It is. If you wish to continue down the path of adding more and more code, you 
could make a MY::xs_o that overrides and doesn’t have that. That may be 
temporarily necessary here, although EUMM 7.05_05 is a dev release.

When EUMM 7.06 comes out, it ought to have XSMULTI available, which will enable 
a vastly simpler build system.

From: kmx 
Sent: Wednesday, January 07, 2015 2:26 PM
To: pdl-porters@jach.hawaii.edu 
Subject: [Pdl-porters] dmake + EUMM-7.05_05 trouble

Hi Rob,

(moving the topic to a different thread)

Further testing revealed that EUMM-7.05_05 adds this new section into generated 
Makefiles (with EUMM-7.04 this section is empty):

# --- MakeMaker xs_o section:
.xs$(OBJ_EXT) :
    $(XSUBPPRUN) $(XSPROTOARG) $(XSUBPPARGS) $*.xs > $*.xsc
    $(MV) $*.xsc $*.c
    $(CCCMD) $(CCCDLFLAGS) "-I$(PERL_INC)" $(PASTHRU_DEFINE) $(DEFINE) $*.c -o 
$*$(OBJ_EXT)

Then the dmake warning is true, there are 2 ambiguous build paths from *.xs to 
*.o

1/ .xs.c: rule + .c$(OBJ_EXT): rule

2/ direct .xs$(OBJ_EXT): rule

So it is clearly a bug in EUMM introduced in 7.05_05

I suspect this commit:
https://github.com/Perl-Toolchain-Gang/ExtUtils-MakeMaker/commit/8e4059c6
Ed, isn't it yours?

--
kmx


On 6.1.2015 18:20, kmx wrote:

  My PDL-2.007_06 + ExtUtils-MakeMaker-7.05_05 fails exactly as yours.

  Unfortunately I am short of time right now to investigate more.

  --
  kmx


  On 6.1.2015 14:26, kmx wrote:

    Rob, 

    For me  PDL-2.007_06 builds fine with ExtUtils::MakeMaker-7.04 

    Will try EU-MM-7.0505 later today. 

    -- 
    kmx 

    On 6.1.2015 13:14, sisyph...@optusnet.com.au wrote: 

      From: kmx 
      Sent: Tuesday, January 06, 2015 10:20 AM 
      To: pdl-porters 
      Subject: [Pdl-porters] Simplified Makefile.PL (better EUMM integration) 


        I have a suggestion for improved ExtUtils::MakeMaker support in PDL. 


      kmx, 

      Have you built PDL-2.007_06 on Windows using EU-MM-7 by any chance ? 
      No problems with EU-MM-6 but, for me (EU-MM-7.0505), 'dmake test' 
terminates very noisily - see attached test_termination.txt. 
      There's no actual failure - it's merely dmake noise. 

      Initially I thought it was going to be caused by the postamble, but both 
the quiet (version 6) Makefile and the noisy (version 7) Makefile contain 
identical 'postamble' sections so I'm now thinking it must be something else. 
      See the attached diff.txt for the differences between the 2 Makefiles 
that were generated for Lib/GSL/SF/zeta (which is one of the affected modules). 

      Anyway - just wanted to mention this in case it ties in with anything 
you're looking at. 
      If it's a separate issue then I'll try to sort it out - but I don't want 
to start messing with it if that's going to interfere with your plans. 

      Cheers, 
      Rob 







--------------------------------------------------------------------------------
_______________________________________________
PDL-porters mailing list
PDL-porters@jach.hawaii.edu
http://mailman.jach.hawaii.edu/mailman/listinfo/pdl-porters
_______________________________________________
PDL-porters mailing list
PDL-porters@jach.hawaii.edu
http://mailman.jach.hawaii.edu/mailman/listinfo/pdl-porters

Reply via email to