Re: [PATCH 3/3] ufs: core: Optimize ufshcd_add_uic_command_trace()

["Peter Wang (王信友)" <peter.wang@mediatek.com>]

On Fri, 2026-04-17 at 14:30 -0700, Bart Van Assche wrote:
> diff --git a/drivers/ufs/core/ufshcd.c b/drivers/ufs/core/ufshcd.c
> index 0ff9d7c2a7ac..12445e012cad 100644
> --- a/drivers/ufs/core/ufshcd.c
> +++ b/drivers/ufs/core/ufshcd.c
> @@ -460,20 +460,11 @@ static void ufshcd_add_uic_command_trace(struct
> ufs_hba *hba,
>                                          const struct uic_command
> *ucmd,
>                                          enum ufs_trace_str_t str_t)
>  {
> -       u32 cmd;
> -
>         if (!trace_ufshcd_uic_command_enabled())
>                 return;
> 
> -       if (str_t == UFS_CMD_SEND)
> -               cmd = ucmd->command;
> -       else
> -               cmd = ufshcd_readl(hba, REG_UIC_COMMAND);
> -
> -       trace_ufshcd_uic_command(hba, str_t, cmd,
> -                                ufshcd_readl(hba,
> REG_UIC_COMMAND_ARG_1),
> -                                ufshcd_readl(hba,
> REG_UIC_COMMAND_ARG_2),
> -                                ufshcd_readl(hba,
> REG_UIC_COMMAND_ARG_3));
> +       trace_ufshcd_uic_command(hba, str_t, ucmd->command, ucmd-
> >argument1,
> +                                ucmd->argument2, ucmd->argument3);
>  }

Hi Bart,

In the complete case, all these commands and arguments 1 to 3 
are filled by hardware. If we do not read them from the hardware, 
how can we be sure of the actual values written by the hardware?

Thanks
Peter