public inbox for gcc@gcc.gnu.org
 help / color / mirror / Atom feed
* Re: Ada does not build (line too long)
@ 2002-05-04 15:56 Robert Dewar
  2002-05-04 16:21 ` Andreas Schwab
  0 siblings, 1 reply; 21+ messages in thread
From: Robert Dewar @ 2002-05-04 15:56 UTC (permalink / raw)
  To: gcc, schwab

<<Ada does not build because almost all source files have lines that are
too long.
>>

This needs more details, since lots of people (including us) have had no
trouble building GNAT.

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: Ada does not build (line too long)
  2002-05-04 15:56 Ada does not build (line too long) Robert Dewar
@ 2002-05-04 16:21 ` Andreas Schwab
  2002-05-05  6:00   ` Florian Weimer
  0 siblings, 1 reply; 21+ messages in thread
From: Andreas Schwab @ 2002-05-04 16:21 UTC (permalink / raw)
  To: Robert Dewar; +Cc: gcc

dewar@gnat.com (Robert Dewar) writes:

|> <<Ada does not build because almost all source files have lines that are
|> too long.
|> >>
|> 
|> This needs more details, since lots of people (including us) have had no
|> trouble building GNAT.

This is on the 3.1 branch.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE GmbH, Deutschherrnstr. 15-19, D-90429 Nürnberg
Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: Ada does not build (line too long)
  2002-05-04 16:21 ` Andreas Schwab
@ 2002-05-05  6:00   ` Florian Weimer
  2002-05-06  3:22     ` Andreas Schwab
  2002-05-06 14:34     ` Andreas Schwab
  0 siblings, 2 replies; 21+ messages in thread
From: Florian Weimer @ 2002-05-05  6:00 UTC (permalink / raw)
  To: Andreas Schwab; +Cc: Robert Dewar, gcc, gcc-patches

[-- Attachment #1: Type: text/plain, Size: 7906 bytes --]

Andreas Schwab <schwab@suse.de> writes:

> dewar@gnat.com (Robert Dewar) writes:
>
> |> <<Ada does not build because almost all source files have lines that are
> |> too long.
> |> >>
> |> 
> |> This needs more details, since lots of people (including us) have had no
> |> trouble building GNAT.
>
> This is on the 3.1 branch.

I've got a patch (gzipped version attached) which does the following
for all affected .ads and .adb files (generated by "perl -i.bak -pne
's/(\$Revision.*?) +--/\1/' *.ad[bs]"):

Index: 4aintnam.ads
===================================================================
RCS file: /cvs/gcc/egcs/gcc/ada/4aintnam.ads,v
retrieving revision 1.1.16.1
diff -u -r1.1.16.1 4aintnam.ads
--- 4aintnam.ads	4 May 2002 03:27:12 -0000	1.1.16.1
+++ 4aintnam.ads	5 May 2002 12:45:55 -0000
@@ -6,7 +6,7 @@
 --                                                                          --
 --                                  S p e c                                 --
 --                                                                          --
---                             $Revision: 1.1.16.1 $                             --
+--                             $Revision: 1.1.16.1 $
 --                                                                          --
 --          Copyright (C) 1991-2001 Free Software Foundation, Inc.          --
 --                                                                          --


Bootstrap has passed stage 2.

However, I'd really like to get rid of these $Revision$ lines after
the 3.1 release.  Besides causing all kinds of trouble, they serve no
purpose.

