Sfp and Link status work #39

logicog · 2025-11-04T20:16:49Z

This fixes several issues with the existing SFP code and the link status code. It now provides full support for 2 SFP ports and uses also the link status register for ports 8 and 9 correctly.
This was tested on 5+1, 4+2 and 8+1 devices.
An example output now is e.g.:

> stat

 Port   State   Link    TxGood          TxBad           RxGood          RxBad
5       SFP OK  Down    0x00000000      0x00000000      0x00000000      0x00000000
1       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
2       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
3       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
4       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
6       SFP OK  10G     0x0000002d      0x00000000      0x00000000      0x00000000

> sfp

Slot 0 - Rate: 67  Encoding: 06
OEM             SFP-10G-SR      2   

Slot 1 - Rate: 0d  Encoding: 01
OEM             SFP-GE-SX-MM850

For reasons of code complexity (use of additional variables), the ports are not in sequence on RTL8272 devices, only for RTL8273 this is the case.

vDorst · 2025-12-07T09:16:54Z

@logicog can you rebase this PR, so I can check it again.

logicog · 2025-12-07T09:54:25Z

On it. Need to push my work on IGMP aside, first.

logicog · 2025-12-07T10:00:15Z

Done.

vDorst · 2025-12-07T11:19:18Z

Website and console don't match.
Console stat is correct.
I do have a SFP, 10Gbit link and RJ45, 1gbit, connected.

status.json

[{"portNum":5,"isSFP":1,"enabled":1,"link":4,"txG":"0x056","txB":"0x0","rxG":"0x033","rxB":"0x0"},{"portNum":1,"isSFP":0,"enabled":1,"link":0,"txG":"0x00","txB":"0x0","rxG":"0x00","rxB":"0x0"},{"portNum":2,"isSFP":0,"enabled":1,"link":2,"txG":"0x0e1","txB":"0x0","rxG":"0x0165","rxB":"0x0"},{"portNum":3,"isSFP":0,"enabled":1,"link":0,"txG":"0x00","txB":"0x0","rxG":"0x00","rxB":"0x0"},{"portNum":4,"isSFP":0,"enabled":1,"link":0,"txG":"0x00","txB":"0x0","rxG":"0x00","rxB":"0x0"},{"portNum":6,"isSFP":1,"enabled":0,"link":1,"txG":"0x00","txB":"0x0","rxG":"0x00","rxB":"0x0"}]

Console

stat

 Port   State   Link    TxGood          TxBad           RxGood          RxBad
5       SFP OK  10G     0x00000035      0x00000000      0x0000005a      0x00000000
1       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
2       On      1000M   0x000001ce      0x00000000      0x00000134      0x00000000
3       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
4       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
6       NO SFP  Down    0x00000000      0x00000000      0x00000000      0x00000000

logicog · 2025-12-07T11:21:46Z

Was the website correct before? The code hopefully only changes the behaviour of the console.

vDorst · 2025-12-07T15:29:37Z

Nope. Result from the current main fccafe8

status.json:

[{"portNum":5,"isSFP":1,"enabled":1,"link":4,"txG":"0x0d","txB":"0x0","rxG":"0x017","rxB":"0x0"},{"portNum":1,"isSFP":0,"enabled":1,"link":0,"txG":"0x00","txB":"0x0","rxG":"0x00","rxB":"0x0"},{"portNum":2,"isSFP":0,"enabled":1,"link":2,"txG":"0x06b","txB":"0x0","rxG":"0x08b","rxB":"0x0"},{"portNum":3,"isSFP":0,"enabled":1,"link":0,"txG":"0x00","txB":"0x0","rxG":"0x00","rxB":"0x0"},{"portNum":4,"isSFP":0,"enabled":1,"link":0,"txG":"0x00","txB":"0x0","rxG":"0x00","rxB":"0x0"},{"portNum":6,"isSFP":1,"enabled":0,"link":1,"txG":"0x00","txB":"0x0","rxG":"0x00","rxB":"0x0"}]

Console stat

> stat

 Port   State   Link    TxGood          TxBad           RxGood          RxBad
4       Off     Up      0x0000003b      0x00000000      0x00000015      0x00000000
5       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
6       On      1000M   0x000000d1      0x00000000      0x000000c2      0x00000000
7       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
8       NO SFP  Down    0x00000000      0x00000000      0x00000000      0x00000000
9       NO SFP  Down    0x00000000      0x00000000      0x00000000      0x00000000

logicog · 2025-12-07T16:23:40Z

I'll add a fix for the website, too, then, now that we know that the new code in principle works. I'll try to do this by tomorrow.

