netdev - Re: [PATCH iproute2-next 1/3] tc: p_ip6: Support pedit of IPv6 dsfield

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:   Mon, 30 Mar 2020 10:32:13 +0200
From:   Petr Machata <petrm@...lanox.com>
To:     Stephen Hemminger <stephen@...workplumber.org>
Cc:     netdev@...r.kernel.org, David Ahern <dsahern@...il.com>
Subject: Re: [PATCH iproute2-next 1/3] tc: p_ip6: Support pedit of IPv6 dsfield


Stephen Hemminger <stephen@...workplumber.org> writes:

>> diff --git a/tc/p_ip6.c b/tc/p_ip6.c
>> index 7cc7997b..b6fe81f5 100644
>> --- a/tc/p_ip6.c
>> +++ b/tc/p_ip6.c
>> @@ -56,6 +56,22 @@ parse_ip6(int *argc_p, char ***argv_p,
>>  		res = parse_cmd(&argc, &argv, 4, TU32, 0x0007ffff, sel, tkey);
>>  		goto done;
>>  	}
>> +	if (strcmp(*argv, "traffic_class") == 0 ||
>> +	    strcmp(*argv, "tos") == 0 ||
>> +	    strcmp(*argv, "dsfield") == 0) {
>> +		NEXT_ARG();
>> +		tkey->off = 1;
>> +		res = parse_cmd(&argc, &argv, 1, TU32, RU8, sel, tkey);
>> +
>> +		/* Shift the field by 4 bits on success. */
>> +		if (!res) {
>> +			int nkeys = sel->sel.nkeys;
>> +			struct tc_pedit_key *key = &sel->sel.keys[nkeys - 1];
>> +			key->mask = htonl(ntohl(key->mask) << 4 | 0xf);
>> +			key->val = htonl(ntohl(key->val) << 4);
>> +		}
>> +		goto done;
>> +	}
> Why in the middle of the list?

Because that's the order IPv4 code does them.

> Why three aliases for the same value?
> Since this is new code choose one and make it match what IPv6 standard
> calls that field.

TOS because flower also calls it TOS, even if it's the IPv6 field.
dsfield, because the IPv4 pedit also accepts this. I'm fine with just
accepting traffic_class though.