[PATCH] Fix type-mismatch in value profiling (PR22525)

Roger Sayle roger@eyesopen.com
Tue Apr 25 15:59:00 GMT 2006

Hi Richard,

On Tue, 25 Apr 2006, Richard Guenther wrote:
> > This is OK for mainline.  Thanks.
> I noticed that val-prof-2.c is now failing due to the different
> signedness in the final transformation.  Fixed with the following
> patch as obvious.

There seem to be differing opinions on the precise meaning of the
statement "Bootstrapped and regtested on x86_64-unknown-linux-gnu."

I think this issue needs to be resolved before I can approve any
further patches from contributors where it doesn't strictly mean
that the changes were fully bootstrapped in an otherwise clean-tree
from the top-level, including all default languages and run-time
libraries and regression tested with a top-level "make -k check"
including the new test cases, with no new failures in any language
or runtime testsuite.

It's not for the love of increasing internet bandwidth that I
precisely state how I've tested my own patch submissions when
posting to gcc-patches!

Of course, people sometimes make mistakes.  However your three
mistakes in the past 24 hours indicate either a particularly
unreliable rate of human error, or that you don't perform the
testing implied by your postings.  But again I could be over
interpreting "bootstrapped and regression tested".

Let's work out how best to restore the trust that's been lost.


More information about the Gcc-patches mailing list