This fixes handling of the SFP ports in the stat command. The output of the command now is for physical ports, which however are not shown in sequence.

The RTL8272/3 use 2 32 bit registers to indicate the MAC link speed, the second register is used for ports 8 and 9, where 9 is the CPU port. Use both registers.

This adds the speed status display of other link speeds: 100M 500M 1000M: Green 2nd LED 2.5G, 5G, 10G: Orange 2nd LED

logicog · 2025-12-08T07:44:12Z

The website seems to work now, too. There had been no support for the 2nd SFP slot and then also the 10G link speed display had been forgotten. I also added the now identifiable link speeds of 5G and 500M, which the switches support in HW (but not in code due to the lack of a 5G module, however one could test 500M connections, but that is an RTL proprietary link speed and hardly found anywhere, so I am not overly enthusiastic).

logicog · 2025-12-08T11:29:29Z

I also added SFP information in the send_basic_info() function, which is shown on the Switch Configuration web-page:

vDorst · 2025-12-08T14:05:33Z

Do you think the switch also supports the 2.5gbit-lite version?

I like the idea but I think that I like a seprat windows/tooltip better.
Example: I saw that Ubuiquiti UDM Pro Max seems to show it when you hover over the port ICON like this

Daniel has I think most of the code already #35 (comment)

logicog · 2025-12-08T14:55:45Z

Yes, the switch does support 2.5GBit-lite, as it is a feature of the RTL8221B and RTL8224 which is probably at the heart of the RTL8372. At the SerDes one gets 1000MBit, the speed setting does not change. However, there is a difference in the SDS configuration. But that is an RTL proprietary format, like 500MBit so not sure about the usefulness.

What I know is that 2.5GBit modules work. One needs the right modules, though. It should also support 5GBit, but I do not have such a module to test, they are rare and expensive. It requires different SDS settings, so without some additional code, it will probably not work.

Admittedly, the tool-tip looks seriously cool, if not to say geeky. I already had though about tool-tips for the ports. I'll work on it...

vDorst · 2025-12-08T21:17:15Z

SFP ports are switched.

> stat

 Port   State   Link    TxGood          TxBad           RxGood          RxBad
5       NO SFP  Down    0x00000000      0x00000000      0x00000000      0x00000000
1       On      Down    0x00000000      0x00000000      0x00000001      0x00000000
2       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
3       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
4       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
6       SFP OK  10G     0x00000331      0x00000000      0x00000223      0x00000000

Left port is used right now.

logicog · 2025-12-09T06:20:40Z

Not for me. The SFP-Module is in the left slot of keeplink KP-9000-6XHML-X2:

> stat

 Port   State   Link    TxGood          TxBad           RxGood          RxBad
5       SFP OK  10G     0x00000099      0x00000000      0x00000415      0x00000000
1       On      1000M   0x000001e8      0x00000000      0x00000476      0x00000000
2       On      1000M   0x0000035d      0x00000000      0x00000162      0x00000000
3       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
4       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
6       NO SFP  Down    0x00000000      0x00000000      0x00000000      0x00000000

Note, that on the web, for you sfp_slot_0, which is the 0th I2C connection is correct, but for me, not. So at least in both cases logical port 5 is associated with I2C- connection 1, So for now, we only have a problem with geometry, not with association of I2C-port to SDS-lane or to the GPIOs.
So what about introducing say a MACHINE_KL_9000_6XHML_X2, and MACHINE_SWGT024_V2_0 which use a different geometry? Something like

In rtl837x_common.h:
#define MACHINE_KL_9000_6XHML_X2
// #define MACHINE_SWGT024_V2_0
#ifdef MACHINE_KL_9000_6XHML_X2
#define GEOMETRY_PHY_3_IS_PORT_5
#endif

