From: Heiner Kallweit <hkallweit1@gmail.com>
To: Joe Salmeri <jmscdba@gmail.com>,
"netdev@vger.kernel.org" <netdev@vger.kernel.org>
Subject: Re: Kernel Module r8169 and the Realtek 8126 PCIe 5 G/bps WIRED ethernet adapter
Date: Tue, 30 Jan 2024 20:59:00 +0100 [thread overview]
Message-ID: <208a69de-af5b-4624-85d5-86e87dfe6272@gmail.com> (raw)
In-Reply-To: <64b65025-792c-43c9-8ae5-22030264e374@gmail.com>
On 30.01.2024 17:34, Joe Salmeri wrote:
> On 1/29/24 17:19, Heiner Kallweit wrote:
>> On 29.01.2024 19:31, Joe Salmeri wrote:
>>> Hi,
>>>
>>> I recently built a new PC using the Asus z790 Maximus Formula motherboard.
>>>
>>> The z790 Formula uses the Realtek 8126 PCIe 5 G/bps WIRED ethernet adapter.
>>>
>>> I am using openSUSE Tumbleweed build 20231228 with kernel 6.6.7-1
>>>
>>> There does not seem to be a driver for the Realtek 8126.
>>>
>>> Here is the device info from "lspci | grep -i net"
>>>
>>> 04:00.0 Network controller: Intel Corporation Device 272b (rev 1a)
>>> 05:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. Device 8126 (rev 01)
>>>
>>> So it is detects the 8126 just fine it just doesn't have a driver for it.
>>>
>>> I checked realtek.com and found
>>>
>>> https://www.realtek.com/en/component/zoo/category/network-interface-controllers-10-100-1000m-gigabit-ethernet-pci-express-software
>>>
>>> The download link still says 8125 ( and kernel 6.4 ), but I compiled the source and since I have Secure boot enabled, I signed the
>>> resulting module file.
>>>
>>> The driver loads successfully and I now have wired networking and it has worked flawlessly for the last 2 months.
>>>
>>> I submitted a bug in Tumbleweed requesting support for the Realtek 8126 be added and was informed that the r8169 kernel module
>>> is what is used to support the older Realtek 8125 device.
>>>
>>> Since the drivers from Realtek seem to support both the r8125 and my newer r8126, the Tumbleweed support prepared a test
>>> kernel 6.6.7-1 for me where they added the PCI entry for the r8126 and I installed and tested it out.
>>>
>>> Although it does now load the r8169 module with their test kernel, the r8126 device still does not work.
>>>
>>> The only 2 lines that reference the r8169 in the dmesg log are these 2 lines:
>>>
>>> [ 3.237151] r8169 0000:05:00.0: enabling device (0000 -> 0003)
>>> [ 3.237289] r8169 0000:05:00.0: error -ENODEV: unknown chip XID 649, contact r8169 maintainers (see MAINTAINERS file)
>>>
>>> I reported the results of the test to Tumbleweed support and they said that additional tweaks will be needed for the r8169
>>> module to support the r8126 wired network adapter and thatn I should request to you to add support.
>>>
>>> The details of the openSUSE bug report on the issue can be found here:
>>>
>>> https://bugzilla.suse.com/show_bug.cgi?id=1217417
>>>
>>> Could we please get support added for the r8126 - Realtek 8126 PCIe 5 G/bps WIRED ethernet adapter added to the kernel ?
>>>
>> Thanks for the report. Actually it's not a bug report but a feature request.
>> Realtek provides no information about new chip versions and no data sheets, therefore the only
>> source of information is the r8125 vendor driver. Each chip requires a lot of version-specific
>> handling, therefore the first steps you described go in the right direction, but are by far not
>> sufficient. Patch below applies on linux-next, please test whether it works for you, and report back.
>>
>> Disclaimer:
>> r8125 references a firmware file that hasn't been provided to linux-firmware by Realtek yet.
>> Typically the firmware files tune PHY parameters to deal with compatibility issues.
>> In addition r8125 includes a lot of PHY tuning for RTL8126A.
