TEXT   21

unshare system call

Guest on 1st July 2022 07:23:02 PM

  1.  
  2. unshare system call:
  3. --------------------
  4. This document describes the new system call, unshare. The document
  5. provides an overview of the feature, why it is needed, how it can
  6. be used, its interface specification, design, implementation and
  7. how it can be tested.
  8.  
  9. Change Log:
  10. -----------
  11. version 0.1  Initial document, Janak Desai .
  12.  
  13. Contents:
  14. ---------
  15.         1) Overview
  16.         2) Benefits
  17.         3) Cost
  18.         4) Requirements
  19.         5) Functional Specification
  20.         6) High Level Design
  21.         7) Low Level Design
  22.         8) Test Specification
  23.         9) Future Work
  24.  
  25. 1) Overview
  26. -----------
  27. Most legacy operating system kernels support an abstraction of threads
  28. as multiple execution contexts within a process. These kernels provide
  29. special resources and mechanisms to maintain these "threads". The Linux
  30. kernel, in a clever and simple manner, does not make distinction
  31. between processes and "threads". The kernel allows processes to share
  32. resources and thus they can achieve legacy "threads" behavior without
  33. requiring additional data structures and mechanisms in the kernel. The
  34. power of implementing threads in this manner comes not only from
  35. its simplicity but also from allowing application programmers to work
  36. outside the confinement of all-or-nothing shared resources of legacy
  37. threads. On Linux, at the time of thread creation using the clone system
  38. call, applications can selectively choose which resources to share
  39. between threads.
  40.  
  41. unshare system call adds a primitive to the Linux thread model that
  42. allows threads to selectively 'unshare' any resources that were being
  43. shared at the time of their creation. unshare was conceptualized by
  44. Al Viro in the August of 2000, on the Linux-Kernel mailing list, as part
  45. of the discussion on POSIX threads on Linux.  unshare augments the
  46. usefulness of Linux threads for applications that would like to control
  47. shared resources without creating a new process. unshare is a natural
  48. addition to the set of available primitives on Linux that implement
  49. the concept of process/thread as a virtual machine.
  50.  
  51. 2) Benefits
  52. -----------
  53. unshare would be useful to large application frameworks such as PAM
  54. where creating a new process to control sharing/unsharing of process
  55. resources is not possible. Since namespaces are shared by default
  56. when creating a new process using fork or clone, unshare can benefit
  57. even non-threaded applications if they have a need to disassociate
  58. from default shared namespace. The following lists two use-cases
  59. where unshare can be used.
  60.  
  61. 2.1 Per-security context namespaces
  62. -----------------------------------
  63. unshare can be used to implement polyinstantiated directories using
  64. the kernel's per-process namespace mechanism. Polyinstantiated directories,
  65. such as per-user and/or per-security context instance of /tmp, /var/tmp or
  66. per-security context instance of a user's home directory, isolate user
  67. processes when working with these directories. Using unshare, a PAM
  68. module can easily setup a private namespace for a user at login.
  69. Polyinstantiated directories are required for Common Criteria certification
  70. with Labeled System Protection Profile, however, with the availability
  71. of shared-tree feature in the Linux kernel, even regular Linux systems
  72. can benefit from setting up private namespaces at login and
  73. polyinstantiating /tmp, /var/tmp and other directories deemed
  74. appropriate by system administrators.
  75.  
  76. 2.2 unsharing of virtual memory and/or open files
  77. -------------------------------------------------
  78. Consider a client/server application where the server is processing
  79. client requests by creating processes that share resources such as
  80. virtual memory and open files. Without unshare, the server has to
  81. decide what needs to be shared at the time of creating the process
  82. which services the request. unshare allows the server an ability to
  83. disassociate parts of the context during the servicing of the
  84. request. For large and complex middleware application frameworks, this
  85. ability to unshare after the process was created can be very
  86. useful.
  87.  
  88. 3) Cost
  89. -------
  90. In order to not duplicate code and to handle the fact that unshare
  91. works on an active task (as opposed to clone/fork working on a newly
  92. allocated inactive task) unshare had to make minor reorganizational
  93. changes to copy_* functions utilized by clone/fork system call.
  94. There is a cost associated with altering existing, well tested and
  95. stable code to implement a new feature that may not get exercised
  96. extensively in the beginning. However, with proper design and code
  97. review of the changes and creation of an unshare test for the LTP
  98. the benefits of this new feature can exceed its cost.
  99.  
  100. 4) Requirements
