All of lore.kernel.org
 help / color / mirror / Atom feed
From: Bin Meng <bmeng.cn@gmail.com>
To: Alistair Francis <alistair23@gmail.com>
Cc: Alistair Francis <Alistair.Francis@wdc.com>,
	"open list:RISC-V" <qemu-riscv@nongnu.org>,
	"Alan Quey-Liang Kao\(\(\(\(\(\(\(\(\(\(\)"
	<alankao@andestech.com>,
	Sagar Karandikar <sagark@eecs.berkeley.edu>,
	Bastian Koppelmann <kbastian@mail.uni-paderborn.de>,
	Dylan Jhong <dylan@andestech.com>,
	"qemu-devel@nongnu.org Developers" <qemu-devel@nongnu.org>,
	Palmer Dabbelt <palmer@dabbelt.com>,
	"Ruinland Chuan-Tzu Tsa\(\(\(\(\(\(\(\(\(\(\)"
	<ruinland@andestech.com>
Subject: Re: [PATCH 3/3] Andes AE350 RISC-V Machine
Date: Fri, 12 Mar 2021 09:07:58 +0800	[thread overview]
Message-ID: <CAEUhbmXkf3k6XN2JNYahA+TvoeQgp034mNFAQmJd+A++nkO+RQ@mail.gmail.com> (raw)
In-Reply-To: <CAKmqyKPdyEjMdp36PjD-zP6LvnAzAi-00+4zASi3bRyRp=+k2A@mail.gmail.com>

On Thu, Mar 11, 2021 at 11:47 PM Alistair Francis <alistair23@gmail.com> wrote:
>
> On Thu, Mar 11, 2021 at 2:03 AM Dylan Jhong <dylan@andestech.com> wrote:
> >
> > On Wed, Mar 10, 2021 at 02:15:25PM +0800, Bin Meng wrote:
> > > On Wed, Mar 10, 2021 at 11:36 AM Dylan Jhong <dylan@andestech.com> wrote:
> > > >
> > > > This provides a RISC-V Board based on Andes's AE350 specification.
> > > > The following machine is implemented:
> > > >
> > > > - 'andes_ae350'; PLIC, PLICSW, PLMT, 16550a UART, VirtIO MMIO, device-tree
> > >
> > > Is this a virtual target because virtio is added? Or does the hardware
> > > provide the virtio programming interface?
> >
> > Andes ae350 is an FPGA evaluation board with many Andes's peripheral devices,
> > but we only provide the most basic functions in the qemu version of ae350.
> > Because we hope that customers can quickly develop and evaluate before getting the real ae350 board,
> > so we use virtio to replace some peripheral devices.
>
> I don't think that's a good idea. The QEMU model should match the
> hardware device.
>

Agree. It is very confusing to add a virtio to a machine that has a
real world board.

