> When testing a extension, it is often necessary for a certain program not to > need some kind of extension, such as the bitmanip extension, to evaluate the > performance or codesize of the extension. However, the current multilib rules > will report an error when it is not a superset of the MULTILIB_REQUIRED list, > which will cause the program to be unable to link normally, thus failing to > achieve the expected purpose. > > Therefore, the compilation option is added to avoid riscv_multi_lib_check() > interruption of compilation.
I think it's dangerous to remove the check, but I can understand there are some cases where we really do not want the `security lock`, so I am OK with introducing a new option to disable that. but plz add documentation to the option into gcc/doc/invoke.texi and mention it will just use default multilib and MIGHT not be correct in some cases. > --- a/gcc/config/riscv/riscv.opt > +++ b/gcc/config/riscv/riscv.opt > @@ -295,3 +295,7 @@ Enum(riscv_autovec_lmul) String(m8) Value(RVV_M8) > -param=riscv-autovec-lmul= > Target RejectNegative Joined Enum(riscv_autovec_lmul) > Var(riscv_autovec_lmul) Init(RVV_M1) > -param=riscv-autovec-lmul=<string> Set the RVV LMUL of > auto-vectorization in the RISC-V port. > + > +mdisable-multilib-check I would like to rename that to mno-multilib-check since that is more like all other gcc option conventions. > +Target Bool Var(riscv_disable_multilib_check) Init(0) You don't need Var(riscv_disable_multilib_check) and Init(0) since no one check the value And you need RejectNegative to prevent gcc accept -mno-disable-multilib-check > +Disable multilib checking by riscv_multi_lib_check(). `riscv_multi_lib_check()` is meanless for users since it's implementation detail. `Disable multilib checking; use the default multilib if a compatible one is not found.`
