For all those who can cursed the inability of gprconfig to find the ada runtime, a fix has been pushed into the gcc sources, https://github.com/gcc-mirror/gcc/commit/9467e7331188705ec16c086b77e1809c5b0aab7d.
Basically, with 'gcc --dumpmachine' would return something like 'x86_64-pc-linux', but the runtime file s-oscons.ads said it was 'x86_64-pc-linux-gnu' giving a mismatch and the runtime never being found.
After modifying compilers.xml to hunt in ../lib(.+)?/ since files are 64bit under lib64, I've built gcc-11, gprbuild & xmlada all without issues. Hooray...
Must be one of the longest gubs in gnat...
On 10/03/2022 16:44, 196....com wrote:
For all those who can cursed the inability of gprconfig to find the ada runtime, a fix has been pushed into the gcc sources, https://github.com/gcc-mirror/gcc/commit/9467e7331188705ec16c086b77e1809c5b0aab7d.
Basically, with 'gcc --dumpmachine' would return something like 'x86_64-pc-linux', but the runtime file s-oscons.ads said it was 'x86_64-pc-linux-gnu' giving a mismatch and the runtime never being found.
After modifying compilers.xml to hunt in ../lib(.+)?/ since files are 64bit under lib64, I've built gcc-11, gprbuild & xmlada all without issues. Hooray...
Must be one of the longest gubs in gnat...On Gentoo, it is x86_64-pc-linux-gnu:
ls /usr/x86_64-pc-linux-gnu/
bin binutils-bin gcc-bin i686-w64-mingw32 lib x86_64-w64-mingw32
gcc -dumpmachine
x86_64-pc-linux-gnu
How does this affect us?
On Thursday, 10 March 2022 at 17:55:30 UTC, Luke A. Guest wrote:
On 10/03/2022 16:44, 196....com wrote:
For all those who can cursed the inability of gprconfig to find the ada runtime, a fix has been pushed into the gcc sources, https://github.com/gcc-mirror/gcc/commit/9467e7331188705ec16c086b77e1809c5b0aab7d.On Gentoo, it is x86_64-pc-linux-gnu:
Basically, with 'gcc --dumpmachine' would return something like 'x86_64-pc-linux', but the runtime file s-oscons.ads said it was 'x86_64-pc-linux-gnu' giving a mismatch and the runtime never being found.
After modifying compilers.xml to hunt in ../lib(.+)?/ since files are 64bit under lib64, I've built gcc-11, gprbuild & xmlada all without issues. Hooray...
Must be one of the longest gubs in gnat...
ls /usr/x86_64-pc-linux-gnu/
bin binutils-bin gcc-bin i686-w64-mingw32 lib x86_64-w64-mingw32
gcc -dumpmachine
x86_64-pc-linux-gnu
How does this affect us?
On suse it's x86_64-suse-linux. If you were to build on redhat, it would be x86_64-redhat-linux. They would fail. I'm sure there are others.
On 10/03/2022 18:19, 196...@googlemail.com wrote:
On Thursday, 10 March 2022 at 17:55:30 UTC, Luke A. Guest wrote:
On 10/03/2022 16:44, 196....com wrote:
For all those who can cursed the inability of gprconfig to find the ada runtime, a fix has been pushed into the gcc sources, https://github.com/gcc-mirror/gcc/commit/9467e7331188705ec16c086b77e1809c5b0aab7d.On Gentoo, it is x86_64-pc-linux-gnu:
Basically, with 'gcc --dumpmachine' would return something like 'x86_64-pc-linux', but the runtime file s-oscons.ads said it was 'x86_64-pc-linux-gnu' giving a mismatch and the runtime never being found.
After modifying compilers.xml to hunt in ../lib(.+)?/ since files are 64bit under lib64, I've built gcc-11, gprbuild & xmlada all without issues. Hooray...
Must be one of the longest gubs in gnat...
ls /usr/x86_64-pc-linux-gnu/
bin binutils-bin gcc-bin i686-w64-mingw32 lib x86_64-w64-mingw32
gcc -dumpmachine
x86_64-pc-linux-gnu
How does this affect us?
On suse it's x86_64-suse-linux. If you were to build on redhat, it would be x86_64-redhat-linux. They would fail. I'm sure there are others.That doesn't answer my question. I asked whether this change will flip
it in the other direction and fail for distributions which use the -gnu
part at the end like gentoo does.
If so, that's not a fix.
| Sysop: | Keyop |
|---|---|
| Location: | Huddersfield, West Yorkshire, UK |
| Users: | 307 |
| Nodes: | 16 (2 / 14) |
| Uptime: | 41:22:38 |
| Calls: | 6,910 |
| Files: | 12,376 |
| Messages: | 5,429,141 |