> >
> > >
> > > >
> > > > Signed-off-by: Dylan Jhong <dylan@andestech.com>
> > > > Signed-off-by: Ruinland ChuanTzu Tsai <ruinland@andestech.com>
> > > > ---
> > > >  default-configs/devices/riscv32-softmmu.mak |   1 +
> > > >  default-configs/devices/riscv64-softmmu.mak |   1 +
> > > >  hw/riscv/Kconfig                            |   7 +
> > > >  hw/riscv/andes_ae350.c                      | 501 ++++++++++++++++++++
> > > >  hw/riscv/meson.build                        |   1 +
> > > >  include/hw/riscv/andes_ae350.h              |  93 ++++
> > > >  6 files changed, 604 insertions(+)
> > > >  create mode 100644 hw/riscv/andes_ae350.c
> > > >  create mode 100644 include/hw/riscv/andes_ae350.h
> > > >
> > > > diff --git a/default-configs/devices/riscv32-softmmu.mak b/default-configs/devices/riscv32-softmmu.mak
> > > > index d847bd5692..a268007e72 100644
> > > > --- a/default-configs/devices/riscv32-softmmu.mak
> > > > +++ b/default-configs/devices/riscv32-softmmu.mak
> > > > @@ -8,6 +8,7 @@ CONFIG_ARM_COMPATIBLE_SEMIHOSTING=y
> > > >
> > > >  # Boards:
> > > >  #
> > > > +CONFIG_ANDES_AE350=y
> > > >  CONFIG_SPIKE=y
> > > >  CONFIG_SIFIVE_E=y
> > > >  CONFIG_SIFIVE_U=y
> > > > diff --git a/default-configs/devices/riscv64-softmmu.mak b/default-configs/devices/riscv64-softmmu.mak
> > > > index d5eec75f05..9a37dfd8c0 100644
> > > > --- a/default-configs/devices/riscv64-softmmu.mak
> > > > +++ b/default-configs/devices/riscv64-softmmu.mak
> > > > @@ -8,6 +8,7 @@ CONFIG_ARM_COMPATIBLE_SEMIHOSTING=y
> > > >
> > > >  # Boards:
> > > >  #
> > > > +CONFIG_ANDES_AE350=y
> > > >  CONFIG_SPIKE=y
> > > >  CONFIG_SIFIVE_E=y
> > > >  CONFIG_SIFIVE_U=y
> > > > diff --git a/hw/riscv/Kconfig b/hw/riscv/Kconfig
> > > > index d139074b02..04f6369ab7 100644
> > > > --- a/hw/riscv/Kconfig
> > > > +++ b/hw/riscv/Kconfig
> > > > @@ -1,6 +1,13 @@
> > > >  config IBEX
> > > >      bool
> > > >
> > > > +config ANDES_AE350
> > >
> > > This needs to be sorted in alphabetical order
> > >
> >
> > Thanks, this will be fixed in V2.
> >
> > > > +    bool
> > > > +    select SERIAL
> > > > +    select VIRTIO_MMIO
> > > > +    select ANDES_PLIC
> > > > +    select ANDES_PLMT
> > > > +
> > > >  config MICROCHIP_PFSOC
> > > >      bool
> > > >      select CADENCE_SDHCI
> > > > diff --git a/hw/riscv/andes_ae350.c b/hw/riscv/andes_ae350.c
> > > > new file mode 100644
> > > > index 0000000000..ed5f9701ad
> > > > --- /dev/null
> > > > +++ b/hw/riscv/andes_ae350.c
> > > > @@ -0,0 +1,501 @@
> > > > +/*
> > > > + * Andes RISC-V AE350 Board
> > > > + *
> > > > + * Copyright (c) 2021 Andes Tech. Corp.
> > > > + *
> > > > + * Andes AE350 Board supports ns16550a UART and VirtIO MMIO.
> > > > + * The interrupt controllers are andes PLIC and andes PLICSW.
> > > > + * Timer is Andes PLMT.
> > > > + *
> > > > + * This program is free software; you can redistribute it and/or modify it
> > > > + * under the terms and conditions of the GNU General Public License,
> > > > + * version 2 or later, as published by the Free Software Foundation.
> > > > + *
> > > > + * This program is distributed in the hope it will be useful, but WITHOUT
> > > > + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
> > > > + * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
> > > > + * more details.
> > > > + *
> > > > + * You should have received a copy of the GNU General Public License along with
> > > > + * this program.  If not, see <http://www.gnu.org/licenses/>.
> > > > + */
> > > > +
> > > > +#include "qemu/osdep.h"
> > > > +#include "qemu/units.h"
> > > > +#include "qemu/log.h"
> > > > +#include "qemu/error-report.h"
> > > > +#include "qapi/error.h"
> > > > +#include "hw/boards.h"
> > > > +#include "hw/loader.h"
> > > > +#include "hw/sysbus.h"
> > > > +#include "hw/qdev-properties.h"
> > > > +#include "hw/char/serial.h"
> > > > +#include "target/riscv/cpu.h"
> > > > +#include "hw/riscv/riscv_hart.h"
> > > > +#include "hw/riscv/boot.h"
> > > > +#include "hw/riscv/numa.h"
> > > > +#include "chardev/char.h"
> > > > +#include "sysemu/arch_init.h"
> > > > +#include "sysemu/device_tree.h"
> > > > +#include "sysemu/sysemu.h"
> > > > +#include "hw/pci/pci.h"
> > > > +#include "hw/pci-host/gpex.h"
> > > > +
> > > > +#include "hw/intc/andes_plic.h"
> > > > +#include "hw/timer/andes_plmt.h"
> > > > +#include "hw/riscv/andes_ae350.h"
> > > > +
> > > > +# define BIOS_FILENAME ""
> > > > +
> > > > +static const struct MemmapEntry {
> > > > +    hwaddr base;
> > > > +    hwaddr size;
> > > > +} andes_ae350_memmap[] = {
> > > > +    [ANDES_AE350_DEBUG]     = { 0x00000000,      0x100 },
> > > > +    [ANDES_AE350_DRAM]      = { 0x00000000, 0x80000000 },
> > > > +    [ANDES_AE350_MROM]      = { 0xb0000000,   0x100000 },
> > > > +    [ANDES_AE350_MAC]       = { 0xe0100000,   0x100000 },
> > > > +    [ANDES_AE350_GEM]       = { 0xe0200000,   0x100000 },
> > > > +    [ANDES_AE350_PLIC]      = { 0xe4000000,   0x400000 },
> > > > +    [ANDES_AE350_PLMT]      = { 0xe6000000,   0x100000 },
> > > > +    [ANDES_AE350_PLICSW]    = { 0xe6400000,   0x400000 },
> > > > +    [ANDES_AE350_UART1]     = { 0xf0200000,      0x100 },
> > > > +    [ANDES_AE350_UART2]     = { 0xf0300000,      0x100 },
> > > > +    [ANDES_AE350_PIT]       = { 0xf0400000,   0x100000 },
> > > > +    [ANDES_AE350_SDC]       = { 0xf0e00000,   0x100000 },
> > > > +    [ANDES_AE350_VIRTIO]    = { 0xfe000000,     0x1000 },
> > > > +};
> > > > +
> > > > +static void
> > > > +create_fdt(AndesAe350BoardState *bs, const struct MemmapEntry *memmap,
> > > > +    uint64_t mem_size, const char *cmdline)
> > > > +{
> > > > +    AndesAe350SocState *s = &bs->soc;
> > > > +    MachineState *ms = MACHINE(qdev_get_machine());
> > > > +    void *fdt;
> > > > +    int cpu, i;
> > > > +    uint64_t mem_addr;
> > > > +    uint32_t *plic_irq_ext, *plicsw_irq_ext, *plmt_irq_ext;
> > > > +    unsigned long plic_addr, plicsw_addr, plmt_addr;
> > > > +    char *plic_name, *plicsw_name, *plmt_name;
> > > > +    uint32_t intc_phandle = 0, plic_phandle = 0;
> > > > +    uint32_t phandle = 1;
> > > > +    char *isa_name, *mem_name, *cpu_name, *intc_name, *uart_name, *virtio_name;
> > > > +
> > > > +    if (ms->dtb) {
> > > > +        fdt = bs->fdt = load_device_tree(ms->dtb, &bs->fdt_size);
> > > > +        if (!fdt) {
> > > > +            error_report("load_device_tree() failed");
> > > > +            exit(1);
> > > > +        }
> > > > +        goto update_bootargs;
