Thread (18 messages) 18 messages, 5 authors, 2015-07-29

RE: [PATCH 03/10] dpaa_eth: add configurable bpool thresholds

flat view

From: Madalin-Cristian Bucur <hidden>
Date: 2015-07-27 12:54:23
Also in: linuxppc-dev, lkml

-----Original Message-----
From: David Miller [mailto:davem@davemloft.net]
Sent: Monday, July 27, 2015 2:35 AM
To: Bucur Madalin-Cristian-B32716
Cc: joe@perches.com; netdev@vger.kernel.org; linuxppc-
dev@lists.ozlabs.org; linux-kernel@vger.kernel.org; Wood Scott-B07421;
Liberman Igal-B31950; ppc@mindchasers.com; pebolle@tiscali.nl;
joakim.tjernlund@transmode.se
Subject: Re: [PATCH 03/10] dpaa_eth: add configurable bpool thresholds

From: Madalin-Cristian Bucur <redacted>
Date: Fri, 24 Jul 2015 15:49:39 +0000
quoted
quoted
-----Original Message-----
From: Joe Perches [mailto:joe@perches.com]
On Wed, 2015-07-22 at 19:16 +0300, Madalin Bucur wrote:
quoted
Allow the user to tweak the refill threshold and the total number
of buffers in the buffer pool. The provided values are for one CPU.
Any value in making these module parameters instead?
I expect one would (hardly ever) change these to improve some corner
cases then use them with the new values. It may help in the tuning process
but afterwards the bloat to the bootcmd would probably be  a nuisance.
I think these should be controlled by the existing ethtool infrastructure.

Neither the Kconfig mechanism nor module parameters are appropriate, at
all.
The existing ethtool options are for ring based drivers (ethtool -g / -G).
I would not use those as we are not using rings (they do not map well anyway).

We could introduce special options for our non-ring devices but for these
parameters in particular I'd just resort to defines in the code as it's
improbable one would want to change them.

Madalin
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help