Constant "SLPlugin.exe has stopped working" errors - Firestorm / LL Second Life Release 7.1.9

[missyrideout]

Frequently, when right clicking an object to edit it, I am getting a "SLPlugin.exe has stopped working" error. Everything is whitelisted properly. For what it's worth I don't see this in Alchemy viewer. According to Windows Activity Viewer it's related to libcef.dll./ All I've found in googling libcef is in relation to Chrome, Steam, or Visual Studio 2015. I uninstalled and reinstalled those programs and ran a system file check for corruption.  I can't figure out if I need to find a new libcef.dll or where one could even safely get that.

This is the event

And the corresponding cef_log

[0831/215943.156:WARNING:policy_logger.cc(148)] :components\enterprise\browser\controller\chrome_browser_cloud_management_controller.cc(88) Could not create policy manager as CBCM is not enabled.
[0831/215946.415:WARNING:sqlite_persistent_store_backend_base.cc(198)] Failed to post task from FlushAndNotifyInBackground@net\extras\sqlite\sqlite_persistent_store_backend_base.cc:246 to client_task_runner_.
[0831/215946.416:WARNING:sqlite_persistent_store_backend_base.cc(198)] Failed to post task from FlushAndNotifyInBackground@net\extras\sqlite\sqlite_persistent_store_backend_base.cc:246 to client_task_runner_.
[0831/215947.719:WARNING:policy_logger.cc(148)] :components\enterprise\browser\controller\chrome_browser_cloud_management_controller.cc(88) Could not create policy manager as CBCM is not enabled.
[0831/215947.719:WARNING:policy_logger.cc(148)] :components\enterprise\browser\controller\chrome_browser_cloud_management_controller.cc(88) Could not create policy manager as CBCM is not enabled.
[0831/215947.720:WARNING:policy_logger.cc(148)] :components\enterprise\browser\controller\chrome_browser_cloud_management_controller.cc(88) Could not create policy manager as CBCM is not enabled.
[0831/215947.740:ERROR:tcp_socket_win.cc(357)] bind() returned an error: Only one usage of each socket address (protocol/network address/port) is normally permitted. (0x2740)
[0831/215947.740:ERROR:devtools_http_handler.cc(311)] Cannot start http server for devtools.
[0831/215947.740:ERROR:tcp_socket_win.cc(357)] bind() returned an error: Only one usage of each socket address (protocol/network address/port) is normally permitted. (0x2740)
[0831/215947.740:ERROR:devtools_http_handler.cc(311)] Cannot start http server for devtools.

Based on the second part of this, something may be using the port that the plugin is trying to use? But it's not immediately obvious to me what port that is. Or is it that the first part, "Could not create policy manager as CBCM is not enabled" means that the plugin is trying to do something, and the mechanism that elevates it is not working so it thinks there is no port available? Far beyond my understanding at this point.  I have the latest version of chrome and all the services that come with it are running according to services.msc

Any suggestions?

Edited September 1 by missyrideout
spelling

[Tiniyn]

I have the issue with SLPlugin.exe for years. Every red X (critical event) marks a day where I ran SL and did more than just sitting around. I asked around in the Fierstorm groups, but to no avail. The issue is not new, there is this old forum thread https://community.secondlife.com/forums/topic/473266-firestorm-slpluginexe-has-stopped-working/

Perhaps there is a version of this library that has been corrected and just needs to be included in the installation package?

[missyrideout]

Thanks for posting. For me it only started with the 7.1.9 viewers. If try to edit twice, it's a guaranteed error on the second click. If I revert to 6.6.17, it never happens. I've tried the latest Firestorm alpha and it's still a problem. No issue with Alchemy Beta 7.1.9.2492. Wonder what they do differently, don't know enough to guess.

[Qie Niangao]

This started for me on some Linden viewer just before the first PBR viewer I tested and has since been a constant aggravation on every Linden viewer and the Firestorm viewers that included the PBR code. It's what first made me switch to Alchemy, which magically doesn't have the problem.

I reported it on a jira back in November while there still was jira. They promised migration to Canny wouldn't lose track of ongoing problems, but I have no idea how to check on it now. Given all the attention on PBR fixes and new glTF features, I doubt anybody is even looking at older viewer bugs, so somebody new filing a Canny might possibly get some attention.

I'm super grateful to see I'm not the only one having this problem. I wonder if we have anything specifically similar in our machine configurations.

[Tiniyn]

From Firestorm Help I gatherd this information. This might be helpful to do a simple check for similarities. 

