AMD Drivers for Debian Jessie - Installing not successful

Message boards : GPUs : AMD Drivers for Debian Jessie - Installing not successful
Message board moderation

To post messages, you must log in.

1 · 2 · Next

AuthorMessage
Ken Linder

Send message
Joined: 27 Jul 11
Posts: 18
United States
Message 73356 - Posted: 16 Oct 2016, 19:10:55 UTC

Before contacting the vendor and opening a support ticket, thought I would ask the experts here.

I have been using BOINC for quite a while but this is my first attempt to use a GPU on a Linux platform. I bought an inexpensive AMD 645x card but the Linux drivers are giving me all sorts of problems. My main goal is to use this AMD GPU as a BOINC data processor only; no video processing. The AMD APU and mobo will handle the video.

Hardware:
CPU:AMD C-70
Motherboard: ASROCK C70M1
RAM: 4GB

Video card: XFX Radeon HD6450 with 1GB RAM
from lspci: 01:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Caicos [Radeon HD 6450/7450/8450 / R5 230 OEM]

OS: Debian Jessie
Windows Manager: LXDE

After installing fglrx-driver and the AMD Catalyst drivers (tried several Catalyst drivers with similar results), on reboot, the GUI boot fails with a message "Failed to start Light Display Manager" I followed that trail for about two hours before giving up. Uninstalling the drivers through a text console will let me reboot back into the LXDE GUI.

So, before I really dig in and try to figure this out, is there anyone out there why might possibly know how to properly install the drivers for this card and have BOINC recognize it?

Thanks!
-Ken
ID: 73356 · Report as offensive
Profile Jord
Volunteer tester
Help desk expert
Avatar

Send message
Joined: 29 Aug 05
Posts: 15477
Netherlands
Message 73357 - Posted: 16 Oct 2016, 19:51:40 UTC - in response to Message 73356.  
Last modified: 16 Oct 2016, 19:53:46 UTC

Do you have a monitor attached to the GPU, or a VGA dummy connector? As without one of either, I don't think you can get it to be detected.

Edit: If you're not technically handy, they come as pre-made.
ID: 73357 · Report as offensive
Profile Agentb
Avatar

Send message
Joined: 30 May 15
Posts: 265
United Kingdom
Message 73358 - Posted: 16 Oct 2016, 21:26:17 UTC - in response to Message 73356.  

I have not used Jessie and fgrlx, nor running a single AMD GPU headless.

There are two links in this message about running headless Radeon R9 290X not detected by BOINC that might help.

But, if you want to get it running quickly i would suggest getting the GPU running in a normal fashion. fgrlx is not easy to work with,especially without X running.

Good luck...
ID: 73358 · Report as offensive
Ken Linder

Send message
Joined: 27 Jul 11
Posts: 18
United States
Message 73364 - Posted: 17 Oct 2016, 15:14:37 UTC
Last modified: 17 Oct 2016, 15:15:05 UTC

Thanks Jord and Agentb.

I appreciate all of that advice. :-) Removed all files and remnants from fgrlx and old install attempts of Catalyst and associated drivers. Then I connected the monitor to the video card. From text mode I installed the drivers and Catalyst. Going to not worry about making it headless for now. I just want to make it work.

And it booted into X/LXDE... I am one step closer but not quite there yet.

The drivers and Catalyst are installed. But, BOINC doesn't seem to recognize the graphics card. BOINC Manager shows "GPU Missing, Waiting to run (0.00936 CPUs + 1 AMD/ATI GPU)

Here are the important lines from BOINC's Event Log:
Mon 17 Oct 2016 09:49:25 AM CDT |  | GPU specified in cc_config.xml: 1 ATI
Mon 17 Oct 2016 09:49:25 AM CDT |  | [coproc] launching child process at /usr/bin/boinc
Mon 17 Oct 2016 09:49:25 AM CDT |  | [coproc] relative to directory /var/lib/boinc-client
Mon 17 Oct 2016 09:49:25 AM CDT |  | [coproc] with data directory /var/lib/boinc-client
Mon 17 Oct 2016 09:49:25 AM CDT |  | OpenCL CPU: AMD C-70 APU with Radeon(tm) HD Graphics (OpenCL driver vendor: Advanced Micro Devices, Inc., driver version 1800.11 (sse2), device version OpenCL 1.2 AMD-APP (1800.11))
Mon 17 Oct 2016 09:49:25 AM CDT |  | No NVIDIA library found
Mon 17 Oct 2016 09:49:25 AM CDT |  | calInit() returned 1
Mon 17 Oct 2016 09:49:25 AM CDT |  | App version needs OpenCL but GPU doesn't support it
Mon 17 Oct 2016 09:49:25 AM CDT | SETI@home | Application uses missing ATI GPU
Mon 17 Oct 2016 09:49:25 AM CDT | SETI@home | Missing coprocessor for task 15jl16aa.21470.19703.6.33.214_0


