What cycle is it ending at? If it's a huge number (like 2^63-1) then what that really means is that the CPU stalled and stopped scheduling events, which causes you to rapidly move forward to the dummy event at MAXINT that keeps time from rolling over into negative values.
Steve On 11/13/07, Rick Strong <[EMAIL PROTECTED]> wrote: > Hi, > > When I attempt to resume an ammp checkpoint for ALPHA_SE AtomicSimpleCPU > at location 59*100E6 instructions into the binary, I am able to run in > AtomicSimpleCPU, but if I switch to O3CPU, I am able to run 248,483 > instructions when suddenly i get the "Exiting @ cycle ... because > simulate() limit reached" which is the normal simulator exit event. As I > set things up to exit at a particular instruction count of 100E6, I an > uncertain what could be causing this early exit. Any ideas? I looked at > a trace of "O3CPUAll" and it seemed fine. (Side note: this seems to be > happening with any ammp checkpoint I take, where as others appears to be > fine). > > -Richard > > _______________________________________________ > m5-users mailing list > m5-users@m5sim.org > http://m5sim.org/cgi-bin/mailman/listinfo/m5-users > _______________________________________________ m5-users mailing list m5-users@m5sim.org http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
