Messages in this thread | | | Date | Fri, 14 Mar 2014 14:43:11 -0400 (EDT) | Subject | Re: [PATCH net-next v3 1/2] r8152:addRTL8152_EARLY_AGG_TIMEOUT_SUPER | From | David Miller <> |
| |
From: hayeswang <hayeswang@realtek.com> Date: Fri, 14 Mar 2014 15:24:50 +0800
> Besides, I don't wish to modify the setting by ethtool when re-loading > the driver or rebooting every time.
You have code to reset the driver, you can do it when the user asks for the setting to be changed via ethtool. I do not see this as a problem.
The ethtool change can occur while the driver is already up, you'll just need to reset the chip and make the new configuration, this is not a problem.
> Excuse me. Why is it not accepted for being a Kconfig parameter. > It let the manufactuers of some platforms, especially the embedded > system, could tune their performance. Should I give up this patch?
We want to discourage such hard-coding, not encourage it.
I think you are limiting yourself unnecessarily by refusing to consider that there may be ways other than Kconfig to adjust this setting.
| |