2002-05-05  Florian Weimer  <fw@deneb.enyo.de>

	* 4aintnam.ads, 4cintnam.ads, 4dintnam.ads, 4hintnam.ads,
	4mintnam.ads, 4nintnam.ads, 4ointnam.ads, 4onumaux.ads,
	4pintnam.ads, 4rintnam.ads, 4sintnam.ads, 4uintnam.ads,
	4vcalend.ads, 4vintnam.ads, 4wintnam.ads, 4zintnam.ads,
	4znumaux.ads, 5avxwork.ads, 5etpopse.adb, 5gintman.adb,
	5gproinf.ads, 5gtasinf.adb, 5ninmaop.adb, 5nintman.adb,
	5oparame.adb, 5pvxwork.ads, 5qvxwork.ads, 5rparame.adb,
	5sintman.adb, 5stasinf.adb, 5stpopse.adb, 5svxwork.ads,
	5uintman.adb, 5vinmaop.adb, 5vintman.adb, 5vtpopde.adb,
	5vtpopde.ads, 5wosprim.adb, 6vinterf.ads, 7sinmaop.adb,
	7sosprim.adb, 7stpopsp.adb, 86numaux.ads, 9drpc.adb, a-astaco.adb,
	a-astaco.ads, a-caldel.ads, a-calend.ads, a-chahan.ads,
	a-charac.ads, a-chlat1.ads, a-colien.adb, a-colien.ads,
	a-decima.adb, a-decima.ads, a-diocst.adb, a-diocst.ads,
	a-direio.adb, a-dynpri.ads, a-excpol.adb, a-filico.ads,
	a-finali.ads, a-flteio.ads, a-fwteio.ads, a-inteio.ads,
	a-interr.adb, a-intnam.ads, a-intsig.adb, a-intsig.ads,
	a-ioexce.ads, a-iwteio.ads, a-lfteio.ads, a-lfwtio.ads,
	a-liteio.ads, a-liwtio.ads, a-llftio.ads, a-llfwti.ads,
	a-llitio.ads, a-lliwti.ads, a-ncelfu.ads, a-ngcoty.ads,
	a-ngelfu.ads, a-nlcefu.ads, a-nlcoty.ads, a-nlelfu.ads,
	a-nllcef.ads, a-nllcty.ads, a-nllefu.ads, a-nscefu.ads,
	a-nscoty.ads, a-nselfu.ads, a-nucoty.ads, a-nuflra.adb,
	a-nuflra.ads, a-numaux.ads, a-numeri.ads, a-sequio.ads,
	a-sfteio.ads, a-sfwtio.ads, a-siocst.adb, a-siocst.ads,
	a-siteio.ads, a-siwtio.ads, a-ssicst.adb, a-ssicst.ads,
	a-ssitio.ads, a-ssiwti.ads, a-storio.adb, a-storio.ads,
	a-stream.ads, a-strfix.ads, a-string.ads, a-strsea.adb,
	a-strsea.ads, a-strunb.ads, a-ststio.ads, a-stunau.adb,
	a-stunau.ads, a-stwibo.ads, a-stwifi.adb, a-stwifi.ads,
	a-stwima.ads, a-stwise.adb, a-stwise.ads, a-sytaco.ads,
	a-tags.ads, a-tasatt.ads, a-teioed.ads, a-ticoau.adb,
	a-ticoau.ads, a-ticoio.adb, a-ticoio.ads, a-tideau.ads,
	a-tideio.ads, a-tienau.ads, a-tifiio.ads, a-tiflau.adb,
	a-tiflau.ads, a-tiflio.ads, a-tiinau.ads, a-tiinio.ads,
	a-timoio.adb, a-tiocst.adb, a-tiocst.ads, a-titest.adb,
	a-titest.ads, a-unccon.ads, a-uncdea.ads, a-wtcoau.adb,
	a-wtcoau.ads, a-wtcoio.adb, a-wtcoio.ads, a-wtcstr.adb,
	a-wtcstr.ads, a-wtdeio.ads, a-wtedit.ads, a-wtenau.ads,
	a-wtfiio.ads, a-wtflau.adb, a-wtflau.ads, a-wtflio.ads,
	a-wtinau.adb, a-wtinau.ads, a-wtinio.ads, a-wtmoau.adb,
	a-wtmoau.ads, a-wtmoio.adb, a-wtmoio.ads, a-wttest.adb,
	a-wttest.ads, ada.ads, binde.ads, bindgen.ads, bindusg.ads,
	calendar.ads, ceinfo.adb, debug_a.adb, debug_a.ads, directio.ads,
	elists.ads, exp_attr.ads, exp_ch10.ads, exp_ch12.ads,
	exp_ch13.ads, exp_ch2.ads, exp_ch6.ads, exp_code.ads,
	exp_disp.ads, exp_dist.ads, exp_fixd.ads, exp_imgv.ads,
	exp_intr.ads, exp_prag.ads, exp_smem.ads, exp_vfpt.adb,
	exp_vfpt.ads, frontend.ads, g-busora.adb, g-busorg.adb,
	g-casuti.adb, g-casuti.ads, g-curexc.ads, g-debuti.adb,
	g-debuti.ads, g-hesora.adb, g-hesorg.adb, g-io_aux.ads,
	g-moreex.adb, g-speche.ads, g-spitbo.adb, g-spitbo.ads,
	g-sptabo.ads, g-sptain.ads, g-sptavs.ads, get_targ.adb,
	gnat1drv.ads, gnatbind.ads, gnatcmd.ads, gnatlink.ads, gnatls.ads,