> > > > +    } else {
> > > > +        fdt = bs->fdt = create_device_tree(&bs->fdt_size);
> > > > +        if (!fdt) {
> > > > +            error_report("create_device_tree() failed");
> > > > +            exit(1);
> > > > +        }
> > > > +    }
> > > > +
> > > > +    qemu_fdt_setprop_string(fdt, "/", "model", "Andes AE350 Board");
> > > > +    qemu_fdt_setprop_string(fdt, "/", "compatible", "andestech,ae350");
> > > > +    qemu_fdt_setprop_cell(fdt, "/", "#size-cells", 0x2);
> > > > +    qemu_fdt_setprop_cell(fdt, "/", "#address-cells", 0x2);
> > > > +
> > > > +    qemu_fdt_add_subnode(fdt, "/soc");
> > > > +    qemu_fdt_setprop(fdt, "/soc", "ranges", NULL, 0);
> > > > +    qemu_fdt_setprop_string(fdt, "/soc", "compatible", "simple-bus");
> > > > +    qemu_fdt_setprop_cell(fdt, "/soc", "#size-cells", 0x2);
> > > > +    qemu_fdt_setprop_cell(fdt, "/soc", "#address-cells", 0x2);
> > > > +
> > > > +    qemu_fdt_add_subnode(fdt, "/cpus");
> > > > +    qemu_fdt_setprop_cell(fdt, "/cpus", "timebase-frequency",
> > > > +                          ANDES_PLMT_TIMEBASE_FREQ);
> > > > +    qemu_fdt_setprop_cell(fdt, "/cpus", "#size-cells", 0x0);
> > > > +    qemu_fdt_setprop_cell(fdt, "/cpus", "#address-cells", 0x1);
> > > > +    qemu_fdt_add_subnode(fdt, "/cpus/cpu-map");
> > > > +
> > > > +    plic_irq_ext = g_new0(uint32_t, s->cpus.num_harts * 4);
> > > > +    plicsw_irq_ext = g_new0(uint32_t, s->cpus.num_harts * 2);
> > > > +    plmt_irq_ext = g_new0(uint32_t, s->cpus.num_harts * 2);
> > > > +
> > > > +    for (cpu = 0; cpu < s->cpus.num_harts; cpu++) {
> > > > +        intc_phandle = phandle++;
> > > > +
> > > > +        cpu_name = g_strdup_printf("/cpus/cpu@%d",
> > > > +            s->cpus.hartid_base + cpu);
> > > > +        qemu_fdt_add_subnode(fdt, cpu_name);
> > > > +#if defined(TARGET_RISCV32)
> > > > +        qemu_fdt_setprop_string(fdt, cpu_name, "mmu-type", "riscv,sv32");
> > > > +#else
> > > > +        qemu_fdt_setprop_string(fdt, cpu_name, "mmu-type", "riscv,sv39");
> > > > +#endif
> > > > +        isa_name = riscv_isa_string(&s->cpus.harts[cpu]);
> > > > +        qemu_fdt_setprop_string(fdt, cpu_name, "riscv,isa", isa_name);
> > > > +        g_free(isa_name);
> > > > +        qemu_fdt_setprop_string(fdt, cpu_name, "compatible", "riscv");
> > > > +        qemu_fdt_setprop_string(fdt, cpu_name, "status", "okay");
> > > > +        qemu_fdt_setprop_cell(fdt, cpu_name, "reg",
> > > > +            s->cpus.hartid_base + cpu);
> > > > +        qemu_fdt_setprop_string(fdt, cpu_name, "device_type", "cpu");
> > > > +
> > > > +        intc_name = g_strdup_printf("%s/interrupt-controller", cpu_name);
> > > > +        qemu_fdt_add_subnode(fdt, intc_name);
> > > > +        qemu_fdt_setprop_cell(fdt, intc_name, "phandle", intc_phandle);
> > > > +        qemu_fdt_setprop_string(fdt, intc_name, "compatible",
> > > > +            "riscv,cpu-intc");
> > > > +        qemu_fdt_setprop(fdt, intc_name, "interrupt-controller", NULL, 0);
> > > > +        qemu_fdt_setprop_cell(fdt, intc_name, "#interrupt-cells", 1);
> > > > +
> > > > +        plic_irq_ext[cpu * 4 + 0] = cpu_to_be32(intc_phandle);
> > > > +        plic_irq_ext[cpu * 4 + 1] = cpu_to_be32(IRQ_M_EXT);
> > > > +        plic_irq_ext[cpu * 4 + 2] = cpu_to_be32(intc_phandle);
> > > > +        plic_irq_ext[cpu * 4 + 3] = cpu_to_be32(IRQ_S_EXT);
> > > > +
> > > > +        plicsw_irq_ext[cpu * 2 + 0] = cpu_to_be32(intc_phandle);
> > > > +        plicsw_irq_ext[cpu * 2 + 1] = cpu_to_be32(IRQ_M_SOFT);
> > > > +
> > > > +        plmt_irq_ext[cpu * 2 + 0] = cpu_to_be32(intc_phandle);
> > > > +        plmt_irq_ext[cpu * 2 + 1] = cpu_to_be32(IRQ_M_TIMER);
> > > > +
> > > > +        g_free(intc_name);
> > > > +    }
> > > > +
> > > > +    mem_addr = memmap[ANDES_AE350_DRAM].base;
> > > > +    mem_name = g_strdup_printf("/memory@%lx", (long)mem_addr);
> > > > +    qemu_fdt_add_subnode(fdt, mem_name);
> > > > +    qemu_fdt_setprop_cells(fdt, mem_name, "reg",
> > > > +        mem_addr >> 32, mem_addr, mem_size >> 32, mem_size);
> > > > +    qemu_fdt_setprop_string(fdt, mem_name, "device_type", "memory");
> > > > +    g_free(mem_name);
> > > > +
> > > > +    /* create plic */
> > > > +    plic_phandle = phandle++;
> > > > +    plic_addr = memmap[ANDES_AE350_PLIC].base;
> > > > +    plic_name = g_strdup_printf("/soc/interrupt-controller@%lx", plic_addr);
> > > > +    qemu_fdt_add_subnode(fdt, plic_name);
> > > > +    qemu_fdt_setprop_cell(fdt, plic_name,
> > > > +        "#address-cells", 0x2);
> > > > +    qemu_fdt_setprop_cell(fdt, plic_name,
> > > > +        "#interrupt-cells", 0x2);
> > > > +    qemu_fdt_setprop_string(fdt, plic_name, "compatible", "riscv,plic0");
> > >
> > > This suggests PLIC is the same as the SiFive one. So why do we have a
> > > different implementation of the PLIC model?
> > >
> >
> > The difference of these two PLICs, please refer to my reply of "[PATCH 1/3] Andes RISC-V PLIC"
> >
> > > > +    qemu_fdt_setprop(fdt, plic_name, "interrupt-controller", NULL, 0);
> > > > +    qemu_fdt_setprop(fdt, plic_name, "interrupts-extended",
> > > > +        plic_irq_ext, s->cpus.num_harts * sizeof(uint32_t) * 4);
> > > > +    qemu_fdt_setprop_cells(fdt, plic_name, "reg",
> > > > +        0x0, plic_addr, 0x0, memmap[ANDES_AE350_PLIC].size);
> > > > +    qemu_fdt_setprop_cell(fdt, plic_name, "riscv,ndev", 0x47);
> > > > +    qemu_fdt_setprop_cell(fdt, plic_name, "phandle", plic_phandle);
> > > > +    g_free(plic_name);
> > > > +    g_free(plic_irq_ext);
> > > > +
> > > > +    /* create plicsw */
> > > > +    plicsw_addr = memmap[ANDES_AE350_PLICSW].base;
> > > > +    plicsw_name = g_strdup_printf("/soc/interrupt-controller@%lx", plicsw_addr);
> > > > +    qemu_fdt_add_subnode(fdt, plicsw_name);
> > > > +    qemu_fdt_setprop_cell(fdt, plicsw_name,
> > > > +        "#address-cells", 0x2);
> > > > +    qemu_fdt_setprop_cell(fdt, plicsw_name,
> > > > +        "#interrupt-cells", 0x2);
> > > > +    qemu_fdt_setprop_string(fdt, plicsw_name, "compatible", "riscv,plic1");
> > >
> > > Is this bindings in the Linux kernel upstream? I can't find any
> > > reference in the kernel tree.
> > >
> >
> > Currently only supports andes Linux BSP,
> > we have plans to push to linux upstream in the future.
>
> This raises a good question about device tree bindings being generated
> from QEMU.
>
> Do we want QEMU to generate a DT that doesn't have matching bindings?

