[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20181101.122856.227553243699206230.davem@davemloft.net>
Date:   Thu, 01 Nov 2018 12:28:56 -0700 (PDT)
From:   David Miller <davem@...emloft.net>
To:     acme@...nel.org
Cc:     yhs@...com, daniel@...earbox.net, jolsa@...hat.com, kafai@...com,
        alexei.starovoitov@...il.com, netdev@...r.kernel.org
Subject: Re: Help with the BPF verifier
From: Arnaldo Carvalho de Melo <acme@...nel.org>
Date: Thu, 1 Nov 2018 16:13:10 -0300
> Nope... this is inside an if:
> 
>         if (filename_arg <= 5) {
>                 augmented_args.filename.reserved = 0;
>                 augmented_args.filename.size = probe_read_str(&augmented_args.filename.value,
>                                                               sizeof(augmented_args.filename.value),
>                                                               (const void *)args->args[filename_arg]);
>                 if (augmented_args.filename.size < sizeof(augmented_args.filename.value)) {
>                         len -= sizeof(augmented_args.filename.value) - augmented_args.filename.size;
>                         len &= sizeof(augmented_args.filename.value) - 1;
>                 }
>         } else {
> 
> I use 6 to mean "hey, this syscall doesn't have any string argument, don't
> bother with it".
Really weird.  And it's unsigned so I can't imagine it wants you to
check that it's >= 0...
Maybe Deniel or someone else can figure it out.
Powered by blists - more mailing lists
 
