[regression] r8169 without realtek_phy

[regression] r8169 without realtek_phy

[Jouni Mettälä]

There is network regression for me. 4.18 was good. 4.18+ is bad. There
was some phy changes in r8169 driver. Fortunately adding
CONFIG_REALTEK_PHY=m to kernel config fixed the regression. 

Should r8169 depend on realtek_phy? Does that breaks something else?

Network doesn't work with Generic PHY (output of dmesg)
Generic PHY r8169-300:00: attached PHY driver [Generic PHY]
(mii_bus:phy_addr=r8169-300:00, irq=IGNORE)

When realtek_phy is compiled, r8169 automatically uses it.
RTL8211B Gigabit Ethernet r8169-300:00: attached PHY driver [RTL8211B
Gigabit Ethernet] (mii_bus:phy_addr=r8169-300:00, irq=IGNORE)

Here is Ethernet controller's lspci for reference:
03:00.0 Ethernet controller [0200]: Realtek Semiconductor Co., Ltd.
RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller [10ec:8168]
(rev 01)
	Subsystem: ABIT Computer Corp. RTL8111/8168/8411 PCI Express
Gigabit Ethernet Controller [147b:1078]
	Flags: bus master, fast devsel, latency 0, IRQ 27
	I/O ports at ce00 [size=256]
	Memory at fddff000 (64-bit, non-prefetchable) [size=4K]
	[virtual] Expansion ROM at fdd00000 [disabled] [size=64K]
	Capabilities: [40] Power Management version 2
	Capabilities: [48] Vital Product Data
	Capabilities: [50] MSI: Enable+ Count=1/2 Maskable- 64bit+
	Capabilities: [60] Express Endpoint, MSI 00
	Capabilities: [84] Vendor Specific Information: Len=4c <?>
	Capabilities: [100] Advanced Error Reporting
	Capabilities: [12c] Virtual Channel
	Capabilities: [148] Device Serial Number 28-00-00-00-00-00-00-
00
	Capabilities: [154] Power Budgeting <?>
	Kernel driver in use: r8169
	Kernel modules: r8169

Re: [regression] r8169 without realtek_phy

[Florian Fainelli]

+Heiner,

On 08/17/2018 01:33 AM, Jouni Mettälä wrote:
> There is network regression for me. 4.18 was good. 4.18+ is bad. There
> was some phy changes in r8169 driver. Fortunately adding
> CONFIG_REALTEK_PHY=m to kernel config fixed the regression. 
> 
> Should r8169 depend on realtek_phy? Does that breaks something else?

That would be reasonable given that there is now a hard dependency on
the r8169 driver having the right PHY driver, since the Generic PHY
driver likely wont' do all the workarounds. Heiner, what do you think?

> 
> Network doesn't work with Generic PHY (output of dmesg)
> Generic PHY r8169-300:00: attached PHY driver [Generic PHY]
> (mii_bus:phy_addr=r8169-300:00, irq=IGNORE)
> 
> When realtek_phy is compiled, r8169 automatically uses it.
> RTL8211B Gigabit Ethernet r8169-300:00: attached PHY driver [RTL8211B
> Gigabit Ethernet] (mii_bus:phy_addr=r8169-300:00, irq=IGNORE)
> 
> Here is Ethernet controller's lspci for reference:
> 03:00.0 Ethernet controller [0200]: Realtek Semiconductor Co., Ltd.
> RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller [10ec:8168]
> (rev 01)
> 	Subsystem: ABIT Computer Corp. RTL8111/8168/8411 PCI Express
> Gigabit Ethernet Controller [147b:1078]
> 	Flags: bus master, fast devsel, latency 0, IRQ 27
> 	I/O ports at ce00 [size=256]
> 	Memory at fddff000 (64-bit, non-prefetchable) [size=4K]
> 	[virtual] Expansion ROM at fdd00000 [disabled] [size=64K]
> 	Capabilities: [40] Power Management version 2
> 	Capabilities: [48] Vital Product Data
> 	Capabilities: [50] MSI: Enable+ Count=1/2 Maskable- 64bit+
> 	Capabilities: [60] Express Endpoint, MSI 00
> 	Capabilities: [84] Vendor Specific Information: Len=4c <?>
> 	Capabilities: [100] Advanced Error Reporting
> 	Capabilities: [12c] Virtual Channel
> 	Capabilities: [148] Device Serial Number 28-00-00-00-00-00-00-
> 00
> 	Capabilities: [154] Power Budgeting <?>
> 	Kernel driver in use: r8169
> 	Kernel modules: r8169
> 


-- 
Florian

Re: [regression] r8169 without realtek_phy

[Heiner Kallweit]

On 17.08.2018 19:39, Florian Fainelli wrote:
> +Heiner,
> 
> On 08/17/2018 01:33 AM, Jouni Mettälä wrote:
>> There is network regression for me. 4.18 was good. 4.18+ is bad. There
>> was some phy changes in r8169 driver. Fortunately adding
>> CONFIG_REALTEK_PHY=m to kernel config fixed the regression. 
>>
>> Should r8169 depend on realtek_phy? Does that breaks something else?
> 
> That would be reasonable given that there is now a hard dependency on
> the r8169 driver having the right PHY driver, since the Generic PHY
> driver likely wont' do all the workarounds. Heiner, what do you think?
> 
Good catch, indeed we need a Kconfig dependency now. I missed this
because I had Realtek PHY support enabled anyway. I will submit a
fix adding this dependency.

Heiner
 
>
>> Network doesn't work with Generic PHY (output of dmesg)
>> Generic PHY r8169-300:00: attached PHY driver [Generic PHY]
>> (mii_bus:phy_addr=r8169-300:00, irq=IGNORE)
>>
>> When realtek_phy is compiled, r8169 automatically uses it.
>> RTL8211B Gigabit Ethernet r8169-300:00: attached PHY driver [RTL8211B
>> Gigabit Ethernet] (mii_bus:phy_addr=r8169-300:00, irq=IGNORE)
>>
>> Here is Ethernet controller's lspci for reference:
>> 03:00.0 Ethernet controller [0200]: Realtek Semiconductor Co., Ltd.
>> RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller [10ec:8168]
>> (rev 01)
>> 	Subsystem: ABIT Computer Corp. RTL8111/8168/8411 PCI Express
>> Gigabit Ethernet Controller [147b:1078]
>> 	Flags: bus master, fast devsel, latency 0, IRQ 27
>> 	I/O ports at ce00 [size=256]
>> 	Memory at fddff000 (64-bit, non-prefetchable) [size=4K]
>> 	[virtual] Expansion ROM at fdd00000 [disabled] [size=64K]
>> 	Capabilities: [40] Power Management version 2
>> 	Capabilities: [48] Vital Product Data
>> 	Capabilities: [50] MSI: Enable+ Count=1/2 Maskable- 64bit+
>> 	Capabilities: [60] Express Endpoint, MSI 00
>> 	Capabilities: [84] Vendor Specific Information: Len=4c <?>
>> 	Capabilities: [100] Advanced Error Reporting
>> 	Capabilities: [12c] Virtual Channel
>> 	Capabilities: [148] Device Serial Number 28-00-00-00-00-00-00-
>> 00
>> 	Capabilities: [154] Power Budgeting <?>
>> 	Kernel driver in use: r8169
>> 	Kernel modules: r8169
>>
> 
>