From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ben Hutchings Subject: Re: [RFC PATCH 1/2] net: ethtool: Add capability to retrieve plug-in module EEPROM Date: Thu, 12 Apr 2012 19:41:58 +0100 Message-ID: <1334256118.2497.9.camel@bwh-desktop.uk.solarflarecom.com> References: <4F71FE0A.3060203@solarflare.com> <1333389160.2623.30.camel@bwh-desktop.uk.solarflarecom.com> <4F85B639.5090900@solarflare.com> <1334168217.2552.5.camel@bwh-desktop.uk.solarflarecom.com> <1334187728.2552.7.camel@bwh-desktop.uk.solarflarecom.com> <4F869DF0.1000709@solarflare.com> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: , , , , , , , To: Stuart Hodgson Return-path: In-Reply-To: <4F869DF0.1000709@solarflare.com> Sender: linux-kernel-owner@vger.kernel.org List-Id: netdev.vger.kernel.org On Thu, 2012-04-12 at 10:18 +0100, Stuart Hodgson wrote: > On 12/04/12 00:42, Ben Hutchings wrote: [...] > > Maybe we should start by refactoring ethtool_get_eeprom() so we can > > reuse most of its code in ethtool_get_module_eeprom(), rather than > > having to worry about what the maximum size of a module EEPROM might be > > and whether we need a loop: > > > > Subject: ethtool: Split ethtool_get_eeprom() to allow for additional EEPROM accessors [...] > > @@ -771,7 +770,7 @@ static int ethtool_get_eeprom(struct net_device *dev, void __user *useraddr) > > return -EINVAL; > > > > /* Check for exceeding total eeprom len */ > > - if (eeprom.offset + eeprom.len> ops->get_eeprom_len(dev)) > > + if (eeprom.offset + eeprom.len> total_len) > > return -EINVAL; > > > > data = kmalloc(PAGE_SIZE, GFP_USER); > > Should this not be eeprom.len? [...] No, because this function loops over PAGE_SIZE chunks. That's presumably necessary for large NIC EEPROMs, and if we reuse it we won't have to wory about whether it's ever necessary for large module EEPROMs. Ben. -- Ben Hutchings, Staff Engineer, Solarflare Not speaking for my employer; that's the marketing department's job. They asked us to note that Solarflare product names are trademarked.