Hi Michael, Apologies for the delayed response to your question about qemu-stable. Yes, I believe this patchset is appropriate for qemu-stable material. The changes fix the VLEN minimum value to properly align with the RISC-V specification requirements (VLEN >= ELEN), which could affect the correctness of RISC-V vector extension behavior in QEMU.
Thank you for picking this up for the stable 10.0 and 10.1 releases. This series is based on the riscv-to-apply.next branch (the VERSION file shows 10.1.50). Thanks, Max On Wed, Oct 8, 2025 at 3:13 PM Michael Tokarev <[email protected]> wrote: > On 10/4/25 10:44, Michael Tokarev wrote: > > On 9/23/25 12:07, Max Chou wrote: > >> According to the RISC-V unprivileged specification, the VLEN should be > >> greater > >> or equal to the ELEN. This patchset provides following modifications: > >> > >> * Replace the checkings of standard V with the checkings of Zve32x > >> * Introduces a check rule for VLEN and ELEN > >> * Modifies the minimum VLEN based on the vector extensions > >> > >> Extension Minimum VLEN > >> V 128 > >> Zve64[d|f|x] 64 > >> Zve32[f|x] 32 > >> > >> v1: [email protected] > >> - Rebase to riscv-to-apply.next branch > >> - Add patch 1 to replace checking RVV by checking Zve32x > >> > >> Max Chou (2): > >> target/riscv: rvv: Replace checking V by checking Zve32x > >> target/riscv: rvv: Modify minimum VLEN according to enabled vector > >> extensions > > > > Is this a qemu-stable material? > > (these changes does not apply directly to 10.1.x, probably the > > MonitorDef change in the first patch here can be dropped) > > Hi! > > I've picked this series for qemu-stable 10.0 and 10.1 series. > I still haven't received any reply from y previous email asking > about these, so I'm a bit uncomfortable by picking this up for > stable. But I'm releasing two stable releases today with these > patches in. > > Thanks, > > /mjt >
