On 2021-04-18 17:54, Heiner Kallweit wrote:

On 18.04.2021 12:36, gregkh@linuxfoundation.org wrote:

...

The patch below does not apply to the 5.10-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to stable@vger.kernel.org.

thanks,

greg k-h

------------------ original commit in Linus's tree ------------------

...

From 453a77894efa4d9b6ef9644d74b9419c47ac427c Mon Sep 17 00:00:00 2001

From: Heiner Kallweit hkallweit1@gmail.com Date: Wed, 14 Apr 2021 10:47:10 +0200 Subject: [PATCH] r8169: don't advertise pause in jumbo mode

It has been reported [0] that using pause frames in jumbo mode impacts performance. There's no available chip documentation, but vendor drivers r8168 and r8125 don't advertise pause in jumbo mode. So let's do the same, according to Roman it fixes the issue.

[0] https://bugzilla.kernel.org/show_bug.cgi?id=212617

Fixes: 9cf9b84cc701 ("r8169: make use of phy_set_asym_pause") Reported-by: Roman Mamedov rm+bko@romanrm.net Tested-by: Roman Mamedov rm+bko@romanrm.net Signed-off-by: Heiner Kallweit hkallweit1@gmail.com Cc: stable@vger.kernel.org Signed-off-by: David S. Miller davem@davemloft.net

diff --git a/drivers/net/ethernet/realtek/r8169_main.c b/drivers/net/ethernet/realtek/r8169_main.c index 581a92fc3292..1df2c002c9f6 100644 --- a/drivers/net/ethernet/realtek/r8169_main.c +++ b/drivers/net/ethernet/realtek/r8169_main.c @@ -2350,6 +2350,13 @@ static void rtl_jumbo_config(struct rtl8169_private *tp) if (pci_is_pcie(tp->pci_dev) && tp->supports_gmii) pcie_set_readrq(tp->pci_dev, readrq);

• /* Chip doesn't support pause in jumbo mode */
• linkmode_mod_bit(ETHTOOL_LINK_MODE_Pause_BIT,

• 	 tp->phydev->advertising, !jumbo);
  

• linkmode_mod_bit(ETHTOOL_LINK_MODE_Asym_Pause_BIT,

• 	 tp->phydev->advertising, !jumbo);
  

• phy_start_aneg(tp->phydev); }

DECLARE_RTL_COND(rtl_chipcmd_cond) @@ -4630,8 +4637,6 @@ static int r8169_phy_connect(struct rtl8169_private *tp) if (!tp->supports_gmii) phy_set_max_speed(phydev, SPEED_100);

• phy_support_asym_pause(phydev);
• phy_attached_info(phydev);

return 0;

...

From ed6aaf67921a7ed44a3e07785c2125835ca5a969 Mon Sep 17 00:00:00 2001

From: Heiner Kallweit hkallweit1@gmail.com Date: Sun, 18 Apr 2021 17:50:06 +0200 Subject: [PATCH] r8169: don't advertise pause in jumbo mode

It has been reported [0] that using pause frames in jumbo mode impacts performance. There's no available chip documentation, but vendor drivers r8168 and r8125 don't advertise pause in jumbo mode. So let's do the same, according to Roman it fixes the issue.

[0] https://bugzilla.kernel.org/show_bug.cgi?id=212617

Fixes: 9cf9b84cc701 ("r8169: make use of phy_set_asym_pause") Reported-by: Roman Mamedov rm+bko@romanrm.net Tested-by: Roman Mamedov rm+bko@romanrm.net Signed-off-by: Heiner Kallweit hkallweit1@gmail.com

---

drivers/net/ethernet/realtek/r8169_main.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/realtek/r8169_main.c b/drivers/net/ethernet/realtek/r8169_main.c index d634da20b4f9..feca75eb75e4 100644 --- a/drivers/net/ethernet/realtek/r8169_main.c +++ b/drivers/net/ethernet/realtek/r8169_main.c @@ -2419,6 +2419,13 @@ static void rtl_jumbo_config(struct rtl8169_private *tp) if (!jumbo && pci_is_pcie(tp->pci_dev) && tp->supports_gmii) pcie_set_readrq(tp->pci_dev, 4096);

• /* Chip doesn't support pause in jumbo mode */
• linkmode_mod_bit(ETHTOOL_LINK_MODE_Pause_BIT,

• 	 tp->phydev->advertising, !jumbo);
  

• linkmode_mod_bit(ETHTOOL_LINK_MODE_Asym_Pause_BIT,

• 	 tp->phydev->advertising, !jumbo);
  

• phy_start_aneg(tp->phydev); }

DECLARE_RTL_COND(rtl_chipcmd_cond) @@ -4710,8 +4717,6 @@ static int r8169_phy_connect(struct rtl8169_private *tp) if (!tp->supports_gmii) phy_set_max_speed(phydev, SPEED_100);

• phy_support_asym_pause(phydev);
• phy_attached_info(phydev);

return 0;

I've *just* been looking into this as well (new machine with an r8169 onboard..sigh) and found that simply applying 5e00e16cb989 ("r8169: tweak max read request size for newer chips also in jumbo mtu mode") makes it apply cleanly. Said patch also seems (to me) make sense by itself, so maybe just pick that instead?

thanks, Holger