| Message ID | 20240820055244.128644-1-d.kandybka@gmail.com (mailing list archive) |
|---|---|
| State | Deferred |
| Delegated to: | Ping-Ke Shih |
| Headers | show |
| Series | wifi: rtw88: coex: remove rf4ce unused code | expand |
On Tue, 2024-08-20 at 08:52 +0300, Dmitry Kandybka wrote: > In 'rtw_coex_run_coex', 'rf4ce_en' is hardcoded to false, > so 'rtw_coex_action_rf4ce(rtwdev)' is never executed. > Assuming that rf4ce was never fully implemented, > remove lookalike leftovers. Compile tested only. > > Found by Linux Verification Center (linuxtesting.org) with SVACE. > > Fixes: a9359faaa47d ("rtw88: coex: add the mechanism for RF4CE") > Signed-off-by: Dmitry Kandybka <d.kandybka@gmail.com> > --- > drivers/net/wireless/realtek/rtw88/coex.c | 30 +---------------------- > 1 file changed, 1 insertion(+), 29 deletions(-) > > diff --git a/drivers/net/wireless/realtek/rtw88/coex.c > b/drivers/net/wireless/realtek/rtw88/coex.c > index de3332eb7a22..1fbcf701e7b7 100644 > --- a/drivers/net/wireless/realtek/rtw88/coex.c > +++ b/drivers/net/wireless/realtek/rtw88/coex.c > @@ -1591,31 +1591,6 @@ static void rtw_coex_action_freerun(struct rtw_dev > *rtwdev) > rtw_coex_tdma(rtwdev, false, 100); > } > > -static void rtw_coex_action_rf4ce(struct rtw_dev *rtwdev) > -{ > - const struct rtw_chip_info *chip = rtwdev->chip; > - struct rtw_efuse *efuse = &rtwdev->efuse; > - u8 table_case, tdma_case; > - > - rtw_dbg(rtwdev, RTW_DBG_COEX, "[BTCoex], %s()\n", __func__); > - > - rtw_coex_set_ant_path(rtwdev, false, COEX_SET_ANT_2G); > - rtw_coex_set_rf_para(rtwdev, chip->wl_rf_para_rx[0]); > - > - if (efuse->share_ant) { > - /* Shared-Ant */ > - table_case = 9; > - tdma_case = 16; > - } else { > - /* Non-Shared-Ant */ > - table_case = 100; > - tdma_case = 100; > - } > - > - rtw_coex_table(rtwdev, false, table_case); > - rtw_coex_tdma(rtwdev, false, tdma_case); > -} > - > static void rtw_coex_action_bt_whql_test(struct rtw_dev *rtwdev) > { > const struct rtw_chip_info *chip = rtwdev->chip; > @@ -2531,7 +2506,6 @@ static void rtw_coex_run_coex(struct rtw_dev > *rtwdev, u8 reason) > struct rtw_coex *coex = &rtwdev->coex; > struct rtw_coex_dm *coex_dm = &coex->dm; > struct rtw_coex_stat *coex_stat = &coex->stat; > - bool rf4ce_en = false; > > lockdep_assert_held(&rtwdev->mutex); > > @@ -2587,9 +2561,7 @@ static void rtw_coex_run_coex(struct rtw_dev > *rtwdev, u8 reason) > coex_stat->wl_coex_mode = COEX_WLINK_2G1PORT; > > if (coex_stat->bt_disabled) { > - if (coex_stat->wl_connected && rf4ce_en) > - rtw_coex_action_rf4ce(rtwdev); > - else if (!coex_stat->wl_connected) > + if (!coex_stat->wl_connected) > rtw_coex_action_wl_not_connected(rtwdev); > else > rtw_coex_action_wl_only(rtwdev); Hello Ping-Ke Shih, This is kindly reminder. Could you pay some attention to this patch and clarify if the rf4ce is actual for this moment and future?
> This is kindly reminder. Could you pay some attention to this patch and > clarify if the rf4ce is actual for this moment and future? Yes. Our coex developers want to keep this chunk. For me, this kind of cleanup patch is not very help to driver, but I and developers need much time to confirm and judge if we keep or remove them, so I would want to ignore this kind of patches...
