On 11/18/2019 3:37 PM, David Marchand wrote:
> We currently do not check that a non option string has been passed to
> testpmd.
>
> Example:
> $ ./master/app/testpmd --no-huge -m 512 --vdev net_null0 \
> --vdev net_null1 -- -i nb-cores=2 --total-num-mbuf 2048
> [...]
> testpmd> show config fwd
> io packet forwarding - ports=2 - cores=1 - streams=2 - NUMA support
> enabled, MP allocation mode: native
> Logical Core 1 (socket 0) forwards packets on 2 streams:
> RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
> RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
>
> Here nb-cores=2 is just ignored, while the (probably sleepy) user did not
> notice this.
>
> Validate that all strings passed to testpmd are part of a known option.
>
> After this patch:
> $ ./master/app/testpmd --no-huge -m 512 --vdev net_null0 \
> --vdev net_null1 -- -i nb-cores=2 --total-num-mbuf 2048
> [...]
> Invalid parameter: nb-cores=2
> EAL: Error - exiting with code: 1
> Cause: Command line incorrect
>
> While at it, when passing an unknown option, print the string that gets
> refused by getopt_long to help the user.
>
> Fixes: af75078fece3 ("first public release")
> Cc: sta...@dpdk.org
>
> Signed-off-by: David Marchand <david.march...@redhat.com>
> Reviewed-by: Ferruh Yigit <ferruh.yi...@intel.com>
> ---
> This seems a bit dangerous to take this kind of change this late.
> Some "working fine" scripts might now report failures from testpmd because
> of garbage in the command line.
>
> Sending the patch anyway to see what others think about it.
>
> Changelog since v1:
> - fixed example in commitlog,
>
> ---
> app/test-pmd/parameters.c | 7 +++++++
> 1 file changed, 7 insertions(+)
>
> diff --git a/app/test-pmd/parameters.c b/app/test-pmd/parameters.c
> index deca7a6828..2e7a504415 100644
> --- a/app/test-pmd/parameters.c
> +++ b/app/test-pmd/parameters.c
> @@ -1363,12 +1363,19 @@ launch_args_parse(int argc, char** argv)
> break;
> default:
> usage(argv[0]);
> + printf("Invalid option: %s\n", argv[optind]);
> rte_exit(EXIT_FAILURE,
> "Command line is incomplete or incorrect\n");
> break;
> }
> }
>
> + if (optind != argc) {
I hope 'optind' works as expected [1] in Windows too (Anatoly verified the
FreeBSD one).
@Pallavi, @Ranjit, Can you please confirm it?
[1]
https://linux.die.net/man/3/optind
"
If there are no more option characters, getopt() returns -1. Then optind is the
index in argv of the first argv-element that is not an option.
"
