svn commit: r296637 - in head: contrib/bmake contrib/bmake/mk contrib/bmake/unit-tests share/mk usr.bin/bmake

[Bryan Drewery]

On 3/13/2016 3:32 PM, Simon J. Gerraty wrote:
> Bryan Drewery <bdrewery at FreeBSD.org> wrote:
>>>  # MACHINE specific depend files are supported, but *not* default.
>>>  # If any already exist, we should follow suit.
>>>  _aml = ${ALL_MACHINE_LIST:Uarm amd64 i386 powerpc:N${MACHINE}} ${MACHINE}
>>> -# MACHINE must be the last entry in _aml ;-)
>>> +# make sure we restore MACHINE
>>> +_m := ${MACHINE}
>>>  _e := ${_aml:@MACHINE@${.MAKE.DEPENDFILE_PREFERENCE:@m@${exists($m):?$m:}@}@}
>>> +MACHINE := ${_m}
>>>  .if !empty(_e)
>>>  .MAKE.DEPENDFILE ?= ${.MAKE.DEPENDFILE_PREFERENCE:M*${MACHINE}:[1]}
>>>  .endif
>>
>>
>> What changed in bmake to require this? Now in older checkouts I don't
> 
> deleting the :@ loop iterator variable.
> 
>> get MACHINE defined with DIRDEPS_BUILD without this change.  That's not
>> ideal since it breaks older builds.
> 
> Yes, I should mfc the above change? it is harmess with older bmake's.
> 

This code is only in head. My problem is now bmake can only build
commits after this one which breaks bisecting and older project branches
which haven't or can't merge in head yet.

Is the :@ iterator variable deletion change in bmake change necessary?


-- 
Regards,
Bryan Drewery

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 473 bytes
Desc: OpenPGP digital signature
URL: <http://lists.freebsd.org/pipermail/svn-src-all/attachments/20160313/8ab85e59/attachment.sig>