[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: permissions (and other) problems

From: Branko Čibej <brane_at_xbc.nu>
Date: 2002-09-14 02:36:47 CEST

Garrett Rooney wrote:

> there are actually two states we need to worry about. the first is
> 'have we been canceled', which tells SVN_ERR and SVN_ERR_W to return
> an SVN_ERR_CANCELED error, and the second is 'are we in the process of
> backing out because we received a cancelation', which causes
> svn_error_clear_all to return an SVN_ERR_CANCELED error.

You only need the one state. Here's the pseudocode:

    # global state
    apr_atomic_t svn_cancelled = false
    apr_mutex_t svn_cancellation_mutex
    apr_pool_t *svn_cancellation_pool = null

    # public interfaces
    def svn_async_cancel(pool):
      svn_cancellation_mutex.lock()
      apr_atoomic_set(svn_canceled, true)
      if svn_cancellation_pool is not null:
        svn_cancellation_pool = pool;
        pool.register_cleanup(lamba x: *x = null, &svn_cancellation_pool)
      svn_cancellation_mutex.unlock()

    def svn_async_cancel_clear():
      svn_cancellation_mutex.lock()
      apr_atoomic_set(svn_canceled, false)
      svn_cancellation_pool = null
      svn_cancellation_mutex.unlock()

    def SVN_ERR(err):
      if apr_atomic_get(svn_canceled):
        return svn_error_cancel(err)
      if err is not null:
        return err
    
    def SVN_ERR_W(err, msg):
      if apr_atomic_get(svn_canceled):
        return svn_error_cancel(err)
      if err is not null:
        return svn_error_quick_wrap(err, msg)

    def svn_error_clear_all(err):
      if apr_atomic_get(svn_canceled):
        return svn_error_cancel(err)
      # clean up errors, etc.

    def svn_error_cancel(err):
      if err is not null:
        # Just propagate cancellation errors
        if err->svn_err is SVN_ERR_CANCELLED:
          return err
        return svn_error_create(SVN_ERR_CANCELLED, err->pool)
      svn_cancellation_mutex.lock()
        if svn_cancellation_pool is null:
          # Oops, a race condition? help! WE have to document that when
          # you call svn_async_cancel, you should _not_ call svn_async_clear
          # until you're absolutely sure your working threads have processed
          # the cancellation. You can't cancel a cancel!
          abort()
        err = svn_error_create(SVN_ERR_CANCELLED, svn_cancellation_pool)
      svn_cancellation_mutex.unlock()
      return err

Regarding the implementation of svn_err_cancel and error locations: Look
at how the other error creation routines are implemented, and note that
svn_error.c undefines the macro wrappers first. With this
implementation, the SVN_ERR_CANCELLED errors will have correct location
info (i.e., they'll come from the SVN_ERR macros). It would make sense
to create a location wrapper for svn_error_clear_all, too; and note that
you should pass the incoming error to svn_error_create_all, so that we
don't get duplicate SVN_ERR_CANCELLED errors.

-- 
Brane Čibej   <brane_at_xbc.nu>   http://www.xbc.nu/brane/
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Sat Sep 14 02:37:30 2002

This is an archived mail posted to the Subversion Dev mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.