netdev - Re: [iproute2] concern regarding the color option usage in "bridge" & "tc" cmd

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]

Message-ID: <e1cfb620-7c87-ce63-2bb4-6e9b3df0863e@gmail.com>
Date:   Tue, 31 Aug 2021 20:41:01 -0700
From:   David Ahern <dsahern@...il.com>
To:     Gokul Sivakumar <gokulkumar792@...il.com>,
        Stephen Hemminger <stephen@...workplumber.org>,
        David Ahern <dsahern@...nel.org>,
        Roopa Prabhu <roopa@...dia.com>
Cc:     netdev@...r.kernel.org
Subject: Re: [iproute2] concern regarding the color option usage in "bridge" &
 "tc" cmd

On 8/29/21 2:49 AM, Gokul Sivakumar wrote:
> Hi Stephen, David,
> 
> Recently I have added a commit 82149efee9 ("bridge: reorder cmd line arg parsing
> to let "-c" detected as "color" option") in iproute2 tree bridge.c which aligns
> the behaviour of the "bridge" cmd with the "bridge" man page description w.r.t
> the color option usage. Now I have stumbled upon a commit f38e278b8446 ("bridge:
> make -c match -compressvlans first instead of -color") that was added back in
> 2018 which says that "there are apps and network interface managers out there
> that are already using -c to prepresent compressed vlans".
> 
> So after finding the commit f38e278b8446, now I think the man page should have
> fixed instead of changing the bridge.c to align the behaviour of the "bridge" cmd
> with the man page. Do you think we can revert the bridge.c changes 82149efee9,
> so that the "bridge" cmd detects "-c" as "-compressedvlans" instead of "-color"?
> 
> If we are reverting the commit 82149efee9, then "-c" will be detected as
> "-compressedvlans" and I will send out a patch to change the "bridge" man page
> to reflect the new "bridge" cmd behaviour. If we are not reverting the commit
> 82149efee9, then "-c" will be detected as "-color" and I will send a out a patch
> to change the "bridge" cmd help menu to reflect the current "bridge" cmd behaviour.
> Please share your thoughts.
> 
> And also regarding the "tc" cmd, in the man/man8/tc.8 man page, the "-c" option
> is mentioned to be used as a shorthand option for "-color", but instead it is
> detected as "-conf". So here also, we need to decide between fixing the man page
> and fixing the "tc" cmd behaviour w.r.t to color option usage.
> 
> I understand that "matches()" gives a lot of trouble and I see that you both are
> now preferring full "strcmp()" over "matches()" for newly added cmd line options.
> 


Stephen: This should be reverted for 5.14 release given the change in
behavior. I will take a wild guess that ifupdown2 is the interface
manager that will notice.