Re: [v2 2/5] target/riscv: Add cycle & instret privilege mode filtering properties

[Atish Kumar Patra <atishp@rivosinc.com>]

On Wed, Jan 3, 2024 at 12:11 PM Daniel Henrique Barboza
<dbarboza@ventanamicro.com> wrote:
>
>
>
> On 12/28/23 21:49, Atish Patra wrote:
> > From: Kaiwen Xue <kaiwenx@rivosinc.com>
> >
> > This adds the properties for ISA extension smcntrpmf. Patches
> > implementing it will follow.
> >
> > Signed-off-by: Atish Patra <atishp@rivosinc.com>
> > Signed-off-by: Kaiwen Xue <kaiwenx@rivosinc.com>
> > ---
> >   target/riscv/cpu.c     | 3 ++-
> >   target/riscv/cpu_cfg.h | 1 +
> >   2 files changed, 3 insertions(+), 1 deletion(-)
> >
> > diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
> > index 83c7c0cf07be..da3f05cd5373 100644
> > --- a/target/riscv/cpu.c
> > +++ b/target/riscv/cpu.c
> > @@ -148,6 +148,7 @@ const RISCVIsaExtData isa_edata_arr[] = {
> >       ISA_EXT_DATA_ENTRY(smstateen, PRIV_VERSION_1_12_0, ext_smstateen),
> >       ISA_EXT_DATA_ENTRY(ssaia, PRIV_VERSION_1_12_0, ext_ssaia),
> >       ISA_EXT_DATA_ENTRY(sscofpmf, PRIV_VERSION_1_12_0, ext_sscofpmf),
> > +    ISA_EXT_DATA_ENTRY(smcntrpmf, PRIV_VERSION_1_12_0, ext_smcntrpmf),
> >       ISA_EXT_DATA_ENTRY(sstc, PRIV_VERSION_1_12_0, ext_sstc),
> >       ISA_EXT_DATA_ENTRY(svadu, PRIV_VERSION_1_12_0, ext_svadu),
> >       ISA_EXT_DATA_ENTRY(svinval, PRIV_VERSION_1_12_0, ext_svinval),
> > @@ -1296,6 +1297,7 @@ const char *riscv_get_misa_ext_description(uint32_t bit)
> >   const RISCVCPUMultiExtConfig riscv_cpu_extensions[] = {
> >       /* Defaults for standard extensions */
> >       MULTI_EXT_CFG_BOOL("sscofpmf", ext_sscofpmf, false),
> > +    DEFINE_PROP_BOOL("smcntrpmf", RISCVCPU, cfg.ext_smcntrpmf, false),
>
> This will end up breaking the build since this macro is adding a Property object
> inside a RISCVCPUMultiExtConfig array. Patch 3 is then fixing it by removing this
> line, so in the end the build works fine. But having a patch that doesn't build
> can make future bisects unpleasant.
>

This was a rebase error. I will fix it in the next version.

> I don't see a problem adding right now the actual flag:
>
> +    MULTI_EXT_CFG_BOOL("smcntrpmf", ext_smcntrpmf, false),
>
> The flag will do nothing, sure, but the commit msg already mentions "Patches
> implementing it will follow", so it's fine to me.
>
>
>
> Thanks,
>
> Daniel
>
>
> >       MULTI_EXT_CFG_BOOL("zifencei", ext_zifencei, true),
> >       MULTI_EXT_CFG_BOOL("zicsr", ext_zicsr, true),
> >       MULTI_EXT_CFG_BOOL("zihintntl", ext_zihintntl, true),
> > @@ -1308,7 +1310,6 @@ const RISCVCPUMultiExtConfig riscv_cpu_extensions[] = {
> >       MULTI_EXT_CFG_BOOL("zve64f", ext_zve64f, false),
> >       MULTI_EXT_CFG_BOOL("zve64d", ext_zve64d, false),
> >       MULTI_EXT_CFG_BOOL("sstc", ext_sstc, true),
> > -
> >       MULTI_EXT_CFG_BOOL("smepmp", ext_smepmp, false),
> >       MULTI_EXT_CFG_BOOL("smstateen", ext_smstateen, false),
> >       MULTI_EXT_CFG_BOOL("svadu", ext_svadu, true),
> > diff --git a/target/riscv/cpu_cfg.h b/target/riscv/cpu_cfg.h
> > index f4605fb190b9..00c34fdd3209 100644
> > --- a/target/riscv/cpu_cfg.h
> > +++ b/target/riscv/cpu_cfg.h
> > @@ -72,6 +72,7 @@ struct RISCVCPUConfig {
> >       bool ext_zihpm;
> >       bool ext_smstateen;
> >       bool ext_sstc;
> > +    bool ext_smcntrpmf;
> >       bool ext_svadu;
> >       bool ext_svinval;
> >       bool ext_svnapot;