[Bug 81896] New: GPU reset when running some "John the Ripper" (+ jumbo patch, from Git) OpenCL tests

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Priority medium
Bug ID 81896
Assignee dri-devel@lists.freedesktop.org
Summary GPU reset when running some "John the Ripper" (+ jumbo patch, from Git) OpenCL tests
Severity normal
Classification Unclassified
OS Linux (All)
Reporter lbdkmjdf@sharklasers.com
Hardware x86-64 (AMD64)
Status NEW
Version git
Component Drivers/Gallium/r600
Product Mesa

Created attachment 103669 [details]
mixed outputs

I get GPU reset when running some opencl tests from this program from TTY. If
I'm running a DE when the lockup happens, the system stops responding.
Attached are:
- the mixed output from /var/log/syslog and this command:
$ R600_DEBUG=cs LWS=128 GWS=9216 ./john --test
--format=PBKDF2-HMAC-SHA256-opencl

- an attempt to clean it, only leaving the output from "R600_DEBUG=cs"

Also notice, there is random garbage after "Build log:", which comes from
clGetProgramBuildInfo (
https://github.com/magnumripper/JohnTheRipper/blob/bleeding-jumbo/src/common-opencl.c#L754
, prints "Build log:" in line 768)

My system:
Hardware: HP dv6, CPU: AMD A6-3400M, GPUs: Radeon HD 6520g (evergreen) + 6750M
(north islands)
OS: Kubuntu 14.04
Kernel: 3.15.7, and 3.13, both amd64
Mesa 10.3~git1407290730.9a53f9 (from oibaf's PPA,
https://launchpad.net/~oibaf/+archive/graphics-drivers ), compiled with LLVM
3.5-rc1.

________________
To reproduce, here's a mini "how to run" for "John The Ripper" + "jumbo"
patches:
$ git clone https://github.com/magnumripper/JohnTheRipper
$ cd JohnTheRipper/src
$ ./configure && make
$ cd ../run
$ ./john --help # prints usage help
$ # the following command prints a list of hash format tests that use opencl,
separated by spaces. remove the awk part to get one per line:
$ ./john --help | grep -Eo '[a-zA-Z0-9-]*-opencl' | awk '{printf "%s ", $0}';
echo

To run a test:
$ ./john --test --format=$format
where $format is one of the formats listed from the previous command

The program does verify the output. When it fails, it should print something
like "FAILED (get_hash[0](0))".


You are receiving this mail because:
_______________________________________________
dri-devel mailing list
dri-devel@xxxxxxxxxxxxxxxxxxxxx
http://lists.freedesktop.org/mailman/listinfo/dri-devel

[Index of Archives]     [Linux DRI Users]     [Linux Intel Graphics]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [XFree86]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [XFree86]
  Powered by Linux