RESOLVED -Linux - GPU WU - Mangled path causes failure. PLEASE fix.

jay
jay
Joined: 25 Jan 07
Posts: 99
Credit: 84044023
RAC: 0
Topic 227046

Greetings!!

An old problem ( on Radeon/ATI GPU wu)  still has not been fixed.

from the WU status:

OpenCL device has no FP64 support
OpenCL compiling FAILED! : -11 . Error message: fatal error: cannot open file '/usr/local//usr/lib/clc/verde-amdgcn-mesa-mesa3d.bc': No such file or directory

 

This is mangled.

And causes all WU I get to fail.

The leading "/usr/local/" should be deleted.

In Linux (all that I have seen in 30 years) , there is no "/usr/local//usr/lib"

((I will come back to add a link to the status.))

Thanks!!

PS. Yes, I can go in and mangle my Linux, create false directories and add a link; but, that gets wiped out on updates.

 

PPS . I did not see other reports on the forum.

anyone know where/how to report on github or other means?

thanks in advance.

 

 

 

Ian&Steve C.
Ian&Steve C.
Joined: 19 Jan 20
Posts: 4045
Credit: 48050706091
RAC: 34873899

This is not an issue with

This is not an issue with Einstein. It’s an issue with the Mesa package. You need to bring it up with the package maintainer(s). 
 

or switch to using the AMDGPU-Pro drivers which don’t have this issue. 

_________________________________________________________________________

jay
jay
Joined: 25 Jan 07
Posts: 99
Credit: 84044023
RAC: 0

Additional Info Task ID: 

Additional Info


Task ID:  https://einsteinathome.org/task/1234582729


Computer ID: https://einsteinathome.org/host/12201025

Application:
   Gamma-ray pulsar binary search #1 on GPUs v1.18 (FGRPopencl1K-ati)
x86_64-pc-linux-gnu

 

GPU portion of my BOINC Log at startup:

Mon 21 Feb 2022 08:15:53 AM EST |  | Starting BOINC client version 7.16.6 for x86_64-pc-linux-gnu
Mon 21 Feb 2022 08:15:53 AM EST |  | log flags: file_xfer, sched_ops, task
Mon 21 Feb 2022 08:15:53 AM EST |  | Libraries: libcurl/7.68.0 OpenSSL/1.1.1f zlib/1.2.11 brotli/1.0.7 libidn2/2.2.0 libpsl/0.21.0 (+libidn2/2.2.0) libssh/0.9.3/openssl/zlib nghttp2/1.40.0 librtmp/2.3
Mon 21 Feb 2022 08:15:53 AM EST |  | Data directory: /var/lib/boinc-client
Mon 21 Feb 2022 08:15:54 AM EST |  | OpenCL: AMD/ATI GPU 0: AMD VERDE (DRM 2.50.0, 5.13.0-30-generic, LLVM 12.0.0) (driver version 21.2.6, device version OpenCL 1.1 Mesa 21.2.6, 2048MB, 2048MB available, 512 GFLOPS peak)
Mon 21 Feb 2022 08:15:54 AM EST |  | libc: Ubuntu GLIBC 2.31-0ubuntu9.2 version 2.31
Mon 21 Feb 2022 08:15:54 AM EST |  | Host name: pc-14
Mon 21 Feb 2022 08:15:54 AM EST |  | Processor: 8 AuthenticAMD AMD FX(tm)-8150 Eight-Core Processor [Family 21 Model 1 Stepping 2]
Mon 21 Feb 2022 08:15:54 AM EST |  | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf pni pclmulqdq monitor ssse3 cx16 sse4_1 sse4_2 popcnt aes xsave avx lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs xop skinit wdt fma4 nodeid_msr topoext perfctr_core perfctr_nb cpb hw_pstate ssbd ibpb vmmcall arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold
Mon 21 Feb 2022 08:15:54 AM EST |  | OS: Linux Ubuntu: Ubuntu 20.04.4 LTS [5.13.0-30-generic|libc 2.31 (Ubuntu GLIBC 2.31-0ubuntu9.2)]
Mon 21 Feb 2022 08:15:54 AM EST |  | Memory: 11.53 GB physical, 9.31 GB virtual
Mon 21 Feb 2022 08:15:54 AM EST |  | Disk: 91.17 GB total, 84.75 GB free
Mon 21 Feb 2022 08:15:54 AM EST |  | Local time is UTC -5 hours
Mon 21 Feb 2022 08:15:54 AM EST |  | Config: GUI RPCs allowed from:
Mon 21 Feb 2022 08:15:54 AM EST | World Community Grid | General prefs: from World Community Grid (last modified 21-Feb-2022 01:36:45)
Mon 21 Feb 2022 08:15:54 AM EST | World Community Grid | Host location: none

