Re: [Gems-users] splash2 benchmarks compilation

[Greg Byrd <gbyrd@xxxxxxxx>]

You are using the "null" version of the macros that come with the Splash-2 distribution (codes/null_macros/c.m4.null).  The CREATE macro prints the message that you see.  You need a version of the macros that uses pthreads (or something) to create threads and locks, and to implement barriers.

...Greg


srikant wrote:

hello everyone

i have a question regarding running splash2 benchmarks with simics+ruby, i have created a 4 processor checkpoint using the $num_cpus attribute with simics, i started simics in the -stall mode, followed all the steps according to the quick start page
i have set ruby0.setparam g_NUM_PROCESSORS 4 and ruby0.setparam g_MEMORY_SIZE_BYTES to  4294967296. i am using MOSI_SMP_bcast protocol.

when the compile the benchmark FFT as ./FFT -m10 -p4 i get the following message

FFT with Blocking Transpose
   1024 Complex Doubles
   4 Processors
   65536 Cache lines
   16 Byte line size
   4096 Bytes per page

No more processors -- this is a uniprocessor version!

every thing works fine with a single processor(./FFT -m18 -p1), i get the ruby statistics with a single processor
could anyone of you please suggest what mistake i am making or suggest any documentaion to go through to figure out a way ( to get ruby statistics for splash2 benchmark program for more > than 1 processor) as i am new to using gems and simics

thanks in advance
sri

_______________________________________________
Gems-users mailing list
Gems-users@xxxxxxxxxxx
https://lists.cs.wisc.edu/mailman/listinfo/gems-users
Use Google to search the GEMS Users mailing list by adding "site:https://lists.cs.wisc.edu/archive/gems-users/" to your search.