On Jun 25, 2008, at 7:16 PM, prannav shrestha wrote: > i changed the MAX_STREAM value in both exolex.cc and exolex.l file > to 256 and then recompiled the m5, but the problem is still the > same. so what to do next? You get familiar with GDB and try to figure out why you're still hitting that fatal if there are enough streams.
> And Steve, i switched to m5-0b3. This version doesn't have the > problem. Initially i was using m5-0b5 You may have worked around one problem like that, but b3 has lots of bugs that have been fixed in b5 which you'll probably start running into. It's probably best to find & fix the problem in the most current code (the mercurial repository) rather than use an older version. Ali > > > regards > prannav > > On Wed, Jun 25, 2008 at 6:10 PM, Ali Saidi <[EMAIL PROTECTED]> wrote: > You'll need edit the exolex.cc file since we don't generate the cc > file in the build script because it would require users to have flex/ > lex installed. > > Ali > > On Jun 25, 2008, at 4:37 PM, Steve Reinhardt wrote: > > > As the error message indicates, it's coming from the eio code. The > > limitation is on the number of simultaneously open EIO traces. The > > error message also tells you how to fix it. > > > > Does this mean you got past the error you were having previously > where > > the second thread never seemed to commit any instructions? If so, > can > > you let us know what the problem was in case others run into it? > > > > Steve > > > > On Wed, Jun 25, 2008 at 12:53 PM, prannav shrestha > > <[EMAIL PROTECTED]> wrote: > >> hi, > >> i am trying to run 8 core with each having 4 threads, using O3CPU. > >> the > >> number of instruction to be run by each thread is 10M. the threads > >> are > >> loaded with speccpu2000 benchmarks. but while rub the simulation, i > >> am > >> getting error like > >> > >> fatal: out of lex buffer streams, increase MAX_STREAMS > >> @ cycle 0 > >> [yy_setstream:/z/stever/bk/m5/eio/exolex.l, line 201] > >> > >> the simulation ran successfully for 4 core 2 thread and 8 core 1 > >> thread > >> system. What may be the reason for this? is there any limitation on > >> the > >> number of threads m5 can handle? > >> would appreaciate if someone help me to solve this problem > >> > >> thank u > >> > >> prannav > >> > >> > >> _______________________________________________ > >> m5-users mailing list > >> [email protected] > >> http://m5sim.org/cgi-bin/mailman/listinfo/m5-users > >> > > _______________________________________________ > > m5-users mailing list > > [email protected] > > http://m5sim.org/cgi-bin/mailman/listinfo/m5-users > > > > _______________________________________________ > m5-users mailing list > [email protected] > http://m5sim.org/cgi-bin/mailman/listinfo/m5-users > > _______________________________________________ > m5-users mailing list > [email protected] > http://m5sim.org/cgi-bin/mailman/listinfo/m5-users _______________________________________________ m5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
