* [PATCH net v3 0/2] r8169: fix DASH devices network lost issue @ 2023-11-09 16:43 ChunHao Lin 2023-11-09 16:43 ` [PATCH net v3 1/2] r8169: add handling DASH when DASH is disabled ChunHao Lin 2023-11-09 16:43 ` [PATCH net v3 2/2] r8169: fix network lost after resume on DASH systems ChunHao Lin 0 siblings, 2 replies; 6+ messages in thread From: ChunHao Lin @ 2023-11-09 16:43 UTC (permalink / raw) To: hkallweit1 Cc: nic_swsd, davem, edumazet, kuba, pabeni, netdev, linux-kernel, ChunHao Lin This series are used to fix network lost issue on systems that support DASH. It has been tested on rtl8168ep and rtl8168fp. V2 -> V3: Add 'Fixes' tag and correct indentation. V1 -> V2: Change variable and function name. And update DASH info message. ChunHao Lin (2): r8169: add handling DASH when DASH is disabled r8169: fix network lost after resume on DASH systems drivers/net/ethernet/realtek/r8169_main.c | 41 +++++++++++++++++------ 1 file changed, 31 insertions(+), 10 deletions(-) -- 2.39.2 ^ permalink raw reply [flat|nested] 6+ messages in thread
* [PATCH net v3 1/2] r8169: add handling DASH when DASH is disabled 2023-11-09 16:43 [PATCH net v3 0/2] r8169: fix DASH devices network lost issue ChunHao Lin @ 2023-11-09 16:43 ` ChunHao Lin 2023-11-09 17:14 ` Heiner Kallweit 2023-11-09 16:43 ` [PATCH net v3 2/2] r8169: fix network lost after resume on DASH systems ChunHao Lin 1 sibling, 1 reply; 6+ messages in thread From: ChunHao Lin @ 2023-11-09 16:43 UTC (permalink / raw) To: hkallweit1 Cc: nic_swsd, davem, edumazet, kuba, pabeni, netdev, linux-kernel, ChunHao Lin, stable For devices that support DASH, even DASH is disabled, there may still exist a default firmware that will influence device behavior. So driver needs to handle DASH for devices that support DASH, no matter the DASH status is. This patch also prepares for "fix network lost after resume on DASH systems". Fixes: ee7a1beb9759 ("r8169:call "rtl8168_driver_start" "rtl8168_driver_stop" only when hardware dash function is enabled") Cc: stable@vger.kernel.org Signed-off-by: ChunHao Lin <hau@realtek.com> --- drivers/net/ethernet/realtek/r8169_main.c | 35 ++++++++++++++++------- 1 file changed, 25 insertions(+), 10 deletions(-) diff --git a/drivers/net/ethernet/realtek/r8169_main.c b/drivers/net/ethernet/realtek/r8169_main.c index 0c76c162b8a9..4954ff0f72b1 100644 --- a/drivers/net/ethernet/realtek/r8169_main.c +++ b/drivers/net/ethernet/realtek/r8169_main.c @@ -624,6 +624,7 @@ struct rtl8169_private { unsigned supports_gmii:1; unsigned aspm_manageable:1; + unsigned dash_enabled:1; dma_addr_t counters_phys_addr; struct rtl8169_counters *counters; struct rtl8169_tc_offsets tc_offset; @@ -1253,14 +1254,26 @@ static bool r8168ep_check_dash(struct rtl8169_private *tp) return r8168ep_ocp_read(tp, 0x128) & BIT(0); } -static enum rtl_dash_type rtl_check_dash(struct rtl8169_private *tp) +static bool rtl_dash_is_enabled(struct rtl8169_private *tp) +{ + switch (tp->dash_type) { + case RTL_DASH_DP: + return r8168dp_check_dash(tp); + case RTL_DASH_EP: + return r8168ep_check_dash(tp); + default: + return false; + } +} + +static enum rtl_dash_type rtl_get_dash_type(struct rtl8169_private *tp) { switch (tp->mac_version) { case RTL_GIGA_MAC_VER_28: case RTL_GIGA_MAC_VER_31: - return r8168dp_check_dash(tp) ? RTL_DASH_DP : RTL_DASH_NONE; + return RTL_DASH_DP; case RTL_GIGA_MAC_VER_51 ... RTL_GIGA_MAC_VER_53: - return r8168ep_check_dash(tp) ? RTL_DASH_EP : RTL_DASH_NONE; + return RTL_DASH_EP; default: return RTL_DASH_NONE; } @@ -1453,7 +1466,7 @@ static void __rtl8169_set_wol(struct rtl8169_private *tp, u32 wolopts) device_set_wakeup_enable(tp_to_dev(tp), wolopts); - if (tp->dash_type == RTL_DASH_NONE) { + if (!tp->dash_enabled) { rtl_set_d3_pll_down(tp, !wolopts); tp->dev->wol_enabled = wolopts ? 