It appears to me BOINC is only seeing the GPU embedded in the CPU, and I know that will not work with BOINC.

clinfo seems to see both GPUs:

  Device Type:                                   CL_DEVICE_TYPE_GPU
  Vendor ID:                                     1002h
  Board name:                                    AMD Radeon HD 6450  
  Device Topology:                               PCI[ B#1, D#0, F#0 ]
...
  Device Type:                                   CL_DEVICE_TYPE_CPU
  Vendor ID:                                     1002h
  Board name:                                    


I tried several things from this post, but the things I tried don't help. Here are a list of my attempts: SLEEP 6 in the startup script; run BOINC as root; several changes to the cc_config.xml (including <use_all_gpus>1</use_all_gpus> and specifying which GPU to use in <coproc>.)

At a loss for the moment, but need to get to work & pay the bills.

Thoughts would be appreciated. Thanks!!!

-Ken
ID: 73364 · Report as offensive
Profile Agentb
Avatar

Send message
Joined: 30 May 15
Posts: 265
United Kingdom
Message 73371 - Posted: 17 Oct 2016, 18:32:58 UTC - in response to Message 73364.  
Last modified: 17 Oct 2016, 18:33:28 UTC

It's not clear what version of boinc you are running, the later version are MUCH better at GPU detection and logging.

Can you set coproc_debug in event log options, and post the results.

Thank you.
ID: 73371 · Report as offensive
Ken Linder

Send message
Joined: 27 Jul 11
Posts: 18
United States
Message 73373 - Posted: 17 Oct 2016, 19:42:14 UTC

