04:21 <davidlt[m]> aurel32: is that with GCC 12 and binutils 2.38?

04:22 <davidlt[m]> "-march=rv64gc -march=rv64imafdc" that looks wrong

05:02 <davidlt[m]> aurel32: I pinged Kito, but could you also create a bug report if you think this is valid bug ?

05:03 <davidlt[m]> Having bug reports in bugzilla might be a better way to track all of this, especially as GCC 12 in a late stage where only bug fixes allowed

07:41 <aurel32> nope that was with gcc 11 (but i don't think binutils should require a new gcc)

07:41 <aurel32> i'll try with gcc 12 but i still need to get it built

07:45 <davidlt[m]> aurel32: Kito is looking into it

07:52 <davidlt[m]> aurel32: the suggestion is to use --with-isa-spec=2.2 if you have GCC 11 + binutils 2.38 setup

07:54 <aurel32> ok, thanks for the feedback

08:01 <davidlt[m]> aurel32: in general I feel that your work testing all of this is more beneficial compared to going with --with-isa-spec=2.2 :)

08:03 <davidlt[m]> aurel32: I see you sent OpenSBI patch too, are you planning to look into U-Boot too?

08:09 <davidlt[m]> U-Boot has: ARCH_FLAGS = -march=$(ARCH_BASE)$(ARCH_A)$(ARCH_C) -mabi=$(ABI) \

08:10 <davidlt[m]> GRUB has:

08:10 <davidlt[m]> CFLAGS="$TARGET_CFLAGS -march=rv64imac -mabi=lp64 -Werror"

08:10 <davidlt[m]> if test "x$target_cpu" = xriscv64; then

08:10 <davidlt[m]> coreboot, oreboot, etc should have issues too

08:12 <davidlt[m]> I think it's fine

08:15 <aurel32> yes, glibc is fine, it was a different issue

08:16 <aurel32> i sent patches for the kernel and opensbi