-----

On my Linux - Ubuntu Mate:

cd /usr/local
jay@pc-14:/usr/local$ ls -al
total 44
drwxr-xr-x 11 root root 4096 Jan 13 18:18 .
drwxr-xr-x 15 root root 4096 Jan 14 14:47 ..
drwxr-xr-x  2 root root 4096 Aug 19  2021 bin
drwxr-xr-x  2 root root 4096 Aug 19  2021 etc
drwxr-xr-x  2 root root 4096 Aug 19  2021 games
drwxr-xr-x  2 root root 4096 Aug 19  2021 include
drwxr-xr-x  3 root root 4096 Aug 19  2021 lib
lrwxrwxrwx  1 root root    9 Jan 13 16:46 man -> share/man
drwxr-xr-x  2 root root 4096 Aug 19  2021 sbin
drwxr-xr-x  7 root root 4096 Aug 19  2021 share
drwxr-xr-x  2 root root 4096 Aug 19  2021 src
drwxr-xr-x  3 root root 4096 Jan 13 18:18 usr

and

 cd /usr/local/lib
jay@pc-14:/usr/local/lib$ ls -al
total 12
drwxr-xr-x  3 root root  4096 Aug 19  2021 .
drwxr-xr-x 11 root root  4096 Jan 13 18:18 ..
drwxrwsr-x  3 root staff 4096 Aug 19  2021 python3.8

 

But on the right path....

cd /usr/lib/clc
jay@pc-14:/usr/lib/clc$ ls -al
total 65776
drwxr-xr-x   2 root root    4096 Feb 10 15:07 .
drwxr-xr-x 142 root root    4096 Feb 21 00:23 ..
-rw-r--r--   1 root root 8173504 Sep  1 06:28 amdgcn--amdhsa.bc
lrwxrwxrwx   1 root root      16 Sep  1 06:28 aruba-r600--.bc -> cayman-r600--.bc
-rw-r--r--   1 root root 4463000 Sep  1 06:28 barts-r600--.bc
lrwxrwxrwx   1 root root      18 Sep  1 06:28 bonaire-amdgcn--.bc -> tahiti-amdgcn--.bc
lrwxrwxrwx   1 root root      15 Sep  1 06:28 caicos-r600--.bc -> barts-r600--.bc
lrwxrwxrwx   1 root root      18 Sep  1 06:28 carrizo-amdgcn--.bc -> tahiti-amdgcn--.bc
-rw-r--r--   1 root root 4450452 Sep  1 06:28 cayman-r600--.bc
-rw-r--r--   1 root root 4463000 Sep  1 06:28 cedar-r600--.bc
-rw-r--r--   1 root root 4450560 Sep  1 06:28 cypress-r600--.bc
lrwxrwxrwx   1 root root      18 Sep  1 06:28 fiji-amdgcn--.bc -> tahiti-amdgcn--.bc
lrwxrwxrwx   1 root root      18 Sep  1 06:28 gfx900-amdgcn--.bc -> tahiti-amdgcn--.bc
lrwxrwxrwx   1 root root      18 Sep  1 06:28 gfx902-amdgcn--.bc -> tahiti-amdgcn--.bc
lrwxrwxrwx   1 root root      18 Sep  1 06:28 gfx904-amdgcn--.bc -> tahiti-amdgcn--.bc
lrwxrwxrwx   1 root root      18 Sep  1 06:28 gfx906-amdgcn--.bc -> tahiti-amdgcn--.bc
lrwxrwxrwx   1 root root      18 Sep  1 06:28 hainan-amdgcn--.bc -> tahiti-amdgcn--.bc
lrwxrwxrwx   1 root root      18 Sep  1 06:28 hawaii-amdgcn--.bc -> tahiti-amdgcn--.bc
lrwxrwxrwx   1 root root      17 Sep  1 06:28 hemlock-r600--.bc -> cypress-r600--.bc
lrwxrwxrwx   1 root root      18 Sep  1 06:28 iceland-amdgcn--.bc -> tahiti-amdgcn--.bc
lrwxrwxrwx   1 root root      15 Sep  1 06:28 juniper-r600--.bc -> cedar-r600--.bc
lrwxrwxrwx   1 root root      18 Sep  1 06:28 kabini-amdgcn--.bc -> tahiti-amdgcn--.bc
lrwxrwxrwx   1 root root      18 Sep  1 06:28 kaveri-amdgcn--.bc -> tahiti-amdgcn--.bc
lrwxrwxrwx   1 root root      18 Sep  1 06:28 mullins-amdgcn--.bc -> tahiti-amdgcn--.bc
-rw-r--r--   1 root root 8297292 Sep  1 06:28 nvptx64--.bc
-rw-r--r--   1 root root 8308140 Sep  1 06:28 nvptx64--nvidiacl.bc
-rw-r--r--   1 root root 8278740 Sep  1 06:28 nvptx--.bc
-rw-r--r--   1 root root 8280684 Sep  1 06:28 nvptx--nvidiacl.bc
lrwxrwxrwx   1 root root      18 Sep  1 06:28 oland-amdgcn--.bc -> tahiti-amdgcn--.bc
lrwxrwxrwx   1 root root      15 Sep  1 06:28 palm-r600--.bc -> cedar-r600--.bc
lrwxrwxrwx   1 root root      18 Sep  1 06:28 pitcairn-amdgcn--.bc -> tahiti-amdgcn--.bc
lrwxrwxrwx   1 root root      18 Sep  1 06:28 polaris10-amdgcn--.bc -> tahiti-amdgcn--.bc
lrwxrwxrwx   1 root root      18 Sep  1 06:28 polaris11-amdgcn--.bc -> tahiti-amdgcn--.bc
lrwxrwxrwx   1 root root      15 Sep  1 06:28 redwood-r600--.bc -> cedar-r600--.bc
lrwxrwxrwx   1 root root      18 Sep  1 06:28 stoney-amdgcn--.bc -> tahiti-amdgcn--.bc
lrwxrwxrwx   1 root root      15 Sep  1 06:28 sumo2-r600--.bc -> cedar-r600--.bc
lrwxrwxrwx   1 root root      15 Sep  1 06:28 sumo-r600--.bc -> cedar-r600--.bc
-rw-r--r--   1 root root 8161396 Sep  1 06:28 tahiti-amdgcn--.bc
lrwxrwxrwx   1 root root      18 Sep  1 06:28 tonga-amdgcn--.bc -> tahiti-amdgcn--.bc
lrwxrwxrwx   1 root root      15 Sep  1 06:28 turks-r600--.bc -> barts-r600--.bc
lrwxrwxrwx   1 root root      18 Sep  1 06:28 verde-amdgcn--.bc -> tahiti-amdgcn--.bc