Sure! Here is the entire log immediately after restart.
Mon 17 Oct 2016 01:59:35 PM CDT |  | Starting BOINC client version 7.4.23 for x86_64-pc-linux-gnu
Mon 17 Oct 2016 01:59:35 PM CDT |  | log flags: file_xfer, sched_ops, task, coproc_debug
Mon 17 Oct 2016 01:59:35 PM CDT |  | Libraries: libcurl/7.38.0 OpenSSL/1.0.1k zlib/1.2.8 libidn/1.29 libssh2/1.4.3 librtmp/2.3
Mon 17 Oct 2016 01:59:35 PM CDT |  | Data directory: /var/lib/boinc-client
Mon 17 Oct 2016 01:59:35 PM CDT |  | [coproc] launching child process at /usr/bin/boinc
Mon 17 Oct 2016 01:59:35 PM CDT |  | [coproc] relative to directory /var/lib/boinc-client
Mon 17 Oct 2016 01:59:35 PM CDT |  | [coproc] with data directory /var/lib/boinc-client
Mon 17 Oct 2016 01:59:35 PM CDT |  | OpenCL CPU: AMD C-70 APU with Radeon(tm) HD Graphics (OpenCL driver vendor: Advanced Micro Devices, Inc., driver version 1800.11 (sse2), device version OpenCL 1.2 AMD-APP (1800.11))
Mon 17 Oct 2016 01:59:35 PM CDT |  | No NVIDIA library found
Mon 17 Oct 2016 01:59:35 PM CDT |  | calInit() returned 1
Mon 17 Oct 2016 01:59:35 PM CDT |  | No usable GPUs found
Mon 17 Oct 2016 01:59:35 PM CDT |  | app version refers to missing GPU type ATI
Mon 17 Oct 2016 01:59:35 PM CDT | SETI@home | Application uses missing ATI GPU
Mon 17 Oct 2016 01:59:35 PM CDT | SETI@home | Missing coprocessor for task 15jl16aa.21470.19703.6.33.214_0
Mon 17 Oct 2016 01:59:35 PM CDT |  | Host name: luna
Mon 17 Oct 2016 01:59:35 PM CDT |  | Processor: 2 AuthenticAMD AMD C-70 APU with Radeon(tm) HD Graphics [Family 20 Model 2 Stepping 0]
Mon 17 Oct 2016 01:59:35 PM CDT |  | 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 extd_apicid aperfmperf pni monitor ssse3 cx16 popcnt lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch ibs skinit wdt arat cpb hw_pstate npt lbrv svm_lock nrip_save pausefilter vmmcall
Mon 17 Oct 2016 01:59:35 PM CDT |  | OS: Linux: 3.16.0-4-amd64
Mon 17 Oct 2016 01:59:35 PM CDT |  | Memory: 3.85 GB physical, 9.33 GB virtual
Mon 17 Oct 2016 01:59:35 PM CDT |  | Disk: 45.83 GB total, 37.57 GB free
Mon 17 Oct 2016 01:59:35 PM CDT |  | Local time is UTC -5 hours
Mon 17 Oct 2016 01:59:35 PM CDT |  | Config: use all coprocessors
Mon 17 Oct 2016 01:59:35 PM CDT |  | Config: GUI RPCs allowed from:
Mon 17 Oct 2016 01:59:35 PM CDT | Milkyway@Home | URL http://milkyway.cs.rpi.edu/milkyway/; Computer ID 688777; resource share 100
Mon 17 Oct 2016 01:59:35 PM CDT | SETI@home | URL http://setiathome.berkeley.edu/; Computer ID 7971203; resource share 100
Mon 17 Oct 2016 01:59:35 PM CDT | SETI@home | General prefs: from SETI@home (last modified 29-Jul-2016 13:50:35)
Mon 17 Oct 2016 01:59:35 PM CDT | SETI@home | Host location: none
Mon 17 Oct 2016 01:59:35 PM CDT | SETI@home | General prefs: using your defaults
Mon 17 Oct 2016 01:59:35 PM CDT |  | Reading preferences override file
Mon 17 Oct 2016 01:59:35 PM CDT |  | Preferences:
Mon 17 Oct 2016 01:59:35 PM CDT |  | max memory usage when active: 2761.84MB
Mon 17 Oct 2016 01:59:35 PM CDT |  | max memory usage when idle: 3550.94MB
Mon 17 Oct 2016 01:59:35 PM CDT |  | max disk usage: 36.62GB
Mon 17 Oct 2016 01:59:35 PM CDT |  | (to change preferences, visit a project web site or select Preferences in the Manager)
Mon 17 Oct 2016 01:59:35 PM CDT |  | gui_rpc_auth.cfg is empty - no GUI RPC password protection
Mon 17 Oct 2016 01:59:35 PM CDT |  | Not using a proxy
Mon 17 Oct 2016 02:40:49 PM CDT |  | Re-reading cc_config.xml
Mon 17 Oct 2016 02:40:49 PM CDT |  | Not using a proxy
Mon 17 Oct 2016 02:40:49 PM CDT |  | Config: use all coprocessors
Mon 17 Oct 2016 02:40:49 PM CDT |  | Config: GUI RPCs allowed from:
Mon 17 Oct 2016 02:40:49 PM CDT |  | log flags: file_xfer, sched_ops, task, coproc_debug

ID: 73373 · Report as offensive
Profile Agentb
Avatar

Send message
Joined: 30 May 15
Posts: 265
United Kingdom
Message 73374 - Posted: 17 Oct 2016, 20:31:21 UTC - in response to Message 73373.  
Last modified: 17 Oct 2016, 20:46:42 UTC

OK i see that you are running 7.4.23 which is a bit old.

LocutusOfBorg boinc PPA should get a later version.

and then again try the coproc_debug flag.

Edit: also how exactly are you stopping and starting the boinc client? Have you tried restarting the boinc-client from within a terminal window. Assuming Jessie is not systrmd - it will be
/etc/init.d/boinc-client start
ID: 73374 · Report as offensive
Juha
Volunteer developer
Volunteer tester
Help desk expert

Send message
Joined: 20 Nov 12
Posts: 801
Finland
Message 73377 - Posted: 17 Oct 2016, 20:55:51 UTC - in response to Message 73374.  