>> Depending on cabling, link partner etc. the patch may work for you, or you may experience
>> link instability or worst case no link at all.
>>
>> Maybe RTL8126a also has a new integrated PHY version that isn't supported yet.
>> In this case the driver will complain with the following message and I'd need the PHY ID.
>> "no dedicated PHY driver found for PHY ID xxx"
>
> Thanks very much for your quick response.
>
> I forward your patch to the openSUSE people I have been working and they prepared a new test kernel 6.7.2 with the patches for for me to test.
>
> I just installed the test kernel provided with the patches but just as you expected it complains about no dedicated PHY driver found.
>
> Here is the dmesg | grep 8169 output with the information you requested
>
> [ 3.176753] r8169 0000:05:00.0: enabling device (0000 -> 0003)
> [ 3.184887] r8169 0000:05:00.0: no dedicated PHY driver found for PHY ID 0x001cc862, maybe realtek.ko needs to be added to initramfs?
> [ 3.184912] r8169: probe of 0000:05:00.0 failed with error -49
>
> Thank you for your efforts.
>
> Please let me know if you need any further details.
>
>> ---
>> drivers/net/ethernet/realtek/r8169.h | 1 +
>> drivers/net/ethernet/realtek/r8169_main.c | 91 +++++++++++++++----
>> .../net/ethernet/realtek/r8169_phy_config.c | 1 +
>> 3 files changed, 77 insertions(+), 16 deletions(-)
>>
>> diff --git a/drivers/net/ethernet/realtek/r8169.h b/drivers/net/ethernet/realtek/r8169.h
>> index 81567fcf3..c921456ed 100644
>> --- a/drivers/net/ethernet/realtek/r8169.h
>> +++ b/drivers/net/ethernet/realtek/r8169.h
>> @@ -68,6 +68,7 @@ enum mac_version {
>> /* support for RTL_GIGA_MAC_VER_60 has been removed */
>> RTL_GIGA_MAC_VER_61,
>> RTL_GIGA_MAC_VER_63,
>> + RTL_GIGA_MAC_VER_65,
>> RTL_GIGA_MAC_NONE
>> };
>> diff --git a/drivers/net/ethernet/realtek/r8169_main.c b/drivers/net/ethernet/realtek/r8169_main.c
>> index e0abdbcfa..ebf7a3b13 100644
>> --- a/drivers/net/ethernet/realtek/r8169_main.c
>> +++ b/drivers/net/ethernet/realtek/r8169_main.c
>> @@ -55,6 +55,7 @@
>> #define FIRMWARE_8107E_2 "rtl_nic/rtl8107e-2.fw"
>> #define FIRMWARE_8125A_3 "rtl_nic/rtl8125a-3.fw"
>> #define FIRMWARE_8125B_2 "rtl_nic/rtl8125b-2.fw"
>> +#define FIRMWARE_8126A_2 "rtl_nic/rtl8126a-2.fw"
>> #define TX_DMA_BURST 7 /* Maximum PCI burst, '7' is unlimited */
>> #define InterFrameGap 0x03 /* 3 means InterFrameGap = the shortest one */
>> @@ -136,6 +137,7 @@ static const struct {
>> [RTL_GIGA_MAC_VER_61] = {"RTL8125A", FIRMWARE_8125A_3},
>> /* reserve 62 for CFG_METHOD_4 in the vendor driver */
>> [RTL_GIGA_MAC_VER_63] = {"RTL8125B", FIRMWARE_8125B_2},
