On Fri, 2010-04-23, Stefan Sperling wrote:
> On Fri, Apr 23, 2010 at 01:01:24PM -0000, stsp_at_apache.org wrote:
> > Author: stsp
> > Date: Fri Apr 23 13:01:23 2010
> > New Revision: 937275
> > URL: http://svn.apache.org/viewvc?rev=937275&view=rev
> > Log:
> > Pool usage fixes in the translation stream.
> > Do not maintain a private pool for each translation stream,
> > because the user cannot control unbound growth of this "secret" pool.
> > Instead, rely on callers to provide pools with sufficient lifetime
> > when the stream is created. "make check" agrees.
> > Suggested by: gstein
> > * subversion/include/svn_subst.h
> > (svn_subst_stream_translated): Rename POOL argument to RESULT_POOL.
> > Document pool lifetime requirements for EOL_STR and KEYWORDS parameters.
> It is unclear whether changing pool lifetime expectations of this API
> is acceptable. This problem was known before commit. Greg wanted to
> comment on this post-commit.
> > --- subversion/trunk/subversion/include/svn_subst.h (original)
> > +++ subversion/trunk/subversion/include/svn_subst.h Fri Apr 23 13:01:23 2010
> > @@ -301,7 +301,9 @@ svn_subst_translate_stream(svn_stream_t
> > * if @a repair is @c TRUE, convert any line ending to @a eol_str.
> > * Recognized line endings are: "\n", "\r", and "\r\n".
> > *
> > - * The stream returned is allocated in @a pool.
> > + * The stream returned is allocated in @a result_pool.
> > + * @a eol_str and @a keywords are expected to be allocated in a pool
> > + * with sufficient lifetime for use by the stream.
It seems to me that adding this requirement now is unsafe, because our
API users would not expect this. I think it would be best to revert
that part of the change, restoring the code that copies EOL_STR and
KEYWORDS into what is now RESULT_POOL.
Received on 2010-04-26 12:55:01 CEST