I'd prefer new bindings are sent to kernel upstream to get it reviewed
there first. It might not just be the compatible string, but some
other properties that may change.

>
> I guess if the bindings change during the upstream process we can just
> add a second compatible string.
>
> Also if there is no upstream support how can I test the board?
>

Regards,
Bin


WARNING: multiple messages have this Message-ID (diff)
From: Bin Meng <bmeng.cn@gmail.com>
To: Alistair Francis <alistair23@gmail.com>
Cc: Dylan Jhong <dylan@andestech.com>,
	"open list:RISC-V" <qemu-riscv@nongnu.org>,
	 "Alan Quey-Liang Kao(((((((((()" <alankao@andestech.com>,
	Sagar Karandikar <sagark@eecs.berkeley.edu>,
	 Bastian Koppelmann <kbastian@mail.uni-paderborn.de>,
	 "qemu-devel@nongnu.org Developers" <qemu-devel@nongnu.org>,
	Palmer Dabbelt <palmer@dabbelt.com>,
	 "Ruinland Chuan-Tzu Tsa(((((((((()" <ruinland@andestech.com>,
	Alistair Francis <Alistair.Francis@wdc.com>
Subject: Re: [PATCH 3/3] Andes AE350 RISC-V Machine
Date: Fri, 12 Mar 2021 09:07:58 +0800	[thread overview]
Message-ID: <CAEUhbmXkf3k6XN2JNYahA+TvoeQgp034mNFAQmJd+A++nkO+RQ@mail.gmail.com> (raw)
In-Reply-To: <CAKmqyKPdyEjMdp36PjD-zP6LvnAzAi-00+4zASi3bRyRp=+k2A@mail.gmail.com>