>> + [RTL_GIGA_MAC_VER_65] = {"RTL8126A", FIRMWARE_8126A_2},
>> };
>> static const struct pci_device_id rtl8169_pci_tbl[] = {
>> @@ -158,6 +160,7 @@ static const struct pci_device_id rtl8169_pci_tbl[] = {
>> { PCI_VENDOR_ID_LINKSYS, 0x1032, PCI_ANY_ID, 0x0024 },
>> { 0x0001, 0x8168, PCI_ANY_ID, 0x2410 },
>> { PCI_VDEVICE(REALTEK, 0x8125) },
>> + { PCI_VDEVICE(REALTEK, 0x8126) },
>> { PCI_VDEVICE(REALTEK, 0x3000) },
>> {}
>> };
>> @@ -327,8 +330,12 @@ enum rtl8168_registers {
>> };
>> enum rtl8125_registers {
>> + INT_CFG0_8125 = 0x34,
>> +#define INT_CFG0_ENABLE_8125 BIT(0)
>> +#define INT_CFG0_CLKREQEN BIT(3)
>> IntrMask_8125 = 0x38,
>> IntrStatus_8125 = 0x3c,
>> + INT_CFG1_8125 = 0x7a,
>> TxPoll_8125 = 0x90,
>> MAC0_BKP = 0x19e0,
>> EEE_TXIDLE_TIMER_8125 = 0x6048,
>> @@ -1139,7 +1146,7 @@ static void rtl_writephy(struct rtl8169_private *tp, int location, int val)
>> case RTL_GIGA_MAC_VER_31:
>> r8168dp_2_mdio_write(tp, location, val);
>> break;
>> - case RTL_GIGA_MAC_VER_40 ... RTL_GIGA_MAC_VER_63:
>> + case RTL_GIGA_MAC_VER_40 ... RTL_GIGA_MAC_VER_65:
>> r8168g_mdio_write(tp, location, val);
>> break;
>> default:
>> @@ -1154,7 +1161,7 @@ static int rtl_readphy(struct rtl8169_private *tp, int location)
>> case RTL_GIGA_MAC_VER_28:
>> case RTL_GIGA_MAC_VER_31:
>> return r8168dp_2_mdio_read(tp, location);
>> - case RTL_GIGA_MAC_VER_40 ... RTL_GIGA_MAC_VER_63:
>> + case RTL_GIGA_MAC_VER_40 ... RTL_GIGA_MAC_VER_65:
>> return r8168g_mdio_read(tp, location);
>> default:
>> return r8169_mdio_read(tp, location);
>> @@ -1507,7 +1514,7 @@ static void __rtl8169_set_wol(struct rtl8169_private *tp, u32 wolopts)
>> break;
>> case RTL_GIGA_MAC_VER_34:
>> case RTL_GIGA_MAC_VER_37:
>> - case RTL_GIGA_MAC_VER_39 ... RTL_GIGA_MAC_VER_63:
>> + case RTL_GIGA_MAC_VER_39 ... RTL_GIGA_MAC_VER_65:
>> if (wolopts)
>> rtl_mod_config2(tp, 0, PME_SIGNAL);
>> else
>> @@ -2073,6 +2080,9 @@ static enum mac_version rtl8169_get_mac_version(u16 xid, bool gmii)
>> u16 val;
>> enum mac_version ver;
>> } mac_info[] = {
>> + /* 8126A family. */
>> + { 0x7cf, 0x649, RTL_GIGA_MAC_VER_65 },
>> +
>> /* 8125B family. */
>> { 0x7cf, 0x641, RTL_GIGA_MAC_VER_63 },
>> @@ -2343,6 +2353,7 @@ static void rtl_init_rxcfg(struct rtl8169_private *tp)
>> RTL_W32(tp, RxConfig, RX_FETCH_DFLT_8125 | RX_DMA_BURST);
>> break;
>> case RTL_GIGA_MAC_VER_63:
>> + case RTL_GIGA_MAC_VER_65:
>> RTL_W32(tp, RxConfig, RX_FETCH_DFLT_8125 | RX_DMA_BURST |
>> RX_PAUSE_SLOT_ON);
>> break;
>> @@ -2772,7 +2783,7 @@ static void rtl_enable_exit_l1(struct rtl8169_private *tp)
>> case RTL_GIGA_MAC_VER_37 ... RTL_GIGA_MAC_VER_38:
>> rtl_eri_set_bits(tp, 0xd4, 0x0c00);
>> break;
>> - case RTL_GIGA_MAC_VER_40 ... RTL_GIGA_MAC_VER_63:
>> + case RTL_GIGA_MAC_VER_40 ... RTL_GIGA_MAC_VER_65:
>> r8168_mac_ocp_modify(tp, 0xc0ac, 0, 0x1f80);
>> break;
>> default:
>> @@ -2786,7 +2797,7 @@ static void rtl_disable_exit_l1(struct rtl8169_private *tp)
>> case RTL_GIGA_MAC_VER_34 ... RTL_GIGA_MAC_VER_38:
>> rtl_eri_clear_bits(tp, 0xd4, 0x1f00);
>> break;
>> - case RTL_GIGA_MAC_VER_40 ... RTL_GIGA_MAC_VER_63:
>> + case RTL_GIGA_MAC_VER_40 ... RTL_GIGA_MAC_VER_65:
>> r8168_mac_ocp_modify(tp, 0xc0ac, 0x1f80, 0);
>> break;
>> default:
>> @@ -2796,6 +2807,8 @@ static void rtl_disable_exit_l1(struct rtl8169_private *tp)
>> static void rtl_hw_aspm_clkreq_enable(struct rtl8169_private *tp, bool enable)
>> {
>> + u8 val8;
>> +
>> if (tp->mac_version < RTL_GIGA_MAC_VER_32)
>> return;
>> @@ -2809,11 +2822,19 @@ static void rtl_hw_aspm_clkreq_enable(struct rtl8169_private *tp, bool enable)
>> return;
>> rtl_mod_config5(tp, 0, ASPM_en);
>> - rtl_mod_config2(tp, 0, ClkReqEn);
>> + switch (tp->mac_version) {
>> + case RTL_GIGA_MAC_VER_65:
>> + val8 = RTL_R8(tp, INT_CFG0_8125) | INT_CFG0_CLKREQEN;
>> + RTL_W8(tp, INT_CFG0_8125, val8);
>> + break;
>> + default:
>> + rtl_mod_config2(tp, 0, ClkReqEn);
>> + break;
>> + }
>> switch (tp->mac_version) {
>> case RTL_GIGA_MAC_VER_46 ... RTL_GIGA_MAC_VER_48:
>> - case RTL_GIGA_MAC_VER_61 ... RTL_GIGA_MAC_VER_63:
>> + case RTL_GIGA_MAC_VER_61 ... RTL_GIGA_MAC_VER_65:
>> /* reset ephy tx/rx disable timer */
>> r8168_mac_ocp_modify(tp, 0xe094, 0xff00, 0);
>> /* chip can trigger L1.2 */
>> @@ -2825,14 +2846,22 @@ static void rtl_hw_aspm_clkreq_enable(struct rtl8169_private *tp, bool enable)
>> } else {
>> switch (tp->mac_version) {
>> case RTL_GIGA_MAC_VER_46 ... RTL_GIGA_MAC_VER_48:
>> - case RTL_GIGA_MAC_VER_61 ... RTL_GIGA_MAC_VER_63:
>> + case RTL_GIGA_MAC_VER_61 ... RTL_GIGA_MAC_VER_65:
>> r8168_mac_ocp_modify(tp, 0xe092, 0x00ff, 0);
>> break;
>> default:
>> break;
>> }
>> - rtl_mod_config2(tp, ClkReqEn, 0);
>> + switch (tp->mac_version) {
>> + case RTL_GIGA_MAC_VER_65:
>> + val8 = RTL_R8(tp, INT_CFG0_8125) & ~INT_CFG0_CLKREQEN;
>> + RTL_W8(tp, INT_CFG0_8125, val8);
>> + break;
>> + default:
>> + rtl_mod_config2(tp, ClkReqEn, 0);
>> + break;
>> + }
>> rtl_mod_config5(tp, ASPM_en, 0);
>> }
>> }
>> @@ -3545,10 +3574,15 @@ static void rtl_hw_start_8125_common(struct rtl8169_private *tp)
>> /* disable new tx descriptor format */
>> r8168_mac_ocp_modify(tp, 0xeb58, 0x0001, 0x0000);
>> - if (tp->mac_version == RTL_GIGA_MAC_VER_63)
>> + if (tp->mac_version == RTL_GIGA_MAC_VER_65)
