- add TLS (thread local storage) support - DONE. 
- fix section headers - DONE. RH9 very sensitive to correct section headers.
  add section with strings to the statified exe  - DONE 
				(added all non-allocated section)
- rework tls_test - I'll need to specify interpreter from real exe,
  instead of hardcoded one.
	tls_test and set_thread_area_addr where dropped altogether in favor
        of my_gdb
- do not use objdump for finding _dl variables - it's not work
  for striped /lib/ld-linux (for example some SuSE versions) - DONE
- add support for RTLD environment variables. - DONE
- support linux on the platforms other than x86, x86_64 and alpha
- support other unixes.
- minimize usage of external programs: readelf, etc.
  ( awk, sh, sed looks ok,  aren't they ?)
- think about gdb. From one side it's provide an easy
  and (I hope !) portable enough way for dump memory mappings
  and registers. From another side - it's one more dependencies.
	I finally get rid of gdb. DONE
- write de-statifier, which will re-create original exe file
- write more tech documentation. - DONE
- what about output file license

- handle case when there is no native room for starter.
- adjust AT_PHDR value
- support -PIE executables
- update documentation about memory layout. - DONE

- now statified executable got (if any)  PT_GNU_STACK segment from original one.
  usually it's ok, but if original exe has PT_GNU_STACK segment with permission
  rw AND (at least) one of the executables DLL or preloaded DLL have no 
  this segment or it has rwx permission (and really need executable stack)
  It'll be a trouble: stack will be set to rw (as required by executable's
  PT_GNU_STACK) and during library initialization / functions execution
  should occure SEGFAULT. For now let's wait and see is it a real scenario.
