Re: [lmi] MinGW gcc-4.3, and .tar.lzma

[Greg Chicares]

On 2008-08-30 01:08Z, Greg Chicares wrote:
[...]
>>  Just FYI: I recently saw a link to experimental 4.3.2 packages available
>> on http://www.tdragon.net/recentgcc/index.php

BTW, I tried a similar gcc-4.x a few months ago. I think it would
be a mistake to try to make every conceivable diagnostic go away
by contorting the code. IIRC, the worst one was that it disliked

  if(0.0 == d) // Yes, indeed: test for *exact* zero! Really!

so I think I turned that one off. Here's the relevant part of an
unapplied local patch; I don't recall the reason for inhibiting
each of these warnings, but IIRC the reasons will become obvious
if you don't inhibit them.

+ifeq (4.3.0,$(gcc_version))
+  # Suppress spurious gcc-4.3.0 warnings:
+  gcc_version_specific_warnings := \
+    -Wno-array-bounds \
+    -Wno-attributes \
+    -Wno-conversion \
+    -Wno-parentheses \
+    -Wno-uninitialized \
+    -Wno-unused \
+endif

Oh, and there's some complaint about (IIRC) boost::bind having
an unused identifier like '_1'. I thought we might work around
that by telling gcc that boost '.hpp' files are "system headers",
but I never figured out how to make that work, and then I started
getting distracted by lots of other stuff.

So it'd be great if you could pick this up. We can't keep using
a 2005 compiler forever.