| Message ID | ZEswqUHBXPHC/znL@ns.kevlo.org (mailing list archive) |
|---|---|
| State | Rejected |
| Delegated to: | Kalle Valo |
| Headers | show |
| Series | [v2] wifi: rtl8xxxu: 8188f: no need to initialize interrupts | expand |
Kevin Lo <kevlo@kevlo.org> writes: > There's no need to initialize interrupts for RTL8188FTV like the vendor driver. > > Signed-off-by: Kevin Lo <kevlo@kevlo.org> Why do this, what's the benefit? Or is this just a theoretical fix. The commit log should explain this.
On Fri, Apr 28, 2023 at 08:40:34AM +0300, Kalle Valo wrote: > > Kevin Lo <kevlo@kevlo.org> writes: > > > There's no need to initialize interrupts for RTL8188FTV like the vendor driver. > > > > Signed-off-by: Kevin Lo <kevlo@kevlo.org> > > Why do this, what's the benefit? Or is this just a theoretical fix. The > commit log should explain this. If I understand correctly, just like rtw88 usb devices, rtl8xxxu has no hardware interrupt.
> -----Original Message----- > From: Kevin Lo <kevlo@kevlo.org> > Sent: Friday, April 28, 2023 2:35 PM > To: Kalle Valo <kvalo@kernel.org> > Cc: linux-wireless@vger.kernel.org; Bitterblue Smith <rtl8821cerfe2@gmail.com>; Jes Sorensen > <Jes.Sorensen@gmail.com>; Ping-Ke Shih <pkshih@realtek.com> > Subject: Re: [PATCH v2] wifi: rtl8xxxu: 8188f: no need to initialize interrupts > > On Fri, Apr 28, 2023 at 08:40:34AM +0300, Kalle Valo wrote: > > > > Kevin Lo <kevlo@kevlo.org> writes: > > > > > There's no need to initialize interrupts for RTL8188FTV like the vendor driver. > > > > > > Signed-off-by: Kevin Lo <kevlo@kevlo.org> > > > > Why do this, what's the benefit? Or is this just a theoretical fix. The > > commit log should explain this. > > If I understand correctly, just like rtw88 usb devices, rtl8xxxu has no > hardware interrupt. > With/without hardware interrupt, it depends on IMR (interrupt mask register) setting. If it is enabled, driver can receive C2H, PS indicators, and beacon TBTT events via certain USB endpoint. But, normally vendor driver doesn't use them neither. Why it initializes interrupt in this patch is to reset ISR (interrupt status register) to avoid unexpected interrupts are received. So, if it doesn't really enable interrupt by IMR, I suppose it should work well even without this patch. Kevin, could I know the real problem you met? Ping-Ke
On Tue, May 02, 2023 at 09:10:04AM +0000, Ping-Ke Shih wrote: > > > > > -----Original Message----- > > From: Kevin Lo <kevlo@kevlo.org> > > Sent: Friday, April 28, 2023 2:35 PM > > To: Kalle Valo <kvalo@kernel.org> > > Cc: linux-wireless@vger.kernel.org; Bitterblue Smith <rtl8821cerfe2@gmail.com>; Jes Sorensen > > <Jes.Sorensen@gmail.com>; Ping-Ke Shih <pkshih@realtek.com> > > Subject: Re: [PATCH v2] wifi: rtl8xxxu: 8188f: no need to initialize interrupts > > > > On Fri, Apr 28, 2023 at 08:40:34AM +0300, Kalle Valo wrote: > > > > > > Kevin Lo <kevlo@kevlo.org> writes: > > > > > > > There's no need to initialize interrupts for RTL8188FTV like the vendor driver. > > > > > > > > Signed-off-by: Kevin Lo <kevlo@kevlo.org> > > > > > > Why do this, what's the benefit? Or is this just a theoretical fix. The > > > commit log should explain this. > > > > If I understand correctly, just like rtw88 usb devices, rtl8xxxu has no > > hardware interrupt. > > > > With/without hardware interrupt, it depends on IMR (interrupt mask register) > setting. If it is enabled, driver can receive C2H, PS indicators, and beacon > TBTT events via certain USB endpoint. But, normally vendor driver doesn't > use them neither. > > Why it initializes interrupt in this patch is to reset ISR (interrupt status > register) to avoid unexpected interrupts are received. So, if it doesn't > really enable interrupt by IMR, I suppose it should work well even without > this patch. > > Kevin, could I know the real problem you met? Ping-Ke, many thanks for the clear explanation. I didn't encounter any problems. As you mentioned, while the vendor driver doesn't have IMR settings, there is no harm in initializing interrupts. Please drop this patch, thanks. > Ping-Ke Regards, Kevin
diff --git a/drivers/net/wireless/realtek/rtl8xxxu/rtl8xxxu_core.c b/drivers/net/wireless/realtek/rtl8xxxu/rtl8xxxu_core.c index 620a5cc2bfdd..7f620dc47ee0 100644 --- a/drivers/net/wireless/realtek/rtl8xxxu/rtl8xxxu_core.c +++ b/drivers/net/wireless/realtek/rtl8xxxu/rtl8xxxu_core.c @@ -4023,9 +4023,6 @@ static int rtl8xxxu_init_device(struct ieee80211_hw *hw) if (priv->rtl_chip == RTL8192E) { rtl8xxxu_write32(priv, REG_HIMR0, 0x00); rtl8xxxu_write32(priv, REG_HIMR1, 0x00); - } else if (priv->rtl_chip == RTL8188F) { - rtl8xxxu_write32(priv, REG_HISR0, 0xffffffff); - rtl8xxxu_write32(priv, REG_HISR1, 0xffffffff); } else if (priv->rtl_chip == RTL8188E) { rtl8xxxu_write32(priv, REG_HISR0, 0xffffffff); val32 = IMR0_PSTIMEOUT | IMR0_TBDER | IMR0_CPWM | IMR0_CPWM2; @@ -4035,7 +4032,7 @@ static int rtl8xxxu_init_device(struct ieee80211_hw *hw) val8 = rtl8xxxu_read8(priv, REG_USB_SPECIAL_OPTION); val8 |= USB_SPEC_INT_BULK_SELECT; rtl8xxxu_write8(priv, REG_USB_SPECIAL_OPTION, val8); - } else { + } else if (priv->rtl_chip != RTL8188F) { /* * Enable all interrupts - not obvious USB needs to do this */
There's no need to initialize interrupts for RTL8188FTV like the vendor driver. Signed-off-by: Kevin Lo <kevlo@kevlo.org> --- v2: - Fix my email address. ---