Re: [PATCH net-next v9 7/9] libeth: add Rx buffer management

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



From: Kees Cook <keescook@xxxxxxxxxxxx>
Date: Tue, 9 Apr 2024 09:17:53 -0700

> On Mon, Apr 08, 2024 at 11:45:32AM +0200, Alexander Lobakin wrote:
>> From: Alexander Lobakin <aleksander.lobakin@xxxxxxxxx>
>> Date: Mon, 8 Apr 2024 11:11:12 +0200
>>
>>> From: Jakub Kicinski <kuba@xxxxxxxxxx>
>>> Date: Fri, 5 Apr 2024 21:25:13 -0700
>>>
>>>> On Thu,  4 Apr 2024 17:44:00 +0200 Alexander Lobakin wrote:
>>>>> +/**
>>>>> + * struct libeth_fq - structure representing a buffer queue
>>>>> + * @fp: hotpath part of the structure
>>>>
>>>> Second time this happens this week, so maybe some tooling change in 6.9
>>>> but apparently kdoc does not want to know about the tagged struct:
>>>>
>>>> include/net/libeth/rx.h:69: warning: Excess struct member 'fp' description in 'libeth_fq'
>>>
>>> Oh no, maybe we should teach kdoc to parse struct_group*()?
>>
>> scripts/kernel-doc apparently can handle them...
>>
>> + Kees
>>
> 
> Ah, hm, scripts/kernel-doc throws away the early arguments of
> struct_group_tagged, but I suspect it needs to create a synthetic member
> for the tag. i.e. instead of:
> 
> 	struct_group_tagged(tag, name, members...)
> 
> becoming
> 
> 	members...
> 
> it needs to become
> 
> 	struct tag name;
> 	members...
> 
> It seems this is the first place anyone has tried to document the tagged
> struct name! :)

It makes sense and TBH I expected kdoc to warn that an element
description is missing :D

> 
> Does this work? I haven't tested it...
> 
> diff --git a/scripts/kernel-doc b/scripts/kernel-doc
> index 967f1abb0edb..64a19228d5dd 100755
> --- a/scripts/kernel-doc
> +++ b/scripts/kernel-doc
> @@ -1151,7 +1151,8 @@ sub dump_struct($$) {
>          # - first eat non-declaration parameters and rewrite for final match
>          # - then remove macro, outer parens, and trailing semicolon
>          $members =~ s/\bstruct_group\s*\(([^,]*,)/STRUCT_GROUP(/gos;
> -        $members =~ s/\bstruct_group_(attr|tagged)\s*\(([^,]*,){2}/STRUCT_GROUP(/gos;
> +        $members =~ s/\bstruct_group_attr\s*\(([^,]*,){2}/STRUCT_GROUP(/gos;
> +        $members =~ s/\bstruct_group_tagged\s*\(([^,]*,)([^,]*,)/struct $1 $2; STRUCT_GROUP(/gos;

This one does not. We need to exclude ','s from the groups...
      +        $members =~
s/\bstruct_group_tagged\s*\(([^,]*),([^,]*),/struct $1 $2;
STRUCT_GROUP(/gos;

That one is fine. $members:

include/net/libeth/rx.h:91: warning: struct libeth_fq_fp  fp;
STRUCT_GROUP( struct page_pool        *pp; struct libeth_fqe
*fqes; u32                     truesize; u32                     count;
); enum libeth_fqe_type    type:2; bool                    hsplit:1;
bool                    xdp:1; u32                     buf_len; int
               nid;

So you almost fixed it :D

Which tree this should go through? Should I include this patch to this
series with libeth or it's better to push this through kees/linux and
then pull to net-next?


>          $members =~ s/\b__struct_group\s*\(([^,]*,){3}/STRUCT_GROUP(/gos;
>          $members =~ s/\bSTRUCT_GROUP(\(((?:(?>[^)(]+)|(?1))*)\))[^;]*;/$2/gos;

Thanks,
Olek




[Index of Archives]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Bugtraq]     [Linux OMAP]     [Linux MIPS]     [eCos]     [Asterisk Internet PBX]     [Linux API]

  Powered by Linux