  101. ---------------
  102. unshare reverses sharing that was done using clone(2) system call,
  103. so unshare should have a similar interface as clone(2). That is,
  104. since flags in clone(int flags, void *stack) specifies what should
  105. be shared, similar flags in unshare(int flags) should specify
  106. what should be unshared. Unfortunately, this may appear to invert
  107. the meaning of the flags from the way they are used in clone(2).
  108. However, there was no easy solution that was less confusing and that
  109. allowed incremental context unsharing in future without an ABI change.
  110.  
  111. unshare interface should accommodate possible future addition of
  112. new context flags without requiring a rebuild of old applications.
  113. If and when new context flags are added, unshare design should allow
  114. incremental unsharing of those resources on an as needed basis.
  115.  
  116. 5) Functional Specification
  117. ---------------------------
  118. NAME
  119.         unshare - disassociate parts of the process execution context
  120.  
  121. SYNOPSIS
  122.         #include <sched.h>
  123.  
  124.         int unshare(int flags);
  125.  
  126. DESCRIPTION
  127.         unshare allows a process to disassociate parts of its execution
  128.         context that are currently being shared with other processes. Part
  129.         of execution context, such as the namespace, is shared by default
  130.         when a new process is created using fork(2), while other parts,
  131.         such as the virtual memory, open file descriptors, etc, may be
  132.         shared by explicit request to share them when creating a process
  133.         using clone(2).
  134.  
  135.         The main use of unshare is to allow a process to control its
  136.         shared execution context without creating a new process.
  137.  
  138.         The flags argument specifies one or bitwise-or'ed of several of
  139.         the following constants.
  140.  
  141.         CLONE_FS
  142.                 If CLONE_FS is set, file system information of the caller
  143.                 is disassociated from the shared file system information.
  144.  
  145.         CLONE_FILES
  146.                 If CLONE_FILES is set, the file descriptor table of the
  147.                 caller is disassociated from the shared file descriptor
  148.                 table.
  149.  
  150.         CLONE_NEWNS
  151.                 If CLONE_NEWNS is set, the namespace of the caller is
  152.                 disassociated from the shared namespace.
  153.  
  154.         CLONE_VM
  155.                 If CLONE_VM is set, the virtual memory of the caller is
  156.                 disassociated from the shared virtual memory.
  157.  
  158. RETURN VALUE
  159.         On success, zero returned. On failure, -1 is returned and errno is
  160.  
  161. ERRORS
  162.         EPERM   CLONE_NEWNS was specified by a non-root process (process
  163.                 without CAP_SYS_ADMIN).
  164.  
  165.         ENOMEM  Cannot allocate sufficient memory to copy parts of caller's
  166.                 context that need to be unshared.
  167.  
  168.         EINVAL  Invalid flag was specified as an argument.
  169.  
  170. CONFORMING TO
  171.         The unshare() call is Linux-specific and  should  not be used
  172.         in programs intended to be portable.
  173.  
  174. SEE ALSO
  175.         clone(2), fork(2)
  176.  
  177. 6) High Level Design
  178. --------------------
  179. Depending on the flags argument, the unshare system call allocates
  180. appropriate process context structures, populates it with values from
  181. the current shared version, associates newly duplicated structures
  182. with the current task structure and releases corresponding shared
  183. versions. Helper functions of clone (copy_*) could not be used
  184. directly by unshare because of the following two reasons.
  185.   1) clone operates on a newly allocated not-yet-active task
  186.      structure, where as unshare operates on the current active
  187.      task. Therefore unshare has to take appropriate task_lock()
  188.      before associating newly duplicated context structures
  189.   2) unshare has to allocate and duplicate all context structures
  190.      that are being unshared, before associating them with the
  191.      current task and releasing older shared structures. Failure
  192.      do so will create race conditions and/or oops when trying
  193.      to backout due to an error. Consider the case of unsharing
  194.      both virtual memory and namespace. After successfully unsharing
  195.      vm, if the system call encounters an error while allocating
  196.      new namespace structure, the error return code will have to
  197.      reverse the unsharing of vm. As part of the reversal the
  198.      system call will have to go back to older, shared, vm
  199.      structure, which may not exist anymore.
  200.  
  201. Therefore code from copy_* functions that allocated and duplicated
  202. current context structure was moved into new dup_* functions. Now,
  203. copy_* functions call dup_* functions to allocate and duplicate
  204. appropriate context structures and then associate them with the