In rtl837x_port.c:
#ifdef GEOMETRY_PHY_3_IS_PORT_5
__code uint8_t log_to_phys_port[9] = {
	0, 0, 0, 5, 1, 2, 3, 4, 6
};
#else
__code uint8_t log_to_phys_port[9] = {
	0, 0, 0, 6, 1, 2, 3, 4, 5
};
#endif

The Problem is that the web-interface also has such definitions in port.js:

const logToPhysPort = [0, 0, 0, 5, 1, 2, 3, 4, 6];
const physToLogPort = [	4, 5, 6, 7, 3, 8];

So this is a bigger mess. And I now know why I had a terrible feeling when introducing PHY-port mappings to the web-application.
We could run the javascript through the C-Preprocessor when compiling the image...
Or we could give a geometry hint in the status.json...

vDorst · 2025-12-09T10:45:24Z

Note, that on the web, for you sfp_slot_0, which is the 0th I2C connection is correct, but for me, not. So at least in both cases logical port 5 is associated with I2C- connection 1, So for now, we only have a problem with geometry, not with association of I2C-port to SDS-lane or to the GPIOs.

I can provide the link-changes prints, so we can figureout which port is connected which sds and also document it on the wiki.

So what about introducing say a MACHINE_KL_9000_6XHML_X2, and MACHINE_SWGT024_V2_0 which use a different geometry?

Can we put characteristic in a struct instead of define so it is more unified?
I see examples in the linux kernel like this
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/drivers/net/ethernet/mediatek/mtk_eth_soc.c?h=v6.17.11#n5593 and https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/drivers/net/ethernet/mediatek/mtk_eth_soc.c?h=v6.17.11#n5570

Or we could give a geometry hint in the status.json...
This is good idea.

logicog · 2025-12-09T10:59:38Z

I can provide the link-changes prints, so we can figureout which port is connected which sds and also document it on the wiki.
At this point I believe this is not the problem, unless one of both of the 2 SFP-ports does not work for you.

I like the structures. This is allows compile-time verification. It may make the code a bit larger, but since we do not have an issue with the flash-size, this is irrelevant.

I will provide a proposal using structures, later today (finally got IGMP working, need to wrap that up while it is fresh in my mind). Note that we will still need some defines, because we cannot auto-detect geometry, this has to be input to the compiler, the structures in the kernel you mention are based on auto-detection of the SoC. Device-tree btw would probably be the ideal solution, as it creates structures in the build image, which are compile-time dependent, but this comes with an enormous amount of additionally required infrastructure.

Because we go for the structure, I will also add information on geometry to the .json sent to the client which ends up in a struct there.

vDorst · 2025-12-09T11:12:55Z

The "auto-detect" can just be a config-setting.
When the device information is to much we can put it in a bank, and only copy the used settings in to XDATA memory or program it in the "CODE"-bank.
And frequently used settings we can add flags in SRAM, so that the code-size to fetch a setting don't explode or consuming to much cycles.

vDorst · 2025-12-09T19:00:44Z

Here the link status log.

Started left at the first RJ45 port.
insert and remove the cable and SFP port also sfp-module.

Detecting CPU: RTL8372
Software version: v0.1.0-g06fbf85

Starting up...
  Flash controller
NIC reset

rtl8372_init called
Configuring 2nd SFP+ port

phy_config: 8
  phy config done

phy_config: 3
  phy config done

rtl8372_init done

vlan_setup called 
vlan_setup done 

port_l2_setup called

port_l2_forget called
port_l2_forget done

port_l2_setup done
uip_arp_init called
A minimal prompt to explore the RTL8372:

Clock register: 0x00001101
Register 0x7b20/RTL837X_REG_SDS_MODES: 0x00000bff
Verifying PHY settings:

 Port   State   Link    TxGood          TxBad           RxGood          RxBad
5       NO SFP  Down    0x00000000      0x00000000      0x00000000      0x00000000
1       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
2       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
3       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
4       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
6       NO SFP  Down    0x00000000      0x00000000      0x00000000      0x00000000
Got ip command: c0a80af7
Got gw command: c0a80a01
Got netmask command: ffffff00

> 
<new link: 2000000000, was 0000000000>


> 
<new link: 2000020000, was 2000000000>
ip: invalid version or header length.ip: invalid version or header length.

