Re: [Qemu-devel] [PATCH v3 03/20] arm: add Faraday FTAHBC020 support

[Peter Crosthwaite]

Hi Dante,

On Wed, Feb 6, 2013 at 7:45 PM, Kuo-Jung Su <address@hidden> wrote:
> From: Kuo-Jung Su <address@hidden>
>
> It's used to perform AHB remap and also QEMU RAM initialization
> when SDRAM is initialized before AHB remap process activated.
>
> Signed-off-by: Kuo-Jung Su <address@hidden>
> ---
>  hw/arm/Makefile.objs  |    1 +
>  hw/arm/faraday_a369.c |    6 ++
>  hw/arm/ftahbc020.c    |  185 
> +++++++++++++++++++++++++++++++++++++++++++++++++
>  3 files changed, 192 insertions(+)
>  create mode 100644 hw/arm/ftahbc020.c
>
> diff --git a/hw/arm/Makefile.objs b/hw/arm/Makefile.objs
> index 02d1a7b..5825c63 100644
> --- a/hw/arm/Makefile.objs
> +++ b/hw/arm/Makefile.objs
> @@ -35,3 +35,4 @@ obj-$(CONFIG_FDT) += ../device_tree.o
>  obj-y := $(addprefix ../,$(obj-y))
>  obj-y += faraday_a360.o faraday_a360_pmu.o
>  obj-y += faraday_a369.o faraday_a369_scu.o faraday_a369_keypad.o
> +obj-y += ftahbc020.o
> diff --git a/hw/arm/faraday_a369.c b/hw/arm/faraday_a369.c
> index e32dc7f..ae6c445 100644
> --- a/hw/arm/faraday_a369.c
> +++ b/hw/arm/faraday_a369.c
> @@ -54,6 +54,12 @@ a369_device_init(A369State *s)
>
>      /* ftkbc010 */
>      sysbus_create_simple("a369.keypad", 0x92f00000, NULL);
> +
> +    /* ftahbc020 */
> +    s->ahbc = qdev_create(NULL, "ftahbc020");
> +    qdev_prop_set_ptr(s->ahbc, "mach", s);
> +    qdev_init_nofail(s->ahbc);
> +    sysbus_mmio_map(SYS_BUS_DEVICE(s->ahbc), 0, 0x94000000);
>  }
>
>  static void
> diff --git a/hw/arm/ftahbc020.c b/hw/arm/ftahbc020.c
> new file mode 100644
> index 0000000..d68676c
> --- /dev/null
> +++ b/hw/arm/ftahbc020.c
> @@ -0,0 +1,185 @@
> +/*
> + * Faraday AHB controller
> + *
> + * Copyright (c) 2012 Faraday Technology
> + * Written by Dante Su <address@hidden>
> + *
> + * This code is licensed under GNU GPL v2+
> + */
> +
> +#include <hw/hw.h>
> +#include <hw/sysbus.h>
> +#include <hw/devices.h>
> +#include <ui/console.h>
> +#include <qemu/timer.h>
> +#include <sysemu/sysemu.h>
> +
> +#include "faraday.h"
> +
> +#define REG_SLAVE(n)    (n * 4) /* Slave device config (base & size) */
> +#define REG_PRIR        0x80    /* Priority register */
> +#define REG_IDLECR      0x84    /* IDLE count register */
> +#define REG_CR          0x88    /* Control register */
> +#define REG_REVR        0x8c    /* Revision register */
> +
> +#define TYPE_FTAHBC020  "ftahbc020"
> +
> +typedef struct Ftahbc020State {
> +    SysBusDevice busdev;
> +    MemoryRegion iomem;
> +    void *mach;
> +
> +    /* HW register cache */
> +    uint32_t slave4;
> +    uint32_t slave6;
> +    uint32_t cr;
> +} Ftahbc020State;
> +
> +#define FTAHBC020(obj) \
> +    OBJECT_CHECK(Ftahbc020State, obj, TYPE_FTAHBC020)
> +
> +static uint64_t
> +ftahbc020_mem_read(void *opaque, hwaddr addr, unsigned size)
> +{
> +    Ftahbc020State *s = FTAHBC020(opaque);
> +    uint64_t ret = 0;
> +
> +    switch (addr) {
> +    case REG_SLAVE(4):
> +        ret = s->slave4;
> +        break;
> +    case REG_SLAVE(6):
> +        ret = s->slave6;
> +        break;
> +    case REG_CR:
> +        ret = s->cr;
> +        break;
> +    case REG_REVR:
> +        ret = 0x00010301;
> +        break;
> +    }
> +
> +    return ret;
> +}
> +
> +static void
> +ftahbc020_mem_write(void *opaque, hwaddr addr, uint64_t val, unsigned size)
> +{
> +    Ftahbc020State *s = FTAHBC020(opaque);
> +    FaradayMachState *mach = s->mach;

I think this breaks the device model and QOM abstractions. You device
has this ad-hoc total machine visibility through these structs. I
think it would be better to use links from device to device that
better reflect you machines interconnect.

