Jump to content
Corsair Community

Corsair Link 4.5.0.55 crashing randomly


HawK2g

Recommended Posts

The program randomly stops working. Everything that it monitors is frozen at whatever value it was at the time of the crash. temps, volts, fans, all stuck at the last read value. This is a huge problem for me running a custom fan profile that is set to low RPM at idle temps and then 80-100% fan speeds when nearing 80C. I do overclock my CPU and the fans MUST go faster when I start playing games. The only quick temp fix is to exit the application from the tray bar and then re launch it. Sometimes it just displays nothing under the "HOME", like the other thread "Corsair link completely blank,..".

 

I have reinstalled, uninstalled, removed all files manually, removed other hardware monitoring software such as Core Temp. It still occurs. It crashes when I web browser, idle or stress test.

Right now I have set a fixed fan speed and reinstalled Core Temp.

Specs:

Windows 7 Ultimate x64, i7 960 @3.8Ghz, GTX970 MSI, Samsung 840evo 250GB.

Running applications:

- Logitech Gaming Software 8.82.151

- Steam

- Cannon IJ Network Scanner Selector FX

- NetWorx 5.5.4 64-bit

- SteelSeries Engine 3.92

 

*EDIT*

Forgot to re-add this after multiple edits before posting.

- Never had this problem on any of the v3.2 versions. I only updated it because the fan speed options were limited with fan curves and it was annoying switching between profiles.

- CL v4.3.0.154 also had this problem which I installed this year, so didn't wait long for v4.5

- CL v4.3.0.154 check for update doesn't work either, as it said I had the latest version, but only found out v4.5.0.55 after going to this forum to report this bug.

Link to comment
Share on other sites

First you should add your full PC specs to your profile rather than posting some of them.

 

Next you need to specify what CL hardware you have.

 

When quoting other threads you should specify their URL and you should also post the same debug screen shots as specified in those threads.

 

When it 'stops working' what % CPU time are the two CL4 processes using? Attach the task manager .PNG screen shot that shows this. What state is the CL4 service in? See http://forum.corsair.com/forums/showthread.php?p=888790 for one way to check.

Link to comment
Share on other sites

Opps, it's a H100i 240mm radiator. The process of the program is running, it hasn't stopped. The CPU usuage changes from 0% to 8% depending what I'm doing with it. I can interact with it and make changes, but those changes don't apply to the cooler.

For example, if it crashed and I went in and manually set the fans to 100%, they would stay at whatever it was. If I exit the program, load it back up, it will set them to 100% as it's saved to the profile.

 

I'll post up some screenshots when it happens again.

Link to comment
Share on other sites

it's a H100i 240mm radiator.

 

Please confirm this is a H100i rather than a H100iV2 or H100iGTX. Given it's an X58 based system I guess it is, but feel it's best to check.

 

Next time it happens after taking the screen shots exit CL4, run my SIV utility and post screen shots of the initial screen + [Link Devices] + [Link Status] + [Link Fans] panels. These will read and report the actual H100i firmware state/configuration and may provide further insight. It would be helpful to also have these when the H100i is working as expected so we can see if anything is different.

 

Oops again, what about the PSU + Case + memory size?

Link to comment
Share on other sites

I guess these are from when the system is working OK, are they?

 

Change the Task Manager to select Show processes for all users as "C:\Program Files (x86)\CorsairLink4\CorsairLink4.Service.exe" runs as system, add CPU Time + Handles + Threads to the TM selected columns and sort on CPU Time descending which should get the two CL4 processes near to the top. Posting Menu->Windows->Services->CL4 Control for OK + wonky should also provide insight.

 

For the compare post a screen with CL4 [Home] + SIV [status] so I can see what SIV reports for all the H80i information etc.. Both CL4 + HWM use the same CPUID SDK which is not ideal, further HWM does not report the H80i at all.

 

It's strange that SIV doesn't report the disk temperatures and I am wondering why. Do any utilities? If you post Menu->Devices->SMART Drives I suspect this will provide some insight. SIV also reports 910 rather than 960 so if you e-mail me/attach the save files I will fix this, see http://forum.corsair.com/forums/showthread.php?p=788225 for what I would like/need. I suspect the overclock setup you have triggers this issue.

Link to comment
Share on other sites

Nah, those pictures were taken when the software crashed. So CL displays temps when I was playing a game, and when I exit out of the game and had it idle for 5 minutes (ignore the RAM graph as that dip is from closing firefox so I could achieve the lowest CPU usuage for the screenshot), the temps went down as HWMonitor displays, while CL stays stuck.

I can only load up SIV once I exit CL if it crashes, or else SIV doesn't load at all and I have to end process it if I want to load it after exiting CL.

 

As for the processes under Task Manager, the service CPU Time kept going up while CorsairLink4.exe only got up to 5 seconds during the 2 minutes.

 

I have CrystalDiskInfo and HD Tune that both display the current SSD's temps. I sent the files via email. I doubt the overclock is the cause but I'll set it back to stock anyway as it usually crashes under 30 minutes.