Firestorm 7.1.9 (74745) Jun 20 2024 10:49:53 (64bit / SSE2) (Firestorm-Releasex64) with Havok support
CPU: AMD Ryzen 7 5700X 8-Core Processor              (3400 MHz)
Memory: 32679 MB (Used: 1188 MB)
Concurrency: 16
OS Version: Microsoft Windows 10 64-bit (Build 19045.4780)
Graphics Card: NVIDIA GeForce RTX 3060 Ti/PCIe/SSE2
Graphics Card Memory: 8024 MB
Windows Graphics Driver Version: 32.0.15.6081
OpenGL Version: 4.6.0 NVIDIA 560.81

RestrainedLove API: (disabled)
libcurl Version: libcurl/7.54.1 OpenSSL/1.1.1q zlib/1.2.11.zlib-ng nghttp2/1.40.0
J2C Decoder Version: KDU v8.4.1
Audio Driver Version: FMOD Studio 2.02.20
Dullahan: 1.14.0.202310131404
  CEF: 118.4.1+g3dd6078+chromium-118.0.5993.54
  Chromium: 118.0.5993.54
LibVLC Version: 3.0.16
Voice Server Version: Not Connected
Settings mode: Firestorm
Viewer Skin: AnsaStorm (Bright Blue)
Window size: 1920x1017 px
Font Used: Deja Vu (96 dpi)
Font Size Adjustment: 0 pt
UI Scaling: 1
Draw distance: 240 m
Bandwidth: 3000 kbit/s
LOD factor: 3
Render quality: High-Ultra (6/7)
Disk cache: Max size 6016.0 MB (72.4% used)
Built with MSVC version 1940
Packets Lost: 0/2.855 (0,0%)
September 01 2024 04:38:57 SLT

 

[Qie Niangao]

Firestorm 7.1.9 (74745) Jun 20 2024 10:49:53 (64bit / SSE2) (Firestorm-Releasex64) with Havok support
Release Notes

CPU: 11th Gen Intel(R) Core(TM) i5-11600K @ 3.90GHz (3912 MHz)
Memory: 16236 MB (Used: 595 MB)
Concurrency: 12
OS Version: Microsoft Windows 11 64-bit (Build 26100.1591)
Graphics Card Vendor: NVIDIA Corporation
Graphics Card: NVIDIA GeForce RTX 3060/PCIe/SSE2
Graphics Card Memory: 12115 MB
Graphics Card Memory (Detected): 12115 MB
Graphics Card Memory (Budget): Unlimited

Windows Graphics Driver Version: 32.0.15.6094
OpenGL Version: 4.6.0 NVIDIA 560.94

RestrainedLove API: (disabled)
libcurl Version: libcurl/7.54.1 OpenSSL/1.1.1q zlib/1.2.11.zlib-ng nghttp2/1.40.0
J2C Decoder Version: KDU v8.4.1
Audio Driver Version: FMOD Studio 2.02.20
Dullahan: 1.14.0.202310131404
  CEF: 118.4.1+g3dd6078+chromium-118.0.5993.54
  Chromium: 118.0.5993.54
LibVLC Version: 3.0.16
Voice Server Version: Not Connected
Settings mode: Firestorm
Viewer Skin: MetaHarper Modern (CoolOcean)
Window size: 2692x1744 px
Font Used: Deja Vu (96 dpi)
Font Size Adjustment: 0 pt
UI Scaling: 1
Draw distance: 96 m
Bandwidth: 3000 kbit/s
LOD factor: 3
Render quality: Ultra (7/7)
Disk cache: Max size 2048.0 MB (70.7% used)
Built with MSVC version 1940
September 01 2024 04:40:08 SLT

Of course Alchemy has a lot of differences. I notice that it's using LibVLC 3.0.20 when both the Linden and Firestorm viewers use 3.0.16, but we really need to find a thing that's common to our setups but distinct from all those who escape the SLPlugin.exe plague.

[Love Zhaoying]

Please remind me, isn't SLPlugin.exe the old Voice component?

[Qie Niangao]

10 minutes ago, Love Zhaoying said:

Please remind me, isn't SLPlugin.exe the old Voice component?

A viewer dev could answer that directly, but my limited understanding is that SLPlugin handles both SLVoice and the media stuff ("Dullahan" and the Chromium Embedded Framework or "CEF"). I have no idea what I could mean by "handles" in that sentence, however.

[Rowan Amore]

22 minutes ago, Love Zhaoying said:

Please remind me, isn't SLPlugin.exe the old Voice component?

 

[Love Zhaoying]

54 minutes ago, Rowan Amore said:

 

Thanks, so if it's used for multiple things, then crashing is "bad".  However (with knowledge gleaned from your link), the crashing could also be occurring as individual instances are terminated (not as bad)..

 

[Qie Niangao]

Other than annoyance, I've never seen any ill effects of SLPlugin.exe stopping, presumably because a new instance is spawned right away. The nuisance factor is high, though, because the error is displayed in a modal system window that definitely demands attention… and because a new instance is spawned there's never a shortage of new nuisances and it's only a matter of time—usually a few minutes, sometimes less—before it happens again.

