Re: [PATCH 3/3] staging: rtl8192e: remove dead code

[Michael Straube]

Am 26.05.24 um 16:31 schrieb Nam Cao:
> On Sun, May 26, 2024 at 01:19:28PM +0200, Michael Straube wrote:
> > Remove two else-if arms that do nothing.
> >
> > Signed-off-by: Michael Straube <straube.linux@xxxxxxxxx>
> > ---
> >   drivers/staging/rtl8192e/rtl8192e/rtl_dm.c | 6 ------
> >   1 file changed, 6 deletions(-)
> >
> > diff --git a/drivers/staging/rtl8192e/rtl8192e/rtl_dm.c b/drivers/staging/rtl8192e/rtl8192e/rtl_dm.c
> > index 5392d2daf870..4e03eb100175 100644
> > --- a/drivers/staging/rtl8192e/rtl8192e/rtl_dm.c
> > +++ b/drivers/staging/rtl8192e/rtl8192e/rtl_dm.c
> > @@ -1370,9 +1370,6 @@ static void _rtl92e_dm_rx_path_sel_byrssi(struct net_device *dev)
> >   					tmp_sec_rssi = cur_rf_rssi;
> >   					sec_rssi_index = i;
> >   				}
> > -			} else if ((cur_rf_rssi < tmp_sec_rssi) &&
> > -					(cur_rf_rssi > tmp_min_rssi)) {
> > -				;
> >   			} else if (cur_rf_rssi == tmp_min_rssi) {
> >   				if (tmp_sec_rssi == tmp_min_rssi) {
> >   					tmp_min_rssi = cur_rf_rssi;
> > @@ -1426,9 +1423,6 @@ static void _rtl92e_dm_rx_path_sel_byrssi(struct net_device *dev)
> >   						tmp_cck_sec_pwdb = cur_cck_pwdb;
> >   						cck_rx_ver2_sec_index = i;
> >   					}
> > -				} else if ((cur_cck_pwdb < tmp_cck_sec_pwdb) &&
> > -						(cur_cck_pwdb > tmp_cck_min_pwdb)) {
> > -					;
> >   				} else if (cur_cck_pwdb == tmp_cck_min_pwdb) {
> >   					if (tmp_cck_sec_pwdb == tmp_cck_min_pwdb)
> >   						tmp_cck_min_pwdb = cur_cck_pwdb;
>
> I would be careful with these changes. These else-if do prevent the
> execution of the other else-if, so the code do not behave the same anymore.
>
> The only case this patch doesn't change anything functionally is when the
> condition of the removed if-else is mutually exclusive with the conditions
> of the following if-else. Are you sure this is the case?

Ah yes, I had not thought about that. Thanks for pointing out.
I'll have a closer look and resend the series. Either without this patch
or, if it's safe to remove, state it in the commit message.

Thanks,
Michael