Good link but wrong distro :) jessie-backports works better here.

@Ken Notice that you have an OpenCL CPU which is listed in both clinfo and BOINC output.
ID: 73377 · Report as offensive
Ken Linder

Send message
Joined: 27 Jul 11
Posts: 18
United States
Message 73378 - Posted: 17 Oct 2016, 22:06:06 UTC
Last modified: 17 Oct 2016, 22:06:23 UTC

@Agentb - Didn't realize my version was that old. After my last batch of work finishes I will upgrade. (finger's crossed)

I restart Boinc like this:
/etc/init.d/boinc-client restart


@Juha - Thanks for that. I did notice the OpenCL CPU was listed in both. I'm new to GPU processing in a Linux environment so no good or bad things popped up in my brain regarding that. :-)
ID: 73378 · Report as offensive
Profile Agentb
Avatar

Send message
Joined: 30 May 15
Posts: 265
United Kingdom
Message 73379 - Posted: 17 Oct 2016, 22:36:49 UTC - in response to Message 73378.  

@Agentb - Didn't realize my version was that old. After my last batch of work finishes I will upgrade. (finger's crossed)


About two years... significant improvements to GPU detection (waves to Juha) happened at 7.6.22


I restart Boinc like this:
/etc/init.d/boinc-client restart


@Juha - Thanks for that. I did notice the OpenCL CPU was listed in both. I'm new to GPU processing in a Linux environment so no good or bad things popped up in my brain regarding that. :-)


clinfo will report both CPU and GPU so that's a good sign.

I guess you have tried
sudo xhost +SI:localuser:boinc
before restarting?
ID: 73379 · Report as offensive
Ken Linder

Send message
Joined: 27 Jul 11
Posts: 18
United States
Message 73453 - Posted: 19 Oct 2016, 21:04:20 UTC - in response to Message 73379.  
Last modified: 19 Oct 2016, 21:06:30 UTC

Yes, I tried xhost to add boinc to the local user ACL, but the client would still not start a GPU workunit.

I did pull down version 7.6.33 from the backport page and built it. But, now can't start the client. Create a seperate thread for that problem here http://boinc.berkeley.edu/dev/forum_thread.php?id=11262

At the least, it seems I am making progress.

[edited] Guess I could have pulled down the binaries for AMD64, maybe would have been easier.
ID: 73453 · Report as offensive
Ken Linder

Send message
Joined: 27 Jul 11
Posts: 18
United States
Message 73456 - Posted: 19 Oct 2016, 21:56:03 UTC - in response to Message 73453.  

OK, after installing binaries from the backport repository, now the BOINC client doesn't recognize any GPUs

Wed 19 Oct 2016 04:47:07 PM CDT |  | Starting BOINC client version 7.6.33 for x86_64-pc-linux-gnu
Wed 19 Oct 2016 04:47:07 PM CDT |  | log flags: file_xfer, sched_ops, task, coproc_debug
Wed 19 Oct 2016 04:47:07 PM CDT |  | Libraries: libcurl/7.38.0 OpenSSL/1.0.1t zlib/1.2.8 libidn/1.29 libssh2/1.4.3 librtmp/2.3
Wed 19 Oct 2016 04:47:07 PM CDT |  | Data directory: /var/lib/boinc-client
Wed 19 Oct 2016 04:47:07 PM CDT |  | [coproc] launching child process at /usr/bin/boinc
Wed 19 Oct 2016 04:47:07 PM CDT |  | [coproc] relative to directory /
Wed 19 Oct 2016 04:47:07 PM CDT |  | [coproc] with data directory /var/lib/boinc-client
Wed 19 Oct 2016 04:47:07 PM CDT |  | [coproc] NVIDIA: libcuda.so: cannot open shared object file: No such file or directory
Wed 19 Oct 2016 04:47:07 PM CDT |  | [coproc] calInit() returned 1
Wed 19 Oct 2016 04:47:07 PM CDT |  | [coproc] clGetDeviceInfo failed to get CL_DEVICE_SIMD_PER_COMPUTE_UNIT_AMD for device 0
Wed 19 Oct 2016 04:47:07 PM CDT |  | [coproc] OpenCL library present but no OpenCL-capable devices found
Wed 19 Oct 2016 04:47:07 PM CDT |  | No usable GPUs found
Wed 19 Oct 2016 04:47:07 PM CDT |  | app version refers to missing GPU type ATI
Wed 19 Oct 2016 04:47:07 PM CDT | SETI@home | Application uses missing ATI GPU
Wed 19 Oct 2016 04:47:07 PM CDT | SETI@home | Missing coprocessor for task 15jl16aa.21470.19703.6.33.214_0
Wed 19 Oct 2016 04:47:07 PM CDT |  | Host name: luna
Wed 19 Oct 2016 04:47:07 PM CDT |  | Processor: 2 AuthenticAMD AMD C-70 APU with Radeon(tm) HD Graphics [Family 20 Model 2 Stepping 0]
Wed 19 Oct 2016 04:47:07 PM CDT |  | 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 extd_apicid aperfmperf pni monitor ssse3 cx16 popcnt lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch ibs skinit wdt arat cpb hw_pstate npt lbrv svm_lock nrip_save pausefilter vmmcall
Wed 19 Oct 2016 04:47:07 PM CDT |  | OS: Linux: 3.16.0-4-amd64


