On 15 February 2011 10:49, Adam Lackorzynski <a...@os.inf.tu-dresden.de> wrote:
> Fix selection of target list filter mode.
>
> Signed-off-by: Adam Lackorzynski <a...@os.inf.tu-dresden.de>
> ---
> hw/arm_gic.c | 4 ++--
> 1 files changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/hw/arm_gic.c b/hw/arm_gic.c
> index e6b1953..0e934ec 100644
> --- a/hw/arm_gic.c
> +++ b/hw/arm_gic.c
> @@ -549,10 +549,10 @@ static void gic_dist_writel(void *opaque,
> target_phys_addr_t offset,
> mask = (value >> 16) & ALL_CPU_MASK;
> break;
> case 1:
> - mask = 1 << cpu;
> + mask = ALL_CPU_MASK ^ (1 << cpu);
> break;
> case 2:
> - mask = ALL_CPU_MASK ^ (1 << cpu);
> + mask = 1 << cpu;
> break;
> default:
> DPRINTF("Bad Soft Int target filter\n");
This looks like a good fix based on the TRM; however I don't have an
A9-based image to hand to test with so I've asked a colleague to give
it a quick smoke test before I formally mark it as reviewed-by me
(probably some time next week).
thanks
-- PMM
