Re: [PATCH, tree-dce]: DCE enh to handle dead calls (not pure)

["Richard Guenther" <richard dot guenther at gmail dot com>]

On Sun, May 18, 2008 at 7:46 PM, Xinliang David Li <davidxl@google.com> wrote:
> Steven/Richard:
>
> Wow, what a big mess I have created :( SORRY about that.
>
> Steven, I now understand where you anger came from. You probably felt
> that I checked in the code anyway after you said 'avoiding checking ..
> with small issues' -- there was unfortunate timing -- the change was
> approved by Diego (maybe I forgot to mention it in the first patch
> email?), and I checked in the patch right after I sent out the
> gcc-patch email before your comments.
>
> I planned to address issues addressed by you and Richard (the only
> functional one is the optimization size). The first step planned is
> the one which I thought does not need approval:
>
> " Free for all
>
> The following changes can be made by everyone with SVN write access:
>
> Fixes for obvious typos in ChangeLog files, docs, web pages, comments
> and similar stuff. Just check in the fix and copy it to gcc-patches.
> We don't want to get overly anal-retentive about checkin policies.
>
> Similarly, no outside approval is needed to revert a patch that you
> checked in."
>
> My understanding for this paragraph may be wrong, but I got the
> impression that  no approval is needed.
>
> This is a certainly a very unusual learning experience for me (a
> little shocked actually), especially on the emphasis of coding style
> (and code quality is judged on that?). This (coding style) used to be
> the least concern of me (I emphasize more on assertion, defensive
> programming etc) -- ok when it is similar enough (Diego actually
> helped me find lots of violations).  Now I realize this is inherited
> deeply and is part of the gcc culture (probably one of the reasons
> that makes gcc great and last so long), and I need to adapt myself to
> that -- even though I may continue to make silly coding style mistakes
> in the future.

Note that coding-style violations make it hard to review patches.  And to be
honest I didn't look at the implementation at all as there were so many
issues with the form of the submission.  Apart from the general comments
I made about the implementation and how it should be more generic to
make other parts of gcc benefit.

> I will continue to make changes to make sure the previous changes
> matches the coding standard, but if you do insist I revert it, I can
> do that too.

Yes, please revert both commits and resubmit the original patch with the
requested changes.

Richard.