From: Hal Rosenstock <hal@dev.mellanox.co.il>
To: Michael Wang <yun.wang@profitbricks.com>
Cc: Roland Dreier <roland@kernel.org>,
Sean Hefty <sean.hefty@intel.com>,
Hal Rosenstock <hal.rosenstock@gmail.com>,
linux-rdma@vger.kernel.org, linux-kernel@vger.kernel.org,
Tom Tucker <tom@opengridcomputing.com>,
Steve Wise <swise@opengridcomputing.com>,
Hoang-Nam Nguyen <hnguyen@de.ibm.com>,
Christoph Raisch <raisch@de.ibm.com>,
Mike Marciniszyn <infinipath@intel.com>,
Eli Cohen <eli@mellanox.com>,
Faisal Latif <faisal.latif@intel.com>,
Jack Morgenstein <jackm@dev.mellanox.co.il>,
Or Gerlitz <ogerlitz@mellanox.com>,
Haggai Eran <haggaie@mellanox.com>,
Ira Weiny <ira.weiny@intel.com>, Tom Talpey <tom@talpey.com>,
Jason Gunthorpe <jgunthorpe@obsidianresearch.com>,
Doug Ledford <dledford@redhat.com>
Subject: Re: [PATCH v4 14/27] IB/Verbs: Reform cma_acquire_dev()
Date: Thu, 16 Apr 2015 09:41:09 -0400 [thread overview]
Message-ID: <552FBBF5.1010605@dev.mellanox.co.il> (raw)
In-Reply-To: <552FBA93.1030304@profitbricks.com>
On 4/16/2015 9:35 AM, Michael Wang wrote:
>
>
> On 04/16/2015 03:19 PM, Hal Rosenstock wrote:
> [snip]
>>>
>>> +static inline int cma_validate_port(struct ib_device *device, u8 port,
>>> + union ib_gid *gid, int dev_type)
>>> +{
>>> + u8 found_port;
>>> + int ret = -ENODEV;
>>> +
>>> + if ((dev_type == ARPHRD_INFINIBAND) && !rdma_tech_ib(device, port))
>>> + return ret;
>>> +
>>> + if ((dev_type != ARPHRD_INFINIBAND) && rdma_tech_ib(device, port))
>>> + return ret;
>>> +
>>> + ret = ib_find_cached_gid(device, gid, &found_port, NULL);
>>> +
>>> + if (!ret && (port == found_port))
>>> + return 0;
>>> +
>>> + return ret;
>>
>> Should the case where ret = 0 and port != found_port need to be handled
>> the same as currently ? It looks different to me since in this case the
>> port will be saved into id_priv->id.port_num whereas currently it isn't.
>
> I get your point :-) what about:
>
> ret = ib_find_cached_gid(device, gid, &found_port, NULL);
> if (port != found_port)
> return -ENODEV;
>
> return ret;
Yes, that looks to me to be consistent with the current implementation.
-- Hal
> Regards,
> Michael Wang
>
>>
>> -- Hal
>>
>>> +}
>>> +
>>> static int cma_acquire_dev(struct rdma_id_private *id_priv,
>>> struct rdma_id_private *listen_id_priv)
>>> {
>>> struct rdma_dev_addr *dev_addr = &id_priv->id.route.addr.dev_addr;
>>> struct cma_device *cma_dev;
>>> - union ib_gid gid, iboe_gid;
>>> + union ib_gid gid, iboe_gid, *gidp;
>>> int ret = -ENODEV;
>>> - u8 port, found_port;
>>> - enum rdma_link_layer dev_ll = dev_addr->dev_type == ARPHRD_INFINIBAND ?
>>> - IB_LINK_LAYER_INFINIBAND : IB_LINK_LAYER_ETHERNET;
>>> + u8 port;
>>>
>>> - if (dev_ll != IB_LINK_LAYER_INFINIBAND &&
>>> + if (dev_addr->dev_type != ARPHRD_INFINIBAND &&
>>> id_priv->id.ps == RDMA_PS_IPOIB)
>>> return -EINVAL;
>>>
>>> @@ -391,41 +409,36 @@ static int cma_acquire_dev(struct rdma_id_private *id_priv,
>>>
>>> memcpy(&gid, dev_addr->src_dev_addr +
>>> rdma_addr_gid_offset(dev_addr), sizeof gid);
>>> - if (listen_id_priv &&
>>> - rdma_port_get_link_layer(listen_id_priv->id.device,
>>> - listen_id_priv->id.port_num) == dev_ll) {
>>> +
>>> + if (listen_id_priv) {
>>> cma_dev = listen_id_priv->cma_dev;
>>> port = listen_id_priv->id.port_num;
>>> - if (rdma_node_get_transport(cma_dev->device->node_type) == RDMA_TRANSPORT_IB &&
>>> - rdma_port_get_link_layer(cma_dev->device, port) == IB_LINK_LAYER_ETHERNET)
>>> - ret = ib_find_cached_gid(cma_dev->device, &iboe_gid,
>>> - &found_port, NULL);
>>> - else
>>> - ret = ib_find_cached_gid(cma_dev->device, &gid,
>>> - &found_port, NULL);
>>> + gidp = rdma_tech_iboe(cma_dev->device, port) ?
