| Message ID | 20260817090503.2104998-1-peter.lin@sifive.com |
|---|---|
| Headers | show |
| Series | Add RISC-V Worlds ISA support to OpenSBI | expand |
On 17/08/2026 11:04, Yu-Chien Peter Lin wrote: > This patch series adds support for RISC-V Worlds ISA extensions, > enabling hardware-enforced isolation boundaries based on World IDs > by extending the OpenSBI domain framework. > > Trust Model and Boot-time Roles > -------------------------------- > > The implementation follows the two-phase M-mode trust model: > > - RoT M-mode phase: Prior boot stage (ROM/SPL) that may program mwid and > mlwidlist and lock mwid before handing off to OpenSBI. > > - Regular M-mode phase (OpenSBI): Treats pmwid/pmwidlist/pmlwidlist as > input policy from hardware/RoT. Locks mwid to prevent later M-mode code > from changing its own WID. Programs per-domain mlwid/mwiddeleg based on > DT configuration. > > Device Tree Bindings > --------------------- > > Per-hart properties defined in linux kernel DT bindings v3 [1]: > - riscv,pmwid: M-mode World ID > - riscv,pmwidlist: M-mode permitted WID bitmap > - riscv,pmlwidlist: S/U-mode permitted WID bitmap> > Per-domain properties (see domain_support.md): > - wid: S-mode WID for this domain > - widdeleg: WID delegation bitmap for this domain I don't understand why I am cc-ed here, why DT is described here but not in the bindings and why do you reference some RFC (which means: not READY for review) patchset in a different patchset which is ready for review. Best regards, Krzysztof
Hi Krzysztof, On Mon, Sep 28, 2026 at 01:33:17PM +0200, Krzysztof Kozlowski wrote: > On 17/08/2026 11:04, Yu-Chien Peter Lin wrote: > > This patch series adds support for RISC-V Worlds ISA extensions, > > enabling hardware-enforced isolation boundaries based on World IDs > > by extending the OpenSBI domain framework. > > > > Trust Model and Boot-time Roles > > -------------------------------- > > > > The implementation follows the two-phase M-mode trust model: > > > > - RoT M-mode phase: Prior boot stage (ROM/SPL) that may program mwid and > > mlwidlist and lock mwid before handing off to OpenSBI. > > > > - Regular M-mode phase (OpenSBI): Treats pmwid/pmwidlist/pmlwidlist as > > input policy from hardware/RoT. Locks mwid to prevent later M-mode code > > from changing its own WID. Programs per-domain mlwid/mwiddeleg based on > > DT configuration. > > > > Device Tree Bindings > > --------------------- > > > > Per-hart properties defined in linux kernel DT bindings v3 [1]: > > - riscv,pmwid: M-mode World ID > > - riscv,pmwidlist: M-mode permitted WID bitmap > > - riscv,pmlwidlist: S/U-mode permitted WID bitmap> > > Per-domain properties (see domain_support.md): > > - wid: S-mode WID for this domain > > - widdeleg: WID delegation bitmap for this domain > > > I don't understand why I am cc-ed here, OpenSBI consumes the riscv,pmwid/pmwidlist/pmlwidlist properties. Sorry for the confusion; I will remove you from the CC list and provide links to the associated OpenSBI patches in the kernel patch cover letter. > why DT is described here but not > in the bindings and why do you reference some RFC (which means: not > READY for review) patchset in a different patchset which is ready for > review. The kernel patch series is currently pending and cannot be merged until the RISC-V Worlds specification has landed with an associated commit hash. Therefore, I used a TBD placeholder in PATCH 2 [1]. Best regards, Peter Lin [1] https://patchwork.kernel.org/project/linux-riscv/patch/20260812122131.1534906-2-peter.lin@sifive.com/ > > Best regards, > Krzysztof