[email protected]
[Top] [All Lists]

Re: Fix for PR/39548

Subject: Re: Fix for PR/39548
From: Jakub Jelinek
Date: Wed, 25 Mar 2009 08:55:38 +0100
On Tue, Mar 24, 2009 at 09:57:13PM -0700, Xinliang David Li wrote:
> This is the new version of the change -- the check is moved out of the
> loop. I did not add the assertion in set_copy_of_val as this ended up
> with lots of failures in regression test. May need investitagion
> separately (real problem or interface design issue).
> compiler bootstrapped on linux/i686. Regression test is redone.

The testcase size is way over the usual testcase sizes.
Either distill a smaller testcase from it (use delta, or manually trim it
down), or don't include it at all.


<Prev in Thread] Current Thread [Next in Thread>