COMPILING 3.5 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384
  1. General Information
  2. ~~~~~~~~~~~~~~~~~~~
  3. To compile this you need a couple things
  4. - A working POSIX system with working POSIX gcc, g++, make (GNU),
  5. ar, sh, awk and sed in the path
  6. - GNU Make 3.74 or so, -- normal UNIX make will NOT work
  7. * Note 3.77 is broken.
  8. - A working ANSI C++ compiler, this is not g++ 2.7.*
  9. g++ 2.8 works OK and newer egcs work well also. Nobody has tried it
  10. on other compilers :< You will need a properly working STL as well.
  11. - A C library with the usual POSIX functions and a BSD socket layer.
  12. If you OS conforms to the Single Unix Spec then you are fine:
  13. http://www.opengroup.org/onlinepubs/7908799/index.html
  14. ** NOTICE **
  15. The C++ global constructors do not link correctly when using non-shared
  16. libaries. This is probably the correct behavior of the linker, but I have
  17. not yet had time to devise a work around for it. The correct thing to
  18. do is add a reference to debSystem in apt-pkg/init.cc,
  19. assert(&debSystem == 0) would be fine for instance.
  20. Guidelines
  21. ~~~~~~~~~~
  22. I am not interested in making 'ultra portable code'. I will accept patches
  23. to make the code that already exists conform more to SUS or POSIX, but
  24. I don't really care if your not-SUS OS doesn't work. It is simply too
  25. much work to maintain patches for dysfunctional OSs. I highly suggest you
  26. contact your vendor and express intrest in a conforming C library.
  27. That said, there are lots of finniky problems that must be delt with even
  28. between the supported OS's. Primarily the path I choose to take is to put
  29. a shim header file in build/include that transparently adds the required
  30. functionality. Patches to make autoconf detect these cases and generate the
  31. required shims are OK.
  32. Current shims:
  33. * C99 integer types 'inttypes.h'
  34. * sys/statvfs.h to convert from BSD/old-glibc statfs to SUS statvfs
  35. * rfc2553 hostname resolution (methods/rfc*), shims to normal gethostbyname.
  36. The more adventerous could steal the KAME IPv6 enabled resolvers for those
  37. OS's with IPv6 support but no rfc2553 (why?)
  38. * define _XOPEN_EXTENDED_SOURCE to bring in h_errno on HP-UX
  39. * socklen_t shim in netdb.h if the OS does not have socklen_t
  40. The only completely non-shimed OS is Linux with glibc2.1, glibc2.0 requires
  41. the first three shims.
  42. Platform Notes
  43. ~~~~~~~~~~~~~~
  44. Debian GNU Linux 2.1 'slink'
  45. Debian GNU Linux 'potato'
  46. Debian GNU Linux 'woody'
  47. * All Archs
  48. - Works flawlessly
  49. - You will want to have debiandoc-sgml and docbook2man installed to get
  50. best results.
  51. - No IPv6 Support in glibc's < 2.1.
  52. Sun Solaris
  53. SunOS cab101 5.7 Generic_106541-04 sun4u sparc
  54. SunOS csu201 5.8 Generic_108528-04 sun4u sparc
  55. - Works fine
  56. - Note, no IPv6 Support, OS lacks RFC 2553 hostname resolution
  57. OpenBSD
  58. OpenBSD gsb086 2.5 CMPUT#0 i386 unknown
  59. OpenBSD csu101 2.7 CMPUT#1 i386 unknown
  60. - OS needs 'ranlib' to generate the symbol table after 'ar'.. (not using
  61. GNU ar with the gnu tool chain :<)
  62. - '2.5' does not have RFC 2553 hostname resolution, but '2.7' does
  63. - Testing on '2.7' suggests the OS has a bug in its handling of
  64. ftruncate on files that have been written via mmap. It fills the page
  65. that crosses the truncation boundary with 0's.
  66. HP-UX
  67. HP-UX nyquist B.10.20 C 9000/780 2016574337 32-user license
  68. - Evil OS, does not conform very well to SUS
  69. 1) snprintf exists but is not prototyped, ignore spurios warnings
  70. 2) No socklen_t
  71. 3) Requires -D_XOPEN_SOURCE_EXTENDED for h_errno
  72. configure should fix the last two (see above)
  73. - Note, no IPv6 Support, OS lacks RFC 2553 hostname resolution