	gnatmake.adb, gnatmake.ads, gnatprep.ads, gnatpsys.adb, hlo.adb,
	hlo.ads, i-c.ads, i-cexten.ads, i-fortra.adb, i-os2err.ads,
	i-os2lib.ads, i-os2syn.ads, i-os2thr.ads, i-pacdec.ads,
	interfac.ads, ioexcept.ads, itypes.ads, krunch.ads, lib-sort.adb,
	live.ads, machcode.ads, makeusg.ads, math_lib.adb, mdllfile.adb,
	nlists.adb, nlists.ads, par-ch2.adb, par-labl.adb, par.ads,
	prj-com.ads, s-addima.adb, s-addima.ads, s-arit64.ads,
	s-assert.adb, s-assert.ads, s-asthan.adb, s-asthan.ads,
	s-atacco.adb, s-chepoo.ads, s-direio.ads, s-errrep.adb,
	s-errrep.ads, s-exnflt.ads, s-exngen.ads, s-exnint.ads,
	s-exnlfl.ads, s-exnlin.ads, s-exnllf.ads, s-exnlli.ads,
	s-exnsfl.ads, s-exnsin.ads, s-exnssi.ads, s-expflt.ads,
	s-expgen.ads, s-expint.ads, s-explfl.ads, s-explin.ads,
	s-expllf.ads, s-explli.ads, s-expllu.adb, s-expllu.ads,
	s-expmod.adb, s-expmod.ads, s-expsfl.ads, s-expsin.ads,
	s-expssi.ads, s-expuns.adb, s-expuns.ads, s-fatflt.ads,
	s-fatlfl.ads, s-fatllf.ads, s-fatsfl.ads, s-finimp.ads,
	s-fore.adb, s-fore.ads, s-gloloc.adb, s-imgbiu.adb, s-imgboo.adb,
	s-imgcha.adb, s-imgint.adb, s-imgllb.adb, s-imglld.adb,
	s-imglli.adb, s-imgllu.adb, s-imgllw.adb, s-imguns.adb,
	s-imgwch.adb, s-imgwiu.adb, s-inmaop.ads, s-io.ads, s-mantis.adb,
	s-mantis.ads, s-pack03.ads, s-pack05.ads, s-pack07.ads,
	s-pack09.ads, s-pack11.ads, s-pack13.ads, s-pack15.ads,
	s-pack17.ads, s-pack19.ads, s-pack21.ads, s-pack23.ads,
	s-pack25.ads, s-pack27.ads, s-pack29.ads, s-pack31.ads,
	s-pack33.ads, s-pack35.ads, s-pack37.ads, s-pack39.ads,
	s-pack41.ads, s-pack43.ads, s-pack45.ads, s-pack47.ads,
	s-pack49.ads, s-pack51.ads, s-pack53.ads, s-pack55.ads,
	s-pack57.ads, s-pack59.ads, s-pack61.ads, s-pack63.ads,
	s-pooglo.ads, s-pooloc.ads, s-poosiz.ads, s-proinf.adb,
	s-proinf.ads, s-rpc.ads, s-sequio.adb, s-sequio.ads, s-shasto.ads,
	s-sopco3.adb, s-sopco4.adb, s-sopco5.adb, s-stoele.adb,
	s-stoele.ads, s-stopoo.ads, s-stratt.adb, s-stratt.ads,
	s-strops.adb, s-taenca.ads, s-tasinf.adb, s-tasren.ads,
	s-tasuti.ads, s-vaflop.ads, s-valboo.adb, s-valboo.ads,
	s-valcha.adb, s-valdec.adb, s-valdec.ads, s-valint.adb,
	s-valint.ads, s-vallld.adb, s-vallld.ads, s-vallli.adb,
	s-vallli.ads, s-valllu.adb, s-valllu.ads, s-valrea.ads,
	s-valuns.adb, s-valuns.ads, s-valuti.ads, s-valwch.adb,
	s-vercon.adb, s-vercon.ads, s-vmexta.ads, s-wchcon.ads,
	s-wchjis.adb, s-wchjis.ads, s-wchstw.ads, s-wchwts.adb,
	s-wchwts.ads, s-widboo.adb, s-widboo.ads, s-widcha.adb,
	s-widcha.ads, s-widlli.adb, s-widlli.ads, s-widllu.adb,
	s-widllu.ads, s-widwch.adb, s-widwch.ads, s-wwdcha.adb,
	s-wwdcha.ads, s-wwdwch.ads, scn-nlit.adb, sdefault.ads,
	sem_aggr.ads, sem_ch10.ads, sem_ch11.ads, sem_ch2.adb,
	sem_ch2.ads, sem_ch4.ads, sem_ch5.ads, sem_ch6.ads, sem_ch7.ads,
	sem_ch8.ads, sem_disp.ads, sem_elim.ads, sem_intr.ads,
	sem_maps.adb, sem_mech.ads, sem_prag.ads, sem_smem.ads,
	sem_vfpt.ads, sequenio.ads, sinput-l.ads, snames.adb, stand.adb,
	switch.ads, text_io.ads, tree_gen.ads, tree_in.ads, tree_io.ads,
	treepr.ads, uname.ads, unchconv.ads, unchdeal.ads, urealp.ads,
	usage.ads, widechar.ads: Adjust Revision: lines which were too
	long.


