Please fix it right!

Please fix it right!

[Stefan Monnier]

Andreas, undo the following change *NOW* and replace it with something
constructive, which addresses the comment you remove.

revision 1.115
date: 2008-04-27 05:47:00 -0400;  author: schwab;  state: Exp;  lines: +1 -3;  commitid: J0UjWj30wxmGmJ0t;
Unbreak bootstrap.
----------------------------


        Stefan



Index: Makefile.in
===================================================================
RCS file: /sources/emacs/emacs/lisp/Makefile.in,v
retrieving revision 1.114
retrieving revision 1.115
diff -u -u -b -r1.114 -r1.115
--- Makefile.in	23 Apr 2008 18:01:46 -0000	1.114
+++ Makefile.in	27 Apr 2008 09:47:00 -0000	1.115
@@ -81,9 +81,7 @@
 
 emacs = EMACSLOADPATH=$(lisp) LC_ALL=C $(EMACS) $(EMACSOPT)
 # Prerequisites for running $(emacs)
-# This should not be necessary because in order to have ../src/emacs
-# we must have built subdirs.el already.
-# emacs-deps = $(lisp)/subdirs.el
+emacs-deps = $(lisp)/subdirs.el
 
 # Common command to find subdirectories

Re: Please fix it right!

[Andreas Schwab]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

> Andreas, undo the following change *NOW* and replace it with something
> constructive, which addresses the comment you remove.

Why?  The comment is just wrong.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux Products GmbH, Maxfeldstraße 5, 90409 Nürnberg, Germany
PGP key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: Please fix it right!

[Stefan Monnier]

>> Andreas, undo the following change *NOW* and replace it with something
>> constructive, which addresses the comment you remove.

> Why?  The comment is just wrong.

Because I believe the comment is right, so please explain.


        Stefan

Re: Please fix it right!

[Andreas Schwab]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

>>> Andreas, undo the following change *NOW* and replace it with something
>>> constructive, which addresses the comment you remove.
>
>> Why?  The comment is just wrong.
>
> Because I believe the comment is right, so please explain.

How can it be right when that dependency does not exist?

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux Products GmbH, Maxfeldstraße 5, 90409 Nürnberg, Germany
PGP key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: Please fix it right!

[Stefan Monnier]

>>>> Andreas, undo the following change *NOW* and replace it with something
>>>> constructive, which addresses the comment you remove.
>> 
>>> Why?  The comment is just wrong.
>> 
>> Because I believe the comment is right, so please explain.
> How can it be right when that dependency does not exist?

Doesn't matter, all you have to do is *explain*.
Let's see... gee... maybe you could show a scenario where src/emacs
ends up existing while lisp/subdirs.el doesn't.

I don't claim your problem doesn't exist, but I think your solution is
wrong.  So explain your problem and maybe you'll convince me or someone
will come up with a better solution (one that doesn't involve
recomputing all the foo/bar-loaddefs.el files whenever subdirs.el is
updated).


        Stefan

Re: Please fix it right!

[Andreas Schwab]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

> I don't claim your problem doesn't exist, but I think your solution is
> wrong.

In which way is it wrong?

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux Products GmbH, Maxfeldstraße 5, 90409 Nürnberg, Germany
PGP key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: Please fix it right!

[Stefan Monnier]

>> I don't claim your problem doesn't exist, but I think your solution is
>> wrong.
> In which way is it wrong?

In the way explained in the comment: those files do not depend on
subdirs.el.  They only depend on a working `emacs' executable.

I.e. my comment is right.  Maybe it doesn't reflect the state of the
code, but then it's a bug in the code.


        Stefan

Re: Please fix it right!

[Stefan Monnier]

>> I don't claim your problem doesn't exist, but I think your solution is
>> wrong.

> In which way is it wrong?

Are you going to give a concrete scenario this year, or are you going to
keep overwriting my changes until I get completely pissed off?


        Stefan

Re: Please fix it right!

[David Kastrup]

Andreas Schwab <schwab@suse.de> writes:

> Stefan Monnier <monnier@iro.umontreal.ca> writes:
>
>> I don't claim your problem doesn't exist, but I think your solution is
>> wrong.
>
> In which way is it wrong?

