From: Roland Dreier <roland@topspin.com>
To: Greg KH <greg@kroah.com>
Cc: netdev@oss.sgi.com, linux-kernel@vger.kernel.org,
openib-general@openib.org
Subject: Re: [PATCH][RFC/v1][8/12] Add IPoIB (IP-over-InfiniBand) driver
Date: Mon, 22 Nov 2004 15:18:07 -0800 [thread overview]
Message-ID: <52k6sdbhhs.fsf@topspin.com> (raw)
In-Reply-To: <20041122223432.GC15634@kroah.com> (Greg KH's message of "Mon, 22 Nov 2004 14:34:32 -0800")
Greg> What's wrong with using the dev_printk() and friends instead
Greg> of your own?
dev_printk expects a struct device, not a net_device.
Greg> And why cast a pointer in a macro, don't you know the type
Greg> of it anyway?
this lets us pass in the return value of netdev_priv() directly
without having to have the cast in the code that uses the macro.
Greg> You're using a separate filesystem to export debug data?
Greg> I'm all for new virtual filesystems, but why not just use
Greg> sysfs for this? What are you doing in here that you can't
Greg> do with another mechanism (netlink, sysfs, sockets, relayfs,
Greg> etc.)?
For each multicast group, we want to export the GID, how long it's
been around, whether our join has completed and whether it's
send-only. It wouldn't be too bad to create a kobject with all those
attributes but getting the info from so many little files is a little
bit of a pain, and so is dealing with kobject lifetime rules. It's
even worse with netlink since then a new tool is required. (AFAIK
relayfs isn't in Linus's kernel).
It's nice to be able to tell someone to just mount ipoib_debugfs and
send the contents of debugfs/ib0_mcg.
The actual filesystem stuff is pretty trivial using everything libfs
provides for us now...
Greg> Why not just use 2 different debug variables for this?
No real reason... I'll fix it up.
>> + +int mcast_debug_level;
Greg> Global?
Good point, I'll move it into ipoib_multicast.c.
- R.
prev parent reply other threads:[~2004-11-22 23:18 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <20041122713.cSeT4UFKGqJDdZ8T@topspin.com>
2004-11-22 15:13 ` [PATCH][RFC/v1][6/12] IPoIB IPv4 multicast Roland Dreier
2004-11-22 15:13 ` [PATCH][RFC/v1][7/12] IPoIB IPv6 support Roland Dreier
2004-11-22 15:14 ` *****SPAM***** [PATCH][RFC/v1][8/12] Add IPoIB (IP-over-InfiniBand) driver Roland Dreier
2004-11-22 19:40 ` Sam Ravnborg
2004-11-22 22:34 ` Greg KH
2004-11-22 23:18 ` Roland Dreier [this message]
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=52k6sdbhhs.fsf@topspin.com \
--to=roland@topspin.com \
--cc=greg@kroah.com \
--cc=linux-kernel@vger.kernel.org \
--cc=netdev@oss.sgi.com \
--cc=openib-general@openib.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;
as well as URLs for NNTP newsgroup(s).