Server Configuration parameters
Oracle Release 184.108.40.206.0 Production
Physical memory on server 4G
Swap space 6G
Being DEV box, I did not configured swap to the value equivalent to 2 x times the size of RAM.
I created a 11gr2 test database with above parameters (memory_max_target 1.7G) which worked fine. But when I created another database with same parameter, I got following error and PMON process terminated the instance.
Exception [type: SIGSEGV, Address not mapped to object] [ADDR:0xFFFFFFFF7FFF1EE0] [PC:0xFFFFFFFF7AFA8FF0, _ndoprnt()+4] [flags: 0x0, count: 1]
Errors in file /oracle/diag/rdbms/dumpdb/dumpdb/trace/dumpdb_pmon_2414.trc (incident=17):
ORA-07445: exception encountered: core dump [_ndoprnt()+4] [SIGSEGV] [ADDR:0xFFFFFFFF7FFF1EE0] [PC:0xFFFFFFFF7AFA8FF0] [Address not mapped to object] 
Incident details in: /oracle/diag/rdbms/dumpdb/dumpdb/incident/incdir_17/dumpdb_pmon_2414_i17.trc
Process startup failed, error stack:
Errors in file /oracle/diag/rdbms/dumpdb/dumpdb/trace/dumpdb_psp0_2430.trc:
ORA-27300: OS system dependent operation:fork failed with status: 12
ORA-27301: OS failure message: Not enough space
ORA-27302: failure occurred at: skgpspawn3
The important message to notice is ORA-27301: OS failure message: Not enough space. This indicates the SWAP problem.
I checked metalink for this problem. I found a Note 470067.1, which was talking about the similar problem & the solution was to increase the swap space. But I just wanted to check, if it can be handled without increasing the swap space.
I tweaked two parameters memory_max_target=1G & shminfo_shmmax=1073741824. This change allowed me to open both the databases.
Remember SHMMAX is the maximum size of a single shared memory segment. SHMMAX allocation was doing the paging, and RAM & SWAP were not capable of handling this.
Now with revised parameters, Oracle paging can be handled by existing resources.