[-- Attachment #2: patch-ada-long-lines.gz --]
[-- Type: application/octet-stream, Size: 13370 bytes --]

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: Ada does not build (line too long)
  2002-05-05  6:00   ` Florian Weimer
@ 2002-05-06  3:22     ` Andreas Schwab
  2002-05-06 14:34     ` Andreas Schwab
  1 sibling, 0 replies; 21+ messages in thread
From: Andreas Schwab @ 2002-05-06  3:22 UTC (permalink / raw)
  To: gcc, gcc-patches; +Cc: Mark Mitchell

Florian Weimer <fw@deneb.enyo.de> writes:

|> Andreas Schwab <schwab@suse.de> writes:
|> 
|> > dewar@gnat.com (Robert Dewar) writes:
|> >
|> > |> <<Ada does not build because almost all source files have lines that are
|> > |> too long.
|> > |> >>
|> > |> 
|> > |> This needs more details, since lots of people (including us) have had no
|> > |> trouble building GNAT.
|> >
|> > This is on the 3.1 branch.
|> 
|> I've got a patch (gzipped version attached) which does the following
|> for all affected .ads and .adb files (generated by "perl -i.bak -pne
|> 's/(\$Revision.*?) +--/\1/' *.ad[bs]"):

Can someone please approve and check in that patch so that we can get back
to bootstrap land?

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE GmbH, Deutschherrnstr. 15-19, D-90429 Nürnberg
Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: Ada does not build (line too long)
  2002-05-05  6:00   ` Florian Weimer
  2002-05-06  3:22     ` Andreas Schwab
@ 2002-05-06 14:34     ` Andreas Schwab
  2002-05-06 14:43       ` Florian Weimer
  2002-05-06 15:04       ` Florian Weimer
  1 sibling, 2 replies; 21+ messages in thread
From: Andreas Schwab @ 2002-05-06 14:34 UTC (permalink / raw)
  To: gcc; +Cc: gcc-patches

Florian Weimer <fw@deneb.enyo.de> writes:

|> Andreas Schwab <schwab@suse.de> writes:
|> 
|> > dewar@gnat.com (Robert Dewar) writes:
|> >
|> > |> <<Ada does not build because almost all source files have lines that are
|> > |> too long.
|> > |> >>
|> > |> 
|> > |> This needs more details, since lots of people (including us) have had no
|> > |> trouble building GNAT.
|> >
|> > This is on the 3.1 branch.
|> 
|> I've got a patch (gzipped version attached) which does the following
|> for all affected .ads and .adb files (generated by "perl -i.bak -pne
|> 's/(\$Revision.*?) +--/\1/' *.ad[bs]"):

I think you should also include *.adt here, and also remove the output of
the "--" trailers in xnmake.adb.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE GmbH, Deutschherrnstr. 15-19, D-90429 Nürnberg
Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: Ada does not build (line too long)
  2002-05-06 14:34     ` Andreas Schwab
@ 2002-05-06 14:43       ` Florian Weimer
  2002-05-06 14:50         ` Geert Bosch
  2002-05-06 22:08         ` Simon Wright
  2002-05-06 15:04       ` Florian Weimer
  1 sibling, 2 replies; 21+ messages in thread
From: Florian Weimer @ 2002-05-06 14:43 UTC (permalink / raw)
  To: Andreas Schwab; +Cc: gcc, dewar, bosch

Andreas Schwab <schwab@suse.de> writes:

> I think you should also include *.adt here, and also remove the output of
> the "--" trailers in xnmake.adb.

Ouch, you are right.  (I wonder why I didn't notice that when
bootstrapping the compiler.  Somehow, this indicates that the wrong
versions of these files are being used.)

Perhaps we should merge the changes from mainline which eliminates the
revision numbers for these files.  Comments?

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: Ada does not build (line too long)
  2002-05-06 14:43       ` Florian Weimer
@ 2002-05-06 14:50         ` Geert Bosch
  2002-05-06 14:53           ` Florian Weimer
  2002-05-06 22:08         ` Simon Wright
  1 sibling, 1 reply; 21+ messages in thread
From: Geert Bosch @ 2002-05-06 14:50 UTC (permalink / raw)
  To: Florian Weimer; +Cc: Andreas Schwab, gcc, dewar


On Monday, May 6, 2002, at 05:43 , Florian Weimer wrote:

> Perhaps we should merge the changes from mainline which eliminates the
> revision numbers for these files.  Comments?

It is too late to merge changes from the mainline now. Originally, this
was the idea, but the front end/back end interface changed after 3.1
branched and that made it impractical to do.

