Bug concerning Proactive Reset

There seems to be a bug concerning proactive reset. This bug may be triggered in a matter of seconds when running the multi-threaded lbzip2 application on linux (32-bit).

Command to run: ./drrun -reset_every_nth_bb_unit 3 -- lbzip2 -c my_file > my_file.bz2

Debug message: DynamoRIO seems to believe that a client it making a raw sysenter.

<Paste into GDB to debug DynamoRIO clients: set confirm off add-symbol-file '/home/john/dynamorio/build/lib32/debug/libdynamorio.so' 0xb7c67bbc

<proactive reset @ 2043 fragments> <get_memory_info mismatch! (can happen if os combines entries in /proc/pid/maps) os says: 0xb7005000-0xb7205000 prot=0x00000000 cache says: 0xb7005000-0xb7206000 prot=0x00000000

<proactive reset @ 3039 fragments> <proactive reset @ 3140 fragments> <Application /usr/bin/lbzip2 (5466) DynamoRIO usage error : Is your client invoking raw system calls via vdso sysenter? While such behavior is not recommended and can create problems, it may work with the -sysenter_is_int80 runtime option.> <Usage error: Is your client invoking raw system calls via vdso sysenter? While such behavior is not recommended and can create problems, it may work with the -sysenter_is_int80 runtime option. (/home/john/dynamorio/core/dispatch.c, line 112)

A quick solution prior to a fix is to increase the vm_size/disable reset.