On Thu, Mar 11, 2021 at 11:47 PM Alistair Francis <alistair23@gmail.com> wrote:
>
> On Thu, Mar 11, 2021 at 2:03 AM Dylan Jhong <dylan@andestech.com> wrote:
> >
> > On Wed, Mar 10, 2021 at 02:15:25PM +0800, Bin Meng wrote:
> > > On Wed, Mar 10, 2021 at 11:36 AM Dylan Jhong <dylan@andestech.com> wrote:
> > > >
> > > > This provides a RISC-V Board based on Andes's AE350 specification.
> > > > The following machine is implemented:
> > > >
> > > > - 'andes_ae350'; PLIC, PLICSW, PLMT, 16550a UART, VirtIO MMIO, device-tree
> > >
> > > Is this a virtual target because virtio is added? Or does the hardware
> > > provide the virtio programming interface?
> >
> > Andes ae350 is an FPGA evaluation board with many Andes's peripheral devices,
> > but we only provide the most basic functions in the qemu version of ae350.
> > Because we hope that customers can quickly develop and evaluate before getting the real ae350 board,
> > so we use virtio to replace some peripheral devices.
>
> I don't think that's a good idea. The QEMU model should match the
> hardware device.
>

Agree. It is very confusing to add a virtio to a machine that has a
real world board.