Andreas, it is not the first time that you are not addressing questions
but instead turn them into a pop-quiz of the questioner's intelligence,
mail after mail.

The time you save writing a few sentences is not worth the time others
have to invest figuring out your meaning.

So please try being less clever and more boring and redundant, by
telling it all.

Thanks,

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: Please fix it right!

[Andreas Schwab]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

> They only depend on a working `emacs' executable.

Exactly.  How can you get a working emacs without it?

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux Products GmbH, Maxfeldstraße 5, 90409 Nürnberg, Germany
PGP key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: Please fix it right!

[Stefan Monnier]

>> They only depend on a working `emacs' executable.

> Exactly.  How can you get a working emacs without it?

I don't know: I just do.  Please show me a *CONCRETE SCENARIO* showing
an *ACTUAL PROBLEM*.  You know, like any normal bug-report.

But FIRST, undo you latest reversion and do not ever try to play this
game again.


        Stefan

Re: Please fix it right!

[Andreas Schwab]

As I already told you the file is required so that emacs can find the
lisp files in the subdirectories.  Since this is the only place where it
is required during build this is the correct way to guarantee it
existence at the right time.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux Products GmbH, Maxfeldstraße 5, 90409 Nürnberg, Germany
PGP key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: Please fix it right!

[Stefan Monnier]

> As I already told you the file is required so that emacs can find the
> lisp files in the subdirectories.  Since this is the only place where it
> is required during build this is the correct way to guarantee it
> existence at the right time.

As I already told, you, I consider it to be a wrong solution.  So please
give me a concrete scenario, not an analysis.


        Stefan

Re: Please fix it right!

[Andreas Schwab]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

> As I already told, you, I consider it to be a wrong solution.

Please expand.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux Products GmbH, Maxfeldstraße 5, 90409 Nürnberg, Germany
PGP key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: Please fix it right!

[Andreas Schwab]

So far my solution was the only working one.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux Products GmbH, Maxfeldstraße 5, 90409 Nürnberg, Germany
PGP key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: Please fix it right!

[Chong Yidong]

Andreas Schwab <schwab@suse.de> writes:

> Chong Yidong <cyd@stupidchicken.com> writes:
>
>> Do you have a test case to demonstrate that this is indeed a
>> problem---yes or no?
>
> Of course, bootstrap can fail without the proper dependencies.

Could you show me an example of this happening?

Re: Please fix it right!

[Stefan Monnier]

> So far my solution was the only working one.

I still haven't seen a concrete scenario showing the problem you're
trying to address.

As someone else said: "If you cannot tell what is "wrong" there is no
way to do it "right"."


        Stefan

Re: Please fix it right!

[Andreas Schwab]

As I already said several times: there is no dependency that creates
subdirs.el at the right time.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux Products GmbH, Maxfeldstraße 5, 90409 Nürnberg, Germany
PGP key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: Please fix it right!

[Jason Rumney]

Andreas Schwab wrote:
> As I already said several times: there is no dependency that creates
> subdirs.el at the right time.
>   

Are you saying that the update-subdirs dependency of the bootstrap 
target is failing to create subdirs.el, or are you saying that 
subdirs.el is required before that point?

Re: Please fix it right!

[Stefan Monnier]

> As I already said several times: there is no dependency that creates
> subdirs.el at the right time.

And as I said several times already, this is not a concrete scenario.
I.e. give us a recipe to reproduce the problem.


        Stefan

RE: Please fix it right!

[Herbert Euler]

> And as I said several times already, this is not a concrete scenario.
> I.e. give us a recipe to reproduce the problem.

I did encounter a bootstrap problem, and with Andreas' patch this
problem went away.  I'm not sure either whether this is the same
problem Andreas tried to address, or whether his patch is the right
one.

