Re: [PATCH bpf-next 1/2] bpf: support symbolic BPF FS delegation mount options

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

 



On Thu, Dec 14, 2023 at 7:58 AM Alexei Starovoitov
<alexei.starovoitov@xxxxxxxxx> wrote:
>
> On Wed, Dec 13, 2023 at 02:23:26PM -0800, Andrii Nakryiko wrote:
> > Besides already supported special "any" value and hex bit mask, support
> > string-based parsing of delegation masks based on exact enumerator
> > names. Utilize BTF information of `enum bpf_cmd`, `enum bpf_map_type`,
> > `enum bpf_prog_type`, and `enum bpf_attach_type` types to find supported
> > symbolic names (ignoring __MAX_xxx guard values). So "BPF_PROG_LOAD" and
> > "BPF_MAP_CREATE" are valid values to specify for delegate_cmds options,
> > "BPF_MAP_TYPE_ARRAY" is among supported for map types, etc.
> >
> > Besides supporting string values, we also support multiple values
> > specified at the same time, using colon (':') separator.
> >
> > There are corresponding changes on bpf_show_options side to use known
> > values to print them in human-readable format, falling back to hex mask
> > printing, if there are any unrecognized bits. This shouldn't be
> > necessary when enum BTF information is present, but in general we should
> > always be able to fall back to this even if kernel was built without BTF.
> >
> > Example below shows various ways to specify delegate_cmds options
> > through mount command and how mount options are printed back:
> >
> >   $ sudo mkdir -p /sys/fs/bpf/token
> >   $ sudo mount -t bpf bpffs /sys/fs/bpf/token \
> >                -o delegate_cmds=BPF_PROG_LOAD \
> >                -o delegate_cmds=BPF_MAP_CREATE \
> >                -o delegate_cmds=BPF_TOKEN_CREATE:BPF_BTF_LOAD:BPF_LINK_CREATE
> >   $ mount | grep token
> >   bpffs on /sys/fs/bpf/token type bpf (rw,relatime,delegate_cmds=BPF_MAP_CREATE:BPF_PROG_LOAD:BPF_BTF_LOAD:BPF_LINK_CREATE:BPF_TOKEN_CREATE)
>
> imo this is too verbose.
> For cmds it doesn't look as bad. "BPF_" prefix is repetitive, but not overly so.
> But for maps and progs it will be bad.
> It will look like:
> delegate_progs=BPF_PROG_TYPE_SOCKET_FILTER:BPF_PROG_TYPE_SOCKET_XDP:BPF_PROG_TYPE_SCHED_CLS
> which is not readable.
> delegate_progs=SOCKET_FILTER:XDP:SCHED_CLS
> is much better.
>
> And I would go further (like libbpf does) and lower case them for output
> while allow both upper and lower for input:
> delegate_progs=socket_filter:xdp:sched_cls.

Ok, I was wondering if someone would complain about this :) Makes
sense, I'll do it more flexibly and succinctly.

>
> Because of stripping the prefix for maps and progs I would strip the prefix for cmds too
> for consistency.
>
> pw-bot: cr
>





[Index of Archives]     [Linux Samsung SoC]     [Linux Rockchip SoC]     [Linux Actions SoC]     [Linux for Synopsys ARC Processors]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]


  Powered by Linux