34408c3122
client: don't start var names with __
2017-04-13 09:16:58 -07:00
ae97e08a95
client: Mac: continue replacing deprecated Mac APIs
2017-02-25 04:56:12 -08:00
dcd9eb3c2e
Client: fix nvidia_driver_version dlopen()
...
On Debian only the versioned object can be opened. It usually is a symlink to the actual file.
2016-04-05 07:43:22 +02:00
31446d1cb9
client: Update commits 5c1fa20 & 5e984c0: when Mac is booted, it can take over 120 seconds before system extension CUDA.kext is loaded (fix occasional failure to recognize CUDA devices when BOINC is launched automatically at login before CUDA has initialized at system boot.)
2016-01-20 04:23:57 -08:00
bc8b9a5d9e
client (Unix): use dlerror() for GPU library failures; shows the filename
2015-12-04 17:54:34 -08:00
4a536c7cba
client: Fix crash bug introduced in commits f0df42e and 1618ce3: do not dlclose(cudalib) until after we are finished using the CUDA library. Also added corresponding FreeLibrary(cudalib) for MS Windows builds.
2015-12-03 02:12:31 -08:00
259ed89c46
client: Fix bug in detecting CUDA GPUs on Macs which was exposed by commits 7be9ff7 and d0005d7.
2015-11-21 05:18:17 -08:00
2df170ebda
client: Fix bug in detecting CUDA GPUs on Macs which was exposed by commits 7be9ff7 and d0005d7.
2015-11-21 04:56:51 -08:00
f0df42efbd
client (Unix): dlclose() dynamic libraries in GPU code
2015-10-21 08:34:02 -07:00
b8bd091a64
client: fix source code indentation from previous commit.
2015-09-28 03:08:00 -07:00
5e984c0778
client: Fix an issue which can sometimes cause client to take a long time to connect, especially on OS 10.10 Yosemite.
2015-09-28 03:04:50 -07:00
5c1fa20254
client: On Macs, fix failure to recognize CUDA devices when BOINC is launched automatically at login before CUDA has initialized at system boot.
2015-08-14 04:03:38 -07:00
408003dd71
client: Allow using NVIDIA GPUs with Compute Capability < 2.0 with CUDA driver 6.5 or later on non-Mac platforms. Future versions of CUDA will drop support for these GPUs on all platforms and presumably not report these GPUs, so we still explain that situation in the Event Log and Notices.
2014-11-05 02:12:41 -08:00
62e70a8e98
client: Handle future versions of CUDA which may not report NVIDIA GPUs with Compute Capability < 2.0 at all; prevent use of those GPUs for OpenCL and show explanation in Event Log and Notices.
2014-10-26 04:12:01 -07:00
5a8b1a2d21
client: If CUDA driver 6.5 or later is installed, prevent use of NVIDIA GPUs with Compute Capability < 2.0 and show explanation in Event Log and Notices.
2014-10-25 06:44:08 -07:00
b2fefc885b
client: If CUDA driver 6.5 or later is installed, prevent use of NVIDIA GPUs with Compute Capability < 2.0 and show explanation in Event Log and Notices.
2014-10-25 05:21:48 -07:00
4726923ef8
client: If CUDA driver 6.5 or later is installed, prevent use of NVIDIA GPUs with Compute Capability < 2.0 and show explanation in Event Log and Notices.
2014-10-25 03:55:29 -07:00
1e2fcb4b68
client/lib: change CONFIG to CC_CONFIG, config to cc_config.
...
Eliminates ambiguity of "config" global var, which is used in server code.
This confuses IDEs that are looking at all the code at once.
2014-05-08 00:51:18 -07:00
b523e16c85
Merge branch 'master' of ssh://boinc.berkeley.edu/boinc-v2
2014-05-02 07:41:41 -07:00
6a8eab73cd
replace tab characters with spaces
2014-05-01 21:03:49 -07:00
4539c96e8c
Merge branch 'master' of ssh://boinc.berkeley.edu/boinc-v2
2014-05-01 19:52:13 -07:00
fab00f1100
Use new (non-deprecated) NvAPI call to get driver version
2014-05-01 19:51:58 -07:00
583c336fd9
Merge branch 'master' of ssh://boinc.berkeley.edu/boinc-v2
2014-05-01 10:37:18 -07:00
ad25a639f9
Fixes for compiling without the sam.h macros.