Link to comment
Share on other sites

I can only load up SIV once I exit CL if it crashes, or else SIV doesn't load at all and I have to end process it if I want to load it after exiting CL.

 

If SIV crashes send me the dump files, Menu->Windows->Parameters->Error Reporting will tell you where they are.

 

It's possible the CL4 service is failing to release the CorsairLinkReadWriteGuardMutex lock which is why SIV can't start. Use the command SIV64X -NOLINK.

 

 

I really need to see Menu->Windows->Services->CL4 Control when it's gone wonky. I can tell the hardware is OK and suspect the CL4 service has failed and may well be spinning.

 

We really need to see how much CPU time CorsairLink4.Service.exe is using.

 

If only I could look for myself...

Link to comment
Share on other sites

I sent the files via email. I doubt the overclock is the cause but I'll set it back to stock anyway as it usually crashes under 30 minutes.

 

The expect overclock is fine and expect the 910 is just a silly bug in SIV.

 

No sign of the files as yet...

Link to comment
Share on other sites

Sorry, Outlook decided not to send them until the next send/receive schedule. Also it just crashed again with everything at stock. Hopefully the files will shed some light.

 

OK, they just arrived. Any SIV dumps will be in "C:\Users\HawK\Desktop\SIV\", are there any?

 

SIV reported 910 rather than 960 as you have set the maximum multiplier to x19 when it should be x24. SIV 5.17 Beta-00 and later will report 960.

 

The disk temperatures are missing as the Marvell Inc. 1.2.0.1047 driver does not allow the SMART data to be read. Many Marvell drivers are like this.

 

Next time it goes wonky before exiting CL4 type sc query CLink4Service and report what it says (post the screen shot).

 

Then from the command window do C:\Users\HawK\Desktop\SIV\SIV64X -NOLINK and post Menu->Windows->Services->CL4 Control

Link to comment
Share on other sites

I can only load up SIV once I exit CL if it crashes, or else SIV doesn't load at all and I have to end process it if I want to load it after exiting CL.

Nah, only the files I sent were created.

 

I am unclear what the first statement is saying. I though you were saying SIV crashed, but given there is no dump it can't have, so what does crash?

 

Below I can see the CL4 service is still running and using CPU time and the earlier SIV screen shots show the H80i hardware was still working so I have no real idea what exactly hay have gone wonky. I suspect it's a software issue within CL4 but as it's not my code I can't really deduce what.

 

All I can suggest it you log a call via https://corsair.secure.force.com/home/home.jsp and ask that the CL4 developers address your issue.

 

attachment.php?attachmentid=27808&d=1487297855
Link to comment
Share on other sites

I am unclear what the first statement is saying. I though you were saying SIV crashed, but given there is no dump it can't have, so what does crash?

 

The program doesn't load. I see it show up in Task Manager under processes with 5,036K memory and no cpu usuage. I guess it hangs more than anythinig, but it will never end up loading even if CL is closed after.

 

Thanks for helping, I'll probably just go back to v3.2.

Link to comment
Share on other sites

The program doesn't load. I see it show up in Task Manager under processes with 5,036K memory and no cpu usuage. I guess it hangs more than anythinig, but it will never end up loading even if CL is closed after.

 

My guess is that SIV is trying to acquire a locking mutex and that CL4 is failing to release it. You could confirm this by doing ad follows:

  1. Closedown CL4
  2. Start SIV64X
  3. Do Menu->Help->Lock States which will show all the lock states.
  4. Start CL4
  5. Wait for CL4 to go wonky
  6. Most of the time all the locks will be in a Released state with the odd one showing as Acquired once in a while.
  7. If it's Acquired it's expected that Global\Access_ASRock_Multiplex will be Acquired all the time.
  8. Once CL4 has gone wonky I expect one of the locks will be Acquired and the time will keep incrementing.
  9. Knowing which lock may also enable me to guess what is going wonky within CL4.

I suspect I could also tell which lock if you used task manager to generate a SIV dump file and sent it to me.

 

After pondering this I decided SIV should report if it's can't acquire a lock in a timely manor so added this to SIV 5.17 Beta-01. Please update the beta (Menu->File->Download->SIV Beta) and if it get's stuck and generates a SIV_DBGOUT.log post it. It's also possible to see these reports interactively by doing as below.

 

D:\SIV>siv64x | find " "

lck_acquire_aio() Status 258 2 for CLCC AIO Mutex in Thread 6216 (AIOs Thread) @ 2017-02-17 12:17:41

lck_acquire_aio() Status 258 122 for CLCC AIO Mutex in Thread 6216 (AIOs Thread) @ 2017-02-17 12:18:11

lck_acquire_aio() Status 258 122 for CLCC AIO Mutex in Thread 6216 (AIOs Thread) @ 2017-02-17 12:18:51

 

attachment.php?attachmentid=27813&stc=1&d=1487327623

1595369640_LockStates.png.22aee8c26293fe230f71b37ba4dcf2ec.png

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...