[missyrideout]

Thanks for posting. "Good" to know that it's happening to someone on Windows 11, was wondering if I needed to rule out Windows10 but really didn't want to install right now.

Here's the details of my Firestorm.

CPU: Intel(R) Core(TM) i5-10600K CPU @ 4.10GHz (4104 MHz)
Memory: 32688 MB (Used: 1233 MB)
Concurrency: 12
OS Version: Microsoft Windows 10 64-bit (Build 19045.4842)
Graphics Card Vendor: NVIDIA Corporation
Graphics Card: NVIDIA GeForce RTX 2060 SUPER/PCIe/SSE2
Graphics Card Memory: 8006 MB
Graphics Card Memory (Detected): 8006 MB
Graphics Card Memory (Budget): Unlimited

Windows Graphics Driver Version: 31.0.15.5212
OpenGL Version: 4.6.0 NVIDIA 552.12

RestrainedLove API: (disabled)
libcurl Version: libcurl/7.54.1 OpenSSL/1.1.1q zlib/1.2.11.zlib-ng nghttp2/1.40.0
J2C Decoder Version: KDU v8.4.1
Audio Driver Version: FMOD Studio 2.02.20
Dullahan: 1.14.0.202310131404
  CEF: 118.4.1+g3dd6078+chromium-118.0.5993.54
  Chromium: 118.0.5993.54
LibVLC Version: 3.0.16
Voice Server Version: Not Connected
Settings mode: Hybrid
Viewer Skin: MetaHarper Modern (CoolOcean)
Window size: 1680x997 px
Font Used: Deja Vu (96 dpi)
Font Size Adjustment: 0 pt
UI Scaling: 1
Draw distance: 128 m
Bandwidth: 1600 kbit/s
LOD factor: 2
Render quality: High (5/7)
Disk cache: Max size 19968.0 MB (21.9% used)
Built with MSVC version 1940
Packets Lost: 0/2,704 (0.0%)
September 01 2024 09:12:42 SLT

[Frionil Fang]

1 hour ago, Qie Niangao said:

The nuisance factor is high, though, because the error is displayed in a modal system window that definitely demands attention

Maybe this is the actual bug, looking at the code for the slplugin it's designed to swallow up exceptions and not have them pop up, but they'll still go into Windows event log as crashes. I was looking up the error and it seems to have happened in other programs too after they switched to a particular version of libcef and it was suggested it has something to do with an unclean shutdown instead of a "true" crash. Maybe there's a way to work around the flaw, or maybe a different version of CEF would work better... all I know is that transplanting a newer version of libcef .dlls to a viewer resulted in slplugin being unable to start, heh.

[filz Camino]

Same here, my error logs on both of my computers are constantly filling up with these errors - since at least a year ago.

i've never noticed an adverse effect on the operation of Firestorm because of it, but it is pretty tiresome to see my logs full of these errors so it would be nice if the bug were to be fixed.

[Tiniyn]

Today I have developed some prototypes with simple prims. I got 34 critical events. You don't notice that a process is dying quietly, again and again. But nevertheless, it is a bug and it should be corrected.

[Evan Pawprint]

I'm also having this issue. I've gone through using the LL Client, Alchemy, and Firestorm. All three viewers give the same error with libcef.dll_unloaded - now it's throwing the error "Incorrect function. "

 

I attempted to use a non-pbr version of the applications in question, and they worked for a bit, but then started throwing bugsplats with the same error message. 

 

I did find a message on microsoft forums about Windows Defender Real Time Protection causing applications to crash to the desktop. I'm testing that now, and *crosses fingers* so far that seems to have helped. 

 

"The constant crashing problem of Windows 11 was due to Windows Defender real-time protection. by turning off real-time protection, system crashing stopped. PC has been running continuously for 24 hours with no issues."

 

[Evan Pawprint]

2 minutes ago, Evan Pawprint said:

I'm also having this issue. I've gone through using the LL Client, Alchemy, and Firestorm. All three viewers give the same error with libcef.dll_unloaded - now it's throwing the error "Incorrect function. "

 

I attempted to use a non-pbr version of the applications in question, and they worked for a bit, but then started throwing bugsplats with the same error message. 

 

I did find a message on microsoft forums about Windows Defender Real Time Protection causing applications to crash to the desktop. I'm testing that now, and *crosses fingers* so far that seems to have helped. 

 

"The constant crashing problem of Windows 11 was due to Windows Defender real-time protection. by turning off real-time protection, system crashing stopped. PC has been running continuously for 24 hours with no issues."

 

Unfortunately this did not fix the issue for me.  