Thanks again!
ID: 73456 · Report as offensive
Profile Agentb
Avatar

Send message
Joined: 30 May 15
Posts: 265
United Kingdom
Message 73457 - Posted: 19 Oct 2016, 22:56:55 UTC - in response to Message 73456.  
Last modified: 19 Oct 2016, 23:04:26 UTC

OK, after installing binaries from the backport repository, now the BOINC client doesn't recognize any GPUs


Has it ever recognized? - i guess it must have, if it has a task.

Check clinfo is running as before from a terminal window (running X) and that is running on the GPU.

OK assuming clinfo is still working, check boinc is a member of the video group
# groups boinc
should show it.

restart boinc from the terminal window.

Edit: probably best to include a pastebin of the full clinfo next, it's not clear to me what drivers were installed.
ID: 73457 · Report as offensive
Profile Agentb
Avatar

Send message
Joined: 30 May 15
Posts: 265
United Kingdom
Message 73458 - Posted: 20 Oct 2016, 0:05:22 UTC

While i think of it.

Assuming you installed via the repo here check that the icd "amd-opencl-icd" loaders are installed. If not install them.
ID: 73458 · Report as offensive
Ken Linder

Send message
Joined: 27 Jul 11
Posts: 18
United States
Message 73464 - Posted: 20 Oct 2016, 13:50:09 UTC - in response to Message 73458.  

Yes, boinc is a member of video.
root@luna:/home/ken# groups boinc
boinc : boinc video


It did recognize the APU's GPU before, downloaded a work unit then complained because the GPU could not support any work units (or something to that effect, the BOINC log is somewhere above).

I checked that I had amd-opencl-icd installed, and I do.
root@luna:/home/ken# apt-get install amd-opencl-icd
Reading package lists... Done
Building dependency tree       
Reading state information... Done
amd-opencl-icd is already the newest version.
0 upgraded, 0 newly installed, 0 to remove and 207 not upgraded.


root@luna:/home/ken# dpkg-query -L amd-opencl-icd
/.
/etc
/etc/OpenCL
/etc/OpenCL/vendors
/etc/OpenCL/vendors/amdocl64.icd
/usr
/usr/share
/usr/share/lintian
/usr/share/lintian/overrides
/usr/share/lintian/overrides/amd-opencl-icd
/usr/share/bug
/usr/share/bug/amd-opencl-icd
/usr/share/bug/amd-opencl-icd/control
/usr/share/bug/amd-opencl-icd/script
/usr/share/doc
/usr/share/doc/amd-opencl-icd
/usr/share/doc/amd-opencl-icd/changelog.Debian.gz
/usr/share/doc/amd-opencl-icd/copyright
/usr/lib
/usr/lib/x86_64-linux-gnu
/usr/lib/x86_64-linux-gnu/libamdocl64.so
/usr/lib/x86_64-linux-gnu/libamdocl12cl64.so


Here is a pastebin of my clinfo results:http://pastebin.com/Mmcvh8k2 Looks to me I may have an older version of openCL. Maybe that's the problem?
ID: 73464 · Report as offensive
Profile Agentb
Avatar

Send message
Joined: 30 May 15
Posts: 265
United Kingdom
Message 73465 - Posted: 20 Oct 2016, 14:49:10 UTC - in response to Message 73464.  
Last modified: 20 Oct 2016, 15:32:41 UTC

OK I have just re-read the thread, the clinfo looks fine to me.