> >
> > >
> > > >
> > > > Signed-off-by: Dylan Jhong <dylan@andestech.com>
> > > > Signed-off-by: Ruinland ChuanTzu Tsai <ruinland@andestech.com>
> > > > ---
> > > >  default-configs/devices/riscv32-softmmu.mak |   1 +
> > > >  default-configs/devices/riscv64-softmmu.mak |   1 +
> > > >  hw/riscv/Kconfig                            |   7 +
> > > >  hw/riscv/andes_ae350.c                      | 501 ++++++++++++++++++++
> > > >  hw/riscv/meson.build                        |   1 +
> > > >  include/hw/riscv/andes_ae350.h              |  93 ++++
> > > >  6 files changed, 604 insertions(+)
> > > >  create mode 100644 hw/riscv/andes_ae350.c
> > > >  create mode 100644 include/hw/riscv/andes_ae350.h
> > > >
> > > > diff --git a/default-configs/devices/riscv32-softmmu.mak b/default-configs/devices/riscv32-softmmu.mak
> > > > index d847bd5692..a268007e72 100644
> > > > --- a/default-configs/devices/riscv32-softmmu.mak
> > > > +++ b/default-configs/devices/riscv32-softmmu.mak
> > > > @@ -8,6 +8,7 @@ CONFIG_ARM_COMPATIBLE_SEMIHOSTING=y
> > > >
> > > >  # Boards:
> > > >  #
> > > > +CONFIG_ANDES_AE350=y
> > > >  CONFIG_SPIKE=y
> > > >  CONFIG_SIFIVE_E=y
> > > >  CONFIG_SIFIVE_U=y
> > > > diff --git a/default-configs/devices/riscv64-softmmu.mak b/default-configs/devices/riscv64-softmmu.mak
> > > > index d5eec75f05..9a37dfd8c0 100644
> > > > --- a/default-configs/devices/riscv64-softmmu.mak
> > > > +++ b/default-configs/devices/riscv64-softmmu.mak
> > > > @@ -8,6 +8,7 @@ CONFIG_ARM_COMPATIBLE_SEMIHOSTING=y
> > > >
> > > >  # Boards:
> > > >  #
> > > > +CONFIG_ANDES_AE350=y
> > > >  CONFIG_SPIKE=y
> > > >  CONFIG_SIFIVE_E=y
> > > >  CONFIG_SIFIVE_U=y
> > > > diff --git a/hw/riscv/Kconfig b/hw/riscv/Kconfig
> > > > index d139074b02..04f6369ab7 100644
> > > > --- a/hw/riscv/Kconfig
> > > > +++ b/hw/riscv/Kconfig
> > > > @@ -1,6 +1,13 @@
> > > >  config IBEX
> > > >      bool
> > > >
> > > > +config ANDES_AE350
> > >
> > > This needs to be sorted in alphabetical order
> > >
> >
> > Thanks, this will be fixed in V2.
> >
> > > > +    bool
> > > > +    select SERIAL
> > > > +    select VIRTIO_MMIO
> > > > +    select ANDES_PLIC
> > > > +    select ANDES_PLMT
> > > > +
> > > >  config MICROCHIP_PFSOC
> > > >      bool
> > > >      select CADENCE_SDHCI
> > > > diff --git a/hw/riscv/andes_ae350.c b/hw/riscv/andes_ae350.c
> > > > new file mode 100644
> > > > index 0000000000..ed5f9701ad
> > > > --- /dev/null
> > > > +++ b/hw/riscv/andes_ae350.c
> > > > @@ -0,0 +1,501 @@
> > > > +/*
> > > > + * Andes RISC-V AE350 Board
> > > > + *
> > > > + * Copyright (c) 2021 Andes Tech. Corp.
> > > > + *
> > > > + * Andes AE350 Board supports ns16550a UART and VirtIO MMIO.
> > > > + * The interrupt controllers are andes PLIC and andes PLICSW.
> > > > + * Timer is Andes PLMT.
> > > > + *
> > > > + * This program is free software; you can redistribute it and/or modify it
> > > > + * under the terms and conditions of the GNU General Public License,
> > > > + * version 2 or later, as published by the Free Software Foundation.
> > > > + *
> > > > + * This program is distributed in the hope it will be useful, but WITHOUT
> > > > + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
> > > > + * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
> > > > + * more details.
> > > > + *
> > > > + * You should have received a copy of the GNU General Public License along with
> > > > + * this program.  If not, see <http://www.gnu.org/licenses/>.
> > > > + */
> > > > +
> > > > +#include "qemu/osdep.h"
> > > > +#include "qemu/units.h"
> > > > +#include "qemu/log.h"
> > > > +#include "qemu/error-report.h"
> > > > +#include "qapi/error.h"
> > > > +#include "hw/boards.h"
> > > > +#include "hw/loader.h"
> > > > +#include "hw/sysbus.h"
> > > > +#include "hw/qdev-properties.h"
> > > > +#include "hw/char/serial.h"
> > > > +#include "target/riscv/cpu.h"
> > > > +#include "hw/riscv/riscv_hart.h"
> > > > +#include "hw/riscv/boot.h"
> > > > +#include "hw/riscv/numa.h"
> > > > +#include "chardev/char.h"
> > > > +#include "sysemu/arch_init.h"
> > > > +#include "sysemu/device_tree.h"
> > > > +#include "sysemu/sysemu.h"
> > > > +#include "hw/pci/pci.h"
> > > > +#include "hw/pci-host/gpex.h"
> > > > +
> > > > +#include "hw/intc/andes_plic.h"
> > > > +#include "hw/timer/andes_plmt.h"
> > > > +#include "hw/riscv/andes_ae350.h"
> > > > +
> > > > +# define BIOS_FILENAME ""
> > > > +
> > > > +static const struct MemmapEntry {
> > > > +    hwaddr base;
> > > > +    hwaddr size;
> > > > +} andes_ae350_memmap[] = {
> > > > +    [ANDES_AE350_DEBUG]     = { 0x00000000,      0x100 },
> > > > +    [ANDES_AE350_DRAM]      = { 0x00000000, 0x80000000 },
> > > > +    [ANDES_AE350_MROM]      = { 0xb0000000,   0x100000 },
> > > > +    [ANDES_AE350_MAC]       = { 0xe0100000,   0x100000 },
> > > > +    [ANDES_AE350_GEM]       = { 0xe0200000,   0x100000 },
> > > > +    [ANDES_AE350_PLIC]      = { 0xe4000000,   0x400000 },
> > > > +    [ANDES_AE350_PLMT]      = { 0xe6000000,   0x100000 },
> > > > +    [ANDES_AE350_PLICSW]    = { 0xe6400000,   0x400000 },
> > > > +    [ANDES_AE350_UART1]     = { 0xf0200000,      0x100 },
> > > > +    [ANDES_AE350_UART2]     = { 0xf0300000,      0x100 },
> > > > +    [ANDES_AE350_PIT]       = { 0xf0400000,   0x100000 },
> > > > +    [ANDES_AE350_SDC]       = { 0xf0e00000,   0x100000 },
> > > > +    [ANDES_AE350_VIRTIO]    = { 0xfe000000,     0x1000 },
> > > > +};
> > > > +
> > > > +static void
> > > > +create_fdt(AndesAe350BoardState *bs, const struct MemmapEntry *memmap,
> > > > +    uint64_t mem_size, const char *cmdline)
> > > > +{
> > > > +    AndesAe350SocState *s = &bs->soc;
> > > > +    MachineState *ms = MACHINE(qdev_get_machine());
> > > > +    void *fdt;
> > > > +    int cpu, i;
> > > > +    uint64_t mem_addr;
> > > > +    uint32_t *plic_irq_ext, *plicsw_irq_ext, *plmt_irq_ext;
> > > > +    unsigned long plic_addr, plicsw_addr, plmt_addr;
> > > > +    char *plic_name, *plicsw_name, *plmt_name;
> > > > +    uint32_t intc_phandle = 0, plic_phandle = 0;
> > > > +    uint32_t phandle = 1;
> > > > +    char *isa_name, *mem_name, *cpu_name, *intc_name, *uart_name, *virtio_name;
> > > > +
> > > > +    if (ms->dtb) {
> > > > +        fdt = bs->fdt = load_device_tree(ms->dtb, &bs->fdt_size);
> > > > +        if (!fdt) {
> > > > +            error_report("load_device_tree() failed");
> > > > +            exit(1);
> > > > +        }
> > > > +        goto update_bootargs;