1 : 0; } @@ -2512,7 +2525,7 @@ static void rtl_wol_enable_rx(struct rtl8169_private *tp) static void rtl_prepare_power_down(struct rtl8169_private *tp) { - if (tp->dash_type != RTL_DASH_NONE) + if (tp->dash_enabled) return; if (tp->mac_version == RTL_GIGA_MAC_VER_32 || @@ -4869,7 +4882,7 @@ static int rtl8169_runtime_idle(struct device *device) { struct rtl8169_private *tp = dev_get_drvdata(device); - if (tp->dash_type != RTL_DASH_NONE) + if (tp->dash_enabled) return -EBUSY; if (!netif_running(tp->dev) || !netif_carrier_ok(tp->dev)) @@ -4896,7 +4909,7 @@ static void rtl_shutdown(struct pci_dev *pdev) rtl_rar_set(tp, tp->dev->perm_addr); if (system_state == SYSTEM_POWER_OFF && - tp->dash_type == RTL_DASH_NONE) { + !tp->dash_enabled) { pci_wake_from_d3(pdev, tp->saved_wolopts); pci_set_power_state(pdev, PCI_D3hot); } @@ -5254,7 +5267,8 @@ static int rtl_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) rc = pci_disable_link_state(pdev, PCIE_LINK_STATE_L1); tp->aspm_manageable = !rc; - tp->dash_type = rtl_check_dash(tp); + tp->dash_type = rtl_get_dash_type(tp); + tp->dash_enabled = rtl_dash_is_enabled(tp); tp->cp_cmd = RTL_R16(tp, CPlusCmd) & CPCMD_MASK; @@ -5325,7 +5339,7 @@ static int rtl_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) /* configure chip for default features */ rtl8169_set_features(dev, dev->features); - if (tp->dash_type == RTL_DASH_NONE) { + if (!tp->dash_enabled) { rtl_set_d3_pll_down(tp, true); } else { rtl_set_d3_pll_down(tp, false); @@ -5365,7 +5379,8 @@ static int rtl_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) "ok" : "ko"); if (tp->dash_type != RTL_DASH_NONE) { - netdev_info(dev, "DASH enabled\n"); + netdev_info(dev, "DASH %s\n", + tp->dash_enabled ? "enabled" : "disabled"); rtl8168_driver_start(tp); } -- 2.39.2 ^ permalink raw reply related [flat|nested] 6+ messages in thread
* Re: [PATCH net v3 1/2] r8169: add handling DASH when DASH is disabled 2023-11-09 16:43 ` [PATCH net v3 1/2] r8169: add handling DASH when DASH is disabled ChunHao Lin @ 2023-11-09 17:14 ` Heiner Kallweit 2023-11-09 17:29 ` Hau 0 siblings, 1 reply; 6+ messages in thread From: Heiner Kallweit @ 2023-11-09 17:14 UTC (permalink / raw) To: ChunHao Lin Cc: nic_swsd, davem, edumazet, kuba, pabeni, netdev, linux-kernel, stable On 09.11.2023 17:43, ChunHao Lin wrote: > For devices that support DASH, even DASH is disabled, there may still > exist a default firmware that will influence device behavior. > So driver needs to handle DASH for devices that support DASH, no > matter the DASH status is. > > This patch also prepares for "fix network lost after resume on DASH > systems". > > Fixes: ee7a1beb9759 ("r8169:call "rtl8168_driver_start" "rtl8168_driver_stop" only when hardware dash function is enabled") > Cc: stable@vger.kernel.org > Signed-off-by: ChunHao Lin <hau@realtek.com> > --- > drivers/net/ethernet/realtek/r8169_main.c | 35 ++++++++++++++++------- > 1 file changed, 25 insertions(+), 10 deletions(-) > > diff --git a/drivers/net/ethernet/realtek/r8169_main.c b/drivers/net/ethernet/realtek/r8169_main.c > index 0c76c162b8a9..4954ff0f72b1 100644 > --- a/drivers/net/ethernet/realtek/r8169_main.c > +++ b/drivers/net/ethernet/realtek/r8169_main.c > @@ -624,6 +624,7 @@ struct rtl8169_private { > > unsigned supports_gmii:1; > unsigned aspm_manageable:1; > + unsigned dash_enabled:1; > dma_addr_t counters_phys_addr; > struct rtl8169_counters *counters; > struct