[PATCH] gcc/m2: Drop references to $(P)

[PATCH] gcc/m2: Drop references to $(P)

[Arsen Arsenović]

$(P) seems to have been a workaround for some old, proprietary make
implementations that we no longer support.  It was removed in
r0-31149-gb8dad04b688e9c.

gcc/m2/ChangeLog:

	* Make-lang.in: Remove references to $(P).
	* Make-maintainer.in: Ditto.
---
Hi,

We spotted that the m2 makefile includes some long-gone compatibility
variable $(P), presumably left over from when m2 was not in the tree
yet.  This induced a build failure on our end:
https://bugs.gentoo.org/904714

Build-tested on x86_64-pc-linux-gnu.  I haven't finished running the
testsuite.  I believe this only ever expands to an empty string (if not
set by the env) in the current build system, so in theory, it should be
safe.

OK for gcc-13 and trunk (with a priority on the former)?

 gcc/m2/Make-lang.in       | 4 ++--
 gcc/m2/Make-maintainer.in | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/gcc/m2/Make-lang.in b/gcc/m2/Make-lang.in
index b34db0d9156..d0fa692e5b8 100644
--- a/gcc/m2/Make-lang.in
+++ b/gcc/m2/Make-lang.in
@@ -514,7 +514,7 @@ GM2_LIBS_BOOT     = m2/gm2-compiler-boot/gm2.a \
 cc1gm2$(exeext): m2/stage1/cc1gm2$(exeext) $(m2.prev)
 	cp -p $< $@
 
-m2/stage2/cc1gm2$(exeext): m2/stage1/cc1gm2$(exeext) m2/gm2-compiler/m2flex.o $(P) \
+m2/stage2/cc1gm2$(exeext): m2/stage1/cc1gm2$(exeext) m2/gm2-compiler/m2flex.o \
                             $(GM2_C_OBJS) $(BACKEND) $(LIBDEPS) $(GM2_LIBS) \
                             m2/gm2-gcc/rtegraph.o plugin/m2rte$(soext)
 	-test -d $(@D) || $(mkinstalldirs) $(@D)
@@ -527,7 +527,7 @@ m2/stage2/cc1gm2$(exeext): m2/stage1/cc1gm2$(exeext) m2/gm2-compiler/m2flex.o $(
 	@$(call LINK_PROGRESS,$(INDEX.m2),end)
 
 m2/stage1/cc1gm2$(exeext): gm2$(exeext) m2/gm2-compiler-boot/m2flex.o \
-                            $(P) $(GM2_C_OBJS) $(BACKEND) $(LIBDEPS) \
+                            $(GM2_C_OBJS) $(BACKEND) $(LIBDEPS) \
                             $(GM2_LIBS_BOOT) $(MC_LIBS) \
                             m2/gm2-gcc/rtegraph.o plugin/m2rte$(soext) \
                             $(m2.prev)
diff --git a/gcc/m2/Make-maintainer.in b/gcc/m2/Make-maintainer.in
index 17bd9a2d37e..a70682673cd 100644
--- a/gcc/m2/Make-maintainer.in
+++ b/gcc/m2/Make-maintainer.in
@@ -753,7 +753,7 @@ GM2_LIBS_PARANOID = m2/gm2-compiler-paranoid/gm2.a \
 gm2.paranoid: m2/m2obj3/cc1gm2$(exeext) gm2.verifyparanoid
 
 m2/m2obj3/cc1gm2$(exeext): m2/m2obj2/cc1gm2$(exeext) m2/gm2-compiler-paranoid/m2flex.o \
-                            $(P) $(GM2_C_OBJS) $(BACKEND) $(LIBDEPS) $(GM2_LIBS_PARANOID) \
+                            $(GM2_C_OBJS) $(BACKEND) $(LIBDEPS) $(GM2_LIBS_PARANOID) \
                             m2/gm2-gcc/rtegraph.o plugin/m2rte$(exeext).so m2/gm2-libs-boot/M2LINK.o
 	-test -d m2/m2obj3 || $(mkinstalldirs) m2/m2obj3
 	@$(call LINK_PROGRESS,$(INDEX.m2),start)
-- 
2.40.0

Re: [PATCH] gcc/m2: Drop references to $(P)

[Gaius Mulley]

Arsen Arsenović <arsen@aarsen.me> writes:

> $(P) seems to have been a workaround for some old, proprietary make
> implementations that we no longer support.  It was removed in
> r0-31149-gb8dad04b688e9c.
>
> gcc/m2/ChangeLog:
>
> 	* Make-lang.in: Remove references to $(P).
> 	* Make-maintainer.in: Ditto.
> ---
> Hi,
>
> We spotted that the m2 makefile includes some long-gone compatibility
> variable $(P), presumably left over from when m2 was not in the tree
> yet.  This induced a build failure on our end:
> https://bugs.gentoo.org/904714
>
> Build-tested on x86_64-pc-linux-gnu.  I haven't finished running the
> testsuite.  I believe this only ever expands to an empty string (if not
> set by the env) in the current build system, so in theory, it should be
> safe.
>
> OK for gcc-13 and trunk (with a priority on the former)?
>
>  gcc/m2/Make-lang.in       | 4 ++--
>  gcc/m2/Make-maintainer.in | 2 +-
>  2 files changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/gcc/m2/Make-lang.in b/gcc/m2/Make-lang.in
> index b34db0d9156..d0fa692e5b8 100644
> --- a/gcc/m2/Make-lang.in
> +++ b/gcc/m2/Make-lang.in
> @@ -514,7 +514,7 @@ GM2_LIBS_BOOT     = m2/gm2-compiler-boot/gm2.a \
>  cc1gm2$(exeext): m2/stage1/cc1gm2$(exeext) $(m2.prev)
>  	cp -p $< $@
>  
> -m2/stage2/cc1gm2$(exeext): m2/stage1/cc1gm2$(exeext) m2/gm2-compiler/m2flex.o $(P) \
> +m2/stage2/cc1gm2$(exeext): m2/stage1/cc1gm2$(exeext) m2/gm2-compiler/m2flex.o \
>                              $(GM2_C_OBJS) $(BACKEND) $(LIBDEPS) $(GM2_LIBS) \
>                              m2/gm2-gcc/rtegraph.o plugin/m2rte$(soext)
>  	-test -d $(@D) || $(mkinstalldirs) $(@D)
> @@ -527,7 +527,7 @@ m2/stage2/cc1gm2$(exeext): m2/stage1/cc1gm2$(exeext) m2/gm2-compiler/m2flex.o $(
>  	@$(call LINK_PROGRESS,$(INDEX.m2),end)
>  
>  m2/stage1/cc1gm2$(exeext): gm2$(exeext) m2/gm2-compiler-boot/m2flex.o \
> -                            $(P) $(GM2_C_OBJS) $(BACKEND) $(LIBDEPS) \
> +                            $(GM2_C_OBJS) $(BACKEND) $(LIBDEPS) \
>                              $(GM2_LIBS_BOOT) $(MC_LIBS) \
>                              m2/gm2-gcc/rtegraph.o plugin/m2rte$(soext) \
>                              $(m2.prev)
> diff --git a/gcc/m2/Make-maintainer.in b/gcc/m2/Make-maintainer.in
> index 17bd9a2d37e..a70682673cd 100644
> --- a/gcc/m2/Make-maintainer.in
> +++ b/gcc/m2/Make-maintainer.in
> @@ -753,7 +753,7 @@ GM2_LIBS_PARANOID = m2/gm2-compiler-paranoid/gm2.a \
>  gm2.paranoid: m2/m2obj3/cc1gm2$(exeext) gm2.verifyparanoid
>  
>  m2/m2obj3/cc1gm2$(exeext): m2/m2obj2/cc1gm2$(exeext) m2/gm2-compiler-paranoid/m2flex.o \
> -                            $(P) $(GM2_C_OBJS) $(BACKEND) $(LIBDEPS) $(GM2_LIBS_PARANOID) \
> +                            $(GM2_C_OBJS) $(BACKEND) $(LIBDEPS) $(GM2_LIBS_PARANOID) \
>                              m2/gm2-gcc/rtegraph.o plugin/m2rte$(exeext).so m2/gm2-libs-boot/M2LINK.o
>  	-test -d m2/m2obj3 || $(mkinstalldirs) m2/m2obj3
>  	@$(call LINK_PROGRESS,$(INDEX.m2),start)

Hi,

yes certainly this is fine.  lgtm.  Thanks for spotting and the patch

regards,
Gaius

Re: [PATCH] gcc/m2: Drop references to $(P)

[Arsen Arsenović]

[-- Attachment #1: Type: text/plain, Size: 233 bytes --]

Hi Gaius,

Gaius Mulley <gaiusmod2@gmail.com> writes:

> yes certainly this is fine.  lgtm.  Thanks for spotting and the patch

Sure.  Will push to master and wait for a RM to weigh in on 13.

Thanks!
-- 
Arsen Arsenović

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 381 bytes --]

Re: [PATCH] gcc/m2: Drop references to $(P)

[Jakub Jelinek]

On Fri, Apr 21, 2023 at 08:27:22PM +0200, Arsen Arsenović wrote:
> Hi Gaius,
> 
> Gaius Mulley <gaiusmod2@gmail.com> writes:
> 
> > yes certainly this is fine.  lgtm.  Thanks for spotting and the patch
> 
> Sure.  Will push to master and wait for a RM to weigh in on 13.

Doesn't fix any regression, so not ok for 13.1 and I wouldn't bother
for 13.2 either.

	Jakub

Re: [PATCH] gcc/m2: Drop references to $(P)

[Arsen Arsenović]

[-- Attachment #1: Type: text/plain, Size: 312 bytes --]


Jakub Jelinek <jakub@redhat.com> writes:

> Doesn't fix any regression, so not ok for 13.1 and I wouldn't bother
> for 13.2 either.

Okay, but it can affect --enable-languages=all in a slim edge case.

Why not 13.2?  It seems sufficiently simple.

Thanks, have a lovely night!
-- 
Arsen Arsenović

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 381 bytes --]

Re: [PATCH] gcc/m2: Drop references to $(P)

[Richard Biener]

On Fri, 21 Apr 2023, Arsen Arsenovi? wrote:

> 
> Jakub Jelinek <jakub@redhat.com> writes:
> 
> > Doesn't fix any regression, so not ok for 13.1 and I wouldn't bother
> > for 13.2 either.
> 
> Okay, but it can affect --enable-languages=all in a slim edge case.
> 
> Why not 13.2?  It seems sufficiently simple.

If it doesn't show any issues after a while on trunk it's OK for 13.2 I
think.

Richard.

Re: [PATCH] gcc/m2: Drop references to $(P)

[Arsen Arsenović]

[-- Attachment #1: Type: text/plain, Size: 222 bytes --]


Richard Biener <rguenther@suse.de> writes:

> If it doesn't show any issues after a while on trunk it's OK for 13.2 I
> think.

Thanks, I'll re-propose it later then.

Have a lovely day!

-- 
Arsen Arsenović

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 381 bytes --]