moving the helper to the header we don't require rte_vdev.h at
>> > build- time of the librte_cryptodev library. This is a preparation to
>> > move the vdev bus into a standalone library.
>> >
>> > Signed-off-by: Jan Blunck <[email protected]>
>>
>> I am seeing some clang errors when applying this patch:
>>
>> lib/librte_cryptodev/rte_cryptodev_vdev.h:88:14: error: implicit
>> declaration of function 'rte_cryptodev_pmd_allocate' is invalid in C99 [-
>> Werror,-Wimplicit-function-declaration]
>>         cryptodev = rte_cryptodev_pmd_allocate(name, socket_id);
>>                     ^
>> lib/librte_cryptodev/rte_cryptodev_vdev.h:88:12: error: incompatible
>> integer to pointer conversion assigning to 'struct rte_cryptodev *' from 
>> 'int'
>> [-Werror,-Wint-conversion]
>>         cryptodev = rte_cryptodev_pmd_allocate(name, socket_id);
>>

Pablo,

I can not reproduce this. There is already an include for
rte_cryptodev_pmd_allocate() in rte_cryptodev_vdev.h.

>> Also, looks like git commit title is not correct, according to 
>> check-git-log.sh:
>>
>> Wrong headline format:
>>         cryptodev: rework PMD init to not require rte_vdev.h
>

This script complains about underscores ...

Tell me what you think,
Jan

Reply via email to