Hello,
I'm developing with other 3 colleagues an extension of GEM5 to support
the injection of 'Stuck@' fault in the BPU for a university project. We
have developed also some simple branch prediction units. We are
validating our work but we are encountering some problems: we don't have
clear how in the O3CPU model is performed the speculation and in
particular how are treated the branches in a speculative path after it
is discovered that was a "false path" (an execution flow caused by a
mispredicted branch).
Thanks in advance,
Antonio
_______________________________________________
gem5-dev mailing list
[email protected]
http://m5sim.org/mailman/listinfo/gem5-dev