2014-05-01 10:15:34 -07:00
386805ebce
client: fix compiler warning on Mac introduced by commit 450f680.
2014-05-01 02:26:23 -07:00
c6995ce06b
These files aren't modified. Git claims otherwise.
...
Subversion was so much better then git. I spend all my effing time dealing
with git's issues. What a PITA.
2014-04-29 07:59:06 -07:00
af4a5acc2e
client: Fix build break on Windows.
2014-04-28 14:48:21 -04:00
4901e27282
client: don't include GPUs that lack OpenCL/Cal/CUDA when main GPU has it
...
E.g.: if the "best" AMD GPU can do OpenCL, don't include AMD GPUs that can't,
even if use_all_gpus is set.
Otherwise lots of jobs will error out.
2014-04-26 09:05:00 -07:00
450f680880
client: detect NVIDIA driver version on Linux; from Matt Harvey
2014-04-24 09:41:23 -07:00
631e236b08
client: tweaks to code for detecting GPUs via a child process.
...
Added safety features requested by Rom Walton:
* Change COPROC_ATI::get_available_ram and COPROC_NVIDIA::get_available_ram to static routines to prevent calling them without first loading CAL or CUDA libraries.
* Add tests for NULL library calls in these routines.
* Add comments warning about need to call from a separate child process on dual-GPU laptops, proper library initialization, etc.
2013-06-27 02:36:20 -07:00
4d74c5abbd
client: tweaks to code for detecting GPUs via a child process and change sprintf calls to safer snprintf.
2013-06-26 05:00:25 -07:00
e2b2370e9d
client: optionally detect GPUs via a child process, for dual_GPU laptops.
...
Some dual-GPU laptops (e.g., Macbook Pro) don't power down the more powerful GPU until all applications which used them exit. To save battery life, the client launches a second instance of the client as a child process to detect and get info about the GPUs.
The child process writes the info to a temp file which our main client then reads.
This option is enabled at compile time by defining USE_CHILD_PROCESS_TO_DETECT_GPUS as non-zero in gpu_detect.cpp
2013-06-25 04:31:34 -07:00
26f9e380d9
- client: Initialize the total mem size_t.
...
svn path=/trunk/boinc/; revision=26116
2012-09-14 18:00:06 +00:00
aff3b7675c
client: Don't support CUDA on dual-GPU MacBooks with automatic GPU switching to save battery power
...
svn path=/trunk/boinc/; revision=26088
2012-09-12 10:05:41 +00:00
2882d5bc29
- client: initialize memfree and memtotal before use for Nvidia
...
cards. It appears that the Nvidia API was only setting 32-bits
of the 64-bit value. The remaining 32-bits were whatever
was on the stack.
client/
gpu_nvidia.cpp
svn path=/trunk/boinc/; revision=26084
2012-09-10 17:56:09 +00:00
446bc4ca28
- client: take GPU exclusions into account when making
...
initial work request to a project
- client: put some casts to double in NVIDIA detect code.
Shouldn't make any difference.
- volunteer storage: truncate file to right size after retrieval
svn path=/trunk/boinc/; revision=26051
2012-08-20 23:41:27 +00:00
68f9880615
- client: remove "device" entry from CUDA_DEVICE_PROP,
...
and change types of mem-size fields from int to double.
These fields are size_t in NVIDIA's version of this;
however, cuDeviceGetAttribute() returns them as int,
so I don't see where this makes any difference.
- client: fix bug in handling of <no_rsc_apps> element.
- scheduler: message tweaks.
Note: [foo] means that the message is enabled by <debug_foo>.
svn path=/trunk/boinc/; revision=25849
2012-07-05 20:24:17 +00:00
133430fa3b
- client: for NVIDIA, make an array of the PCI info
...
for all usable GPUs
svn path=/trunk/boinc/; revision=25775
2012-06-18 21:09:16 +00:00
5cf6321b67
- win compile fixes
...
svn path=/trunk/boinc/; revision=25774
2012-06-18 20:49:26 +00:00
b050deecf7
- client: compile fixes
...
svn path=/trunk/boinc/; revision=25773
2012-06-18 20:41:37 +00:00
5e61c29cc3
- client: split GPU detection code into separate files
...
svn path=/trunk/boinc/; revision=25771
2012-06-18 20:12:30 +00:00
David Anderson