Package Details: mingw-w64-gcc 9.2.0-1

Git Clone URL: https://aur.archlinux.org/mingw-w64-gcc.git (read-only)
Package Base: mingw-w64-gcc
Description: Cross GCC for the MinGW-w64 cross-compiler
Upstream URL: https://gcc.gnu.org
Licenses: GPL, custom, LGPL, FDL
Groups: mingw-w64, mingw-w64-toolchain
Provides: mingw-w64-gcc-base
Submitter: Barthalion
Maintainer: xantares
Last Packager: xantares
Votes: 44
Popularity: 2.088084
First Submitted: 2018-01-07 17:33
Last Updated: 2019-08-15 06:16

Required by (189)

Sources (3)

Pinned Comments

xantares commented on 2018-03-07 17:54

To install this package from source I recommend to use:

aurman -S --noedit --solution_way --pgp_fetch mingw-w64-gcc

For a binary install I recommend to use Martchus ownstuff repo:

https://wiki.archlinux.org/index.php/unofficial_user_repositories#ownstuff

Latest Comments

1 2 3 4 5 6 ... Next › Last »

Martchus commented on 2019-09-07 12:07

I guess I can scrap the plan of untieing the architectures on the next rebuild. It would work in theory to automatically create distinct packages by using some regex magic in my build scripts. However, my repo would then be unusable for anybody who needs additional mingw-64 packages. So I'll just keep them together.

But I still need to mess with the sources to increment the pkgrel of the packages. Previously I sometimes "added 0.1" to the version so it is still lower than the next version the maintainer would possibly update the package to. It will look very ugly if then every package has such a weird pkgrel but it is the best solution I can think of. Other ideas are welcome but I think we can forget that every maintainer will update all of the packages accordingly.

jerry73204 commented on 2019-08-27 03:09

The makepkg failed with configure: error: GNAT is required to build ada. I'm sure my gcc-ada is installed. Did I miss something?

checking build system type... x86_64-pc-linux-gnu
checking host system type... x86_64-pc-linux-gnu
checking target system type... i686-w64-mingw32
checking for a BSD-compatible install... /usr/bin/install -c
checking whether ln works... yes
checking whether ln -s works... yes
checking for a sed that does not truncate output... /usr/bin/sed
checking for gawk... gawk
checking for libatomic support... yes
checking for libitm support... no
checking for libsanitizer support... no
checking for libvtv support... yes
checking for libhsail-rt support... no
checking for libphobos support... no
checking for gcc... gcc
checking whether the C compiler works... yes
checking for C compiler default output file name... a.out
checking for suffix of executables... 
checking whether we are cross compiling... no
checking for suffix of object files... o
checking whether we are using the GNU C compiler... yes
checking whether gcc accepts -g... yes
checking for gcc option to accept ISO C89... none needed
checking for g++... g++
checking whether we are using the GNU C++ compiler... yes
checking whether g++ accepts -g... yes
checking whether g++ accepts -static-libstdc++ -static-libgcc... yes
checking for gnatbind... gnatbind
checking for gnatmake... gnatmake
checking whether compiler driver understands Ada... no
checking how to compare bootstrapped objects... cmp --ignore-initial=16 $$f1 $$f2
checking for objdir... .libs
checking for the correct version of gmp.h... yes
checking for the correct version of mpfr.h... yes
checking for the correct version of mpc.h... yes
checking for the correct version of the gmp/mpfr/mpc libraries... yes
configure: WARNING: using in-tree isl, disabling version check
configure: error: GNAT is required to build ada

CyberShadow commented on 2019-08-25 08:33

Oh, I see. I had the old version of the file cached.

Edit: I see, the old file is just HTML junk because they redirected HTTPS to HTTP, ugh! I suggested using source=('unique_file_name::file_uri') here previously but you probably know about that already and I guess this isn't really the situation for that.

xantares commented on 2019-08-25 08:28

@CyberShadow this is curious, I still get d18ca11:

$ curl -L http://isl.gforge.inria.fr/isl-0.21.tar.bz2|sha256sum

% Total % Received % Xferd Average Speed Time Time Time Current

                             Dload  Upload   Total   Spent    Left  Speed

100 1778k 100 1778k 0 0 1119k 0 0:00:01 0:00:01 --:--:-- 1119k

d18ca11f8ad1a39ab6d03d3dcb3365ab416720fcb65b42d69f34f51bf0a0e859

CyberShadow commented on 2019-08-25 01:41

==> Validating source files with sha256sums...
    gcc-9.2.0.tar.xz ... Passed
    gcc-9.2.0.tar.xz.sig ... Skipped
    isl-0.21.tar.bz2 ... FAILED

I noticed that in dab12c1e4a18 you changed the source from https to http, but also updated the file sha256sum. Did the file contents really actually change for some reason related to whether it's fetched with/without SSL?

Martchus commented on 2019-08-21 14:03

Spring 2020 sounds good. I likely (need) to drop some legacy packages from my repo then (like angleproject and qt4 and everything which doesn't build anymore) so it will be spring cleanup then :-)

I'm also wondering who is using i686 anymore anyways? I haven't used any i686 Windows system anymore for ages. But in some cases producing a 32-bit DLL can still be useful (e.g. for the "multilib" use-case). So somehow I'd really like to untie the architectures. This "sometimes usefulness" does not justify to build each and every package for i686 anymore in my opinion.

xantares commented on 2019-08-21 06:59

@Martchus I propose to switch it for gcc 10 (spring 2020)

Martchus commented on 2019-08-20 09:45

Also, why is fortran enabled?

Besides that it is actually needed like @xantares already pointed out I think the mingw-w64 packages should be configured as closely as possible to their Linux counterpart.

Why do you have --disable-dw2-exceptions?

The previous maintainer (when the package even was still in the official community repo) used this. Maybe it was the best option at this point. Since then I have actually put some thought into it: https://github.com/Martchus/PKGBUILDs/issues/56

It is a huge rebuilding effort for me so I will not do this change soon. (So my binary repo will stay SJLJ even though the AUR package is updated to DW2.)

I would also like to untie the architectures at some point and it would make sense to combine these breaking changes. (Maybe the archs could be easily untied by applying a simple search and replace for each i686-w64-mingw32 x86_64-w64-mingw32 and pkgname='mingw-w64- before building each package. I need a script to update the pkgrel before doing rebuilds anyways.)

Possibly i686 should also be dropped just completely. However, I don't like to drop it in a way which would prevent us from building i686 packages at all. The architecture should not be hard-coded because maybe it is still useful to build at least a subset of the packages for i686 (or even i386).

xantares commented on 2019-08-20 08:19

@tannisroot, yes I've planned to switch to dwarf, I'll probably do it for gcc 10 then

yes, some scientific software package still require fortran (arpack, ...)

tannisroot commented on 2019-08-19 20:32

Also, why is fortran enabled? Does anyone on Arch actually still use it?