I notice you mentioned you added
<coproc>
  Some things...
</coproc>

that is not needed and needs removing.

<use_all_gpus>1</use_all_gpus>

Can be kept.

Edit: might easier to post the cc_config.xml file in total.

I'm still curious about the OpenCL dynamic libraries

ldconfig -p | grep OpenCL

but after that i'm short of ideas (you were detecting one device before, but not now)

Edit2: You haven't mentioned this but the get the GPU working with fglrx - there will have been a "sudo aticonfig --initial" at some point.
ID: 73465 · Report as offensive
Ken Linder

Send message
Joined: 27 Jul 11
Posts: 18
United States
Message 73473 - Posted: 21 Oct 2016, 9:04:36 UTC - in response to Message 73465.  

Odd. I have no <coproc> section and <use_all_gpus> is set to 1. Also, <no_gpus> is set to 0. Here's my cc_config.xml

<cc_config>
    <log_flags>
        <file_xfer>1</file_xfer>
        <sched_ops>1</sched_ops>
        <task>1</task>
        <android_debug>0</android_debug>
        <app_msg_receive>0</app_msg_receive>
        <app_msg_send>0</app_msg_send>
        <async_file_debug>0</async_file_debug>
        <benchmark_debug>0</benchmark_debug>
        <checkpoint_debug>0</checkpoint_debug>
        <coproc_debug>1</coproc_debug>
        <cpu_sched>0</cpu_sched>
        <cpu_sched_debug>0</cpu_sched_debug>
        <cpu_sched_status>0</cpu_sched_status>
        <dcf_debug>0</dcf_debug>
        <disk_usage_debug>0</disk_usage_debug>
        <file_xfer_debug>0</file_xfer_debug>
        <gui_rpc_debug>0</gui_rpc_debug>
        <heartbeat_debug>0</heartbeat_debug>
        <http_debug>0</http_debug>
        <http_xfer_debug>0</http_xfer_debug>
        <idle_detection_debug>0</idle_detection_debug>
        <mem_usage_debug>0</mem_usage_debug>
        <network_status_debug>0</network_status_debug>
        <notice_debug>0</notice_debug>
        <poll_debug>0</poll_debug>
        <priority_debug>0</priority_debug>
        <proxy_debug>0</proxy_debug>
        <rr_simulation>0</rr_simulation>
        <rrsim_detail>0</rrsim_detail>
        <sched_op_debug>0</sched_op_debug>
        <scrsave_debug>0</scrsave_debug>
        <slot_debug>0</slot_debug>
        <state_debug>0</state_debug>
        <statefile_debug>0</statefile_debug>
        <suspend_debug>0</suspend_debug>
        <task_debug>0</task_debug>
        <time_debug>0</time_debug>
        <trickle_debug>0</trickle_debug>
        <unparsed_xml>0</unparsed_xml>
        <work_fetch_debug>0</work_fetch_debug>
    </log_flags>
    <options>
        <abort_jobs_on_exit>0</abort_jobs_on_exit>
        <allow_multiple_clients>0</allow_multiple_clients>
        <allow_remote_gui_rpc>0</allow_remote_gui_rpc>
        <client_version_check_url>http://boinc.berkeley.edu/download.php?xml=1</client_version_check_url>
        <client_new_version_text></client_new_version_text>
        <client_download_url>http://boinc.berkeley.edu/download.php</client_download_url>
        <disallow_attach>0</disallow_attach>
        <dont_check_file_sizes>0</dont_check_file_sizes>
        <dont_contact_ref_site>0</dont_contact_ref_site>
        <lower_client_priority>0</lower_client_priority>
        <dont_suspend_nci>0</dont_suspend_nci>
        <dont_use_vbox>0</dont_use_vbox>
        <exit_after_finish>0</exit_after_finish>
        <exit_before_start>0</exit_before_start>
        <exit_when_idle>0</exit_when_idle>
        <fetch_minimal_work>0</fetch_minimal_work>
        <fetch_on_update>0</fetch_on_update>
        <force_auth>default</force_auth>
        <http_1_0>0</http_1_0>
        <http_transfer_timeout>300</http_transfer_timeout>
        <http_transfer_timeout_bps>10</http_transfer_timeout_bps>
        <max_event_log_lines>2000</max_event_log_lines>
        <max_file_xfers>8</max_file_xfers>
        <max_file_xfers_per_project>2</max_file_xfers_per_project>
        <max_stderr_file_size>0</max_stderr_file_size>
        <max_stdout_file_size>0</max_stdout_file_size>
        <max_tasks_reported>0</max_tasks_reported>
        <ncpus>-1</ncpus>
        <network_test_url>http://www.google.com/</network_test_url>
        <no_alt_platform>0</no_alt_platform>
        <no_gpus>0</no_gpus>
        <no_info_fetch>0</no_info_fetch>
        <no_opencl>0</no_opencl>
        <no_priority_change>0</no_priority_change>
        <os_random_only>0</os_random_only>
        <process_priority>-1</process_priority>
        <process_priority_special>-1</process_priority_special>
<proxy_info>
    <socks_server_name></socks_server_name>
    <socks_server_port>80</socks_server_port>
    <http_server_name></http_server_name>
    <http_server_port>80</http_server_port>
    <socks5_user_name></socks5_user_name>
    <socks5_user_passwd></socks5_user_passwd>
    <http_user_name></http_user_name>
    <http_user_passwd></http_user_passwd>
    <no_proxy></no_proxy>
</proxy_info>
        <rec_half_life_days>10.000000</rec_half_life_days>
        <report_results_immediately>0</report_results_immediately>
        <run_apps_manually>0</run_apps_manually>
        <save_stats_days>30</save_stats_days>
        <skip_cpu_benchmarks>0</skip_cpu_benchmarks>
        <simple_gui_only>0</simple_gui_only>
        <start_delay>0.000000</start_delay>
        <stderr_head>0</stderr_head>
        <suppress_net_info>0</suppress_net_info>
        <unsigned_apps_ok>0</unsigned_apps_ok>
        <use_all_gpus>1</use_all_gpus>
        <use_certs>0</use_certs>
        <use_certs_only>0</use_certs_only>
        <vbox_window>0</vbox_window>
    </options>
</cc_config>