rtl8169_tc_offsets tc_offset; > @@ -1253,14 +1254,26 @@ static bool r8168ep_check_dash(struct rtl8169_private *tp) > return r8168ep_ocp_read(tp, 0x128) & BIT(0); > } > > -static enum rtl_dash_type rtl_check_dash(struct rtl8169_private *tp) > +static bool rtl_dash_is_enabled(struct rtl8169_private *tp) > +{ > + switch (tp->dash_type) { > + case RTL_DASH_DP: > + return r8168dp_check_dash(tp); > + case RTL_DASH_EP: > + return r8168ep_check_dash(tp); > + default: > + return false; > + } > +} > + > +static enum rtl_dash_type rtl_get_dash_type(struct rtl8169_private *tp) > { > switch (tp->mac_version) { > case RTL_GIGA_MAC_VER_28: > case RTL_GIGA_MAC_VER_31: > - return r8168dp_check_dash(tp) ? RTL_DASH_DP : RTL_DASH_NONE; > + return RTL_DASH_DP; > case RTL_GIGA_MAC_VER_51 ... RTL_GIGA_MAC_VER_53: > - return r8168ep_check_dash(tp) ? RTL_DASH_EP : RTL_DASH_NONE; > + return RTL_DASH_EP; > default: > return RTL_DASH_NONE; > } > @@ -1453,7 +1466,7 @@ static void __rtl8169_set_wol(struct rtl8169_private *tp, u32 wolopts) > > device_set_wakeup_enable(tp_to_dev(tp), wolopts); > > - if (tp->dash_type == RTL_DASH_NONE) { > + if (!tp->dash_enabled) { > rtl_set_d3_pll_down(tp, !wolopts); > tp->dev->wol_enabled = wolopts ? 1 : 0; > } > @@ -2512,7 +2525,7 @@ static void rtl_wol_enable_rx(struct rtl8169_private *tp) > > static void rtl_prepare_power_down(struct rtl8169_private *tp) > { > - if (tp->dash_type != RTL_DASH_NONE) > + if (tp->dash_enabled) > return; > > if (tp->mac_version == RTL_GIGA_MAC_VER_32 || > @@ -4869,7 +4882,7 @@ static int rtl8169_runtime_idle(struct device *device) > { > struct rtl8169_private *tp = dev_get_drvdata(device); > > - if (tp->dash_type != RTL_DASH_NONE) > + if (tp->dash_enabled) > return -EBUSY; > > if (!netif_running(tp->dev) || !netif_carrier_ok(tp->dev)) > @@ -4896,7 +4909,7 @@ static void rtl_shutdown(struct pci_dev *pdev) > rtl_rar_set(tp, tp->dev->perm_addr); > > if (system_state == SYSTEM_POWER_OFF && > - tp->dash_type == RTL_DASH_NONE) { > + !tp->dash_enabled) { Why break the line at all? Now the check fits the 80 char line limit. > pci_wake_from_d3(pdev, tp->saved_wolopts); > pci_set_power_state(pdev, PCI_D3hot); > } > @@ -5254,7 +5267,8 @@ static int rtl_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) > rc = pci_disable_link_state(pdev, PCIE_LINK_STATE_L1); > tp->aspm_manageable = !rc; > > - tp->dash_type = rtl_check_dash(tp); > + tp->dash_type = rtl_get_dash_type(tp); > + tp->dash_enabled = rtl_dash_is_enabled(tp); > > tp->cp_cmd = RTL_R16(tp, CPlusCmd) & CPCMD_MASK; > > @@ -5325,7 +5339,7 @@ static int rtl_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) > /* configure chip for default features */ > rtl8169_set_features(dev, dev->features); > > - if (tp->dash_type == RTL_DASH_NONE) { > + if (!tp->dash_enabled) { > rtl_set_d3_pll_down(tp, true); > } else { > rtl_set_d3_pll_down(tp, false); > @@ -5365,7 +5379,8 @@ static int rtl_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) > "ok" : "ko"); > > if (tp->dash_type != RTL_DASH_NONE) { > - netdev_info(dev, "DASH enabled\n"); > + netdev_info(dev, "DASH %s\n", > + tp->dash_enabled ? "enabled" : "disabled"); > rtl8168_driver_start(tp); > } > ^ permalink raw reply [flat|nested] 6+ messages in thread
