On Tue, Apr 23, 2013 at 9:09 PM, Philip Martin <philip.mar...@wandisco.com> wrote: > Ivan Zhakov <i...@visualsvn.com> writes: > >> On Tue, Apr 23, 2013 at 5:04 PM, <phi...@apache.org> wrote: >>> Author: philip >>> Date: Tue Apr 23 13:04:42 2013 >>> New Revision: 1470936 >>> >>> URL: http://svn.apache.org/r1470936 >>> Log: >>> Significantly reduce the number of open files during a typical update >>> over ra_serf by using lazy-opening streams to delay opening until the >>> HTTP response is received. This changes the new-in-1.8 lazy-open API. >>> >> [...] >> >>> --- subversion/trunk/subversion/libsvn_wc/update_editor.c (original) >>> +++ subversion/trunk/subversion/libsvn_wc/update_editor.c Tue Apr 23 >>> 13:04:42 2013 >>> @@ -3525,6 +3525,46 @@ open_file(const char *path, >>> return SVN_NO_ERROR; >>> } >>> >> [...] >> >>> + tb = apr_palloc(handler_pool, sizeof(struct lazy_target_baton)); >> apr_pcalloc() ? Otherwise it looks like needless premature optimization. > > I prefer alloc with explicit initialisation unless we know zero is a > correct initialisation. Adding initialisation to zero makes it harder > to use a tool like valgrind to identify missing initialisations. > Ok, but on the other apr_pcalloc() makes code execution stable and code will crash if not initialized properly instead of accessing garbage.
-- Ivan Zhakov CTO | VisualSVN | http://www.visualsvn.com
