Merge branch 'master' of /home/git/model-checker
avoid warning
merge Merge branch 'master' of ssh://demsky.eecs.uci.edu/home/git/model-checker Conflicts: config.h model.cc
optimization...don't calloc the datarace tables every single time...it is really expensive for rollback
common/config: allow enable/disable ASSERT() easily Now, assertions (ASSERT()) are only checked if the CONFIG_ASSERT macro is defined. We leave this enabled for all builds for now.
common/config: disable ASSERT() by default
small hack to optimize our common case for snapshotting and set config option to turn this back on. hack assumes that same pages will be dirty across different executions so just copies them and doesn't reprotect them
fix bug.... not quite perfect for linux locks, but runs to completion... Revert "model: revert broken bugfix" This reverts commit ae7fcd2e5e499f72d9d1530bdc293f4fbc5f0644.
config: don't enable graph dumping in master repo
(1) add actions for the fence (2) a little more support for cyclegraph -- show rmw edges (3) add extra documentation for norris
add support for dumping cyclegraphs as dot files... also eliminate redundant edges to make them easier to view
move more configurables and add documentation
config: automatically determine BIT48 GCC says we can use _LP64, and LLVM (clang) imitates, so this can be automatically redefined for 32-bit/64-bit architectures. Also, add config.h to Makefile so it triggers a rebuild. This is still a blunt mechanism, but it's fine for now.
datarace: fixup, properly document BIT48 The "48-bit virtual address" macro should be checked with #if, not #ifdef, since it's defined to 1 or 0. You would use #ifdef if it's either "define" or "not defined." Make this clear in the comment as well.
add forgetten file