2019-05-24 11:10 UTC-0700 ~ Andrii Nakryiko <andrii.nakryiko@xxxxxxxxx>
> On Fri, May 24, 2019 at 2:15 AM Quentin Monnet
> <quentin.monnet@xxxxxxxxxxxxx> wrote:
>>
>> 2019-05-23 13:42 UTC-0700 ~ Andrii Nakryiko <andriin@xxxxxx>
>>> Add bash completion for new C btf dump option.
>>>
>>> Cc: Quentin Monnet <quentin.monnet@xxxxxxxxxxxxx>
>>> Signed-off-by: Andrii Nakryiko <andriin@xxxxxx>
>>> ---
>>> tools/bpf/bpftool/bash-completion/bpftool | 25 +++++++++++++++++++----
>>> 1 file changed, 21 insertions(+), 4 deletions(-)
>>>
>>> diff --git a/tools/bpf/bpftool/bash-completion/bpftool b/tools/bpf/bpftool/bash-completion/bpftool
>>> index 50e402a5a9c8..5b65e0309d2a 100644
>>> --- a/tools/bpf/bpftool/bash-completion/bpftool
>>> +++ b/tools/bpf/bpftool/bash-completion/bpftool
>>> @@ -638,11 +638,28 @@ _bpftool()
>>> esac
>>> return 0
>>> ;;
>>> + format)
>>> + COMPREPLY=( $( compgen -W "c raw" -- "$cur" ) )
>>> + ;;
>>> *)
>>> - if [[ $cword == 6 ]] && [[ ${words[3]} == "map" ]]; then
>>> - COMPREPLY+=( $( compgen -W 'key value kv all' -- \
>>> - "$cur" ) )
>>> - fi
>>> + # emit extra options
>>> + case ${words[3]} in
>>> + id|file)
>>> + if [[ $cword > 4 ]]; then
>>
>> Not sure if this "if" is necessary. It seems to me that if $cword is 4
>> then we are just after "id" or "file" in the command line, in which case
>> we hit previous cases and never reach this point?
>
> Yep, you are right, removed.
>
>>
>> Also, reading the completion code I wonder, do we have completion for
>> BTF ids? It seems to me that we have nothing proposed to complete
>> "bpftool btf dump id <tab>". Any chance to get that in a follow-up patch?
>
> We currently don't have a way to iterate all BTFs in a system (neither
> in bpftool, nor in libbpf, AFAICT), but I can do that based on btf_id
> field, dumped as part of `bpftool prog list` command. Would that work?
> I'll post that as a separate patch.
Yes, I suppose it would. Unless there is a plan to add a "bpftool btf
dump" soon, in which case it could wait? But otherwise parsing "bpftool
prog list --pretty" (better parse JSON output, more stable) should be fine.
Thanks!
