On 25.06.2024 20:09, Oleksii wrote: > On Tue, 2024-06-25 at 16:45 +0200, Jan Beulich wrote: >> On 25.06.2024 15:51, Oleksii Kurochko wrote: >>> The following compilation error occurs: >>> ./arch/riscv/include/asm/processor.h: Assembler messages: >>> ./arch/riscv/include/asm/processor.h:70: Error: unrecognized >>> opcode `0x0100000F' >>> In case of the following Binutils: >>> $ riscv64-linux-gnu-as --version >>> GNU assembler (GNU Binutils for Debian) 2.35.2 >> >> In patch 6 you say 2.39. Why is 2.35.2 suddenly becoming of interest? > Andrew has (or had) an older version of binutils and started to face > the issues mentioned in this patch and the next one. As a result, some > changes were suggested. > > The version in the README wasn't changed because, in my opinion, this > requires a separate CI job with a prebuilt or fixed toolchain version. > At the moment, it is supported only the version mentioned in README and > that one I have on my machine.
So from my perspective, if you go to the lengths of making changes to support anything older than what you put into README, you will want to at least briefly mention why this is needed / wanted. Plus, as to "separate CI job": That makes little sense to me, or else we'd need to have separate jobs for each and every compiler version out in the world (and within range of what README says). Not just for RISC-V, but also for other architectures. This imo simply wouldn't scale. Jan
