sim: Add a system-global option to bypass caches

Virtualized CPUs and the fastmem mode of the atomic CPU require direct
access to physical memory. We currently require caches to be disabled
when using them to prevent chaos. This is not ideal when switching
between hardware virutalized CPUs and other CPU models as it would
require a configuration change on each switch. This changeset
introduces a new version of the atomic memory mode,
'atomic_noncaching', where memory accesses are inserted into the
memory system as atomic accesses, but bypass caches.

To make memory mode tests cleaner, the following methods are added to
the System class:

 * isAtomicMode() -- True if the memory mode is 'atomic' or 'direct'.
 * isTimingMode() -- True if the memory mode is 'timing'.
 * bypassCaches() -- True if caches should be bypassed.

The old getMemoryMode() and setMemoryMode() methods should never be
used from the C++ world anymore.

1 files changed, 8 insertions, 0 deletions

diff --git a/src/python/m5/simulate.py b/src/python/m5/simulate.py
index 3583e8264..682104c26 100644
--- a/src/python/m5/simulate.py
+++ b/src/python/m5/simulate.py
@@ -63,6 +63,7 @@ MaxTick = 2**63 - 1
 _memory_modes = {
     "atomic" : objects.params.atomic,
     "timing" : objects.params.timing,
+    "atomic_noncaching" : objects.params.atomic_noncaching,
     }
 
 # The final hook to generate .ini files.  Called from the user script
@@ -288,6 +289,13 @@ def switchCpus(system, cpuList, do_drain=True):
     # Change the memory mode if required. We check if this is needed
     # to avoid printing a warning if no switch was performed.
     if system.getMemoryMode() != memory_mode:
+        # Flush the memory system if we are switching to a memory mode
+        # that disables caches. This typically happens when switching to a
+        # hardware virtualized CPU.
+        if memory_mode == objects.params.atomic_noncaching:
+            memWriteback(system)
+            memInvalidate(system)
+
         _changeMemoryMode(system, memory_mode)
 
     for old_cpu, new_cpu in cpuList: