- exit code -1073741515 (0xc0000135)

Oliver Behnke
Oliver Behnke
Moderator
Administrator
Joined: 4 Sep 07
Posts: 982
Credit: 25,170,813
RAC: 0

Could someone of those of you

Could someone of those of you who experience this issue on Windows 7 64-bit please install the latest stable 32-bit (!) BOINC 7.0.28 client and see if that fixes the problem?

Thanks,
Oliver

Einstein@Home Project

LumenDan
LumenDan
Joined: 30 Apr 05
Posts: 11
Credit: 202,469,585
RAC: 123,997

RE: Could someone of those

Quote:

Could someone of those of you who experience this issue on Windows 7 64-bit please install the latest stable 32-bit (!) BOINC 7.0.28 client and see if that fixes the problem?

Thanks,
Oliver

BOINC 7.0.38 (32 bit) also fails in the same way as the 64 bit version with ABPcuda32nv301.

Just quickly covering what I have found already:
BOINC 6.12.34 (64 bit) also had a similar problem that I managed to circumvent by relaxing my User Account Control setting to "never notify".
The error message from 6.12.34 was slightly more informative. I don't remember the exact message text but it explained that a process failed to start because it required elevated privileges.

I only recently replaced my XP machine Windows 7 64 bit so several factors changed for me literally overnight and its hard for me to pinpoint a single change that may cause this issue.

I have always installed BOINC with "Protected application execution"=FALSE but this error sounds similar to the warning given for the "Protected application execution" option.
It also seems like BOINC 7 handles the issue in a different way to BOINC 6. It behaves like a UAC notification was issued and declined even though I have relaxed UAC to "never notify"

The E@H CPU applications work just fine but E@H seems to lack appropriate privileges to launch the CUDA application.

Best of Luck

astro-marwil
astro-marwil
Joined: 28 May 05
Posts: 529
Credit: 622,266,543
RAC: 1,086,360

Hallo! I also installed

Hallo!
I also installed boinc_7.0.28_windows_intelx86.exe, dated 21 May 2012, and NVIDIA driver 301.42.
All 24 downloaded BRP4 task ended up after Run Time 0.00 with:

Quote:

Exit status-185 (0xffffffffffffff47)

Stderr output
7.0.28

couldn't start Can't link app version file: -108

]]>

The event log shows:

Quote:
12/07/2012 20:39:15 | Einstein@Home | Started download of b2030.20110428.G42.33-00.43.C.b3s0g0.00000_2840.bin4
12/07/2012 20:39:27 | Einstein@Home | Finished download of b2030.20110428.G42.33-00.43.C.b2s0g0.00000_2919.bin4
12/07/2012 20:39:27 | Einstein@Home | Finished download of b2030.20110428.G42.33-00.43.C.b3s0g0.00000_2840.bin4
12/07/2012 20:39:27 | Einstein@Home | Started download of b2030.20110428.G42.33-00.43.C.b3s0g0.00000_2841.bin4
12/07/2012 20:39:27 | Einstein@Home | Started download of b2030.20110428.G42.33-00.43.C.b3s0g0.00000_2842.bin4

12/07/2012 20:39:27 | Einstein@Home | [error] Can't create link file slots/6/einsteinbinary_BRP4_1.25_windows_intelx86__BRP4cuda32nv301.exe
12/07/2012 20:39:33 | Einstein@Home | Computation for task b2030.20110428.G42.33-00.43.C.b2s0g0.00000_2912_1 finished
12/07/2012 20:39:33 | Einstein@Home | Output file b2030.20110428.G42.33-00.43.C.b2s0g0.00000_2912_1_0 for task b2030.20110428.G42.33-00.43.C.b2s0g0.00000_2912_1 absent
12/07/2012 20:39:33 | Einstein@Home | Output file b2030.20110428.G42.33-00.43.C.b2s0g0.00000_2912_1_1 for task b2030.20110428.G42.33-00.43.C.b2s0g0.00000_2912_1 absent
12/07/2012 20:39:33 | Einstein@Home | Output file b2030.20110428.G42.33-00.43.C.b2s0g0.00000_2912_1_2 for task b2030.20110428.G42.33-00.43.C.b2s0g0.00000_2912_1 absent
12/07/2012 20:39:33 | Einstein@Home | Output file b2030.20110428.G42.33-00.43.C.b2s0g0.00000_2912_1_3 for task b2030.20110428.G42.33-00.43.C.b2s0g0.00000_2912_1 absent
12/07/2012 20:39:33 | Einstein@Home | Output file b2030.20110428.G42.33-00.43.C.b2s0g0.00000_2912_1_4 for task b2030.20110428.G42.33-00.43.C.b2s0g0.00000_2912_1 absent
12/07/2012 20:39:33 | Einstein@Home | Output file b2030.20110428.G42.33-00.43.C.b2s0g0.00000_2912_1_5 for task b2030.20110428.G42.33-00.43.C.b2s0g0.00000_2912_1 absent
12/07/2012 20:39:33 | Einstein@Home | Output file b2030.20110428.G42.33-00.43.C.b2s0g0.00000_2912_1_6 for task b2030.20110428.G42.33-00.43.C.b2s0g0.00000_2912_1 absent
12/07/2012 20:39:33 | Einstein@Home | Output file b2030.20110428.G42.33-00.43.C.b2s0g0.00000_2912_1_7 for task b2030.20110428.G42.33-00.43.C.b2s0g0.00000_2912_1 absent
12/07/2012 20:39:40 | Einstein@Home | Finished download of b2030.20110428.G42.33-00.43.C.b3s0g0.00000_2842.bin4