>>> + &iboe_gid : &gid;
>>>
>>> - if (!ret && (port == found_port)) {
>>> - id_priv->id.port_num = found_port;
>>> + ret = cma_validate_port(cma_dev->device, port, gidp,
>>> + dev_addr->dev_type);
>>> + if (!ret) {
>>> + id_priv->id.port_num = port;
>>> goto out;
>>> }
>>> }
>>> +
>>> list_for_each_entry(cma_dev, &dev_list, list) {
>>> for (port = 1; port <= cma_dev->device->phys_port_cnt; ++port) {
>>> if (listen_id_priv &&
>>> listen_id_priv->cma_dev == cma_dev &&
>>> listen_id_priv->id.port_num == port)
>>> continue;
>>> - if (rdma_port_get_link_layer(cma_dev->device, port) == dev_ll) {
>>> - if (rdma_node_get_transport(cma_dev->device->node_type) == RDMA_TRANSPORT_IB &&
>>> - rdma_port_get_link_layer(cma_dev->device, port) == IB_LINK_LAYER_ETHERNET)
>>> - ret = ib_find_cached_gid(cma_dev->device, &iboe_gid, &found_port, NULL);
>>> - else
>>> - ret = ib_find_cached_gid(cma_dev->device, &gid, &found_port, NULL);
>>> -
>>> - if (!ret && (port == found_port)) {
>>> - id_priv->id.port_num = found_port;
>>> - goto out;
>>> - }
>>> +
>>> + gidp = rdma_tech_iboe(cma_dev->device, port) ?
>>> + &iboe_gid : &gid;
>>> +
>>> + ret = cma_validate_port(cma_dev->device, port, gidp,
>>> + dev_addr->dev_type);
>>> + if (!ret) {
>>> + id_priv->id.port_num = port;
>>> + goto out;
>>> }
>>> }
>>> }
>>
>
next prev parent reply other threads:[~2015-04-16 13:41 UTC|newest]
Thread overview: 57+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-04-16 8:04 [PATCH v4 00/27] IB/Verbs: IB Management Helpers Michael Wang
2015-04-16 8:04 ` [PATCH v4 01/27] IB/Verbs: Implement new callback query_transport() Michael Wang
2015-04-16 8:05 ` [PATCH v4 02/27] IB/Verbs: Implement raw management helpers Michael Wang
2015-04-16 8:05 ` [PATCH v4 03/27] IB/Verbs: Reform IB-core mad/agent/user_mad Michael Wang
2015-04-16 12:43 ` Hal Rosenstock
2015-04-16 12:51 ` Michael Wang
2015-04-16 8:05 ` [PATCH v4 04/27] IB/Verbs: Reform IB-core cm Michael Wang
2015-04-16 12:43 ` Hal Rosenstock
2015-04-16 12:55 ` Michael Wang
2015-04-16 8:06 ` [PATCH v4 05/27] IB/Verbs: Reform IB-core sa_query Michael Wang
2015-04-16 8:06 ` [PATCH v4 06/27] IB/Verbs: Reform IB-core multicast Michael Wang
2015-04-16 8:07 ` [PATCH v4 07/27] IB/Verbs: Reform IB-ulp ipoib Michael Wang
2015-04-16 8:07 ` [PATCH v4 08/27] IB/Verbs: Reform IB-ulp xprtrdma Michael Wang
2015-04-16 8:07 ` [PATCH v4 09/27] IB/Verbs: Reform IB-core verbs/uverbs_cmd/sysfs Michael Wang
2015-04-16 8:08 ` [PATCH v4 10/27] IB/Verbs: Reform cm related part in IB-core cma/ucm Michael Wang
2015-04-16 13:10 ` Hal Rosenstock
2015-04-16 13:19 ` Michael Wang
2015-04-16 14:31 ` Hefty, Sean
2015-04-16 15:22 ` Michael Wang
2015-04-16 17:21 ` Tom Talpey
2015-04-16 17:30 ` Hefty, Sean
2015-04-17 8:27 ` Michael Wang
2015-04-17 8:11 ` Michael Wang
2015-04-16 15:58 ` Jason Gunthorpe
2015-04-16 16:55 ` Hefty, Sean
