On Thu, Oct 9, 2025 at 12:13 AM Michael Tokarev <[email protected]> wrote: > > On 10/3/25 06:26, [email protected] wrote: > > From: Alistair Francis <[email protected]> > > > First RISC-V PR for 10.2 > > > > * Fix MSI table size limit > > * Add riscv64 to FirmwareArchitecture > > * Sync RISC-V hwprobe with Linux > > * Implement MonitorDef HMP API > > * Update OpenSBI to v1.7 > > * Fix SiFive UART character drop issue and minor refactors > > * Fix RISC-V timer migration issues > > * Use riscv_cpu_is_32bit() when handling SBI_DBCN reg > > * Use riscv_csrr in riscv_csr_read > > * Align memory allocations to 2M on RISC-V > > * Do not use translator_ldl in opcode_at > > * Minor fixes of RISC-V CFI > > * Modify minimum VLEN rule > > * Fix vslide1[up|down].vx unexpected result when XLEN=32 and SEW=64 > > * Fixup IOMMU PDT Nested Walk > > * Fix endianness swap on compressed instructions > > * Update status of IOMMU kernel support > > Hi! > > Is there anything in there which is worth picking up for the active > stable releases of qemu, which are 10.0.x (lts) and 10.1.x? > > From the patch descriptions, it seems like quite a few changes in > there are worth to be back-ported. For example: > > - target/riscv: rvv: Fix vslide1[up|down].vx unexpected result > when XLEN=32 and SEW=64
Yes > - target/riscv: Fix ssamoswap error handling > - target/riscv: Fix SSP CSR error handling in VU/VS mode > - target/riscv: Fix the mepc when sspopchk triggers the exception Yes > - target/riscv: use riscv_csrr in riscv_csr_read Yes > - hw/char: sifive_uart: Raise IRQ according to the Tx/Rx > watermark thresholds Yes > > This is just from reading the commit messages, - I haven't looked > at the actual changes in the code. Maybe some other changes should > be picked up too. > > I already picked up VLEN fixes, "MSI table size limit" and > "do not use translator_ldl in opcode_at" (and I'm still unsure > about the latter). > > Some changes has been Cc'd qemu-stable before, so I picked these > up too, eg "Fix endianness swap on compressed instructions" and > "Fixup PDT Nested Walk". Thank you. I think that's all that should be back ported > > Some other changes, like the timer array/time delta, does not look > like possibilities for the older series due to incompatibility in > migration they bring. Agreed Alistair > > Thanks, > > /mjt
