SteveHay at planit
Nov 27, 2007, 5:07 AM
Post #5 of 5
(107 views)
Permalink
|
|
RE: Building perl on Win32 with dmake-4.11
[In reply to]
|
|
Steve Hay wrote:
> Steve Hay wrote:
>> Rafael Garcia-Suarez wrote:
>>> On 27/11/2007, Steve Hay <SteveHay [at] planit> wrote:
>>>> Yesterday I released a build of the latest version of dmake (4.11)
>>>> to CPAN, and it seems that it generates numerous warnings when
>>>> building perl which previous versions didn't:
>>>>
>>>> dmake: Warning: -- Target [config.h] was made but the time stamp
>>>> has not been updated. dmake: Warning: -- Target [..\perly.c] was
>>>> made but the time stamp has not been updated. dmake: Warning: --
>>>> Found file corresponding to virtual target [defsubs.h]. dmake:
>>>> Warning: -- Found file corresponding to virtual target [PPPort.pm].
>>>> dmake: Warning: -- Found file corresponding to virtual target
>>>> [ppport.h]. dmake: Warning: -- Found file corresponding to virtual
>>>> target [RealPPPort.xs].
>>>
>>> Sounds noisy -- IIRC those are all generated files, not part of the
>>> common make process.
>>>
>>>> dmake: Warning: -- Found file corresponding to virtual target
>>>> [Errno.pm].
>>>
>>>> plus warnings like these for each extension that gets built:
>>>>
>>>> dmake: Warning: -- Found file corresponding to virtual target
>>>> [..\..\lib\.exists]. dmake: Warning: -- Found file corresponding
>>>> to virtual target [..\..\lib\auto\Win32CORE\.exists].
>>>> dmake: Warning: -- Found file corresponding to virtual target
>>>> [blib\bin\.exists]. dmake: Warning: -- Found file corresponding to
>>>> virtual target [blib\script\.exists]. dmake: Warning: -- Found
>>>> file corresponding to virtual target [blib\man1\.exists]. dmake:
>>>> Warning: -- Found file corresponding to virtual target
>>>> [blib\man3\.exists].
>>>>
>>>> Are these warnings things that we should be worrying about, or is
>>>> dmake going over the top with all these warnings and should perhaps
>>>> relegate them to verbose mode only?
>>>
>>> I don't understand what those warnings mean.
>>
>> Me neither. Here's what the dmake manual says on virtual targets:
>>
>> Dmake allows to define targets with the sole purpose to
>> enforce a dependency chain that are unable to create the target,
>> hence virtual targets. When dmake tries to make a target, but
>> only finds a target definition without recipe lines, it would
>> normally issues a "Don't know how to make ..." error message, but
>> if a target rule is terminated by a semicolon and has no following
>> recipe lines, or if it has no recipe lines, but defines
>> prerequisites, or if the AUGMAKE mode is enabled (see the
>> COMPATIBILITY section for details), the target is treated as a
>> virtual target and the error is suppressed. In addition to this,
>> if the default target does not have recipe lines it is also
>> treated as a vir- tual target.
>>
>> Virtual targets should not have a corresponding file therefore
>> they inherit the time of their newest prerequisite if they have
>> prerequi- sites, otherwise they get the current time assigned
>> when being made. If the virtual target has a corresponding file a
>> warning is issued, but the time stamp of that file is taken into
>> account. The virtual target uses the time stamp of the
>> corresponding file if it is newer than the one determined by the
>> previous rule.
>>
>> Do these '.exists' file exist when building on other OSes?
>>
>> I must admit that I'm somewhat confused as to why dmake thinks, e.g.,
>> ..\..\lib\.exists is a virtual target. The only references to this in
>> the Makefile are as follows (where INST_LIBDIR=..\..\lib and
>> DFSEP=\):
>>
>> blibdirs : $(INST_LIBDIR)$(DFSEP).exists
>> $(INST_ARCHLIB)$(DFSEP).exists $(INST_AUTODIR)$(DFSEP).exists
>> $(INST_ARCHAUTODIR)$(DFSEP).exists $(INST_BIN)$(DFSEP).exists
>> $(INST_SCRIPT)$(DFSEP).exists $(INST_MAN1DIR)$(DFSEP).exists
>> $(INST_MAN3DIR)$(DFSEP).exists
>>
>> $(INST_LIBDIR)$(DFSEP).exists :: Makefile.PL
>> $(NOECHO) $(MKPATH) $(INST_LIBDIR)
>> $(NOECHO) $(CHMOD) 755 $(INST_LIBDIR)
>> $(NOECHO) $(TOUCH) $(INST_LIBDIR)$(DFSEP).exists
>>
>> The target rule "$(INST_LIBDIR)$(DFSEP).exists" is not terminated by
>> a semicolon, it does have following recipe lines, and (as far as I
>> can tell) AUGMAKE mode is not enabled so is dmake mistaken to treat
>> it as a virtual target in the first place?
>>
>> I may have to go back to the dmake maintainers on this one...
>
> Well, it's easily reproducible. This makefile:
>
> all : exists
> @echo Done
>
> exists ::
> @echo here>exists
>
> emits a warning, while this one doesn't:
>
> all : exists
> @echo Done
>
> exists :
> @echo here>exists
>
> (they both still create the file 'exists').
>
> Does that look like a bug in dmake? My understanding of what '::' is
> supposed to do certainly doesn't explain this behaviour.
I think it must have been a bug that has since been fixed in the current
CVS branch, because rebuilding from scratch with 4.12-cvs I now only get
two warnings from dmake:
dmake: Warning: -- Target [config.h] was made but the time stamp has
not been updated.
dmake: Warning: -- Target [..\perly.c] was made but the time stamp has
not been updated.
The first is presumably because config.h is "created" by getting copied
from one of the canned config files (e.g. win32/config_H.vc). Perhaps
config.h should be "touched" after the copy?
Not sure what's going on with perly.c. I thought that was part of the
distro, neither generated nor copied.
(For anyone interested, the dmake issue fixing the bogus warnings is
probably http://www.openoffice.org/issues/show_bug.cgi?id=64572)
|
