From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 16555 invoked by alias); 23 Oct 2007 17:39:03 -0000 Received: (qmail 16463 invoked by uid 48); 23 Oct 2007 17:38:46 -0000 Date: Tue, 23 Oct 2007 17:39:00 -0000 Message-ID: <20071023173846.16462.qmail@sourceware.org> X-Bugzilla-Reason: CC References: Subject: [Bug testsuite/28870] [4.2/4.3 Regression] configuring, over-riding timeout values in testsuite In-Reply-To: Reply-To: gcc-bugzilla@gcc.gnu.org To: gcc-bugs@gcc.gnu.org From: "janis at gcc dot gnu dot org" Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Id: List-Archive: List-Post: List-Help: Sender: gcc-bugs-owner@gcc.gnu.org X-SW-Source: 2007-10/txt/msg02085.txt.bz2 ------- Comment #15 from janis at gcc dot gnu dot org 2007-10-23 17:38 ------- I've been looking at this again and now have a better understanding of how timeouts are handled and how to set variables used within the testsuite. Timeouts for compilation can be controlled by passing --timeout=n in the list of options for target_compile. The GCC testsuite procedures that call it can pass in the value from a target board, or from a global variable set in ${HOME}/.dejagnurc, or a value for a specific test in a (not-yet-existing) dg-timeout directive. Timeouts for execution are controlled in the mudflap testsuite by overriding the DejaGnu proc standard_wait; other testsuites can do the same thing, and can look for new timeout values from a variety of sources. Before I go any further with this, is there interest in a dg-timeout to override timeouts for individual tests, with a target specifier? Should it apply to compilation times and test execution? Is there interest in a timeout value provided in ~/.dejagnurc or site.exp that applies to all tests, or perhaps to a particular testsuite like gcc_timeout or v3_timeout? Should timeout values be in seconds, or in terms of some multiplier also defined for a testsuite, or a target, or by the person running the tests? This is for tests expected to have problems; H-P's method is better for cranking down the amount of time spent for tests on simulators or slow hardware. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28870