9p.txt 6.2 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173
  1. v9fs: Plan 9 Resource Sharing for Linux
  2. =======================================
  3. ABOUT
  4. =====
  5. v9fs is a Unix implementation of the Plan 9 9p remote filesystem protocol.
  6. This software was originally developed by Ron Minnich <rminnich@sandia.gov>
  7. and Maya Gokhale. Additional development by Greg Watson
  8. <gwatson@lanl.gov> and most recently Eric Van Hensbergen
  9. <ericvh@gmail.com>, Latchesar Ionkov <lucho@ionkov.net> and Russ Cox
  10. <rsc@swtch.com>.
  11. The best detailed explanation of the Linux implementation and applications of
  12. the 9p client is available in the form of a USENIX paper:
  13. http://www.usenix.org/events/usenix05/tech/freenix/hensbergen.html
  14. Other applications are described in the following papers:
  15. * XCPU & Clustering
  16. http://xcpu.org/papers/xcpu-talk.pdf
  17. * KVMFS: control file system for KVM
  18. http://xcpu.org/papers/kvmfs.pdf
  19. * CellFS: A New Programming Model for the Cell BE
  20. http://xcpu.org/papers/cellfs-talk.pdf
  21. * PROSE I/O: Using 9p to enable Application Partitions
  22. http://plan9.escet.urjc.es/iwp9/cready/PROSE_iwp9_2006.pdf
  23. USAGE
  24. =====
  25. For remote file server:
  26. mount -t 9p 10.10.1.2 /mnt/9
  27. For Plan 9 From User Space applications (http://swtch.com/plan9)
  28. mount -t 9p `namespace`/acme /mnt/9 -o trans=unix,uname=$USER
  29. For server running on QEMU host with virtio transport:
  30. mount -t 9p -o trans=virtio <mount_tag> /mnt/9
  31. where mount_tag is the tag associated by the server to each of the exported
  32. mount points. Each 9P export is seen by the client as a virtio device with an
  33. associated "mount_tag" property. Available mount tags can be
  34. seen by reading /sys/bus/virtio/drivers/9pnet_virtio/virtio<n>/mount_tag files.
  35. OPTIONS
  36. =======
  37. trans=name select an alternative transport. Valid options are
  38. currently:
  39. unix - specifying a named pipe mount point
  40. tcp - specifying a normal TCP/IP connection
  41. fd - used passed file descriptors for connection
  42. (see rfdno and wfdno)
  43. virtio - connect to the next virtio channel available
  44. (from QEMU with trans_virtio module)
  45. rdma - connect to a specified RDMA channel
  46. uname=name user name to attempt mount as on the remote server. The
  47. server may override or ignore this value. Certain user
  48. names may require authentication.
  49. aname=name aname specifies the file tree to access when the server is
  50. offering several exported file systems.
  51. cache=mode specifies a caching policy. By default, no caches are used.
  52. loose = no attempts are made at consistency,
  53. intended for exclusive, read-only mounts
  54. fscache = use FS-Cache for a persistent, read-only
  55. cache backend.
  56. debug=n specifies debug level. The debug level is a bitmask.
  57. 0x01 = display verbose error messages
  58. 0x02 = developer debug (DEBUG_CURRENT)
  59. 0x04 = display 9p trace
  60. 0x08 = display VFS trace
  61. 0x10 = display Marshalling debug
  62. 0x20 = display RPC debug
  63. 0x40 = display transport debug
  64. 0x80 = display allocation debug
  65. 0x100 = display protocol message debug
  66. 0x200 = display Fid debug
  67. 0x400 = display packet debug
  68. 0x800 = display fscache tracing debug
  69. rfdno=n the file descriptor for reading with trans=fd
  70. wfdno=n the file descriptor for writing with trans=fd
  71. maxdata=n the number of bytes to use for 9p packet payload (msize)
  72. port=n port to connect to on the remote server
  73. noextend force legacy mode (no 9p2000.u or 9p2000.L semantics)
  74. version=name Select 9P protocol version. Valid options are:
  75. 9p2000 - Legacy mode (same as noextend)
  76. 9p2000.u - Use 9P2000.u protocol
  77. 9p2000.L - Use 9P2000.L protocol
  78. dfltuid attempt to mount as a particular uid
  79. dfltgid attempt to mount with a particular gid
  80. afid security channel - used by Plan 9 authentication protocols
  81. nodevmap do not map special files - represent them as normal files.
  82. This can be used to share devices/named pipes/sockets between
  83. hosts. This functionality will be expanded in later versions.
  84. access there are three access modes.
  85. user = if a user tries to access a file on v9fs
  86. filesystem for the first time, v9fs sends an
  87. attach command (Tattach) for that user.
  88. This is the default mode.
  89. <uid> = allows only user with uid=<uid> to access
  90. the files on the mounted filesystem
  91. any = v9fs does single attach and performs all
  92. operations as one user
  93. cachetag cache tag to use the specified persistent cache.
  94. cache tags for existing cache sessions can be listed at
  95. /sys/fs/9p/caches. (applies only to cache=fscache)
  96. RESOURCES
  97. =========
  98. Our current recommendation is to use Inferno (http://www.vitanuova.com/nferno/index.html)
  99. as the 9p server. You can start a 9p server under Inferno by issuing the
  100. following command:
  101. ; styxlisten -A tcp!*!564 export '#U*'
  102. The -A specifies an unauthenticated export. The 564 is the port # (you may
  103. have to choose a higher port number if running as a normal user). The '#U*'
  104. specifies exporting the root of the Linux name space. You may specify a
  105. subset of the namespace by extending the path: '#U*'/tmp would just export
  106. /tmp. For more information, see the Inferno manual pages covering styxlisten
  107. and export.
  108. A Linux version of the 9p server is now maintained under the npfs project
  109. on sourceforge (http://sourceforge.net/projects/npfs). The currently
  110. maintained version is the single-threaded version of the server (named spfs)
  111. available from the same SVN repository.
  112. There are user and developer mailing lists available through the v9fs project
  113. on sourceforge (http://sourceforge.net/projects/v9fs).
  114. A stand-alone version of the module (which should build for any 2.6 kernel)
  115. is available via (http://github.com/ericvh/9p-sac/tree/master)
  116. News and other information is maintained on SWiK (http://swik.net/v9fs)
  117. and the Wiki (http://sf.net/apps/mediawiki/v9fs/index.php).
  118. Bug reports may be issued through the kernel.org bugzilla
  119. (http://bugzilla.kernel.org)
  120. For more information on the Plan 9 Operating System check out
  121. http://plan9.bell-labs.com/plan9
  122. For information on Plan 9 from User Space (Plan 9 applications and libraries
  123. ported to Linux/BSD/OSX/etc) check out http://swtch.com/plan9
  124. STATUS
  125. ======
  126. The 2.6 kernel support is working on PPC and x86.
  127. PLEASE USE THE KERNEL BUGZILLA TO REPORT PROBLEMS. (http://bugzilla.kernel.org)