Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Date: Wed, 02 May 2012 14:05:35 -0300
From: Claudio André <>
Subject: Nvidia compiler verbosity

Hi, now it is possible to enabled verbose on NVIDIA compiler. No side 

Register, local, shared, and constant memory usages are reported by the 
compiler when compiling with the -cl-nv-verbose build option.

Let me know if you guys would like to have this on default version (only 

OpenCL platform 0: NVIDIA CUDA, 1 device(s).
Using device 0: GeForce GTX 570
Compilation log:
ptxas info    : Compiling entry function 'kernel_crypt' for 'sm_20'
ptxas info    : Function properties for kernel_crypt
  144 bytes stack frame, 16 bytes spill stores, 24 bytes spill loads
ptxas info    : Used 63 registers, 52 bytes cmem[0], 640 bytes cmem[2], 
24 bytes cmem[16]
Building the kernel, this will take a while: Done.
Local work size (LWS) 32, Keys per crypt (KPC) 512
Benchmarking: crypt SHA-512 (rounds=5000) [OpenCL]...

Changes are minimum:

     sprintf(include, "-I %s %s %s%d %s %s", path_expand(pathname),
             get_device_type(dev_id) == CL_DEVICE_TYPE_CPU ?
             "-DDEVICE_IS_CPU" : "",
                 "-DDEVICE_INFO=", device_info,
+ =>                gpu_nvidia(device_info) ? "-cl-nv-verbose" : "",
             "-cl-strict-aliasing -cl-mad-enable");

Powered by blists - more mailing lists

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.