   -Geert

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: Ada does not build (line too long)
  2002-05-06 14:50         ` Geert Bosch
@ 2002-05-06 14:53           ` Florian Weimer
  2002-05-06 15:05             ` Geert Bosch
  0 siblings, 1 reply; 21+ messages in thread
From: Florian Weimer @ 2002-05-06 14:53 UTC (permalink / raw)
  To: Geert Bosch; +Cc: Andreas Schwab, gcc, dewar

Geert Bosch <bosch@gnat.com> writes:

> On Monday, May 6, 2002, at 05:43 , Florian Weimer wrote:
>
>> Perhaps we should merge the changes from mainline which eliminates the
>> revision numbers for these files.  Comments?
>
> It is too late to merge changes from the mainline now. Originally, this
> was the idea, but the front end/back end interface changed after 3.1
> branched and that made it impractical to do.

We are talking about xnmake generating comment lines which are so long
that you cannot compile the sources in -gnatg mode.

No code generator involved at all.

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: Ada does not build (line too long)
  2002-05-06 14:34     ` Andreas Schwab
  2002-05-06 14:43       ` Florian Weimer
@ 2002-05-06 15:04       ` Florian Weimer
  1 sibling, 0 replies; 21+ messages in thread
From: Florian Weimer @ 2002-05-06 15:04 UTC (permalink / raw)
  To: Andreas Schwab; +Cc: gcc, gcc-patches

Andreas Schwab <schwab@suse.de> writes:

> Florian Weimer <fw@deneb.enyo.de> writes:
>
> |> Andreas Schwab <schwab@suse.de> writes:
> |> 
> |> > dewar@gnat.com (Robert Dewar) writes:
> |> >
> |> > |> <<Ada does not build because almost all source files have lines that are
> |> > |> too long.
> |> > |> >>
> |> > |> 
> |> > |> This needs more details, since lots of people (including us) have had no
> |> > |> trouble building GNAT.
> |> >
> |> > This is on the 3.1 branch.
> |> 
> |> I've got a patch (gzipped version attached) which does the following
> |> for all affected .ads and .adb files (generated by "perl -i.bak -pne
> |> 's/(\$Revision.*?) +--/\1/' *.ad[bs]"):
>
> I think you should also include *.adt here, and also remove the output of
> the "--" trailers in xnmake.adb.

Okay, here is an (incremental) patch to cover these cases, too.  Could
someone please approve and commit it?

(I won't be able to perform the commit before, errr, 2002-05-07 21:00
UTC or something like that.)

2002-05-07  Florian Weimer  <fw@deneb.enyo.de>

	* nmake.adt, treeprs.adt: Shorten long Revision: line.

	* xnmake.adb, xtreeprs.adb: Do not write comment trailer for
	Revision: lines.

Index: treeprs.adt
===================================================================
RCS file: /cvs/gcc/egcs/gcc/ada/treeprs.adt,v
retrieving revision 1.1.16.1
diff -c -r1.1.16.1 treeprs.adt
*** treeprs.adt	4 May 2002 03:29:23 -0000	1.1.16.1
--- treeprs.adt	6 May 2002 21:59:44 -0000
***************
*** 6,12 ****
  --                                                                          --
  --                             T e m p l a t e                              --
  --                                                                          --
! --                            $Revision: 1.1.16.1 $                             --
  --                                                                          --
  --          Copyright (C) 1992-1997 Free Software Foundation, Inc.          --
  --                                                                          --
--- 6,12 ----
  --                                                                          --
  --                             T e m p l a t e                              --
  --                                                                          --
! --                            $Revision: 1.1.16.1 $
  --                                                                          --
  --          Copyright (C) 1992-1997 Free Software Foundation, Inc.          --
  --                                                                          --
Index: nmake.adt
===================================================================
RCS file: /cvs/gcc/egcs/gcc/ada/nmake.adt,v
retrieving revision 1.1.16.1
diff -c -r1.1.16.1 nmake.adt
*** nmake.adt	4 May 2002 03:28:22 -0000	1.1.16.1
--- nmake.adt	6 May 2002 21:59:44 -0000
***************
*** 6,12 ****
  --                                                                          --
  --                             T e m p l a t e                              --
  --                                                                          --
! --                            $Revision: 1.1.16.1 $                             --
  --                                                                          --
  --          Copyright (C) 1992-2001 Free Software Foundation, Inc.          --
  --                                                                          --
--- 6,12 ----
  --                                                                          --
  --                             T e m p l a t e                              --
  --                                                                          --
! --                            $Revision: 1.1.16.1 $
  --                                                                          --
  --          Copyright (C) 1992-2001 Free Software Foundation, Inc.          --
  --                                                                          --
Index: xtreeprs.adb
===================================================================
RCS file: /cvs/gcc/egcs/gcc/ada/xtreeprs.adb,v
retrieving revision 1.1.16.1
diff -c -r1.1.16.1 xtreeprs.adb
*** xtreeprs.adb	4 May 2002 03:29:26 -0000	1.1.16.1
--- xtreeprs.adb	6 May 2002 21:59:45 -0000
***************
*** 149,155 ****
  begin
     Anchored_Mode := True;
  
