Re: [Evolution-hackers] Deadlock in camel-gpg-context.c

[Jeffrey Stedfast <fejj ximian com>]

On Mon, 2003-04-21 at 13:55, Christophe Saout wrote:
> Am Mon, 2003-04-21 um 19.41 schrieb Jeffrey Stedfast:
> 
> > > In camel/camel-gpg-context.c (evo 1.3 cvs) there is a problem:
> > > 
> > > In gpg_ctx_op_step the file descriptors stdout_fd, stderr_fd and
> > > status_fd are polled for input.
> > > 
> > > But stdout_fd is not read when there is no gpg->ostream to get the data
> > > (if (FD_ISSET (gpg->stdout_fd, &rdset) && gpg->ostream)).
> > > 
> > > So the end of file state cannot be detected and thus gpg->seen_eof1
> > > won't be set.
> > > 
> > > If that happens gpg_ctx_op_complete fails to detect when gpg has
> > > finished and the users of these functions run in an endless loop.
> > 
> > When can this condition ever happen?
> 
> When gpg_ctx_op_step is called with gpg->ostream being NULL.
> gpg_verify and gpg_import_keys don't set them (it actually happened to
> me when trying to verify a signature).

a few things:

1) gpg_import_keys() is not yet in use, so this isn't a real concern.

2) seen_eof1 is used differently in the case of gpg_verify(). Instead of
being used as a eof marker, it is used to denote that we have gotten a
trust metric for the data being verified. ie, we parsed a
TRUST_[NEVER,MARGINAL,FULLY,ULTIMATE] status line from gpg's --status-fd

apparently you have discovered a bug in gpg, although the
camel-gpg-context code is a bit dodgy in this case as well.

Jeff

-- 
Jeffrey Stedfast
Evolution Hacker - Ximian, Inc.
fejj ximian com  - www.ximian.com