Closed f-laurens closed 2 years ago
I think Bullseye toolchains are broken, I'm seeing plenty of issues regarding it. I'll look into it.
Thanks!
@f-laurens Apparently there's a bug with GCC 10.x.x when built with Glibc 2.31 and was fixed in: https://sourceware.org/pipermail/glibc-cvs/2020q1/069150.html and committed here: https://sourceware.org/pipermail/libc-alpha/2020-May/113553.html but never was backported to Glibc 2.31 release.
Here's what bugfix's author said:
I also think that we need this patch backported to glibc 2.31 branch because we are seeing failures for arm-none-linux-gnueabihf target when building GCC 10 with glibc 2.31:
I think we need to apply patch directly to math/Makefile
.
Successfully resolved and merged in commit https://github.com/abhiTronix/raspberry-pi-cross-compilers/commit/e7353d20f6a134a44ab672afee4a301947b521b9
@f-laurens Please wait till updated binaries are release. I'll inform you here for testing again.
@f-laurens There were way too many bugs with both GCC and Glibc. v3.0.9
toolchains are in build and soon be available.
This issue has been resolved in the latest Toolchains version v3.0.9
. Download it from here: https://sourceforge.net/projects/raspberry-pi-cross-compilers/files/Raspberry%20Pi%20GCC%20Cross-Compiler%20Toolchains/
This issue has been resolved in the latest Toolchains version
v3.0.9
.
The new version of the Bullseye cross toolchains indeed solves my problem. Thanks!
I am attempting to use the cross compiler toolchains for Bullseye to build Python, and the Python build 'configure' step detects that the compiler is not able to compile a simple program:
I do not have the problem when using the equivalent toolchains for Buster.