From: Abhinav Jain <jain.abhinav177@gmail.com>
To: leon@kernel.org
Cc: jain.abhinav177@gmail.com, jgg@ziepe.ca,
linux-kernel@vger.kernel.org, linux-rdma@vger.kernel.org
Subject: Re: [PATCH] RDMA/core: Publish node GUID with the uevent for ib_device
Date: Mon, 10 Mar 2025 07:01:56 +0000 [thread overview]
Message-ID: <20250310070156.8068-1-jain.abhinav177@gmail.com> (raw)
In-Reply-To: <20250309192751.GA7027@unreal>
On Sun, 9 Mar 2025 21:27:51 +0200, Leon Romanovsky wrote:
>On Sun, Mar 09, 2025 at 05:57:31PM +0000, Abhinav Jain wrote:
>> As per the comment, modify ib_device_uevent to publish the node
>> GUID alongside device name, upon device state change.
>>
>> Have compiled the file manually to ensure that it builds. Do not have
>> a readily available IB hardware to test. Confirmed with checkpatch
>> that the patch has no errors/warnings.
>
>I'm missing motivation for this patch. Why is this change needed?
>
>Thanks
Originally, I was looking at this function in order to solve a syzkaller
bug. I noticed this comment from Jason and I assumed that the motivation
would be to identify the node on which the event is happening.
With the name, users can identify nodes however Subnet Manager uses
node_guid for discovery and configuration of the nodes. To conclude, I
think just using the node name might not be sufficient for unambiguous
and reliable device management in the network.
>>
>> Signed-off-by: Abhinav Jain <jain.abhinav177@gmail.com>
>> ---
>> drivers/infiniband/core/device.c | 13 +++++++++----
>> 1 file changed, 9 insertions(+), 4 deletions(-)
>>
>> diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c
>> index 0ded91f056f3..1812038f1a91 100644
>> --- a/drivers/infiniband/core/device.c
>> +++ b/drivers/infiniband/core/device.c
>> @@ -499,12 +499,17 @@ static void ib_device_release(struct device *device)
>> static int ib_device_uevent(const struct device *device,
>> struct kobj_uevent_env *env)
>> {
>> - if (add_uevent_var(env, "NAME=%s", dev_name(device)))
>> + const struct ib_device *dev =
>> + container_of(device, struct ib_device, dev);
>> +
>> + if (add_uevent_var(env, "NAME=%s", dev_name(&dev->dev)))
>> return -ENOMEM;
>>
>> - /*
>> - * It would be nice to pass the node GUID with the event...
>> - */
>> + __be64 node_guid_be = dev->node_guid;
>> + u64 node_guid = be64_to_cpu(node_guid_be);
>> +
>> + if (add_uevent_var(env, "NODE_GUID=0x%llx", node_guid))
>> + return -ENOMEM;
>>
>> return 0;
>> }
>> --
>> 2.34.1
>>
next prev parent reply other threads:[~2025-03-10 7:02 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-03-09 17:57 Abhinav Jain
2025-03-09 19:27 ` Leon Romanovsky
2025-03-10 7:01 ` Abhinav Jain [this message]
2025-03-10 10:17 ` Leon Romanovsky
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=20250310070156.8068-1-jain.abhinav177@gmail.com \
--to=jain.abhinav177@gmail.com \
--cc=jgg@ziepe.ca \
--cc=leon@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-rdma@vger.kernel.org \
/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
all inboxes | Powered by JetHome®