This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Bootstrap broken for more than 10 days on cygwin and mingw


Ian Lance Taylor wrote:
> I had already looked at the
> patch and made the same comment that Michael did: to use the enum
> constants rather than the casts.  I probably could have said that more
> clearly.  My apologies.

> problem as soon as it is brought to my attention; in this case there
> was already a patch, and I already reviewed it, 

  I didn't understand your email as a formal review, partly since it was in a
different thread on a different list, and didn't clearly say "do this" but was
worded as if offering a general opinion, rather than a pre-condition of
approval, and also partly because I was more expecting the win32 maintainers
to end up approving it so I was waiting to hear from them too.  I guess I
could have been a little quicker to drive it forward than I was, given that
it's a bootstrap-breaker maybe it would have been appropriate to start pinging
it sooner than I would otherwise do for an ordinary patch?

  Anyhow, all fixed now and no serious harm done.  FX, I appreciate your
desire for stability, but I reckon Ian's right; it just isn't practical to
test everything everywhere, and I reckon it works best to get stuff committed
once it's reasonably-well-tested, and iron out the last few bugs on mainline
during stage 1, when a bit of instability is only to be expected, because it
keeps GCC moving forward.  Yes, that then necessitates a bit of volunteer
effort to look after the other targets, and they risk bitrotting if neglected
for a while, but I think we get from A to B faster this way even if we do take
the occasional mis-step.

    cheers,
      DaveK


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]