> stat

 Port   State   Link    TxGood          TxBad           RxGood          RxBad
5       NO SFP  Down    0x00000000      0x00000000      0x00000000      0x00000000
1       On      1000M   0x0000000e      0x00000000      0x00000000      0x00000000
2       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
3       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
4       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
6       NO SFP  Down    0x00000000      0x00000000      0x00000000      0x00000000

> 
<new link: 2000000000, was 2000020000>

<new link: 2000200000, was 2000000000>


> stat

 Port   State   Link    TxGood          TxBad           RxGood          RxBad
5       NO SFP  Down    0x00000000      0x00000000      0x00000000      0x00000000
1       On      Down    0x0000000e      0x00000000      0x00000000      0x00000000
2       On      1000M   0x00000000      0x00000000      0x00000000      0x00000000
3       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
4       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
6       NO SFP  Down    0x00000000      0x00000000      0x00000000      0x00000000

> 
<new link: 2000000000, was 2000200000>

<new link: 2002000000, was 2000000000>
stat

 Port   State   Link    TxGood          TxBad           RxGood          RxBad
5       NO SFP  Down    0x00000000      0x00000000      0x00000000      0x00000000
1       On      Down    0x0000000e      0x00000000      0x00000000      0x00000000
2       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
3       On      1000M   0x00000000      0x00000000      0x00000000      0x00000000
4       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
6       NO SFP  Down    0x00000000      0x00000000      0x00000000      0x00000000

> 
<new link: 2000000000, was 2002000000>

<new link: 2020000000, was 2000000000>
stat

 Port   State   Link    TxGood          TxBad           RxGood          RxBad
5       NO SFP  Down    0x00000000      0x00000000      0x00000000      0x00000000
1       On      Down    0x0000000e      0x00000000      0x00000000      0x00000000
2       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
3       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
4       On      1000M   0x00000000      0x00000000      0x00000000      0x00000000
6       NO SFP  Down    0x00000000      0x00000000      0x00000000      0x00000000

> 
<new link: 2000000000, was 2020000000>

<MODULE INSERTED>  Rate: 67  Encoding: 06

DELL            WTRD1           A0  
sds_config sds: 01, mode: 1a

RTL837X_REG_SDS_MODES: 0x0000035f

Trying to set SDS mode to 0x1a

<new link: 2400000000, was 2000000000>
ip: invalid version or header length.ip: invalid version or header length.

> stat

 Port   State   Link    TxGood          TxBad           RxGood          RxBad
5       NO SFP  Down    0x00000000      0x00000000      0x00000000      0x00000000
1       On      Down    0x0000000e      0x00000000      0x00000000      0x00000000
2       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
3       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
4       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
6       SFP OK  10G     0x00000012      0x00000000      0x00000000      0x00000000

> 
<new link: 2000000000, was 2400000000>

<MODULE REMOVED>


> 
<MODULE 2 INSERTED>  Rate: 67  Encoding: 06
DELL            WTRD1           A0  
sds_config sds: 00, mode: 1a

RTL837X_REG_SDS_MODES: 0x0000035a

Trying to set SDS mode to 0x1a

<new link: 2000004000, was 2000000000>
ip: invalid version or header length.ip: invalid version or header length.

> stat

 Port   State   Link    TxGood          TxBad           RxGood          RxBad
5       SFP OK  10G     0x00000011      0x00000000      0x00000000      0x00000000
1       On      Down    0x0000000e      0x00000000      0x00000000      0x00000000
2       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
3       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
4       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
6       NO SFP  Down    0x00000012      0x00000000      0x00000000      0x00000000

logicog · 2025-12-10T14:28:52Z

I pushed an initial implementation. It is more for feedback and not very well tested, so far only on the 4+2 device. But it is a bit of work and easy to get wrong, so better get feedback early.
There are lots of things that still can be done, such as using SFP-pins, there is nothing yet about the I2C pins. and looking at #40 there may be the need to have SDS configuration in.

