On Mar 16, 2007, at 11:01 AM, Nicolas Zea wrote:
I'm running an Alpha system emulation compilation, and calling sched_break_cycle in gdb gives an error about it not existing in the context. setTraceFlag and clearTraceFlag work just fine though. Has this function been removed/changed or might I be doing something wrong?
It is now called schedBreakCycle().
Additionally, are trace-flags sometimes disabled during runtime? I've made my own trace flag by adding it to traceflags.py, and then tried to use it in a DPRINTF call. Depending on where I make the call, the traceflag is sometimes on and sometimes off. For example, within BaseDynInst<impl>::read() it prints fine, but within PhysicalMemory::doFunctionalAccess() it won't print, and just skips the DPRINTF statement as if the trace flag were not set.
No they aren't. I've never seen any behavior like that.
I've tried setting the flag both through the --trace-flags="" argument when I run the program and using setTraceFlags from within gdb. Is there some sort of scope to where certain trace flags exist?
There is no scoping of trace flags. Ali _______________________________________________ m5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
