[Skiboot] [PATCH] xive: fix return value of opal_xive_allocate_irq()

[Oliver O'Halloran]

On Fri, Sep 6, 2019 at 1:59 AM Cédric Le Goater <clg at kaod.org> wrote:
>
> When the maximum number of interrupts per chip is reached,
> xive_try_allocate_irq() returns an internal XIVE error:
> XIVE_ALLOC_NO_SPACE. But its value 0xffffffff is interpreted as a
> positive value by its caller opal_xive_allocate_irq() and not as an
> error.
>
> opal_xive_allocate_irq() returns this value to Linux which also
> considers 0xffffffff as a valid interrupt number and tries to get the
> interrupt characteritics using opal_xive_get_irq_info(). This OPAL
> calls finally fails leading to all sort of errors on the host which is
> not prepared for such a scenario. Code impacted are the IPI setup and
> the both XIVE KVM devices.

How do you replicate this bug? It'd be nice if we had some way to
exercise some of the XIVE API in op-test and this bug seems like a
good place to start.

> Fix by returning OPAL_RESOURCE from xive_try_allocate_irq() which is
> consistent with the other errors returned by this routine. This fixes
> the behavior in opal_xive_allocate_irq() and in Linux.
>
> A workaround could be introduced in Linux to consider 0xffffffff as a
> OPAL_RESOURCE value. This assumption is valid with the current XIVE
> IRQ number encoding.

This is probably going to be needed anyway to support old firmware.

> Reported-by: Greg Kurz <groug at kaod.org>
> Signed-off-by: Cédric Le Goater <clg at kaod.org>

Needs a Fixes: tag. Also can you also add Cc:
skiboot-stable at lists.ozlabs.org so Vasant picks it up for the stable
branches.

> ---
>  hw/xive.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/hw/xive.c b/hw/xive.c
> index 76b41a9ee95f..5dc66598ee12 100644
> --- a/hw/xive.c
> +++ b/hw/xive.c
> @@ -5059,7 +5059,7 @@ static int64_t xive_try_allocate_irq(struct xive *x)
>         idx = bitmap_find_zero_bit(*x->ipi_alloc_map, base_idx, max_count);
>         if (idx < 0) {
>                 unlock(&x->lock);
> -               return XIVE_ALLOC_NO_SPACE;
> +               return OPAL_RESOURCE;
>         }
>         bitmap_set_bit(*x->ipi_alloc_map, idx);
>         girq = x->int_base + idx;
> --
> 2.21.0
>