!    Match ("$Revision: 1.1.16.1 $", "$Rev" & "ision: " & Break (' ') * Treeprsrev);
  
     if Argument_Count > 0 then
        Create (OutS, Out_File, Argument (1));
--- 149,156 ----
  begin
     Anchored_Mode := True;
  
!    Match ("$Revision: 1.1.16.1 $",
!           "$Rev" & "ision: " & Break (' ') * Treeprsrev);
  
     if Argument_Count > 0 then
        Create (OutS, Out_File, Argument (1));
***************
*** 201,217 ****
           Put_Line
             (OutS,
              "--                Generated by xtreeprs revision " &
!             Treeprsrev & " using                 --");
  
           Put_Line
             (OutS,
              "--                         sinfo.ads revision " &
!             Sinforev & "                          --");
  
           Put_Line
             (OutS,
              "--                        treeprs.adt revision "
!             & Temprev & "                          --");
  
        else
           --  Skip lines describing the template
--- 202,218 ----
           Put_Line
             (OutS,
              "--                Generated by xtreeprs revision " &
!             Treeprsrev & " using");
  
           Put_Line
             (OutS,
              "--                         sinfo.ads revision " &
!             Sinforev);
  
           Put_Line
             (OutS,
              "--                        treeprs.adt revision "
!             & Temprev);
  
        else
           --  Skip lines describing the template
Index: xnmake.adb
===================================================================
RCS file: /cvs/gcc/egcs/gcc/ada/xnmake.adb,v
retrieving revision 1.2.12.1
diff -c -r1.2.12.1 xnmake.adb
*** xnmake.adb	4 May 2002 03:29:25 -0000	1.2.12.1
--- xnmake.adb	6 May 2002 21:59:45 -0000
***************
*** 207,213 ****
  begin
     --  Capture our revision (following line updated by RCS)
  
!    Match ("$Revision: 1.2.12.1 $", "$Rev" & "ision: " & Break (' ') * XNmake_Rev);
  
     Lineno := 0;
     NWidth := 28;
--- 207,214 ----
  begin
     --  Capture our revision (following line updated by RCS)
  
!    Match ("$Revision: 1.2.12.1 $",
!           "$Rev" & "ision: " & Break (' ') * XNmake_Rev);
  
     Lineno := 0;
     NWidth := 28;
***************
*** 286,303 ****
        if Match (Line, GetT_Rev) then
           WriteBS
             ("--                 Generated by xnmake revision " &
!             XNmake_Rev & " using" &
!             "                  --");
  
           WriteBS
             ("--                         sinfo.ads revision " &
!             Sinfo_Rev &
!             "                         --");
  
           WriteBS
             ("--                         nmake.adt revision " &
!             Temp_Rev &
!             "                          --");
  
        else
           --  Skip lines describing the template
--- 287,301 ----
        if Match (Line, GetT_Rev) then
           WriteBS
             ("--                 Generated by xnmake revision " &
!             XNmake_Rev & " using");
  
           WriteBS
             ("--                         sinfo.ads revision " &
!             Sinfo_Rev);
  
           WriteBS
             ("--                         nmake.adt revision " &
!             Temp_Rev);
  
        else
           --  Skip lines describing the template

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: Ada does not build (line too long)
  2002-05-06 14:53           ` Florian Weimer
@ 2002-05-06 15:05             ` Geert Bosch
  2002-05-06 22:04               ` Florian Weimer
  0 siblings, 1 reply; 21+ messages in thread
From: Geert Bosch @ 2002-05-06 15:05 UTC (permalink / raw)
  To: Florian Weimer; +Cc: Andreas Schwab, gcc, dewar


On Monday, May 6, 2002, at 05:53 , Florian Weimer wrote:

> We are talking about xnmake generating comment lines which are so long
> that you cannot compile the sources in -gnatg mode.
>
> No code generator involved at all.

That should be fine. Patch for that is preapproved.