>> + RTL_W8(tp, 0xD8, RTL_R8(tp, 0xD8) & ~0x02);
>> +
>> + if (tp->mac_version == RTL_GIGA_MAC_VER_65)
>> + r8168_mac_ocp_modify(tp, 0xe614, 0x0700, 0x0400);
>> + else if (tp->mac_version == RTL_GIGA_MAC_VER_63)
>> r8168_mac_ocp_modify(tp, 0xe614, 0x0700, 0x0200);
>> else
>> - r8168_mac_ocp_modify(tp, 0xe614, 0x0700, 0x0400);
>> + r8168_mac_ocp_modify(tp, 0xe614, 0x0700, 0x0300);
>> if (tp->mac_version == RTL_GIGA_MAC_VER_63)
>> r8168_mac_ocp_modify(tp, 0xe63e, 0x0c30, 0x0000);
>> @@ -3561,6 +3595,10 @@ static void rtl_hw_start_8125_common(struct rtl8169_private *tp)
>> r8168_mac_ocp_modify(tp, 0xe056, 0x00f0, 0x0030);
>> r8168_mac_ocp_modify(tp, 0xe040, 0x1000, 0x0000);
>> r8168_mac_ocp_modify(tp, 0xea1c, 0x0003, 0x0001);
>> + if (tp->mac_version == RTL_GIGA_MAC_VER_65)
>> + r8168_mac_ocp_modify(tp, 0xea1c, 0x0300, 0x0000);
>> + else
>> + r8168_mac_ocp_modify(tp, 0xea1c, 0x0004, 0x0000);
>> r8168_mac_ocp_modify(tp, 0xe0c0, 0x4f0f, 0x4403);
>> r8168_mac_ocp_modify(tp, 0xe052, 0x0080, 0x0068);
>> r8168_mac_ocp_modify(tp, 0xd430, 0x0fff, 0x047f);
>> @@ -3575,10 +3613,10 @@ static void rtl_hw_start_8125_common(struct rtl8169_private *tp)
>> rtl_loop_wait_low(tp, &rtl_mac_ocp_e00e_cond, 1000, 10);
>> - if (tp->mac_version == RTL_GIGA_MAC_VER_63)
>> - rtl8125b_config_eee_mac(tp);
>> - else
>> + if (tp->mac_version == RTL_GIGA_MAC_VER_61)
>> rtl8125a_config_eee_mac(tp);
>> + else
>> + rtl8125b_config_eee_mac(tp);
>> rtl_disable_rxdvgate(tp);
>> }
>> @@ -3622,6 +3660,12 @@ static void rtl_hw_start_8125b(struct rtl8169_private *tp)
>> rtl_hw_start_8125_common(tp);
>> }
>> +static void rtl_hw_start_8126a(struct rtl8169_private *tp)
>> +{
>> + rtl_set_def_aspm_entry_latency(tp);
>> + rtl_hw_start_8125_common(tp);
>> +}
>> +
>> static void rtl_hw_config(struct rtl8169_private *tp)
>> {
>> static const rtl_generic_fct hw_configs[] = {
>> @@ -3664,6 +3708,7 @@ static void rtl_hw_config(struct rtl8169_private *tp)
>> [RTL_GIGA_MAC_VER_53] = rtl_hw_start_8117,
>> [RTL_GIGA_MAC_VER_61] = rtl_hw_start_8125a_2,
>> [RTL_GIGA_MAC_VER_63] = rtl_hw_start_8125b,
>> + [RTL_GIGA_MAC_VER_65] = rtl_hw_start_8126a,
>> };
>> if (hw_configs[tp->mac_version])
>> @@ -3674,9 +3719,23 @@ static void rtl_hw_start_8125(struct rtl8169_private *tp)
>> {
>> int i;
>> + RTL_W8(tp, INT_CFG0_8125, 0x00);
>> +
>> /* disable interrupt coalescing */
>> - for (i = 0xa00; i < 0xb00; i += 4)
>> - RTL_W32(tp, i, 0);
>> + switch (tp->mac_version) {
>> + case RTL_GIGA_MAC_VER_61:
>> + for (i = 0xa00; i < 0xb00; i += 4)
>> + RTL_W32(tp, i, 0);
>> + break;
>> + case RTL_GIGA_MAC_VER_63:
>> + case RTL_GIGA_MAC_VER_65:
>> + for (i = 0xa00; i < 0xa80; i += 4)
>> + RTL_W32(tp, i, 0);
>> + RTL_W16(tp, INT_CFG1_8125, 0x0000);
>> + break;
>> + default:
>> + break;
>> + }
>> rtl_hw_config(tp);
>> }
>> diff --git a/drivers/net/ethernet/realtek/r8169_phy_config.c b/drivers/net/ethernet/realtek/r8169_phy_config.c
>> index b50f16786..badf78f81 100644
>> --- a/drivers/net/ethernet/realtek/r8169_phy_config.c
>> +++ b/drivers/net/ethernet/realtek/r8169_phy_config.c
>> @@ -1152,6 +1152,7 @@ void r8169_hw_phy_config(struct rtl8169_private *tp, struct phy_device *phydev,
>> [RTL_GIGA_MAC_VER_53] = rtl8117_hw_phy_config,
>> [RTL_GIGA_MAC_VER_61] = rtl8125a_2_hw_phy_config,
>> [RTL_GIGA_MAC_VER_63] = rtl8125b_hw_phy_config,
>> + [RTL_GIGA_MAC_VER_65] = NULL,
>> };
>> if (phy_configs[ver])
>
>
The followoing adds support for the integrated PHY.
Please apply it on-top and re-test.
---
drivers/net/phy/realtek.c | 10 ++++++++++
1 file changed, 10 insertions(+)
diff --git a/drivers/net/phy/realtek.c b/drivers/net/phy/realtek.c
index 894172a3e..132784321 100644
--- a/drivers/net/phy/realtek.c
+++ b/drivers/net/phy/realtek.c
@@ -1047,6 +1047,16 @@ static struct phy_driver realtek_drvs[] = {
.resume = rtlgen_resume,
.read_page = rtl821x_read_page,
.write_page = rtl821x_write_page,
+ }, {
+ PHY_ID_MATCH_EXACT(0x001cc862),
+ .name = "RTL8251B 5Gbps PHY",
+ .get_features = rtl822x_get_features,
+ .config_aneg = rtl822x_config_aneg,
+ .read_status = rtl822x_read_status,
+ .suspend = genphy_suspend,
+ .resume = rtlgen_resume,
+ .read_page = rtl821x_read_page,
+ .write_page = rtl821x_write_page,
}, {
PHY_ID_MATCH_EXACT(0x001cc961),
.name = "RTL8366RB Gigabit Ethernet",
--
2.43.0
next prev parent reply other threads:[~2024-01-30 19:59 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <edabbc1f-5440-4170-83a4-f436a6d04f76@gmail.com>
2024-01-30 16:34 ` Kernel Module r8169 and the Realtek 8126 PCIe 5 G/bps WIRED ethernet adapter Joe Salmeri
2024-01-30 19:59 ` Heiner Kallweit [this message]
2024-01-31 16:14 ` Joe Salmeri
2024-01-31 20:40 ` Heiner Kallweit
2024-01-31 23:36 ` Joe Salmeri
2024-02-01 7:09 ` Heiner Kallweit
2024-02-01 7:19 ` Takashi Iwai
2024-02-09 22:01 ` Heiner Kallweit
2024-02-10 8:11 ` Takashi Iwai
[not found] <ab516cfd-24e8-4761-93a3-49f96a4e6b8f@gmail.com>
2024-01-29 22:19 ` Heiner Kallweit
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=208a69de-af5b-4624-85d5-86e87dfe6272@gmail.com \
--to=hkallweit1@gmail.com \
--cc=jmscdba@gmail.com \
--cc=netdev@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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 read-only IMAP folder(s) and NNTP newsgroup(s).