Re: [PATCH] hw/intc: cannot clear GICv3 ITS CTLR[Enabled] bit

[Peter Maydell]

On Wed, 24 Nov 2021 at 18:22, Shashi Mallela <shashi.mallela@linaro.org> wrote:
>
> When Enabled bit is cleared in GITS_CTLR,ITS feature continues
> to be enabled.This patch fixes the issue.
>
> Signed-off-by: Shashi Mallela <shashi.mallela@linaro.org>
> ---
>  hw/intc/arm_gicv3_its.c | 7 ++++---
>  1 file changed, 4 insertions(+), 3 deletions(-)
>
> diff --git a/hw/intc/arm_gicv3_its.c b/hw/intc/arm_gicv3_its.c
> index 84bcbb5f56..c929a9cb5c 100644
> --- a/hw/intc/arm_gicv3_its.c
> +++ b/hw/intc/arm_gicv3_its.c
> @@ -896,13 +896,14 @@ static bool its_writel(GICv3ITSState *s, hwaddr offset,
>
>      switch (offset) {
>      case GITS_CTLR:
> -        s->ctlr |= (value & ~(s->ctlr));
> -
> -        if (s->ctlr & ITS_CTLR_ENABLED) {
> +        if (value & R_GITS_CTLR_ENABLED_MASK) {
> +            s->ctlr |= ITS_CTLR_ENABLED;
>              extract_table_params(s);
>              extract_cmdq_params(s);
>              s->creadr = 0;
>              process_cmdq(s);
> +        } else {
> +            s->ctlr &= ~ITS_CTLR_ENABLED;
>          }
>          break;
>      case GITS_CBASER:

The code looks fine, so in that sense
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>

It seems odd that we have two different #defines for the
same bit, though (ITS_CTLR_ENABLED and R_GITS_CTLR_ENABLED_MASK).
We should probably standardize on the latter and drop the
former.

thanks
-- PMM