   -Geert

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: Ada does not build (line too long)
  2002-05-06 15:05             ` Geert Bosch
@ 2002-05-06 22:04               ` Florian Weimer
  0 siblings, 0 replies; 21+ messages in thread
From: Florian Weimer @ 2002-05-06 22:04 UTC (permalink / raw)
  To: Geert Bosch; +Cc: Andreas Schwab, gcc, dewar

Geert Bosch <bosch@gnat.com> writes:

> On Monday, May 6, 2002, at 05:53 , Florian Weimer wrote:
>
>> We are talking about xnmake generating comment lines which are so long
>> that you cannot compile the sources in -gnatg mode.
>>
>> No code generator involved at all.
>
> That should be fine. Patch for that is preapproved.

What about the other changes which reduce line length?  Okay as well?

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: Ada does not build (line too long)
  2002-05-06 14:43       ` Florian Weimer
  2002-05-06 14:50         ` Geert Bosch
@ 2002-05-06 22:08         ` Simon Wright
  2002-05-06 22:15           ` Stan Shebs
  1 sibling, 1 reply; 21+ messages in thread
From: Simon Wright @ 2002-05-06 22:08 UTC (permalink / raw)
  To: fw; +Cc: schwab, gcc, dewar, bosch

> From: Florian Weimer <fw@deneb.enyo.de>

> Perhaps we should merge the changes from mainline which eliminates
> the revision numbers for these files.  Comments?

I'm surprised that $Revision$ is omitted from other parts of GCC, but
no doubt there are very good reasons!

Having machine-updated sections inside carefully laid-out parts of the
code seems a bad idea. Perhaps a compromise would be to use a separate
comment?

--  $Revision$

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: Ada does not build (line too long)
  2002-05-06 22:08         ` Simon Wright
@ 2002-05-06 22:15           ` Stan Shebs
  2002-05-07  5:14             ` Geert Bosch
  0 siblings, 1 reply; 21+ messages in thread
From: Stan Shebs @ 2002-05-06 22:15 UTC (permalink / raw)
  To: Simon Wright; +Cc: fw, schwab, gcc, dewar, bosch

Simon Wright wrote:
> 
> > From: Florian Weimer <fw@deneb.enyo.de>
> 
> > Perhaps we should merge the changes from mainline which eliminates
> > the revision numbers for these files.  Comments?
> 
> I'm surprised that $Revision$ is omitted from other parts of GCC, but
> no doubt there are very good reasons!

Among other things, it's a disaster when people import the FSF
sources into their own repositories.  In the worst cases, CVS
will actually report conflicts needing manual resolution, perhaps
on hundreds of files.  Doesn't take too many incidents like that
to stimulate dark and evil thoughts about the people using
$Revision$... :-)

Stan

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: Ada does not build (line too long)
  2002-05-06 22:15           ` Stan Shebs
@ 2002-05-07  5:14             ` Geert Bosch
  0 siblings, 0 replies; 21+ messages in thread
From: Geert Bosch @ 2002-05-07  5:14 UTC (permalink / raw)
  To: Stan Shebs; +Cc: Simon Wright, fw, schwab, gcc, dewar


On Tuesday, May 7, 2002, at 01:15 , Stan Shebs wrote:

> Among other things, it's a disaster when people import the FSF
> sources into their own repositories.  In the worst cases, CVS
> will actually report conflicts needing manual resolution, perhaps
> on hundreds of files.  Doesn't take too many incidents like that
> to stimulate dark and evil thoughts about the people using
> $Revision$... :-)

This is already removed in the mainline, but not in the branch.
If someone volunteers to do this, that would be great.

   -Geert

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: Ada does not build (line too long)
@ 2002-05-07  5:08 Robert Dewar
  0 siblings, 0 replies; 21+ messages in thread
From: Robert Dewar @ 2002-05-07  5:08 UTC (permalink / raw)
  To: shebs, simon; +Cc: bosch, dewar, fw, gcc, schwab

> Among other things, it's a disaster when people import the FSF
> sources into their own repositories.  In the worst cases, CVS
> will actually report conflicts needing manual resolution, perhaps
> on hundreds of files.  Doesn't take too many incidents like that
> to stimulate dark and evil thoughts about the people using
> $Revision$... :-)

As I noted, the revision lines are essential for reasonable source
control in our configuration managed system, where it is essential
that we can identify any version of any source by looking at it.

But in the GCC environment, which is by nature less controlled, and
which works more by differential patches, with many variations running
around, I agree that the revision lines should be deep sixed!

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: Ada does not build (line too long)
@ 2002-05-05  6:04 Robert Dewar
  0 siblings, 0 replies; 21+ messages in thread
From: Robert Dewar @ 2002-05-05  6:04 UTC (permalink / raw)
  To: fw, schwab; +Cc: dewar, gcc-patches, gcc

<<However, I'd really like to get rid of these $Revision$ lines after
the 3.1 release.  Besides causing all kinds of trouble, they serve no
purpose.
>>

I agree. They are critical to our internal configuration management
process, where every version of every source file is required to be
instantly identifiable regardless of where it resides, but no one
aims at this kind of control in the gcc environment in any case.

