Re: branch bootstrap breakage in new DECL_RTL

["Joseph S. Myers" <jsm28 at cam dot ac dot uk>]

On Fri, 16 Mar 2001, Richard Kenner wrote:

> Over the last two weeks we've been talking about a number of patches that
> broke either the release branch or mainline.  But each has been due to a
> *different* reason.  I had an ia32 patch that was fine, but got misapplied to
> the branch.  Testing the patch as it was meant to be applied (the usual
> procedure) didn't find this, though testing on commit would have.  RTH's
[...]

Some concrete questions about any proposed pre-commit testing:

* What is the security model?  Who would be able to submit patches to the
pre-commit tester?  How would they be authenticated?  Would contributors
have shell access to the test machines to debug problems showing up on
other platforms?

* Suppose a contributor develops a patch, runs it through a
multiple-platform tester, and submits it for review.  It is reviewed two
hours, two days, two weeks, two months or two years later.  In which cases
would it need to be run through the tester again?  Would the answer differ
if some manual adjustments for patch conflicts were required?

* Would it be possible to specify a particular CVS version against which
an automated tester should test a submitted patch (for when the HEAD has
been broken by unforseen interactions of two patches)?

-- 
Joseph S. Myers
jsm28@cam.ac.uk