Reproducing the problem is easy for me.  Just check out a fresh source
tree, comment out the exact line in lisp/Makefile.in, run the
configure script this way:

  $ CFLAGS="-g -g3 -gdwarf-2" ./configure --with-x-toolkit=gtk

and then start the bootstrap process:

  $ make -j2 bootstrap

The bootstrap process fails with the following messages:

Finding pointers to doc strings...
Finding pointers to doc strings...done
Dumping under names emacs and emacs-23.0.60
72218 pure bytes used
mv -f emacs bootstrap-emacs
make[2]: Leaving directory `/data/emacs/emacs-3/src'
(cd lisp; make -w --jobserver-fds=3,4 - --jobserver-fds=3,4 -j bootstrap EMACS=../src/bootstrap-emacs)
make[2]: Entering directory `/data/emacs/emacs-3/lisp'
wd=/data/emacs/emacs-3/lisp; subdirs=`(cd $wd; find . -type d -print)`; for file in $subdirs; do case $file in */Old | */RCS | */CVS | */CVS/* | */.* | */.*/* | */=* ) ;; *) wins="$wins $wd/$file" ;; esac; done; \
        for file in $wins; do \
           /data/emacs/emacs-3/lisp/../update-subdirs $file; \
        done;
EMACSLOADPATH=/data/emacs/emacs-3/lisp LC_ALL=C ../src/bootstrap-emacs -batch --no-site-file --multibyte -l autoload \
           --eval "(setq generate-autoload-cookie \";;;###cal-autoload\")" \
           --eval "(setq generated-autoload-file \"/data/emacs/emacs-3/lisp/calendar/cal-loaddefs.el\")" \
           --eval "(setq make-backup-files nil)" \
           -f batch-update-autoloads /data/emacs/emacs-3/lisp/calendar
Cannot open load file: autoload
make[2]: *** [/data/emacs/emacs-3/lisp/calendar/cal-loaddefs.el] Error 255
make[2]: *** Waiting for unfinished jobs....
make[2]: Leaving directory `/data/emacs/emacs-3/lisp'
make[1]: *** [bootstrap-build] Error 2
make[1]: Leaving directory `/data/emacs/emacs-3'
make: *** [bootstrap] Error 2

Regards,
Guanpeng Xu
_________________________________________________________________
Discover the new Windows Vista
http://search.msn.com/results.aspx?q=windows+vista&mkt=en-US&form=QBRE

Re: Please fix it right!

[David Kastrup]

Herbert Euler <herberteuler@hotmail.com> writes:

>> And as I said several times already, this is not a concrete scenario.
>> I.e. give us a recipe to reproduce the problem.
>
> I did encounter a bootstrap problem, and with Andreas' patch this
> problem went away.  I'm not sure either whether this is the same
> problem Andreas tried to address, or whether his patch is the right
> one.
>
> Reproducing the problem is easy for me.  Just check out a fresh source
> tree, comment out the exact line in lisp/Makefile.in, run the
> configure script this way:
>
>   $ CFLAGS="-g -g3 -gdwarf-2" ./configure --with-x-toolkit=gtk
>
> and then start the bootstrap process:
>
>   $ make -j2 bootstrap

Is the -j2 necessary for reproducing the problem?

-- 
David Kastrup

RE: Please fix it right!

[Herbert Euler]

> Herbert Euler  writes:
> 
>>> And as I said several times already, this is not a concrete scenario.
>>> I.e. give us a recipe to reproduce the problem.
>>
>> I did encounter a bootstrap problem, and with Andreas' patch this
>> problem went away.  I'm not sure either whether this is the same
>> problem Andreas tried to address, or whether his patch is the right
>> one.
>>
>> Reproducing the problem is easy for me.  Just check out a fresh source
>> tree, comment out the exact line in lisp/Makefile.in, run the
>> configure script this way:
>>
>>   $ CFLAGS="-g -g3 -gdwarf-2" ./configure --with-x-toolkit=gtk
>>
>> and then start the bootstrap process:
>>
>>   $ make -j2 bootstrap
> 
> Is the -j2 necessary for reproducing the problem?

Oops sorry I forgot to mention that.  Yes, without -j2 the bootstrap
process does not fail.

Regards,
Guanpeng Xu
_________________________________________________________________
Discover the new Windows Vista
http://search.msn.com/results.aspx?q=windows+vista&mkt=en-US&form=QBRE

Re: Please fix it right!

[Stefan Monnier]

> Oops sorry I forgot to mention that.  Yes, without -j2 the bootstrap
> process does not fail.

Thank you.  I should be able to find a good fix for it, now.


        Stefan