[PATCH] doc: update flow-perf.rst --query-rate option description

[Stephen Hemminger]

On Fri, 26 Dec 2025 09:50:57 +0800
Sunyang Wu <sunyang.wu at jaguarmicro.com> wrote:

> Update the --query-rate option description in flow-perf.rst
> 
> Signed-off-by: Sunyang Wu <sunyang.wu at jaguarmicro.com>
> ---
>  doc/guides/tools/flow-perf.rst | 15 ---------------
>  1 file changed, 15 deletions(-)
> 
> diff --git a/doc/guides/tools/flow-perf.rst b/doc/guides/tools/flow-perf.rst
> index 5d7ae86cb0..a289ba482b 100644
> --- a/doc/guides/tools/flow-perf.rst
> +++ b/doc/guides/tools/flow-perf.rst
> @@ -89,21 +89,6 @@ The command line options are:
>  
>  *	``--query-rate``
>  	Enable query rate calculations.
> -	The application measures the rate of flow rule query operations.
> -	After inserting the specified number of flow rules, the application
> -	queries each rule sequentially and measures the total time taken.
> -	The query rate is calculated as the number of query operations divided
> -	by the total time.
> -
> -	Example usage with query rate calculation:
> -
> -	.. code-block:: console
> -
> -	    sudo ./dpdk-test-flow_perf -a 08:00.0 -- --ingress --ether --ipv4 --queue --rules-count=100000 --query-rate
> -
> -	The query rate measurement follows the same window-based approach as
> -	insertion and deletion rate measurements, using the window size
> -	configured via --rules-batch.
>  
>  *	``--dump-socket-mem``
>  	Dump the memory stats for each socket before the insertion and after.

The description in the guide is awkward to read and unclear.
But dropping the complete description is not the best way to resolve it.

I would recommend rewriting (maybe using AI) and make it clea what the
test measures and what the expected result is.