> > > > +    } else {
> > > > +        fdt = bs->fdt = create_device_tree(&bs->fdt_size);
> > > > +        if (!fdt) {
> > > > +            error_report("create_device_tree() failed");
> > > > +            exit(1);
> > > > +        }
> > > > +    }
> > > > +
> > > > +    qemu_fdt_setprop_string(fdt, "/", "model", "Andes AE350 Board");
> > > > +    qemu_fdt_setprop_string(fdt, "/", "compatible", "andestech,ae350");
> > > > +    qemu_fdt_setprop_cell(fdt, "/", "#size-cells", 0x2);
> > > > +    qemu_fdt_setprop_cell(fdt, "/", "#address-cells", 0x2);
> > > > +
> > > > +    qemu_fdt_add_subnode(fdt, "/soc");
> > > > +    qemu_fdt_setprop(fdt, "/soc", "ranges", NULL, 0);
> > > > +    qemu_fdt_setprop_string(fdt, "/soc", "compatible", "simple-bus");
> > > > +    qemu_fdt_setprop_cell(fdt, "/soc", "#size-cells", 0x2);
> > > > +    qemu_fdt_setprop_cell(fdt, "/soc", "#address-cells", 0x2);
> > > > +
> > > > +    qemu_fdt_add_subnode(fdt, "/cpus");
> > > > +    qemu_fdt_setprop_cell(fdt, "/cpus", "timebase-frequency",
> > > > +                          ANDES_PLMT_TIMEBASE_FREQ);
> > > > +    qemu_fdt_setprop_cell(fdt, "/cpus", "#size-cells", 0x0);
> > > > +    qemu_fdt_setprop_cell(fdt, "/cpus", "#address-cells", 0x1);
> > > > +    qemu_fdt_add_subnode(fdt, "/cpus/cpu-map");
> > > > +
> > > > +    plic_irq_ext = g_new0(uint32_t, s->cpus.num_harts * 4);
> > > > +    plicsw_irq_ext = g_new0(uint32_t, s->cpus.num_harts * 2);
> > > > +    plmt_irq_ext = g_new0(uint32_t, s->cpus.num_harts * 2);
> > > > +
> > > > +    for (cpu = 0; cpu < s->cpus.num_harts; cpu++) {
> > > > +        intc_phandle = phandle++;
> > > > +
> > > > +        cpu_name = g_strdup_printf("/cpus/cpu@%d",
> > > > +            s->cpus.hartid_base + cpu);
> > > > +        qemu_fdt_add_subnode(fdt, cpu_name);
> > > > +#if defined(TARGET_RISCV32)
> > > > +        qemu_fdt_setprop_string(fdt, cpu_name, "mmu-type", "riscv,sv32");
> > > > +#else
> > > > +        qemu_fdt_setprop_string(fdt, cpu_name, "mmu-type", "riscv,sv39");
> > > > +#endif
> > > > +        isa_name = riscv_isa_string(&s->cpus.harts[cpu]);
> > > > +        qemu_fdt_setprop_string(fdt, cpu_name, "riscv,isa", isa_name);
> > > > +        g_free(isa_name);
> > > > +        qemu_fdt_setprop_string(fdt, cpu_name, "compatible", "riscv");
> > > > +        qemu_fdt_setprop_string(fdt, cpu_name, "status", "okay");
> > > > +        qemu_fdt_setprop_cell(fdt, cpu_name, "reg",
> > > > +            s->cpus.hartid_base + cpu);
> > > > +        qemu_fdt_setprop_string(fdt, cpu_name, "device_type", "cpu");
> > > > +
> > > > +        intc_name = g_strdup_printf("%s/interrupt-controller", cpu_name);
> > > > +        qemu_fdt_add_subnode(fdt, intc_name);
> > > > +        qemu_fdt_setprop_cell(fdt, intc_name, "phandle", intc_phandle);
> > > > +        qemu_fdt_setprop_string(fdt, intc_name, "compatible",
> > > > +            "riscv,cpu-intc");
> > > > +        qemu_fdt_setprop(fdt, intc_name, "interrupt-controller", NULL, 0);
> > > > +        qemu_fdt_setprop_cell(fdt, intc_name, "#interrupt-cells", 1);
> > > > +
> > > > +        plic_irq_ext[cpu * 4 + 0] = cpu_to_be32(intc_phandle);
> > > > +        plic_irq_ext[cpu * 4 + 1] = cpu_to_be32(IRQ_M_EXT);
> > > > +        plic_irq_ext[cpu * 4 + 2] = cpu_to_be32(intc_phandle);
> > > > +        plic_irq_ext[cpu * 4 + 3] = cpu_to_be32(IRQ_S_EXT);
> > > > +
> > > > +        plicsw_irq_ext[cpu * 2 + 0] = cpu_to_be32(intc_phandle);
> > > > +        plicsw_irq_ext[cpu * 2 + 1] = cpu_to_be32(IRQ_M_SOFT);
> > > > +
> > > > +        plmt_irq_ext[cpu * 2 + 0] = cpu_to_be32(intc_phandle);
> > > > +        plmt_irq_ext[cpu * 2 + 1] = cpu_to_be32(IRQ_M_TIMER);
> > > > +
> > > > +        g_free(intc_name);
> > > > +    }
> > > > +
> > > > +    mem_addr = memmap[ANDES_AE350_DRAM].base;
> > > > +    mem_name = g_strdup_printf("/memory@%lx", (long)mem_addr);
> > > > +    qemu_fdt_add_subnode(fdt, mem_name);
> > > > +    qemu_fdt_setprop_cells(fdt, mem_name, "reg",
> > > > +        mem_addr >> 32, mem_addr, mem_size >> 32, mem_size);
> > > > +    qemu_fdt_setprop_string(fdt, mem_name, "device_type", "memory");
> > > > +    g_free(mem_name);
> > > > +
> > > > +    /* create plic */
> > > > +    plic_phandle = phandle++;
> > > > +    plic_addr = memmap[ANDES_AE350_PLIC].base;
> > > > +    plic_name = g_strdup_printf("/soc/interrupt-controller@%lx", plic_addr);
> > > > +    qemu_fdt_add_subnode(fdt, plic_name);
> > > > +    qemu_fdt_setprop_cell(fdt, plic_name,
> > > > +        "#address-cells", 0x2);
> > > > +    qemu_fdt_setprop_cell(fdt, plic_name,
> > > > +        "#interrupt-cells", 0x2);
> > > > +    qemu_fdt_setprop_string(fdt, plic_name, "compatible", "riscv,plic0");
> > >
> > > This suggests PLIC is the same as the SiFive one. So why do we have a
> > > different implementation of the PLIC model?
> > >
> >
> > The difference of these two PLICs, please refer to my reply of "[PATCH 1/3] Andes RISC-V PLIC"
> >
> > > > +    qemu_fdt_setprop(fdt, plic_name, "interrupt-controller", NULL, 0);
> > > > +    qemu_fdt_setprop(fdt, plic_name, "interrupts-extended",
> > > > +        plic_irq_ext, s->cpus.num_harts * sizeof(uint32_t) * 4);
> > > > +    qemu_fdt_setprop_cells(fdt, plic_name, "reg",
> > > > +        0x0, plic_addr, 0x0, memmap[ANDES_AE350_PLIC].size);
> > > > +    qemu_fdt_setprop_cell(fdt, plic_name, "riscv,ndev", 0x47);
> > > > +    qemu_fdt_setprop_cell(fdt, plic_name, "phandle", plic_phandle);
> > > > +    g_free(plic_name);
> > > > +    g_free(plic_irq_ext);
> > > > +
> > > > +    /* create plicsw */
> > > > +    plicsw_addr = memmap[ANDES_AE350_PLICSW].base;
> > > > +    plicsw_name = g_strdup_printf("/soc/interrupt-controller@%lx", plicsw_addr);
> > > > +    qemu_fdt_add_subnode(fdt, plicsw_name);
> > > > +    qemu_fdt_setprop_cell(fdt, plicsw_name,
> > > > +        "#address-cells", 0x2);
> > > > +    qemu_fdt_setprop_cell(fdt, plicsw_name,
> > > > +        "#interrupt-cells", 0x2);
> > > > +    qemu_fdt_setprop_string(fdt, plicsw_name, "compatible", "riscv,plic1");
> > >
> > > Is this bindings in the Linux kernel upstream? I can't find any
> > > reference in the kernel tree.
> > >
> >
> > Currently only supports andes Linux BSP,
> > we have plans to push to linux upstream in the future.
>
> This raises a good question about device tree bindings being generated
> from QEMU.
>
> Do we want QEMU to generate a DT that doesn't have matching bindings?

