Re: [PATCH 3/9] hw/core/sysbus: Resolve main_system_bus singleton

[Philippe Mathieu-Daudé]

On 20/9/22 01:17, Bernhard Beschow wrote:
> In QEMU, a machine and the main_system_bus always go togehter. Usually
> the bus is part of the machine which suggsts to host it there.

"together", "suggests"

> Since tere is already a current_machine singleton, all code that
> accesses the main_system_bus can be changed (behind the scenes) to go
> through current_machine. This resolves a singleton. Futhermore, by

"Furthermore"

> reifying it in code, the every-machine-has-exactly-one-main-system-bus
> relationship becomes very obvious.
>
> Note that the main_system_bus attribute is a value rather than a
> pointer. This trades pointer dereferences for pointer arithmetic. The
> idea is to reduce cache misses - a rule of thumb says that
> every pointer dereference causes a cache miss while arithmetic is
> basically free.
>
> Signed-off-by: Bernhard Beschow <shentey@gmail.com>
> ---
>   hw/core/bus.c       |  5 ++++-
>   hw/core/machine.c   |  3 +++
>   hw/core/sysbus.c    | 22 +++++-----------------
>   include/hw/boards.h |  1 +
>   4 files changed, 13 insertions(+), 18 deletions(-)

> diff --git a/include/hw/boards.h b/include/hw/boards.h
> index 311ed17e18..7af940102d 100644
> --- a/include/hw/boards.h
> +++ b/include/hw/boards.h

Likely missing the BusState declaration:

  #include "hw/qdev-core.h"

> @@ -346,6 +346,7 @@ struct MachineState {
>        */
>       MemoryRegion *ram;
>       DeviceMemoryState *device_memory;
> +    BusState main_system_bus;
>   
>       ram_addr_t ram_size;
>       ram_addr_t maxram_size;