> +    uint32_t base;
> +
> +    if (!mach) {
> +        hw_error("ftahbc020: mach is not yet registered!\n");
> +        exit(1);
> +    }
> +
> +    switch (addr) {
> +    case REG_CR:
> +        s->cr = (uint32_t)val;
> +        if (!mach->ahb_remapped && (s->cr & 0x01)) {
> +            /* Remap AHB slave 4 (ROM) & slave 6 (RAM) */
> +            /* 1. Remap ROM to (0x00000000 + size of RAM) */
> +            base = (1 << ((s->slave6 >> 16) & 0x0f)) << 20;
> +            sysbus_mmio_map(SYS_BUS_DEVICE(mach->rom), 0, base);
> +            /* 2. Update slave4(ROM) & slave6(RAM) */
> +            s->slave4 = base | (s->slave4 & 0x000fffff);
> +            s->slave6 = s->slave6 & 0x000fffff;
> +            /* 3. Update SDRAM map if it has been initialized. */
> +            if (mach->ddr_inited) {
> +                memory_region_del_subregion(mach->as, mach->ram_alias);
> +                memory_region_add_subregion(mach->as, 0, mach->ram);
> +            }
> +            mach->ahb_remapped = 1;

Strange, is the device only capable of one-shot remapping or is this
just a limitation of this device due to usage of sysbus_mmio_map?

> +        }
> +        break;
> +    }
> +}
> +
> +static const MemoryRegionOps ftahbc020_mem_ops = {
> +    .read  = ftahbc020_mem_read,
> +    .write = ftahbc020_mem_write,
> +    .endianness = DEVICE_LITTLE_ENDIAN,
> +};
> +
> +static void ftahbc020_reset(DeviceState *ds)
> +{
> +    SysBusDevice *busdev = SYS_BUS_DEVICE(ds);
> +    Ftahbc020State *s = FTAHBC020(FROM_SYSBUS(Ftahbc020State, busdev));
> +    FaradayMachState *mach = s->mach;
> +
> +    if (!mach) {
> +        hw_error("ftahbc020: mach is not yet initialized!\n");
> +        exit(1);
> +    }
> +
> +    if (mach->ahb_remapped) {
> +        sysbus_mmio_map(SYS_BUS_DEVICE(mach->rom), 0, 0x00000000);
> +        mach->ahb_remapped = 0;
> +    }
> +
> +    s->cr = 0;
> +    s->slave4 = mach->ahb_slave4;
> +    s->slave6 = mach->ahb_slave6;
> +
> +    if (s->slave4 == 0 || s->slave6 == 0) {
> +        hw_error("ftahbc020: slave 4 or 6 is not yet initialized!\n");
> +        exit(1);
> +    }
> +}
> +
> +static int ftahbc020_init(SysBusDevice *dev)
> +{
> +    Ftahbc020State *s = FTAHBC020(FROM_SYSBUS(Ftahbc020State, dev));
> +
> +    memory_region_init_io(&s->iomem,
> +                          &ftahbc020_mem_ops,
> +                          s,
> +                          TYPE_FTAHBC020,
> +                          0x1000);
> +    sysbus_init_mmio(dev, &s->iomem);
> +    return 0;
> +}
> +
> +static Property ftahbc020_properties[] = {
> +    DEFINE_PROP_PTR("mach", Ftahbc020State, mach),

DEFINE_PROP_PTR is deprecated and should not be used in new devices.
QOM links are the replacement.

Regards,
Peter

> +    DEFINE_PROP_END_OF_LIST(),
> +};
> +
> +static const VMStateDescription vmstate_ftahbc020 = {
> +    .name = TYPE_FTAHBC020,
> +    .version_id = 1,
> +    .minimum_version_id = 1,
> +    .minimum_version_id_old = 1,
> +    .fields = (VMStateField[]) {
> +        VMSTATE_UINT32(cr, Ftahbc020State),
> +        VMSTATE_END_OF_LIST(),
> +    }
> +};
> +
> +static void ftahbc020_class_init(ObjectClass *klass, void *data)
> +{
> +    SysBusDeviceClass *k = SYS_BUS_DEVICE_CLASS(klass);
> +    DeviceClass *dc = DEVICE_CLASS(klass);
> +
> +    k->init   = ftahbc020_init;
> +    dc->desc  = TYPE_FTAHBC020;
> +    dc->vmsd  = &vmstate_ftahbc020;
> +    dc->props = ftahbc020_properties;
> +    dc->reset = ftahbc020_reset;
> +    dc->no_user = 1;
> +}
> +
> +static const TypeInfo ftahbc020_info = {
> +    .name          = TYPE_FTAHBC020,
> +    .parent        = TYPE_SYS_BUS_DEVICE,
> +    .instance_size = sizeof(Ftahbc020State),
> +    .class_init    = ftahbc020_class_init,
> +};
> +
> +static void ftahbc020_register_types(void)
> +{
> +    type_register_static(&ftahbc020_info);
> +}
> +
> +type_init(ftahbc020_register_types)
> --
> 1.7.9.5
>
>