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

Re: svn commit: r1532572 - in /subversion/trunk/subversion: include/private/svn_subr_private.h libsvn_repos/config_pool.c libsvn_subr/config.c libsvn_subr/config_file.c libsvn_subr/config_impl.h

From: Stefan Fuhrmann <stefan.fuhrmann_at_wandisco.com>
Date: Sun, 20 Oct 2013 23:26:46 +0200

On Fri, Oct 18, 2013 at 3:36 PM, Branko Čibej <brane_at_wandisco.com> wrote:

> On 16.10.2013 00:32, stefan2_at_apache.org wrote:
>
> Author: stefan2
> Date: Tue Oct 15 22:32:44 2013
> New Revision: 1532572
>
> URL: http://svn.apache.org/r1532572
> Log:
> Add support for read-only access to svn_config_t. In read-only mode,
> concurrent multi-threaded access to the same config data structure is
> safe.
>
>
> + /* Ignore write attempts to r/o configurations.
> + *
> + * Since we should never try to modify r/o data, trigger an assertion
> + * in debug mode.
> + */
> + assert(!cfg->read_only);
> + if (cfg->read_only)
> + return;
> +
> remove_expansions(cfg);
>
>
> Please don't use assert like this. You're assuming that what people like
> to call "release" builds are always compiled with -DNDEBUG. I've always
> found that assumption to be naïve at best.
>
> Instead, make the code depend on whether we're in maintainer mode or not;
> the result will be much less ambiguous.
>

From what I can see, there is no macro to test for
(SVN_DEBUG being more or less equivalent to DEBUG).

Should we introduce SVN_MAINTAINER_MODE?

-- Stefan^2.
Received on 2013-10-20 23:27:19 CEST

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