linux-kernel - Re: question about samsung/sxgbe/sxgbe

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Tue, 19 Jan 2016 14:36:28 -0500 (EST)
From:	David Miller <davem@...emloft.net>
To:	julia.lawall@...6.fr
Cc:	shuahkhan@...il.com, bh74.an@...sung.com, ks.giri@...sung.com,
	vipul.pandya@...sung.com, netdev@...r.kernel.org,
	linux-kernel@...r.kernel.org, shuahkh@....samsung.com
Subject: Re: question about samsung/sxgbe/sxgbe_xpcs.c

From: Julia Lawall <julia.lawall@...6.fr>
Date: Tue, 19 Jan 2016 19:54:20 +0100 (CET)

> On Tue, 19 Jan 2016, Shuah Khan wrote:
> 
>> 2016-01-19 9:26 GMT-07:00 Julia Lawall <julia.lawall@...6.fr>:
>> > Hello,
>> >
>> > I was windering what is the purpose of the file
>> > drivers/net/ethernet/samsung/sxgbe/sxgbe_xpcs.c
>>
>> From git log and comments, it appears this driver is for
>> Samsung SoCs.
>>
>> >
>> > I can't find anything that refers to it, and if I remove it from the
>> > makefile:
>> >
>> > @@ -1,4 +1,4 @@
>> >  obj-$(CONFIG_SXGBE_ETH) += samsung-sxgbe.o
>> >  samsung-sxgbe-objs:= sxgbe_platform.o sxgbe_main.o sxgbe_desc.o \
>> >                 sxgbe_dma.o sxgbe_core.o sxgbe_mtl.o  sxgbe_mdio.o \
>> > -               sxgbe_ethtool.o sxgbe_xpcs.o $(samsung-sxgbe-y)
>> > +               sxgbe_ethtool.o $(samsung-sxgbe-y)
>> >
>> > and then make clean and recompile, I see no bad effect.
>> >
>>
>> What is the motivation for the experiment to remove it?
>> This file probably gets used only on Samsung SoCs.
> 
> I just wondered.  I was looking at dependencies between networking files.
> This one stands out because nothing is dependenton it, even the files it
> is compiled with, and it doesn't contain the usual functions,
> register_netdev, etc.

Even with that explanation, this is a bogus situation.

There are no in-tree callers of this code.  It should be removed until there
are in-tree users.

Nobody can figure out if the interface for this is done properly without seeing
the call sites and how they work.  It is therefore impossible to review this
code and judge it's design.

If someone doesn't send me a removal patch, I will remove this code myself.