======================

so,  verde-amdgcn--.bc

is really at

/usr/lib/clc

 

 

ES-Six
Joined: 6 Oct 18
Posts: 18
Credit: 30422791
RAC: 15514

I got the same issue on the

Hello,

 

You can try to check if the mesa-opencl-icd package is installed, but I suspect you are affected by the same issue than me.

I got the same issue on the lastest Ubuntu release, this IS definitely a problem with the Mesa package. Ubuntu need to fix it.

It already appened in the past.

In this case, all similar opencl tasks will fail on our configurations (this issue will affect other projets too).

In my case, this issue occured because I migrated to the lastest version of Ubuntu. The previous Ubuntu version doesn't have this issue as its Mesa package is not broken.

 

Note : in my case, the Mesa package is completely broken as the targeted .bc file in the logs is not present (laptop amd GPU with "oland" architecture).

jay
jay
Joined: 25 Jan 07
Posts: 99
Credit: 84044023
RAC: 0

Greetings !! the 'bad'

Greetings !!

the 'bad' workaround of

  • creating bad directories,
  • putting in a link to the real ----.bc file,
  • substituting  verde-amdgcn-mesa-mesa3d.bc with the replacement: verde-amdgcn--.bc in the link

works.

and I progress to the next problem: verde-amdgcn-mesa-mesa3d.bc

which happens about 26 seconds into the WU.

And that is a different problem to share in the forum.

((wheeee))

Jay

 

Ian&Steve C.
Ian&Steve C.
Joined: 19 Jan 20
Posts: 4045
Credit: 48050706091
RAC: 34873899

Ian&Steve C. wrote: This is

Ian&Steve C. wrote:

This is not an issue with Einstein. It’s an issue with the Mesa package. You need to bring it up with the package maintainer(s). 
 

or switch to using the AMDGPU-Pro drivers which don’t have this issue. 

_________________________________________________________________________

Comment viewing options

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