vDorst · 2025-12-10T15:14:14Z

@logicog Thanks, you did not added machine.h and machine.c to the commit.

logicog · 2025-12-13T18:57:46Z

There were a couple of issues in c5995a4 and 001c52e which somehow cancelled each other out to only break the the RTL8221B-port plus the speed settings for the links, so this was not immediately obvious when testing. The RTL8221B works again with above fix. Now checking the speed settings.

vDorst · 2025-12-14T10:41:40Z

I updated to latest commit and added the fix GPIO settings,readout and document to it.
See https://github.com/vDorst/RTLPlayground/commits/sfp_work5/.
I had to swap the sds ports other I did not get a link on right sfp port.

Both modules do detect properly and get a link.
Also 1G-RJ45-SFP module works a 1Gbit.

There were a couple of issues in....

Great that you found it!

logicog · 2025-12-14T10:47:01Z

This looks great! We are going for fully flexible device support!

I just pushed support for link speed settings, try:

> port 1 1g
> port 1 duplex off
> port 1 duplex on
> port 1 show

I want to read out the SFP module parameters such as temperature, next and send the link state info and the SFP info to the web.,

logicog · 2025-12-14T10:48:12Z

I'll also pull in your fixes for the SWGT024.

logicog · 2025-12-14T10:54:00Z

I think I got the latest changes from you into the current commits.

vDorst · 2025-12-14T10:56:12Z

I just pushed support for link speed settings, try:

Maybe put this in a separate PR?

So we can merge the current work.

logicog · 2025-12-14T11:00:51Z

Can we get the link speed in? This was already in the code, it was just broken, because of incorrect translation of physical port number to logical port number. OK, I added link partner display. But leaving it without the last commit would keep the code broken. I will start a new PR for the Web display and SFP properties, ok?

vDorst · 2025-12-14T11:02:20Z

Ok, that is fine.

logicog · 2025-12-14T11:03:07Z

Cool!

vDorst · 2025-12-14T11:12:29Z

Correct that the SFP module don't show this information yet?

> port 6 show

PORT 08
Forced speed: 0x0000
s1, s2: 00 00
10M
Duplex: 0x0000 enabled: no

Link Partner advertises:

> stat

 Port   State   Link    TxGood          TxBad           RxGood          RxBad
6       SFP OK  10G     0x00000019      0x00000000      0x00000042      0x00000000
1       On      100M    0x00000114      0x00000000      0x000000a7      0x00000000
2       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
3       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
4       On      Down    0x00000000      0x00000000      0x00000000      0x00000000
5       NO SFP  Down    0x0000000e      0x00000000      0x0000000c      0x00000000

logicog · 2025-12-14T11:14:33Z

Yes. This is entirely PHY information. But I need to make sure it is not called for an SFP module. We shall think about Ethernet-SFP modules later ;-)

vDorst · 2025-12-14T11:22:54Z

This means I can't force the SFP port to let say 1gbit or 2.5bit if I want/need to?

logicog · 2025-12-14T11:24:25Z

No, you cannot. If you put a 1Gbit module on one side and the other side has a 2.5Gbit module which is able to support 1Gbit (which is not necessarily the case), you can make the 2.5Gbit module work at 1Gbit.

logicog · 2025-12-14T11:32:12Z

https://superuser.com/questions/1731301/why-does-auto-port-speed-negotiation-not-work-on-sfp-fiber-uplinks
However, you can force an SFP port sometimes to use a different speed. This works by changing the SerDes frequency. The module needs to support this, even if the frequency is lower than its maximum frequency because of things like PLLs that are involved.

vDorst · 2025-12-14T11:37:26Z

I am fine to make that in a separate PR.
It is up to you, when you want this in this this PR.
Because I am ready to merge.

logicog · 2025-12-14T11:40:17Z

Then let's merge.

vDorst · 2025-12-14T11:53:36Z

Done, great job!

logicog · 2025-12-14T11:54:19Z

Thanks the great collaboration on this!

