netdev - Re: [PATCH v5 net-next 4/4] rtnetlink: return EINVAL when request cannot succeed

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: <b8b90d39-9052-c150-18f6-a482686db7b8@wifirst.fr>
Date:   Tue, 19 Apr 2022 09:29:40 +0200
From:   Florent Fourcot <florent.fourcot@...irst.fr>
To:     Stephen Hemminger <stephen@...workplumber.org>
Cc:     netdev@...r.kernel.org, cong.wang@...edance.com,
        edumazet@...gle.com, Brian Baboch <brian.baboch@...irst.fr>
Subject: Re: [PATCH v5 net-next 4/4] rtnetlink: return EINVAL when request
 cannot succeed

>> -		return -ENODEV;
>> +		return -EINVAL;
> 
> Sometimes changing errno can be viewed as ABI change and break applications.

I agree, but I think this one is OK. __rtnl_newlink function has more 
than 20 return, I don't see how an application can have behavior based 
on this specific path.

And actually, patch 1/4 is protecting almost all previous callers, this 
return is now only here for calls without ifindex/ifname/ifgroup, and 
NLM_F_CREATE not set.

If you think that this change can be merged, I can add extack error at 
this place. EINVAL is indeed very hard to parse.

-- 
Florent Fourcot.