> And now the build will break for those who use CC="/path/to/ccache
> gcc", won't it?
Probably, yes. But that might well have happened earlier, too. Living
on the master branch requires some manual adaption now and then. The
alternative would be to make no progress.
--tml
___
Hi Tor, *,
It is pointless to write to the mailinglist for feedback, if you
don't even wait two ours before taking the action you ask to be OK or
not. If you want to do it, just do it, but such "alibi" posts don't
really make much sense then.
On Thu, Sep 20, 2012 at 10:29 AM, Tor Lillqvist wrote
> btw, why doesn't macosx.mk use --sysroot, which is much better than
> -isysroot because it mangles the library search path as well? is the
> gcc too old for it?
man gcc:
-isysroot dir
This option is like the --sysroot option, but applies only
to header files, except for Apple
On 20/09/12 09:08, Stephan Bergmann wrote:
> On 09/20/2012 08:48 AM, Tor Lillqvist wrote:
>> I think there is some fuzziness about what compiler options should be
>> part of the CC/CXX variables and what should be in
>> gb_CFLAGS/gb_CXXFLAGS.
>>
>> In some cases, like my Mac build tree that uses th
> Should the -isysroot flag actually be part ot CC/CXX, not gb_CXXFLAGS?
> Should configure.in and macosx.mk be modified thusly?
That's what I did, and reverted my other attempts from this morning to
get liborcus to build, and it seems that it was a good move, the
tinderboxes are getting greener n
On 09/20/2012 08:48 AM, Tor Lillqvist wrote:
I think there is some fuzziness about what compiler options should be
part of the CC/CXX variables and what should be in
gb_CFLAGS/gb_CXXFLAGS.
In some cases, like my Mac build tree that uses the 10.4 SDK, from
Xcode 3 installed in /Xcode3, gb_CXXFLAG