This was referenced Nov 4, 2025

Link detection registers #38

Open

Testing firmware on OFS-XF2GT4 #35

Open

Do not show non-ascii Transceiver field in SFP module description

32b5539

logicog force-pushed the sfp_work branch from 05e4cd3 to e677346 Compare December 7, 2025 09:59

logicog added 5 commits December 8, 2025 08:39

Correct handling of SFP ports in stat command

e9b8ab5

This fixes handling of the SFP ports in the stat command. The output of the command now is for physical ports, which however are not shown in sequence.

Fix link change handling

101b493

The RTL8272/3 use 2 32 bit registers to indicate the MAC link speed, the second register is used for ports 8 and 9, where 9 is the CPU port. Use both registers.

Detect 2nd SFP-slot RX-Los on KP-9000-6XHML-2X

636981c

Add support for 2nd SFP port in stats

ab6a66a

Add support for 10G 500M and 5G speed display

3c869e9

This adds the speed status display of other link speeds: 100M 500M 1000M: Green 2nd LED 2.5G, 5G, 10G: Orange 2nd LED

logicog force-pushed the sfp_work branch from e677346 to 3c869e9 Compare December 8, 2025 07:40

Add SFP information in send_basic_info()

06fbf85

logicog force-pushed the sfp_work branch from 06e6bb3 to 9aaac1a Compare December 13, 2025 17:28

logicog and others added 5 commits December 13, 2025 22:22

Support keeplink 5+1 device

e6cf97b

Add machine name in send_basic_info()

74364ed

Move RTL8372 SMI address configuration

d71a303

Fix PHY issues

28f832a

Fixed to make SWGT024 V2.0 work

11d2c84

Add more options to Link configuration and status display

2ed9974

logicog force-pushed the sfp_work branch from 9aaac1a to 2ed9974 Compare December 14, 2025 10:46

Fixed to make SWGT024 V2.0 work

0197c9b

Make sure port is not called for an SFP module

b414520

vDorst merged commit 677e01f into main Dec 14, 2025

vDorst deleted the sfp_work branch December 23, 2025 21:12

Sfp and Link status work #39

Sfp and Link status work #39

Uh oh!

Conversation

logicog commented Nov 4, 2025

Uh oh!

vDorst commented Dec 7, 2025

Uh oh!

logicog commented Dec 7, 2025

Uh oh!

logicog commented Dec 7, 2025

Uh oh!

vDorst commented Dec 7, 2025

Uh oh!

logicog commented Dec 7, 2025

Uh oh!

vDorst commented Dec 7, 2025

Uh oh!

logicog commented Dec 7, 2025

Uh oh!

logicog commented Dec 8, 2025

Uh oh!

logicog commented Dec 8, 2025

Uh oh!

vDorst commented Dec 8, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

logicog commented Dec 8, 2025

Uh oh!

vDorst commented Dec 8, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

logicog commented Dec 9, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

vDorst commented Dec 9, 2025

Uh oh!

logicog commented Dec 9, 2025

Uh oh!

vDorst commented Dec 9, 2025

Uh oh!

vDorst commented Dec 9, 2025

Uh oh!

logicog commented Dec 10, 2025

Uh oh!

vDorst commented Dec 10, 2025

Uh oh!

logicog commented Dec 13, 2025

Uh oh!

vDorst commented Dec 14, 2025

Uh oh!

logicog commented Dec 14, 2025

Uh oh!

logicog commented Dec 14, 2025

Uh oh!

logicog commented Dec 14, 2025

Uh oh!

vDorst commented Dec 14, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

logicog commented Dec 14, 2025

Uh oh!

vDorst commented Dec 14, 2025

Uh oh!

logicog commented Dec 14, 2025

Uh oh!

vDorst commented Dec 14, 2025

Uh oh!

logicog commented Dec 14, 2025

Uh oh!

vDorst commented Dec 14, 2025

Uh oh!

logicog commented Dec 14, 2025

Uh oh!

logicog commented Dec 14, 2025

Uh oh!

vDorst commented Dec 14, 2025

Uh oh!

logicog commented Dec 14, 2025

Uh oh!

vDorst commented Dec 8, 2025 •

edited

Loading

vDorst commented Dec 8, 2025 •

edited

Loading

logicog commented Dec 9, 2025 •

edited

Loading

vDorst commented Dec 14, 2025 •

edited

Loading