On Mon, Feb 12, 2024 at 08:17:31PM +0000, Conor Dooley wrote: > On Mon, Feb 12, 2024 at 08:11:21PM +0100, Miguel Ojeda wrote: > > On Mon, Feb 12, 2024 at 8:02 PM Ramon de C Valle <rcvalle@xxxxxxxxxx> wrote: > > > > > > Sorry for the late reply. Sami might be the best person to answer this, but KCFI (not CFI) tests are lowered by passes that are architecture specific (see https://reviews.llvm.org/D119296), so we'd need to add support for RISC-V. There is no additional work required in the Rust compiler besides enabling it for the new target. > > > > Thanks a lot Ramon! > > > > Then for RISC-V let's go for the `depends on` for the moment, and we > > can remove when the support lands for RISC-V (ideally when someone has > > managed to boot it at least under some configuration). > > If all you want is a boot under some configuration, that's not > difficult. After all, I found the original issue by booting a kernel > with CFI_CLANG enabled on the C side... Also, regardless of depends on on RISC-V, things will still be broken on arm64 and x86_64, since KCFI is not enabled in rustc there either? > > There is no additional work required in the Rust compiler besides enabling it for the new target. > > This is not super clear though, it says "in the Rust compiler", not "in > the kernel's buildsystem". I realise I was not clear either. What I meant was that this talks about rustc and not kbuild, so what is meant by "the new target" is not clear. Do arm64 and x86_64 have functional support, so adding RISC-V in rustc is needed, or did you mean for the new target in the kernel?
Attachment:
signature.asc
Description: PGP signature
