Re: PATCH win32/mingw32: jcf_open_exact_case prototype

[Mohan Embar <gnustuff at thisiscool dot com>]

Hi Aaron,

>> I was able to reproduce this and fix this by including both coretypes.h
>> and jcf.h.
>> 
>> ...<i>but</i> if #include "jcf.h" can't be done without #include "coretypes.h",
>> wouldn't it be better to include coretypes.h in jcf.h and eliminate the
>> double-include everywhere else? coretypes.h has an include guard so
>> this shouldn't be a problem.
>
>Sorry I couldn't get it to it myself before you had to be called; I was
>stalled by other work.

No problem. I happened to be playing around with this anyway. I still need
to know the right thing to do before I submit a patch, though.

>> P.S. Aaron: are you building the mainline on MinGW? If so, how are
>> you able to get past the breakage in libmudflap in mf-runtime.c? I
>> had to revert to the 23-06-2004 CVS version to get past this
>> temporarily. I haven't spent much time looking into the causes
>> of this and don't even know who owns libmudflap.
>
>Yes, I build mainline periodically on MinGW.
>
>As I've been told, libmudflap only works on Linux, and will need some
>porting for it to work on MinGW.  On the occasions I've tried,
>libmudflap has never built for me.  Its on my todo list to check this
>out and see how much work it would be to get working properly.  In the
>meantime, I've just been disabling it:
>http://gcc.gnu.org/ml/gcc-patches/2004-05/msg01610.html

Thanks for this. While I've got you here (even though this is probably
OT for this list), what about libbanshee/engine/util.c. vfail() has calls
to sync() and fsync() which don't exist in MinGW. I've hacked around
this, but don't know what an official solution should be (update the
MinGW runtime or put an #ifdef in util.c) or who to go to for it....

-- Mohan
http://www.thisiscool.com/
http://www.animalsong.org/