 

 

- <Event xmlns="http://schemas.microsoft.com/win/2004/08/events/event">

- <System>

  <Provider Name="BugSplat-firestorm_release-Firestorm-Releasex64" />

  <EventID Qualifiers="0">1</EventID>

  <Version>0</Version>

  <Level>2</Level>

  <Task>0</Task>

  <Opcode>0</Opcode>

  <Keywords>0x80000000000000</Keywords>

  <TimeCreated SystemTime="2024-09-01T18:51:20.1387148Z" />

  <EventRecordID>4313</EventRecordID>

  <Correlation />

  <Execution ProcessID="35116" ThreadID="0" />

  <Channel>Application</Channel>

  <Computer>Bear</Computer>

  <Security />

  </System>

- <EventData>

  <Data>firestorm_release</Data>

  <Data>Firestorm-Releasex64</Data>

  <Data>847534</Data>

  </EventData>

  </Event>

[Perrie Juran]

First, as to what SLplugin does:

"slplugin is needed to view any media content in the viewer & the slplugin processes are launched & culled by the viewer as needed. 
There will be one slplugin process for each media instance the viewer is currently using - so for example you will have one slplugin launched to load the login screen image (which is essentially just a web page), another launched  when you open a web profile, another launched when you open web search, another launched if you open the destination guide in the viewer & parcel media or MOAP objects inworld will each have their own slplugin instance. "

 

Second, the usual cause of plugin failure is overaggressive Anti Virus. It is really nuts. You  can be happily going along your SL way and suddenly the AV decides it doesn't like the plugin. This happens randomly with all three of the executables in the viewer, Slplugin, Dullahanhost & Voice. Hence the need to add them to the exclusion list in your AV. There may be other reasons for it failures but AV false flagging is by and large is the most common.

[Love Zhaoying]

54 minutes ago, Perrie Juran said:

This happens randomly with all three of the executables in the viewer, Slplugin, Dullahanhost & Voice. Hence the need to add them to the exclusion list in your AV.

It's possible that I never have (and don't remember ever having) this issue because I use Norton AV, and once I may have told it to trust the main Viewer, which may by default caused it to trust all sub-processes.

[missyrideout]

1 hour ago, Perrie Juran said:

Second, the usual cause of plugin failure is overaggressive Anti Virus. It is really nuts. You  can be happily going along your SL way and suddenly the AV decides it doesn't like the plugin. This happens randomly with all three of the executables in the viewer, Slplugin, Dullahanhost & Voice. Hence the need to add them to the exclusion list in your AV. There may be other reasons for it failures but AV false flagging is by and large is the most common.

I do have everything whitelisted properly. And I only use WIndows Defender. And even with realtime protection temporarily disabled the issue remains.

Edited September 1 by missyrideout

[Love Zhaoying]

..if "SLPlugin.exe" was previously quarantined, it will probably stay that way.

Depending on your viewer, LL or the Firestorm team may be able to help find some cause for the error in technical logs, etc.

[missyrideout]

@Love Zhaoying 

I did post the relevant log in this thread. Got it by clearing the log files, running the viewer and quitting immediately after the error to isolate the exact event at the time it happened.

Maybe @Beq Janus has some ideas. FS in world support didn't, other than to verify that I had everything whitelisted properly.

No quarantined items in my protection history.

[Beq Janus]

I don't beleive that we have any alterations to slplugin/CEF. I know for certain that we use the libCEF built by LL 

Out of interest, do you see this issue when using the LL default viewer?

As other have noted slplugin crashes are not new, nor are they typically problematic but I agree that neither of those facts make it any less annoying.

Let me know if you are able to get this to reproduce on the LL viewer, because we use the LL build of the CEF library I cannot debug it so readily as I might another part of the viewer. I will see if that log extract suggests anything to anyone though.

(as for the port reuse, the plugines are started on a "random" port within a defined range IIRC, it is possible that they are crashing into one another but even that should be a simple startup glitch that resolves itself.

 

 

[missyrideout]

8 minutes ago, Beq Janus said:

Out of interest, do you see this issue when using the LL default viewer?

Yeah, same problem with a fresh, whitelisted install of LL's Second Life Release 7.1.9.10515727195 (64bit). 
Upon logging in, the second time I right click and click "edit" the error is triggered, without fail. After that, the SLPlugin failed popup is sporadic, but frequent.
For all of my testing my avatar is not wearing any attachments (other than the FS bridge in FS.) I've tried toggling various settings, disabling RLV, disabling media--no dice.

Thanks for checking it out!

[Qie Niangao]

13 minutes ago, Beq Janus said:

Out of interest, do you see this issue when using the LL default viewer?

Yes, for me it started happening on LL project/RC viewers just before test viewers for PBR were available, and didn't start on Firestorm until the 7.x (PBR) alpha viewers. On Alchemy it is not a problem, with SLPlugin.exe generating an error maybe once a month rather than several times an hour (and sometimes more than one a minute) on Firestorm and Linden viewers.