Thread (41 messages) read the whole thread 41 messages, 6 authors, 2017-09-04

Re: [PATCH 1/2] eventdev: add event adapter for ethernet Rx queues

From: Jerin Jacob <hidden>
Date: 2017-08-11 09:49:33

-----Original Message-----
Date: Fri, 11 Aug 2017 10:55:38 +0530
From: "Rao, Nikhil" <redacted>
To: Jerin Jacob <redacted>
CC: gage.eads@intel.com, dev@dpdk.org, thomas@monjalon.net,
 bruce.richardson@intel.com, harry.van.haaren@intel.com,
 hemant.agrawal@nxp.com, nipun.gupta@nxp.com, narender.vangati@intel.com,
 Abhinandan Gujjar [off-list ref]
Subject: Re: [PATCH 1/2] eventdev: add event adapter for ethernet Rx queues
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:52.0) Gecko/20100101
 Thunderbird/52.2.1

On 8/1/2017 10:12 PM, Jerin Jacob wrote:
quoted
quoted
I have had the application specify the name. so that it can call

struct rte_service_spec *rte_service_get_by_name(const char *name);
OK. I was thinking like there will be fixed name for the rx_adapter
service like "eth_rx_adapter_%d"(%d for adapter id) which be exposed in
rte_event_eth_rx_adapter.h.
It can help in removing cfg.service_name. I am fine with either way.
OK, I will use a fixed name.
quoted
quoted
quoted
eventdev PMD op api(not as adapter PMD as discussed earlier)
-------------------

1) typedef uint64_t (*eventdev_rx_adap_capa)(struct rte_eventdev *dev,  uint8_t ethdev_id)

Return the adapter capability of a given eventdev when it needs to
connected to a specific ethdev_id
Doesn't the capability of a <eventdev, ethdev> also need to be made
available to the application as an adapter API ?
Yes. Make sense to expose as adapter API also.
Correcting myself, Isn't this really an eventdev API ? An adapter API would
take an adapter id argument which is unnecessary in this case.
Yes, it is eventdev API.
Nikhil
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help