Please be advised that config-patches is a traditional mailing list where we stick to the old convention of "bottom posting", that is, replying to each piece of your message *underneath* that piece. Keep reading for my actual reply. Please use this convention yourself when replying.
On Mon, Dec 16, 2024, at 2:18 AM, 龙雨洁 wrote: > The body of your message contained only a screen shot of the errors you received. As a courtesy to you, I have OCRed the screen shot, corrected the OCR errors, and then cut it down to the most important bit, which I quote below. In the future, please do not send us screen shots; instead, copy and paste the text itself into your message. That way, we won't have to try to read tiny print in a low-resolution image, or correct OCR errors. > This script (version 2024-07-27), has failed to recognize the > operating system you are using. > [...] > config.guess timestamp = 2024-07-27 > uname -m = sw_64 > uname -r = 5.10.0-46.10.uel20.jk.sw_64 > uname -s = Linux > uname -v = #1 SMP Tue Sep 12 17:50:39 CST 2023 So, you're trying to build LLVM for a CPU architecture called "sw_64", which config.guess does not know about. Neither does LLVM itself, nor the Linux kernel, GCC, Binutils, or any other piece of software I checked. You must have a kernel and a C compiler that have been locally modified to support this CPU. Also, I haven't been able to find any official documentation for the "sw_64" architecture, only some notes from people who tried to reverse engineer something called "Sunway" or "ShenWei" that may or may not be the same thing. If you want to get official support for the "sw_64" architecture into the GNU toolchain and the Linux kernel, we're happy to work with you, but it's going to be a long process. The *very first* step of that process is, the CPU's manufacturer, whoever they are, needs to make the architecture manual publicly available. I don't think we can help you any further until that happens. You may have better luck talking to the people who developed the modified Linux kernel and distribution you seem to be using. zw