2015-04-16 16:57 ` Jason Gunthorpe
2015-04-16 17:38 ` Hal Rosenstock
2015-04-16 17:41 ` Weiny, Ira
2015-04-16 17:48 ` Jason Gunthorpe
2015-04-17 8:36 ` Michael Wang
2015-04-16 8:08 ` [PATCH v4 11/27] IB/Verbs: Reform route related part in IB-core cma Michael Wang
2015-04-16 8:08 ` [PATCH v4 12/27] IB/Verbs: Reform mcast " Michael Wang
2015-04-16 8:09 ` [PATCH v4 13/27] IB/Verbs: Reserve legacy transport type in 'dev_addr' Michael Wang
2015-04-16 8:09 ` [PATCH v4 14/27] IB/Verbs: Reform cma_acquire_dev() Michael Wang
2015-04-16 13:19 ` Hal Rosenstock
2015-04-16 13:35 ` Michael Wang
2015-04-16 13:41 ` Hal Rosenstock [this message]
2015-04-16 8:10 ` [PATCH v4 15/27] IB/Verbs: Reform rest part in IB-core cma Michael Wang
2015-04-16 8:10 ` [PATCH v4 16/27] IB/Verbs: Use management helper cap_ib_mad() Michael Wang
2015-04-16 8:11 ` [PATCH v4 17/27] IB/Verbs: Use management helper cap_ib_smi() Michael Wang
2015-04-16 8:12 ` [PATCH v4 18/27] IB/Verbs: Use management helper cap_ib_cm() Michael Wang
2015-04-16 8:12 ` [PATCH v4 19/27] IB/Verbs: Use management helper cap_iw_cm() Michael Wang
2015-04-16 8:12 ` [PATCH v4 20/27] IB/Verbs: Use management helper cap_ib_sa() Michael Wang
2015-04-16 13:36 ` Hal Rosenstock
2015-04-16 13:38 ` Michael Wang
2015-04-16 8:13 ` [PATCH v4 21/27] IB/Verbs: Use management helper cap_ib_mcast() Michael Wang
2015-04-16 8:13 ` [PATCH v4 22/27] IB/Verbs: Use management helper cap_ipoib() Michael Wang
2015-04-16 8:14 ` [PATCH v4 23/27] IB/Verbs: Use management helper cap_read_multi_sge() Michael Wang
2015-04-16 8:14 ` [PATCH v4 24/27] IB/Verbs: Use management helper cap_af_ib() Michael Wang
2015-04-16 8:15 ` [PATCH v4 25/27] IB/Verbs: Use management helper cap_eth_ah() Michael Wang
2015-04-16 8:15 ` [PATCH v4 26/27] IB/Verbs: Clean up rdma_ib_or_iboe() Michael Wang
2015-04-16 8:15 ` [PATCH v4 27/27] IB/Verbs: Cleanup rdma_node_get_transport() Michael Wang
2015-04-16 13:36 ` Hal Rosenstock
2015-04-16 13:41 ` Michael Wang
2015-04-16 13:42 ` Hal Rosenstock
2015-04-16 13:45 ` Michael Wang
2015-04-16 13:51 ` Tom Tucker
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=552FBBF5.1010605@dev.mellanox.co.il \
--to=hal@dev.mellanox.co.il \
--cc=dledford@redhat.com \
--cc=eli@mellanox.com \
--cc=faisal.latif@intel.com \
--cc=haggaie@mellanox.com \
--cc=hal.rosenstock@gmail.com \
--cc=hnguyen@de.ibm.com \
--cc=infinipath@intel.com \
--cc=ira.weiny@intel.com \
--cc=jackm@dev.mellanox.co.il \
--cc=jgunthorpe@obsidianresearch.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-rdma@vger.kernel.org \
--cc=ogerlitz@mellanox.com \
--cc=raisch@de.ibm.com \
--cc=roland@kernel.org \
--cc=sean.hefty@intel.com \
--cc=swise@opengridcomputing.com \
--cc=tom@opengridcomputing.com \
--cc=tom@talpey.com \
--cc=yun.wang@profitbricks.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
Powered by JetHome