* RE: [PATCH net v3 1/2] r8169: add handling DASH when DASH is disabled 2023-11-09 17:14 ` Heiner Kallweit @ 2023-11-09 17:29 ` Hau 0 siblings, 0 replies; 6+ messages in thread From: Hau @ 2023-11-09 17:29 UTC (permalink / raw) To: Heiner Kallweit Cc: nic_swsd, davem@davemloft.net, edumazet@google.com, kuba@kernel.org, pabeni@redhat.com, netdev@vger.kernel.org, linux-kernel@vger.kernel.org, stable@vger.kernel.org > > if (!netif_running(tp->dev) || !netif_carrier_ok(tp->dev)) @@ > > -4896,7 +4909,7 @@ static void rtl_shutdown(struct pci_dev *pdev) > > rtl_rar_set(tp, tp->dev->perm_addr); > > > > if (system_state == SYSTEM_POWER_OFF && > > - tp->dash_type == RTL_DASH_NONE) { > > + !tp->dash_enabled) { > > Why break the line at all? Now the check fits the 80 char line limit. I will correct this. Thanks. ^ permalink raw reply [flat|nested] 6+ messages in thread
* [PATCH net v3 2/2] r8169: fix network lost after resume on DASH systems 2023-11-09 16:43 [PATCH net v3 0/2] r8169: fix DASH devices network lost issue ChunHao Lin 2023-11-09 16:43 ` [PATCH net v3 1/2] r8169: add handling DASH when DASH is disabled ChunHao Lin @ 2023-11-09 16:43 ` ChunHao Lin 2023-11-09 17:15 ` Heiner Kallweit 1 sibling, 1 reply; 6+ messages in thread From: ChunHao Lin @ 2023-11-09 16:43 UTC (permalink / raw) To: hkallweit1 Cc: nic_swsd, davem, edumazet, kuba, pabeni, netdev, linux-kernel, ChunHao Lin, stable Device that support DASH may be reseted or powered off during suspend. So driver needs to handle DASH during system suspend and resume. Or DASH firmware will influence device behavior and causes network lost. Fixes: b646d90053f8 ("r8169: magic.") Cc: stable@vger.kernel.org Signed-off-by: ChunHao Lin <hau@realtek.com> --- drivers/net/ethernet/realtek/r8169_main.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/net/ethernet/realtek/r8169_main.c b/drivers/net/ethernet/realtek/r8169_main.c index 4954ff0f72b1..7e90dac2d97d 100644 --- a/drivers/net/ethernet/realtek/r8169_main.c +++ b/drivers/net/ethernet/realtek/r8169_main.c @@ -4661,10 +4661,16 @@ static void rtl8169_down(struct rtl8169_private *tp) rtl8169_cleanup(tp); rtl_disable_exit_l1(tp); rtl_prepare_power_down(tp); + + if (tp->dash_type != RTL_DASH_NONE) + rtl8168_driver_stop(tp); } static void rtl8169_up(struct rtl8169_private *tp) { + if (tp->dash_type != RTL_DASH_NONE) + rtl8168_driver_start(tp); + pci_set_master(tp->pci_dev); phy_init_hw(tp->phydev); phy_resume(tp->phydev); -- 2.39.2 ^ permalink raw reply related [flat|nested] 6+ messages in thread
* Re: [PATCH net v3 2/2] r8169: fix network lost after resume on DASH systems 2023-11-09 16:43 ` [PATCH net v3 2/2] r8169: fix network lost after resume on DASH systems ChunHao Lin @ 2023-11-09 17:15 ` Heiner Kallweit 0 siblings, 0 replies; 6+ messages in thread From: Heiner Kallweit @ 2023-11-09 17:15 UTC (permalink / raw) To: ChunHao Lin Cc: nic_swsd, davem, edumazet, kuba, pabeni, netdev, linux-kernel, stable On 09.11.2023 17:43, ChunHao Lin wrote: > Device that support DASH may be reseted or powered off during suspend. > So driver needs to handle DASH during system suspend and resume. Or > DASH firmware will influence device behavior and causes network lost. > > Fixes: b646d90053f8 ("r8169: magic.") > Cc: stable@vger.kernel.org > Signed-off-by: ChunHao Lin <hau@realtek.com> Reviewed-by: Heiner Kallweit <hkallweit1@gmail.com> ^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2023-11-09 17:30 UTC | newest] Thread overview: 6+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2023-11-09 16:43 [PATCH net v3 0/2] r8169: fix DASH devices network lost issue ChunHao Lin 2023-11-09 16:43 ` [PATCH net v3 1/2] r8169: add handling DASH when DASH is disabled ChunHao Lin 2023-11-09 17:14 ` Heiner Kallweit 2023-11-09 17:29 ` Hau 2023-11-09 16:43 ` [PATCH net v3 2/2] r8169: fix network lost after resume on DASH systems ChunHao Lin 2023-11-09 17:15 ` Heiner Kallweit
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).