So once more I disabled BRP4 tasks and do crunch instead Milkyway. I have an eye on this thread to see, what´s going on in this regard, hoping to come back soon, after the problems became cured.

Kind regards and happy crunching
Martin

tolafoph
tolafoph
Joined: 14 Sep 07
Posts: 122
Credit: 74,659,937
RAC: 0

After a system change I also

After a system change I also got this error on my win 7 OS. Win Vista didn't have this problem. I changed the Boinc Manager to 6.12.34 (x64) and now Albert@home + Einstein@home CUDA tasks don't crash.

Ivan Terentev
Ivan Terentev
Joined: 16 Oct 11
Posts: 3
Credit: 57,258,942
RAC: 0

Maybe problem caused by Intel

Maybe problem caused by Intel OpenCL SDK? Which in my case was automatically installed on my computer with Intel HD Graphics Driver, because once I uninstalled it , everything works like nothing happened.

Oliver Behnke
Oliver Behnke
Moderator
Administrator
Joined: 4 Sep 07
Posts: 982
Credit: 25,170,813
RAC: 0

FYI, we're still looking into

FYI, we're still looking into this... We're also trying to get 64 bit Windows binaries built soonish...

Oliver

Einstein@Home Project

LumenDan
LumenDan
Joined: 30 Apr 05
Posts: 11
Credit: 202,469,585
RAC: 123,997

Thanks for the update

Thanks for the update

Holmis
Joined: 4 Jan 05
Posts: 1,118
Credit: 1,055,935,564
RAC: 0

RE: Maybe problem caused by

Quote:
Maybe problem caused by Intel OpenCL SDK? Which in my case was automatically installed on my computer with Intel HD Graphics Driver, because once I uninstalled it , everything works like nothing happened.

I'll add my experience of this:

I did an upgrade on the Intel HD Graphics Driver and all BRP GPU-tasks started crashing with -1073741515 (0xffffffffc0000135).
I had version 8.15.10.2696 that was installed through Windows Update and I upgraded to version 8.15.10.2761, downloaded directly from Intel.

However I managed to fix the problem by doing a system restore to before I upgraded the HD Graphics driver and now all seems fine again.

Jord
Joined: 26 Jan 05
Posts: 2,952
Credit: 5,893,653
RAC: 668

RE: I did an upgrade on the

Quote:
I did an upgrade on the Intel HD Graphics Driver and all BRP GPU-tasks started crashing with -1073741515 (0xffffffffc0000135).


In your case, the normal reasons for this error showed.

Are you using the built-into the CPU Intel HD Graphics chipset? If not, then why do you update its drivers? The latest aren't always the best for a running system. And unless you have problems with the present drivers, there's really no need to update to later ones. When they show you all that you can see, it's not necessary.

As for possibly adding OpenCL capability for the HD Graphics, also not necessary as BOINC does not support the Intel HD Graphics yet.

Holmis
Joined: 4 Jan 05
Posts: 1,118
Credit: 1,055,935,564
RAC: 0

Thanks for your reply

Thanks for your reply Ageless.

As I've recently built this machine I've been experimenting with different setups and some over-clocking. As of right now I am using the IGP just to test how powerful it is (or not) and at the same time Boinc (and Einstein) can use the 9600GT all by it self. But as you say the latest drivers aren't always better and it seems to work without problems with the next to newest version, so I'll stay there until there's real reason to upgrade.

As for Boinc detecting the IGP i knew that won't work yet and may take some time.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.