On 2-6-2012 23:10, Alberto Villa wrote:
> On Sat, Jun 2, 2012 at 10:09 PM, Mel Flynn wrote:
>> I'm setting QMAKESPEC in my shell startup, because bsd.qt.mk hints that
>> when one wants to use ccache one should write their own spec and set
>> QMAKESPEC. So all the work USE_GCC is doing is in vain,
On Sat, Jun 2, 2012 at 10:09 PM, Mel Flynn wrote:
> I'm setting QMAKESPEC in my shell startup, because bsd.qt.mk hints that
> when one wants to use ccache one should write their own spec and set
> QMAKESPEC. So all the work USE_GCC is doing is in vain, unless it
> overrides QMAKESPEC.
Unless you
On 2-6-2012 10:21, Andriy Gapon wrote:
>
> bsd.qt.mk sets QMAKESPEC based on CXX. This is done in the "early" stage of
> bsd.port.mk processing (!_POSTMKINCLUDED).
> bsd.gcc.mk may change CXX value based on USE_GCC (or WITH_GCC/WANT_CC in the
> future). But this is done in the "main" stage (_POS
On Sat, Jun 2, 2012 at 10:21 AM, Andriy Gapon wrote:
> Is there a good reason that QMAKESPEC is derived that early?
As far as i remember, no.
> Could that logic be moved to the "main" stage?
Yes, probably. I will look into it. Thanks for notifying!
--
Alberto Villa, FreeBSD committer
http://p
bsd.qt.mk sets QMAKESPEC based on CXX. This is done in the "early" stage of
bsd.port.mk processing (!_POSTMKINCLUDED).
bsd.gcc.mk may change CXX value based on USE_GCC (or WITH_GCC/WANT_CC in the
future). But this is done in the "main" stage (_POSTMKINCLUDED).
Thus, QMAKESPEC may fail to honor
