[PATCH net-next] r8169: disable tx interrupt coalescing on RTL8168

[PATCH net-next] r8169: disable tx interrupt coalescing on RTL8168

[Heiner Kallweit]

In contrast to switching rx irq coalescing off what fixed an issue,
switching tx irq coalescing off is merely a latency optimization,
therefore net-next. As part of this change:

- Remove INTT_0 .. INTT_3 constants, they aren't used.

- Remove comment in rtl_hw_start_8169(), we now have a detailed
  description by the code in rtl_set_coalesce().

- Due to switching irq coalescing off per default we don't need the
  initialization in rtl_hw_start_8168(). If ethtool is used to switch
  on coalescing then rtl_set_coalesce() will configure this register.

For the sake of completeness: This patch just changes the default.
Users still have the option to configure irq coalescing via ethtool.

Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
---
 drivers/net/ethernet/realtek/r8169.c | 20 ++------------------
 1 file changed, 2 insertions(+), 18 deletions(-)

diff --git a/drivers/net/ethernet/realtek/r8169.c b/drivers/net/ethernet/realtek/r8169.c
index 8a8a4f70d..8f065bf3e 100644
--- a/drivers/net/ethernet/realtek/r8169.c
+++ b/drivers/net/ethernet/realtek/r8169.c
@@ -490,10 +490,6 @@ enum rtl_register_content {
 	PCIDAC		= (1 << 4),
 	PCIMulRW	= (1 << 3),
 #define INTT_MASK	GENMASK(1, 0)
-	INTT_0		= 0x0000,	// 8168
-	INTT_1		= 0x0001,	// 8168
-	INTT_2		= 0x0002,	// 8168
-	INTT_3		= 0x0003,	// 8168
 
 	/* rtl8169_PHYstatus */
 	TBI_Enable	= 0x80,
@@ -4702,6 +4698,8 @@ static void rtl_hw_start(struct  rtl8169_private *tp)
 	rtl_set_rx_tx_desc_registers(tp);
 	rtl_lock_config_regs(tp);
 
+	/* disable interrupt coalescing */
+	RTL_W16(tp, IntrMitigate, 0x0000);
 	/* Initially a 10 us delay. Turned it into a PCI commit. - FR */
 	RTL_R8(tp, IntrMask);
 	RTL_W8(tp, ChipCmd, CmdTxEnb | CmdRxEnb);
@@ -4734,12 +4732,6 @@ static void rtl_hw_start_8169(struct rtl8169_private *tp)
 
 	rtl8169_set_magic_reg(tp, tp->mac_version);
 
-	/*
-	 * Undocumented corner. Supposedly:
-	 * (TxTimer << 12) | (TxPackets << 8) | (RxTimer << 4) | RxPackets
-	 */
-	RTL_W16(tp, IntrMitigate, 0x0000);
-
 	RTL_W32(tp, RxMissed, 0);
 }
 
@@ -5456,12 +5448,6 @@ static void rtl_hw_start_8168(struct rtl8169_private *tp)
 {
 	RTL_W8(tp, MaxTxPacketSize, TxPacketMax);
 
-	tp->cp_cmd &= ~INTT_MASK;
-	tp->cp_cmd |= PktCntrDisable | INTT_1;
-	RTL_W16(tp, CPlusCmd, tp->cp_cmd);
-
-	RTL_W16(tp, IntrMitigate, 0x5100);
-
 	/* Work around for RxFIFO overflow. */
 	if (tp->mac_version == RTL_GIGA_MAC_VER_11) {
 		tp->irq_mask |= RxFIFOOver;
@@ -5749,8 +5735,6 @@ static void rtl_hw_start_8101(struct rtl8169_private *tp)
 		rtl_hw_start_8168h_1(tp);
 		break;
 	}
-
-	RTL_W16(tp, IntrMitigate, 0x0000);
 }
 
 static int rtl8169_change_mtu(struct net_device *dev, int new_mtu)
-- 
2.21.0

Re: [PATCH net-next] r8169: disable tx interrupt coalescing on RTL8168

[David Miller]

From: Heiner Kallweit <hkallweit1@gmail.com>
Date: Sun, 31 Mar 2019 15:43:59 +0200

> @@ -5456,12 +5448,6 @@ static void rtl_hw_start_8168(struct rtl8169_private *tp)
>  {
>  	RTL_W8(tp, MaxTxPacketSize, TxPacketMax);
>  
> -	tp->cp_cmd &= ~INTT_MASK;
> -	tp->cp_cmd |= PktCntrDisable | INTT_1;
> -	RTL_W16(tp, CPlusCmd, tp->cp_cmd);
> -
> -	RTL_W16(tp, IntrMitigate, 0x5100);
> -
>  	/* Work around for RxFIFO overflow. */

This patch isn't against net-next.

I'm guessing it's against a tree with the RX coalesce disable changes
applied.

Re: [PATCH net-next] r8169: disable tx interrupt coalescing on RTL8168

[Heiner Kallweit]

On 31.03.2019 20:13, David Miller wrote:
> From: Heiner Kallweit <hkallweit1@gmail.com>
> Date: Sun, 31 Mar 2019 15:43:59 +0200
> 
>> @@ -5456,12 +5448,6 @@ static void rtl_hw_start_8168(struct rtl8169_private *tp)
>>  {
>>  	RTL_W8(tp, MaxTxPacketSize, TxPacketMax);
>>  
>> -	tp->cp_cmd &= ~INTT_MASK;
>> -	tp->cp_cmd |= PktCntrDisable | INTT_1;
>> -	RTL_W16(tp, CPlusCmd, tp->cp_cmd);
>> -
>> -	RTL_W16(tp, IntrMitigate, 0x5100);
>> -
>>  	/* Work around for RxFIFO overflow. */
> 
> This patch isn't against net-next.
> 
> I'm guessing it's against a tree with the RX coalesce disable changes
> applied.
> 
You're right, sorry. It will apply cleanly only once net has been merged
into net-next. However, if I rebase this patch to net-next, then merging
net to net-next will complain. Can we defer applying this patch until net
has been merged?

Re: [PATCH net-next] r8169: disable tx interrupt coalescing on RTL8168

[David Miller]

From: Heiner Kallweit <hkallweit1@gmail.com>
Date: Sun, 31 Mar 2019 20:16:51 +0200

> Can we defer applying this patch until net has been merged?

Yes, but please mention such dependencies next time.

Re: [PATCH net-next] r8169: disable tx interrupt coalescing on RTL8168

[Heiner Kallweit]

On 31.03.2019 20:23, David Miller wrote:
> From: Heiner Kallweit <hkallweit1@gmail.com>
> Date: Sun, 31 Mar 2019 20:16:51 +0200
> 
>> Can we defer applying this patch until net has been merged?
> 
Thanks

> Yes, but please mention such dependencies next time.
> 
OK