On Fri, 2024-10-11 at 00:26 +0000, Ping-Ke Shih wrote: > > > This is kindly reminder. Could you pay some attention to this patch and > > clarify if the rf4ce is actual for this moment and future? > > Yes. Our coex developers want to keep this chunk. For me, this kind of > cleanup > patch is not very help to driver, but I and developers need much time to > confirm and judge if we keep or remove them, so I would want to ignore > this > kind of patches... > > Thanks for you clarification. Have a good work on driver.
Ping-Ke Shih <pkshih@realtek.com> writes: >> This is kindly reminder. Could you pay some attention to this patch and >> clarify if the rf4ce is actual for this moment and future? > > Yes. Our coex developers want to keep this chunk. For me, this kind of cleanup > patch is not very help to driver, but I and developers need much time to > confirm and judge if we keep or remove them, so I would want to ignore this > kind of patches... Yeah, I share your pain. Cleanup patches are most of the time unnecessary extra work for us maintainers. We should try to write that "cleanup policy for wireless subsystem" doc at some point, then we could just point that to everyone submitting cleanup patches.
diff --git a/drivers/net/wireless/realtek/rtw88/coex.c b/drivers/net/wireless/realtek/rtw88/coex.c index de3332eb7a22..1fbcf701e7b7 100644 --- a/drivers/net/wireless/realtek/rtw88/coex.c +++ b/drivers/net/wireless/realtek/rtw88/coex.c @@ -1591,31 +1591,6 @@ static void rtw_coex_action_freerun(struct rtw_dev *rtwdev) rtw_coex_tdma(rtwdev, false, 100); } -static void rtw_coex_action_rf4ce(struct rtw_dev *rtwdev) -{ - const struct rtw_chip_info *chip = rtwdev->chip; - struct rtw_efuse *efuse = &rtwdev->efuse; - u8 table_case, tdma_case; - - rtw_dbg(rtwdev, RTW_DBG_COEX, "[BTCoex], %s()\n", __func__); - - rtw_coex_set_ant_path(rtwdev, false, COEX_SET_ANT_2G); - rtw_coex_set_rf_para(rtwdev, chip->wl_rf_para_rx[0]); - - if (efuse->share_ant) { - /* Shared-Ant */ - table_case = 9; - tdma_case = 16; - } else { - /* Non-Shared-Ant */ - table_case = 100; - tdma_case = 100; - } - - rtw_coex_table(rtwdev, false, table_case); - rtw_coex_tdma(rtwdev, false, tdma_case); -} - static void rtw_coex_action_bt_whql_test(struct rtw_dev *rtwdev) { const struct rtw_chip_info *chip = rtwdev->chip; @@ -2531,7 +2506,6 @@ static void rtw_coex_run_coex(struct rtw_dev *rtwdev, u8 reason) struct rtw_coex *coex = &rtwdev->coex; struct rtw_coex_dm *coex_dm = &coex->dm; struct rtw_coex_stat *coex_stat = &coex->stat; - bool rf4ce_en = false; lockdep_assert_held(&rtwdev->mutex); @@ -2587,9 +2561,7 @@ static void rtw_coex_run_coex(struct rtw_dev *rtwdev, u8 reason) coex_stat->wl_coex_mode = COEX_WLINK_2G1PORT; if (coex_stat->bt_disabled) { - if (coex_stat->wl_connected && rf4ce_en) - rtw_coex_action_rf4ce(rtwdev); - else if (!coex_stat->wl_connected) + if (!coex_stat->wl_connected) rtw_coex_action_wl_not_connected(rtwdev); else rtw_coex_action_wl_only(rtwdev);
In 'rtw_coex_run_coex', 'rf4ce_en' is hardcoded to false, so 'rtw_coex_action_rf4ce(rtwdev)' is never executed. Assuming that rf4ce was never fully implemented, remove lookalike leftovers. Compile tested only. Found by Linux Verification Center (linuxtesting.org) with SVACE. Fixes: a9359faaa47d ("rtw88: coex: add the mechanism for RF4CE") Signed-off-by: Dmitry Kandybka <d.kandybka@gmail.com> --- drivers/net/wireless/realtek/rtw88/coex.c | 30 +---------------------- 1 file changed, 1 insertion(+), 29 deletions(-)