On Fri, Nov 27, 2020 at 10:51:25PM +0000, Peter Maydell wrote:
> openrisc_sim_net_init() attempts to connect the IRQ line from the
> ethernet device to both CPUs in an SMP configuration by simply caling
> sysbus_connect_irq() for it twice. This doesn't work, because the
> second connection simply overrides the first.
>
> Fix this by creating a TYPE_SPLIT_IRQ to split the IRQ in the SMP
> case.
>
> Signed-off-by: Peter Maydell <peter.mayd...@linaro.org>
> ---
> hw/openrisc/openrisc_sim.c | 13 +++++++++++--
> hw/openrisc/Kconfig | 1 +
> 2 files changed, 12 insertions(+), 2 deletions(-)
>
> diff --git a/hw/openrisc/openrisc_sim.c b/hw/openrisc/openrisc_sim.c
> index d752282e675..a8adf6b70d7 100644
> --- a/hw/openrisc/openrisc_sim.c
> +++ b/hw/openrisc/openrisc_sim.c
> @@ -34,6 +34,7 @@
> #include "hw/sysbus.h"
> #include "sysemu/qtest.h"
> #include "sysemu/reset.h"
> +#include "hw/core/split-irq.h"
>
> #define KERNEL_LOAD_ADDR 0x100
>
> @@ -64,8 +65,16 @@ static void openrisc_sim_net_init(hwaddr base, hwaddr
> descriptors,
>
> s = SYS_BUS_DEVICE(dev);
> sysbus_realize_and_unref(s, &error_fatal);
> - for (i = 0; i < num_cpus; i++) {
> - sysbus_connect_irq(s, 0, cpu_irqs[i][irq_pin]);
> + if (num_cpus > 1) {
> + DeviceState *splitter = qdev_new(TYPE_SPLIT_IRQ);
> + qdev_prop_set_uint32(splitter, "num-lines", num_cpus);
> + qdev_realize_and_unref(splitter, NULL, &error_fatal);
> + for (i = 0; i < num_cpus; i++) {
> + qdev_connect_gpio_out(splitter, i, cpu_irqs[i][irq_pin]);
> + }
> + sysbus_connect_irq(s, 0, qdev_get_gpio_in(splitter, 0));
> + } else {
> + sysbus_connect_irq(s, 0, cpu_irqs[0][irq_pin]);
> }
> sysbus_mmio_map(s, 0, base);
> sysbus_mmio_map(s, 1, descriptors);
> diff --git a/hw/openrisc/Kconfig b/hw/openrisc/Kconfig
> index 6c1e86884e2..8f284f3ba04 100644
> --- a/hw/openrisc/Kconfig
> +++ b/hw/openrisc/Kconfig
> @@ -3,3 +3,4 @@ config OR1K_SIM
> select SERIAL
> select OPENCORES_ETH
> select OMPIC
> + select SPLIT_IRQ
> --
> 2.20.1
This looks good to me, I don't think I ever tested networking with SMP. Thanks
for the fix!
Reviewed-by: Stafford Horne <sho...@gmail.com>
Can you help merge the patch? I am not working a queue right now.
