Re: PATCH to rs6000.md, comments?

[Franz Sirl <Franz dot Sirl-kernel at lauterbach dot com>]

Am Thu, 16 Jul 1998 schrieb Jeffrey A Law:
>In message <98071601501700.28356@ns1102.munich.netsurf.de>you write:
>  > >I would recommend against folks spending time on this right now, our
>  > >concentration should be on stabilizing for egcs-1.1, not adding new
>  > >optimizations.
>  > 
>  > Uah, I don't agree!! First I don't think the CONSTANT_P_RTX is
>  > an optimization,
>The new code is deferring the evaluation to a later point in the compiler;
>the warning occurs because the backend does not support deferred evaluation.
>
>The deferred evaluation is an optimization.  Thus it is not critical
>for egcs-1.1.  If it was necessary for correct behavior I would have
>objected to it going in without someone fixing all the backends at
>the same time.

Ah, ok.

>  > I thought it's simply different __builtin_constant handling. If this is not
>  > fixed before 1.1, users will be flooded with the resulting warnings, which
>  > additionally invalidate the whole testsuite! Something _has_ to be done about
>  > this before 1.1!
>It is likely the warning will be removed from the egcs-1.1 branch
>much like the -W -Wall stuff in the Makefile.  There's no need to
>panic the user community over a missed optimization.

You know, I (and others) mentioned this issue a few times on the list with no
response, and then you came around like "this won't be fixed". Probably I had
not enough information to panic, but I tried to get this information ;-)

I will disable the warning to get reasonable testsuite results.


>  > BTW, I see new testcases are only added to the mainline now. Shouldn't the
>  > testcase deadline moved nearer to the 1.1 release date? Especially for c++
>  > testcases...
>In general, once the branch is made, we don't want to be adding new
>files to the release, including testcases.
>
>I'm not going to strongly object to someone adding new tests to the
>branch right now, but the default behavior is to add them to the
>mainline tree only.  This is mostly a time/maintenance issue.  It is
>bad enough just trying to manage patches that have to be installed
>into multiple locations.
>
>I'll probably let things loosely go in "stabilize and bugfix mode" for 
>about another week while I try to get the official testing criteria
>ready.   Once the official testing criteria are available, then I'll
>be approving all patches for the release branch.
>
>
>I do not want the egcs-1.1 to drag on like egcs-1.0 did.  Thus the
>need to say "enough is enough" at some point :-)

Agreed. Actually there's now only one bug specific to
powerpc-unknown-linux-gnu I would like to get resolved before 1.1,
gcc.dg/980526-1.c. I'm _very_ pleased with the current status.

Franz.