On 21/11/14 18:46, Matt Turner wrote:
On Fri, Nov 21, 2014 at 10:42 AM, Ilia Mirkin <imir...@alum.mit.edu> wrote:
What about src/mesa/drivers/windows/gdi/mesa.def -- it claims to be
generated by mesadef.py. I have no idea what it is, but it should
either also be deleted, or the mesadef.py script kept around.
Good point. Maybe Brian knows (Cc'd).
_______________________________________________
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://urldefense.proofpoint.com/v2/url?u=http-3A__lists.freedesktop.org_mailman_listinfo_mesa-2Ddev&d=AAIGaQ&c=Sqcl0Ez6M0X8aeM67LKIiDJAXVeAw-YihVMNtXt-uEs&r=zfmBZnnVGHeYde45pMKNnVyzeaZbdIqVLprmZCM2zzE&m=fmXAA0WbK-5OnmrssLHDausZYYaPM2tl6mSNxDySxYo&s=zkw6gth0NE7nK10rHcksWQXoVMp_WFP1JXeQoJtB-y4&e=
Don't know about mesadef.py, but mesa.def itself is used, as `git grep`
shows:
$ git grep -C1 '\<mesa.def\>'
[...]
--
src/mesa/drivers/windows/gdi/SConscript-sources = [
src/mesa/drivers/windows/gdi/SConscript: 'mesa.def',
src/mesa/drivers/windows/gdi/SConscript- 'wgl.c',
But I think the warning at the top of mesa.def is misleading. That file
has benn manually edited several times now.
Strictly speaking .def aren't necessary -- `declspec(dllexport)` are
sufficient, but I personally like .def files because they cause build
failures if symbols are missing, or sometimes have different types (due
to symbol decorators used by stdcall convention).
In short, let's keep mesa.def, but we should update the comment.
Jose
_______________________________________________
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