I'd prefer new bindings are sent to kernel upstream to get it reviewed
there first. It might not just be the compatible string, but some
other properties that may change.

>
> I guess if the bindings change during the upstream process we can just
> add a second compatible string.
>
> Also if there is no upstream support how can I test the board?
>

Regards,
Bin


  reply	other threads:[~2021-03-12  1:09 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-03-10  3:33 [PATCH 0/3] Support Andes AE350 Platform Dylan Jhong
2021-03-10  3:33 ` Dylan Jhong
2021-03-10  3:33 ` [PATCH 1/3] Andes RISC-V PLIC Dylan Jhong
2021-03-10  3:33   ` Dylan Jhong
2021-03-10  6:05   ` Bin Meng
2021-03-10  6:05     ` Bin Meng
2021-03-11  6:52     ` Dylan Jhong
2021-03-11  6:52       ` Dylan Jhong
2021-03-10  7:50   ` Yixun Lan
2021-03-11 15:42     ` Alistair Francis
2021-03-11 15:42       ` Alistair Francis
2021-03-10  3:33 ` [PATCH 2/3] Andes RISC-V PLMT Dylan Jhong
2021-03-10  3:33   ` Dylan Jhong
2021-03-10  3:33 ` [PATCH 3/3] Andes AE350 RISC-V Machine Dylan Jhong
2021-03-10  3:33   ` Dylan Jhong
2021-03-10  6:15   ` Bin Meng
2021-03-10  6:15     ` Bin Meng
2021-03-11  6:50     ` Dylan Jhong
2021-03-11  6:50       ` Dylan Jhong
2021-03-11 15:46       ` Alistair Francis
2021-03-11 15:46         ` Alistair Francis
2021-03-12  1:07         ` Bin Meng [this message]
2021-03-12  1:07           ` Bin Meng

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=CAEUhbmXkf3k6XN2JNYahA+TvoeQgp034mNFAQmJd+A++nkO+RQ@mail.gmail.com \
    --to=bmeng.cn@gmail.com \
    --cc=Alistair.Francis@wdc.com \
    --cc=alankao@andestech.com \
    --cc=alistair23@gmail.com \
    --cc=dylan@andestech.com \
    --cc=kbastian@mail.uni-paderborn.de \
    --cc=palmer@dabbelt.com \
    --cc=qemu-devel@nongnu.org \
    --cc=qemu-riscv@nongnu.org \
    --cc=ruinland@andestech.com \
    --cc=sagark@eecs.berkeley.edu \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.