  205. task structure that is being constructed. unshare system call on
  206. the other hand performs the following:
  207.   1) Check flags to force missing, but implied, flags
  208.   2) For each context structure, call the corresponding unshare
  209.      helper function to allocate and duplicate a new context
  210.      structure, if the appropriate bit is set in the flags argument.
  211.   3) If there is no error in allocation and duplication and there
  212.      are new context structures then lock the current task structure,
  213.      associate new context structures with the current task structure,
  214.      and release the lock on the current task structure.
  215.   4) Appropriately release older, shared, context structures.
  216.  
  217. 7) Low Level Design
  218. -------------------
  219. Implementation of unshare can be grouped in the following 4 different
  220. items:
  221.   a) Reorganization of existing copy_* functions
  222.   b) unshare system call service function
  223.   c) unshare helper functions for each different process context
  224.   d) Registration of system call number for different architectures
  225.  
  226.   7.1) Reorganization of copy_* functions
  227.        Each copy function such as copy_mm, copy_namespace, copy_files,
  228.        etc, had roughly two components. The first component allocated
  229.        and duplicated the appropriate structure and the second component
  230.        linked it to the task structure passed in as an argument to the copy
  231.        function. The first component was split into its own function.
  232.        These dup_* functions allocated and duplicated the appropriate
  233.        context structure. The reorganized copy_* functions invoked
  234.        their corresponding dup_* functions and then linked the newly
  235.        duplicated structures to the task structure with which the
  236.        copy function was called.
  237.  
  238.   7.2) unshare system call service function
  239.        * Check flags
  240.          Force implied flags. If CLONE_THREAD is set force CLONE_VM.
  241.          If CLONE_VM is set, force CLONE_SIGHAND. If CLONE_SIGHAND is
  242.          set and signals are also being shared, force CLONE_THREAD. If
  243.          CLONE_NEWNS is set, force CLONE_FS.
  244.        * For each context flag, invoke the corresponding unshare_*
  245.          helper routine with flags passed into the system call and a
  246.          reference to pointer pointing the new unshared structure
  247.        * If any new structures are created by unshare_* helper
  248.          functions, take the task_lock() on the current task,
  249.          modify appropriate context pointers, and release the
  250.          task lock.
  251.        * For all newly unshared structures, release the corresponding
  252.          older, shared, structures.
  253.  
  254.   7.3) unshare_* helper functions
  255.        For unshare_* helpers corresponding to CLONE_SYSVSEM, CLONE_SIGHAND,
  256.        and CLONE_THREAD, return -EINVAL since they are not implemented yet.
  257.        For others, check the flag value to see if the unsharing is
  258.        required for that structure. If it is, invoke the corresponding
  259.        dup_* function to allocate and duplicate the structure and return
  260.        a pointer to it.
  261.  
  262.   7.4) Appropriately modify architecture specific code to register the
  263.        new system call.
  264.  
  265. 8) Test Specification
  266. ---------------------
  267. The test for unshare should test the following:
  268.   1) Valid flags: Test to check that clone flags for signal and
  269.         signal handlers, for which unsharing is not implemented
  270.         yet, return -EINVAL.
  271.   2) Missing/implied flags: Test to make sure that if unsharing
  272.         namespace without specifying unsharing of filesystem, correctly
  273.         unshares both namespace and filesystem information.
  274.   3) For each of the four (namespace, filesystem, files and vm)
  275.         supported unsharing, verify that the system call correctly
  276.         unshares the appropriate structure. Verify that unsharing
  277.         them individually as well as in combination with each
  278.         other works as expected.
  279.   4) Concurrent execution: Use shared memory segments and futex on
  280.         an address in the shm segment to synchronize execution of
  281.         about 10 threads. Have a couple of threads execute execve,
  282.         a couple _exit and the rest unshare with different combination
  283.         of flags. Verify that unsharing is performed as expected and
  284.         that there are no oops or hangs.
  285.  
  286. 9) Future Work
  287. --------------
  288. The current implementation of unshare does not allow unsharing of
  289. signals and signal handlers. Signals are complex to begin with and
  290. to unshare signals and/or signal handlers of a currently running
  291. process is even more complex. If in the future there is a specific
  292. need to allow unsharing of signals and/or signal handlers, it can
  293. be incrementally added to unshare without affecting legacy
  294. applications using unshare.

Raw Paste


Login or Register to edit or fork this paste. It's free.