(our here = ACT), I = Robert Dewar :-)

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: Ada does not build (line too long)
@ 2002-05-04 16:57 Robert Dewar
  0 siblings, 0 replies; 21+ messages in thread
From: Robert Dewar @ 2002-05-04 16:57 UTC (permalink / raw)
  To: dewar, schwab; +Cc: gcc

Ah, OK, yes, I see the problem. We will think about how to best handle
this. Perhaps we should ust get rid of the revision lines for the sources
at gnu.org.

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: Ada does not build (line too long)
  2002-05-04 16:30 Robert Dewar
@ 2002-05-04 16:36 ` Andreas Schwab
  0 siblings, 0 replies; 21+ messages in thread
From: Andreas Schwab @ 2002-05-04 16:36 UTC (permalink / raw)
  To: Robert Dewar; +Cc: gcc

dewar@gnat.com (Robert Dewar) writes:

|> can you give more details. What's an example of a too long line?

The line with the $Revision: ... $ tag.  The long revision number is
shifting the line beyond 80 columns.  See gcc/ada/bindgen.ads, for an
example.  Don't forget to update to current branch head.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE GmbH, Deutschherrnstr. 15-19, D-90429 Nürnberg
Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: Ada does not build (line too long)
@ 2002-05-04 16:30 Robert Dewar
  2002-05-04 16:36 ` Andreas Schwab
  0 siblings, 1 reply; 21+ messages in thread
From: Robert Dewar @ 2002-05-04 16:30 UTC (permalink / raw)
  To: dewar, schwab; +Cc: gcc

can you give more details. What's an example of a too long line?

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Re: Ada does not build (line too long)
  2002-05-04  8:32 Andreas Schwab
@ 2002-05-04 12:10 ` Geoff Keating
  0 siblings, 0 replies; 21+ messages in thread
From: Geoff Keating @ 2002-05-04 12:10 UTC (permalink / raw)
  To: Andreas Schwab; +Cc: gcc

Andreas Schwab <schwab@suse.de> writes:

> Ada does not build because almost all source files have lines that are
> too long.

On a branch?  On mainline?

If the problem is on a branch, you need to import this patch:

2002-03-14  Geoffrey Keating  <geoffk@redhat.com>

        Delete all lines containing "$Revision:".
        * xeinfo.adb: Don't look for revision numbers.
        * xnmake.adb: Likewise.
        * xsinfo.adb: Likewise.
        * xsnames.adb: Likewise.
        * xtreeprs.adb: Likewise.

because apparently ACT doesn't use many branches, and so didn't expect that
the Revision keyword could expand to something as large as '1.12.3.10'.

-- 
- Geoffrey Keating <geoffk@geoffk.org> <geoffk@redhat.com>

^ permalink raw reply	[flat|nested] 21+ messages in thread

* Ada does not build (line too long)
@ 2002-05-04  8:32 Andreas Schwab
  2002-05-04 12:10 ` Geoff Keating
  0 siblings, 1 reply; 21+ messages in thread
From: Andreas Schwab @ 2002-05-04  8:32 UTC (permalink / raw)
  To: gcc

Ada does not build because almost all source files have lines that are
too long.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE GmbH, Deutschherrnstr. 15-19, D-90429 Nürnberg
Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

^ permalink raw reply	[flat|nested] 21+ messages in thread

end of thread, other threads:[~2002-05-07 12:11 UTC | newest]

Thread overview: 21+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2002-05-04 15:56 Ada does not build (line too long) Robert Dewar
2002-05-04 16:21 ` Andreas Schwab
2002-05-05  6:00   ` Florian Weimer
2002-05-06  3:22     ` Andreas Schwab
2002-05-06 14:34     ` Andreas Schwab
2002-05-06 14:43       ` Florian Weimer
2002-05-06 14:50         ` Geert Bosch
2002-05-06 14:53           ` Florian Weimer
2002-05-06 15:05             ` Geert Bosch
2002-05-06 22:04               ` Florian Weimer
2002-05-06 22:08         ` Simon Wright
2002-05-06 22:15           ` Stan Shebs
2002-05-07  5:14             ` Geert Bosch
2002-05-06 15:04       ` Florian Weimer
  -- strict thread matches above, loose matches on Subject: below --
2002-05-07  5:08 Robert Dewar
2002-05-05  6:04 Robert Dewar
2002-05-04 16:57 Robert Dewar
2002-05-04 16:30 Robert Dewar
2002-05-04 16:36 ` Andreas Schwab
2002-05-04  8:32 Andreas Schwab
2002-05-04 12:10 ` Geoff Keating

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).