Here's from ldconfig:
root@luna:/home/ken# ldconfig -p | grep OpenCL
	libOpenCL.so.1 (libc6,x86-64) => /usr/lib/x86_64-linux-gnu/libOpenCL.so.1
	libOpenCL.so.1 (libc6,x86-64) => /usr/lib/libOpenCL.so.1
	libOpenCL.so (libc6,x86-64) => /usr/lib/x86_64-linux-gnu/libOpenCL.so


I don't recall the aticonfig, so I ran it; no difference that I could see.
ID: 73473 · Report as offensive
Ken Linder

Send message
Joined: 27 Jul 11
Posts: 18
United States
Message 73475 - Posted: 21 Oct 2016, 11:02:13 UTC - in response to Message 73473.  
Last modified: 21 Oct 2016, 11:02:31 UTC

Made a tiny bit of progress. I did this:
xhost local:boinc


And, now BOINC sees the card, but still gets a failure with clGetDeviceInfo
Fri 21 Oct 2016 05:53:44 AM CDT |  | Starting BOINC client version 7.6.33 for x86_64-pc-linux-gnu
Fri 21 Oct 2016 05:53:44 AM CDT |  | log flags: file_xfer, sched_ops, task, coproc_debug
Fri 21 Oct 2016 05:53:44 AM CDT |  | Libraries: libcurl/7.38.0 OpenSSL/1.0.1t zlib/1.2.8 libidn/1.29 libssh2/1.4.3 librtmp/2.3
Fri 21 Oct 2016 05:53:44 AM CDT |  | Data directory: /var/lib/boinc-client
Fri 21 Oct 2016 05:53:44 AM CDT |  | [coproc] launching child process at /usr/bin/boinc
Fri 21 Oct 2016 05:53:44 AM CDT |  | [coproc] relative to directory /
Fri 21 Oct 2016 05:53:44 AM CDT |  | [coproc] with data directory /var/lib/boinc-client
Fri 21 Oct 2016 05:53:45 AM CDT |  | CAL: ATI GPU 0: AMD Radeon HD 6350/6450/7450/7470/R5 230 series (Caicos) (CAL version 1.4.1848, 1024MB, 909MB available, 400 GFLOPS peak)
Fri 21 Oct 2016 05:53:45 AM CDT |  | OpenCL: AMD/ATI GPU 0: AMD Radeon HD 6350/6450/7450/7470/R5 230 series (Caicos) (driver version 1800.11, device version OpenCL 1.2 AMD-APP (1800.11), 1024MB, 909MB available, 400 GFLOPS peak)
Fri 21 Oct 2016 05:53:45 AM CDT |  | [coproc] NVIDIA: libcuda.so: cannot open shared object file: No such file or directory
Fri 21 Oct 2016 05:53:45 AM CDT |  | [coproc] clGetDeviceInfo failed to get CL_DEVICE_SIMD_PER_COMPUTE_UNIT_AMD for device 0
Fri 21 Oct 2016 05:53:45 AM CDT |  | Host name: luna
Fri 21 Oct 2016 05:53:45 AM CDT |  | Processor: 2 AuthenticAMD AMD C-70 APU with Radeon(tm) HD Graphics [Family 20 Model 2 Stepping 0]
Fri 21 Oct 2016 05:53:45 AM CDT |  | 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 extd_apicid aperfmperf pni monitor ssse3 cx16 popcnt lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch ibs skinit wdt arat cpb hw_pstate npt lbrv svm_lock nrip_save pausefilter vmmcall
Fri 21 Oct 2016 05:53:45 AM CDT |  | OS: Linux: 3.16.0-4-amd64
Fri 21 Oct 2016 05:53:45 AM CDT |  | Memory: 3.85 GB physical, 9.33 GB virtual
Fri 21 Oct 2016 05:53:45 AM CDT |  | Disk: 45.83 GB total, 38.02 GB free
Fri 21 Oct 2016 05:53:45 AM CDT |  | Local time is UTC -5 hours
Fri 21 Oct 2016 05:53:45 AM CDT |  | Config: GUI RPCs allowed from:
Fri 21 Oct 2016 05:53:45 AM CDT |  | Config: use all coprocessors
...

ID: 73475 · Report as offensive
Profile Agentb
Avatar

Send message
Joined: 30 May 15
Posts: 265
United Kingdom
Message 73477 - Posted: 21 Oct 2016, 11:59:33 UTC - in response to Message 73475.  

Made a tiny bit of progress. I did this:
xhost local:boinc


And, now BOINC sees the card, but still gets a failure with clGetDeviceInfo


Hmmm. That error can be ignored, for the moment.

Is it picking up GPU tasks?

What is the output from xhost showing now?
ID: 73477 · Report as offensive
Ken Linder

Send message
Joined: 27 Jul 11
Posts: 18
United States
Message 73478 - Posted: 21 Oct 2016, 14:21:28 UTC - in response to Message 73477.  
Last modified: 21 Oct 2016, 14:23:53 UTC

BINGO!!!!!!!!
My GPU card is now working on a SETI workunit, with one waiting in the queue.

Must have been a combination of things to get here. One thing I did was add
sleep 10
xhost local:boinc

in the boinc-client script, in the start() proc right after this:
    if [ -n "$DISPLAY" -a -x /usr/bin/xhost ]; then
       # grant the boinc client to perform GPU computing
       xhost local:boinc || echo -n "xhost error ignored, GPU computing may not be possible"
    fi


It is odd... In order for it to recognize and do any GPU work, I must still stop and restart the daemon.

One thing I didn't mention before was the pesky message I received from the BIOS before boot:
The system failed to boot several times."
Never saw that before installing the GPU. I reset the BIOS to defaults and now no message on startup.

And, something that I really do NOT like... X appears to stutter occasionally when the client is running with a GPU work unit running. The screen freezes temporarily and overall is non-responsive. Not sure what is happening...

But, it is indeed processing GPU work units now.

Thanks again!!!
-Ken
ID: 73478 · Report as offensive
1 · 2 · Next

Message boards : GPUs : AMD Drivers for Debian Jessie - Installing not successful

Copyright © 2024 University of California.
Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.