linux-kernel - Re: [PATCH net-next] tools: ynl-gen: use big-endian netlink attribute types

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: <376db1bb-8a00-40e5-bf70-f8587d460372@fiberby.net>
Date: Sat, 14 Sep 2024 19:03:38 +0000
From: Asbjørn Sloth Tønnesen <ast@...erby.net>
To: Jakub Kicinski <kuba@...nel.org>
Cc: Eric Dumazet <edumazet@...gle.com>, Paolo Abeni <pabeni@...hat.com>,
 David Ahern <dsahern@...nel.org>, Matthieu Baerts <matttbe@...nel.org>,
 Mat Martineau <martineau@...nel.org>, Geliang Tang <geliang@...nel.org>,
 "David S. Miller" <davem@...emloft.net>,
 Donald Hunter <donald.hunter@...il.com>, netdev@...r.kernel.org,
 mptcp@...ts.linux.dev, linux-kernel@...r.kernel.org
Subject: Re: [PATCH net-next] tools: ynl-gen: use big-endian netlink attribute
 types

Hi Kuba,

On 9/14/24 4:39 AM, Jakub Kicinski wrote:
> Nice improvement! Since it technically missed net-next closing by a few
> hours, let me nit pick a little..

Yeah, sorry about that, first realized that net-next had closed after posting.
I had just waited for my net patch to make its way to net-next before posting, so
MPTCP_PM_ADDR_ATTR_PORT wouldn't change to NLA_BE16.

> On Fri, 13 Sep 2024 08:55:54 +0000 Asbjørn Sloth Tønnesen wrote:
>> diff --git a/tools/net/ynl/ynl-gen-c.py b/tools/net/ynl/ynl-gen-c.py
>> index 717530bc9c52e..e26f2c3c40891 100755
>> --- a/tools/net/ynl/ynl-gen-c.py
>> +++ b/tools/net/ynl/ynl-gen-c.py
>> @@ -48,6 +48,7 @@ class Type(SpecAttr):
>>           self.attr = attr
>>           self.attr_set = attr_set
>>           self.type = attr['type']
>> +        self.nla_type = self.type
> 
> is it worth introducing nla_type as Type attribute just for one user?
> inside a netlink code generator meaning of nla_type may not be crystal
> clear

Maybe not, I just took the same approach as byte_order_comment, and
co-located it with the existing byte-order condition in TypeScalar.

>>           self.checks = attr.get('checks', {})
>>   
>>           self.request = False
>> @@ -157,7 +158,7 @@ class Type(SpecAttr):
>>           return '{ .type = ' + policy + ', }'
>>   
>>       def attr_policy(self, cw):
>> -        policy = c_upper('nla-' + self.attr['type'])
>> +        policy = c_upper('nla-' + self.nla_type)
> 
> We could just swap the type directly here?

That could work too, WDYT?

diff --git a/tools/net/ynl/ynl-gen-c.py b/tools/net/ynl/ynl-gen-c.py
index 717530bc9c52e..e8706f36e5e7b 100755
--- a/tools/net/ynl/ynl-gen-c.py
+++ b/tools/net/ynl/ynl-gen-c.py
@@ -157,7 +157,10 @@ class Type(SpecAttr):
          return '{ .type = ' + policy + ', }'

      def attr_policy(self, cw):
-        policy = c_upper('nla-' + self.attr['type'])
+        policy = f'NLA_{c_upper(self.type)}'
+        if self.attr.get('byte-order', '') == 'big-endian':
+            if self.type in {'u16', 'u32'}:
+                policy = f'NLA_BE{self.type[1:]}'

          spec = self._attr_policy(policy)
          cw.p(f"\t[{self.enum_name}] = {spec},")