Re: [PATCH 2/3] iommu/ipmmu-vmsa: Calculate context registers' offset instead of a macro

[Geert Uytterhoeven <geert@xxxxxxxxxxxxxx>]

Hi Shimoda-san,

On Wed, Oct 9, 2019 at 10:27 AM Yoshihiro Shimoda
<yoshihiro.shimoda.uh@xxxxxxxxxxx> wrote:
> Since we will have changed memory mapping of the IPMMU in the future,
> this patch uses ipmmu_features values instead of a macro to
> calculate context registers offset. No behavior change.
>
> Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@xxxxxxxxxxx>

Thanks for your patch!

> --- a/drivers/iommu/ipmmu-vmsa.c
> +++ b/drivers/iommu/ipmmu-vmsa.c
> @@ -50,6 +50,8 @@ struct ipmmu_features {
>         bool twobit_imttbcr_sl0;
>         bool reserved_context;
>         bool cache_snoop;
> +       u32 ctx_offset_base;
> +       u32 ctx_offset_stride;
>  };
>
>  struct ipmmu_vmsa_device {
> @@ -99,8 +101,6 @@ static struct ipmmu_vmsa_device *to_ipmmu(struct device *dev)
>
>  #define IM_NS_ALIAS_OFFSET             0x800
>
> -#define IM_CTX_SIZE                    0x40
> -
>  #define IMCTR                          0x0000
>  #define IMCTR_TRE                      (1 << 17)
>  #define IMCTR_AFE                      (1 << 16)
> @@ -253,18 +253,25 @@ static void ipmmu_write(struct ipmmu_vmsa_device *mmu, unsigned int offset,
>         iowrite32(data, mmu->base + offset);
>  }
>
> +static u32 ipmmu_ctx_reg(struct ipmmu_vmsa_device *mmu, unsigned int context_id,
> +                        unsigned int reg)
> +{
> +       return mmu->features->ctx_offset_base +
> +              context_id * mmu->features->ctx_offset_stride + reg;
> +}
> +
>  static u32 ipmmu_ctx_read_root(struct ipmmu_vmsa_domain *domain,
>                                unsigned int reg)
>  {
>         return ipmmu_read(domain->mmu->root,
> -                         domain->context_id * IM_CTX_SIZE + reg);
> +                         ipmmu_ctx_reg(domain->mmu, domain->context_id, reg));

For consistency:

    ipmmu_ctx_reg(domain->mmu->root, ...)

but in practice the features for domain->mmu and domain->mmu->root are
identical anyway.

>  }
>
>  static void ipmmu_ctx_write_root(struct ipmmu_vmsa_domain *domain,
>                                  unsigned int reg, u32 data)
>  {
>         ipmmu_write(domain->mmu->root,
> -                   domain->context_id * IM_CTX_SIZE + reg, data);
> +                   ipmmu_ctx_reg(domain->mmu, domain->context_id, reg), data);

Likewise:

    ipmmu_ctx_reg(domain->mmu->root, ...)?

I find these ipmmu_{read,write}() a bit hard too read, with passing the
mmu to both ipmmu_{read,write}() and ipmmu_ctx_reg().

What do you think about providing two helpers ipmmu_ctx_{read,write}(),
so all users can just use e.g.

    ipmmu_ctx_write(mmu, context_id, reg, data);

instead of

    ipmmu_write(mmu, ipmmu_ctx_reg(mmu, context_id, reg), data);

?

Gr{oetje,eeting}s,

                        Geert

-- 
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@xxxxxxxxxxxxxx

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds