All of lore.kernel.org
 help / color / mirror / Atom feed
From: Alistair Francis <alistair23@gmail.com>
To: Damien Hedde <damien.hedde@greensocs.com>
Cc: "qemu-devel@nongnu.org Developers" <qemu-devel@nongnu.org>,
	"Peter Maydell" <peter.maydell@linaro.org>,
	"Michael S. Tsirkin" <mst@redhat.com>,
	"David Hildenbrand" <david@redhat.com>,
	"Peter Xu" <peterx@redhat.com>,
	mirela.grujic@greensocs.com,
	"Alistair Francis" <Alistair.Francis@wdc.com>,
	"Gerd Hoffmann" <kraxel@redhat.com>,
	"Ani Sinha" <ani@anisinha.ca>, "Eric Blake" <eblake@redhat.com>,
	"Stefano Stabellini" <sstabellini@kernel.org>,
	"open list:X86" <xen-devel@lists.xenproject.org>,
	"Paul Durrant" <paul@xen.org>,
	"Markus Armbruster" <armbru@redhat.com>,
	"Anthony Perard" <anthony.perard@citrix.com>,
	"Marc-André Lureau" <marcandre.lureau@redhat.com>,
	"Eduardo Habkost" <ehabkost@redhat.com>,
	"Dr. David Alan Gilbert" <dgilbert@redhat.com>,
	"Eric Auger" <eric.auger@redhat.com>,
	"Paolo Bonzini" <pbonzini@redhat.com>,
	"open list:RISC-V" <qemu-riscv@nongnu.org>,
	"Daniel P. Berrangé" <berrange@redhat.com>,
	"Mark Burton" <mark.burton@greensocs.com>,
	"Edgar Iglesias" <edgari@xilinx.com>,
	"Igor Mammedov" <imammedo@redhat.com>
Subject: Re: [RFC PATCH v2 12/16] add x-sysbus-mmio-map qmp command
Date: Wed, 13 Oct 2021 17:16:03 +1000	[thread overview]
Message-ID: <CAKmqyKPWDL52kC23TgtpdEST_kvVjNWn3-j2YFORN457_cEwnw@mail.gmail.com> (raw)
In-Reply-To: <20210922161405.140018-13-damien.hedde@greensocs.com>

On Thu, Sep 23, 2021 at 2:26 AM Damien Hedde <damien.hedde@greensocs.com> wrote:
>
> This command allows to map an mmio region of sysbus device onto
> the system memory. Its behavior mimics the sysbus_mmio_map()
> function apart from the automatic unmap (the C function unmaps
> the region if it is already mapped).
> For the qmp function we consider it is an error to try to map
> an already mapped function. If unmapping is required, it is
> probably better to add a sysbus-mmip-unmap function.
>
> This command is still experimental (hence the x prefix), as it
> is related to the sysbus device creation through qmp commands.
>
> In future, we expect to have to handle the overlap/priority
> parameter but also multiple mapping of one mmio. For some
> devices, one mmio is mapped several times at different addresses on
> the bus (which is not supported by sysbus_mmio_map() function and
> requires the use of memory region aliases).

I think as is this is a good start. This is a useful feature!

>
> Signed-off-by: Damien Hedde <damien.hedde@greensocs.com>

Acked-by: Alistair Francis <alistair.francis@wdc.com>

> ---
>
> Note: this qmp command is required to be able to build a machine from
> scratch as there is no qmp-way of doing a memory mapping today.
>
> We've added the command into qapi/qdev.json section. It does not seem to
> have any really adequate section yet. Any idea ? should we create for
> example a new one: qapi/sysbus.json or qapi/memory.json ?

I would say leave it in qdev. We don't really want many more sysbus
commands, so qapi/sysbus.json doesn't need it's own file.

Alistair

> ---
>  qapi/qdev.json   | 21 +++++++++++++++++++++
>  hw/core/sysbus.c | 41 +++++++++++++++++++++++++++++++++++++++++
>  2 files changed, 62 insertions(+)
>
> diff --git a/qapi/qdev.json b/qapi/qdev.json
> index ad669ae175..dfc1104197 100644
> --- a/qapi/qdev.json
> +++ b/qapi/qdev.json
> @@ -125,3 +125,24 @@
>  ##
>  { 'event': 'DEVICE_DELETED',
>    'data': { '*device': 'str', 'path': 'str' } }
> +
> +##
> +# @x-sysbus-mmio-map:
> +#
> +# Map a sysbus device mmio onto the main system bus.
> +#
> +# @device: the device's QOM path
> +#
> +# @mmio: The mmio number to be mapped (defaults to 0).
> +#
> +# @addr: The base address for the mapping.
> +#
> +# Since: 6.2
> +#
> +# Returns: Nothing on success
> +#
> +##
> +
> +{ 'command': 'x-sysbus-mmio-map',
> +  'data': {'device': 'str', '*mmio': 'uint8', 'addr': 'uint64'},
> +  'allow-preconfig' : true }
> diff --git a/hw/core/sysbus.c b/hw/core/sysbus.c
> index aaae8e23cc..b0891f37b6 100644
> --- a/hw/core/sysbus.c
> +++ b/hw/core/sysbus.c
> @@ -23,6 +23,7 @@
>  #include "hw/sysbus.h"
>  #include "monitor/monitor.h"
>  #include "exec/address-spaces.h"
> +#include "qapi/qapi-commands-qdev.h"
>
>  static void sysbus_dev_print(Monitor *mon, DeviceState *dev, int indent);
>  static char *sysbus_get_fw_dev_path(DeviceState *dev);
> @@ -154,6 +155,46 @@ static void sysbus_mmio_map_common(SysBusDevice *dev, int n, hwaddr addr,
>      }
>  }
>
> +void qmp_x_sysbus_mmio_map(const char *device, bool has_mmio, uint8_t mmio,
> +                           uint64_t addr, Error **errp)
> +{
> +    Object *obj = object_resolve_path_type(device, TYPE_SYS_BUS_DEVICE, NULL);
> +    SysBusDevice *dev;
> +
> +    if (phase_get() != MACHINE_INIT_PHASE_INITIALIZED) {
> +        error_setg(errp, "The command is permitted only when "
> +                         "the machine is in initialized phase");
> +        return;
> +    }
> +
> +    if (obj == NULL) {
> +        error_setg(errp, "Device '%s' not found", device);
> +        return;
> +    }
> +
> +    dev = SYS_BUS_DEVICE(obj);
> +    if (!has_mmio) {
> +        mmio = 0;
> +    }
> +    if (mmio >= dev->num_mmio) {
> +        error_setg(errp, "MMIO index '%u' is out of range", mmio);
> +        return;
> +    }
> +
> +    if (dev->mmio[mmio].addr != (hwaddr)-1) {
> +        error_setg(errp, "MMIO index '%u' is already mapped", mmio);
> +        return;
> +    }
> +
> +    if (!memory_region_try_add_subregion(get_system_memory(), addr,
> +                                         dev->mmio[mmio].memory, 0,
> +                                         errp)) {
> +        return;
> +    }
> +
> +    dev->mmio[mmio].addr = addr;
> +}
> +
>  void sysbus_mmio_unmap(SysBusDevice *dev, int n)
>  {
>      assert(n >= 0 && n < dev->num_mmio);
> --
> 2.33.0
>
>


WARNING: multiple messages have this Message-ID (diff)
From: Alistair Francis <alistair23@gmail.com>
To: Damien Hedde <damien.hedde@greensocs.com>
Cc: "Peter Maydell" <peter.maydell@linaro.org>,
	"David Hildenbrand" <david@redhat.com>,
	"qemu-devel@nongnu.org Developers" <qemu-devel@nongnu.org>,
	"Peter Xu" <peterx@redhat.com>,
	mirela.grujic@greensocs.com,
	"Alistair Francis" <Alistair.Francis@wdc.com>,
	"Gerd Hoffmann" <kraxel@redhat.com>,
	"Ani Sinha" <ani@anisinha.ca>, "Eric Blake" <eblake@redhat.com>,
	"Stefano Stabellini" <sstabellini@kernel.org>,
	"Michael S. Tsirkin" <mst@redhat.com>,
	"Markus Armbruster" <armbru@redhat.com>,
	"Anthony Perard" <anthony.perard@citrix.com>,
	"Marc-André Lureau" <marcandre.lureau@redhat.com>,
	"Igor Mammedov" <imammedo@redhat.com>,
	"Eduardo Habkost" <ehabkost@redhat.com>,
	"Dr. David Alan Gilbert" <dgilbert@redhat.com>,
	"Paul Durrant" <paul@xen.org>,
	"Eric Auger" <eric.auger@redhat.com>,
	"open list:X86" <xen-devel@lists.xenproject.org>,
	"open list:RISC-V" <qemu-riscv@nongnu.org>,
	"Daniel P. Berrangé" <berrange@redhat.com>,
	"Mark Burton" <mark.burton@greensocs.com>,
	"Edgar Iglesias" <edgari@xilinx.com>,
	"Paolo Bonzini" <pbonzini@redhat.com>
Subject: Re: [RFC PATCH v2 12/16] add x-sysbus-mmio-map qmp command
Date: Wed, 13 Oct 2021 17:16:03 +1000	[thread overview]
Message-ID: <CAKmqyKPWDL52kC23TgtpdEST_kvVjNWn3-j2YFORN457_cEwnw@mail.gmail.com> (raw)
In-Reply-To: <20210922161405.140018-13-damien.hedde@greensocs.com>

On Thu, Sep 23, 2021 at 2:26 AM Damien Hedde <damien.hedde@greensocs.com> wrote:
>
> This command allows to map an mmio region of sysbus device onto
> the system memory. Its behavior mimics the sysbus_mmio_map()
> function apart from the automatic unmap (the C function unmaps
> the region if it is already mapped).
> For the qmp function we consider it is an error to try to map
> an already mapped function. If unmapping is required, it is
> probably better to add a sysbus-mmip-unmap function.
>
> This command is still experimental (hence the x prefix), as it
> is related to the sysbus device creation through qmp commands.
>
> In future, we expect to have to handle the overlap/priority
> parameter but also multiple mapping of one mmio. For some
> devices, one mmio is mapped several times at different addresses on
> the bus (which is not supported by sysbus_mmio_map() function and
> requires the use of memory region aliases).

I think as is this is a good start. This is a useful feature!

>
> Signed-off-by: Damien Hedde <damien.hedde@greensocs.com>

Acked-by: Alistair Francis <alistair.francis@wdc.com>

> ---
>
> Note: this qmp command is required to be able to build a machine from
> scratch as there is no qmp-way of doing a memory mapping today.
>
> We've added the command into qapi/qdev.json section. It does not seem to
> have any really adequate section yet. Any idea ? should we create for
> example a new one: qapi/sysbus.json or qapi/memory.json ?

I would say leave it in qdev. We don't really want many more sysbus
commands, so qapi/sysbus.json doesn't need it's own file.

Alistair

> ---
>  qapi/qdev.json   | 21 +++++++++++++++++++++
>  hw/core/sysbus.c | 41 +++++++++++++++++++++++++++++++++++++++++
>  2 files changed, 62 insertions(+)
>
> diff --git a/qapi/qdev.json b/qapi/qdev.json
> index ad669ae175..dfc1104197 100644
> --- a/qapi/qdev.json
> +++ b/qapi/qdev.json
> @@ -125,3 +125,24 @@
>  ##
>  { 'event': 'DEVICE_DELETED',
>    'data': { '*device': 'str', 'path': 'str' } }
> +
> +##
> +# @x-sysbus-mmio-map:
> +#
> +# Map a sysbus device mmio onto the main system bus.
> +#
> +# @device: the device's QOM path
> +#
> +# @mmio: The mmio number to be mapped (defaults to 0).
> +#
> +# @addr: The base address for the mapping.
> +#
> +# Since: 6.2
> +#
> +# Returns: Nothing on success
> +#
> +##
> +
> +{ 'command': 'x-sysbus-mmio-map',
> +  'data': {'device': 'str', '*mmio': 'uint8', 'addr': 'uint64'},
> +  'allow-preconfig' : true }
> diff --git a/hw/core/sysbus.c b/hw/core/sysbus.c
> index aaae8e23cc..b0891f37b6 100644
> --- a/hw/core/sysbus.c
> +++ b/hw/core/sysbus.c
> @@ -23,6 +23,7 @@
>  #include "hw/sysbus.h"
>  #include "monitor/monitor.h"
>  #include "exec/address-spaces.h"
> +#include "qapi/qapi-commands-qdev.h"
>
>  static void sysbus_dev_print(Monitor *mon, DeviceState *dev, int indent);
>  static char *sysbus_get_fw_dev_path(DeviceState *dev);
> @@ -154,6 +155,46 @@ static void sysbus_mmio_map_common(SysBusDevice *dev, int n, hwaddr addr,
>      }
>  }
>
> +void qmp_x_sysbus_mmio_map(const char *device, bool has_mmio, uint8_t mmio,
> +                           uint64_t addr, Error **errp)
> +{
> +    Object *obj = object_resolve_path_type(device, TYPE_SYS_BUS_DEVICE, NULL);
> +    SysBusDevice *dev;
> +
> +    if (phase_get() != MACHINE_INIT_PHASE_INITIALIZED) {
> +        error_setg(errp, "The command is permitted only when "
> +                         "the machine is in initialized phase");
> +        return;
> +    }
> +
> +    if (obj == NULL) {
> +        error_setg(errp, "Device '%s' not found", device);
> +        return;
> +    }
> +
> +    dev = SYS_BUS_DEVICE(obj);
> +    if (!has_mmio) {
> +        mmio = 0;
> +    }
> +    if (mmio >= dev->num_mmio) {
> +        error_setg(errp, "MMIO index '%u' is out of range", mmio);
> +        return;
> +    }
> +
> +    if (dev->mmio[mmio].addr != (hwaddr)-1) {
> +        error_setg(errp, "MMIO index '%u' is already mapped", mmio);
> +        return;
> +    }
> +
> +    if (!memory_region_try_add_subregion(get_system_memory(), addr,
> +                                         dev->mmio[mmio].memory, 0,
> +                                         errp)) {
> +        return;
> +    }
> +
> +    dev->mmio[mmio].addr = addr;
> +}
> +
>  void sysbus_mmio_unmap(SysBusDevice *dev, int n)
>  {
>      assert(n >= 0 && n < dev->num_mmio);
> --
> 2.33.0
>
>


  reply	other threads:[~2021-10-13  7:16 UTC|newest]

Thread overview: 82+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-09-22 16:13 [RFC PATCH v2 00/16] Initial support for machine creation via QMP Damien Hedde
2021-09-22 16:13 ` Damien Hedde
2021-09-22 16:13 ` [RFC PATCH v2 01/16] rename MachineInitPhase enum constants for QAPI compatibility Damien Hedde
2021-09-22 16:13   ` Damien Hedde
2021-09-22 16:13 ` [RFC PATCH v2 02/16] qapi: Implement query-machine-phase QMP command Damien Hedde
2021-09-22 16:13   ` Damien Hedde
2021-09-22 17:37   ` Philippe Mathieu-Daudé
2021-09-22 17:37     ` Philippe Mathieu-Daudé
2021-09-23 12:43     ` Damien Hedde
2021-09-23 12:43       ` Damien Hedde
2021-09-23 13:46       ` Ani Sinha
2021-09-23 13:46         ` Ani Sinha
2021-10-12 22:08   ` Alistair Francis
2021-10-12 22:08     ` Alistair Francis
2021-09-22 16:13 ` [RFC PATCH v2 03/16] qapi: Implement x-machine-init " Damien Hedde
2021-09-22 16:13   ` Damien Hedde
2021-10-12 22:19   ` Alistair Francis
2021-10-12 22:19     ` Alistair Francis
2021-09-22 16:13 ` [RFC PATCH v2 04/16] softmmu/qdev-monitor: add error handling in qdev_set_id Damien Hedde
2021-09-22 16:13   ` Damien Hedde
2021-10-13  7:10   ` Alistair Francis
2021-10-13  7:10     ` Alistair Francis
2021-10-13 14:30     ` Damien Hedde
2021-10-13 14:30       ` Damien Hedde
2021-09-22 16:13 ` [RFC PATCH v2 05/16] qdev-monitor: prevent conflicts between qmp/device_add and cli/-device Damien Hedde
2021-09-22 16:13   ` Damien Hedde
2021-09-22 16:13 ` [RFC PATCH v2 06/16] qapi: Allow device_add to execute in machine initialized phase Damien Hedde
2021-09-22 16:13   ` Damien Hedde
2021-10-12 22:25   ` Alistair Francis
2021-10-12 22:25     ` Alistair Francis
2021-09-22 16:13 ` [RFC PATCH v2 07/16] hw/core/machine: add machine_class_is_dynamic_sysbus_dev_allowed Damien Hedde
2021-09-22 16:13   ` Damien Hedde
2021-09-23 10:51   ` Ani Sinha
2021-09-23 10:51     ` Ani Sinha
2021-09-23 13:07     ` Damien Hedde
2021-09-23 13:07       ` Damien Hedde
2021-09-22 16:13 ` [RFC PATCH v2 08/16] qdev-monitor: Check sysbus device type before creating it Damien Hedde
2021-09-22 16:13   ` Damien Hedde
2021-10-12 22:42   ` Alistair Francis
2021-10-12 22:42     ` Alistair Francis
2021-09-22 16:13 ` [RFC PATCH v2 09/16] hw/core/machine: Remove the dynamic sysbus devices type check Damien Hedde
2021-09-22 16:13   ` Damien Hedde
2021-10-12 23:07   ` Alistair Francis
2021-10-12 23:07     ` Alistair Francis
2021-09-22 16:13 ` [RFC PATCH v2 10/16] qdev-monitor: allow adding any sysbus device before machine is ready Damien Hedde
2021-09-22 16:13   ` Damien Hedde
2021-09-23 11:04   ` Ani Sinha
2021-09-23 11:04     ` Ani Sinha
2021-09-23 11:55     ` Ani Sinha
2021-09-23 11:55       ` Ani Sinha
2021-09-23 14:04       ` Damien Hedde
2021-09-23 14:04         ` Damien Hedde
2021-09-22 16:14 ` [RFC PATCH v2 11/16] softmmu/memory: add memory_region_try_add_subregion function Damien Hedde
2021-09-22 16:14   ` Damien Hedde
2021-09-22 17:42   ` Philippe Mathieu-Daudé
2021-09-22 17:42     ` Philippe Mathieu-Daudé
2021-09-22 17:49   ` David Hildenbrand
2021-09-22 17:49     ` David Hildenbrand
2021-10-13  7:12   ` Alistair Francis
2021-10-13  7:12     ` Alistair Francis
2021-09-22 16:14 ` [RFC PATCH v2 12/16] add x-sysbus-mmio-map qmp command Damien Hedde
2021-09-22 16:14   ` Damien Hedde
2021-10-13  7:16   ` Alistair Francis [this message]
2021-10-13  7:16     ` Alistair Francis
2021-09-22 16:14 ` [RFC PATCH v2 13/16] hw/mem/system-memory: add a memory sysbus device Damien Hedde
2021-09-22 16:14   ` Damien Hedde
2021-09-22 16:14 ` [RFC PATCH v2 14/16] docs/system: add doc about the initialized machine phase and an example Damien Hedde
2021-09-22 16:14   ` Damien Hedde
2021-09-22 16:14 ` [RFC PATCH v2 15/16] hw/char/ibex_uart: set user_creatable Damien Hedde
2021-09-22 16:14   ` Damien Hedde
2021-09-22 16:14 ` [RFC PATCH v2 16/16] hw/intc/ibex_plic: " Damien Hedde
2021-09-22 16:14   ` Damien Hedde
2021-09-22 17:50 ` [RFC PATCH v2 00/16] Initial support for machine creation via QMP Philippe Mathieu-Daudé
2021-09-22 17:50   ` Philippe Mathieu-Daudé
2021-09-23 11:08   ` Damien Hedde
2021-09-23 11:08     ` Damien Hedde
2021-10-04 15:56 ` Damien Hedde
2021-10-04 15:56   ` Damien Hedde
2021-10-12 22:16 ` Alistair Francis
2021-10-12 22:16   ` Alistair Francis
2021-10-13  5:56   ` Mark Burton
2021-10-13  5:56     ` Mark Burton

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=CAKmqyKPWDL52kC23TgtpdEST_kvVjNWn3-j2YFORN457_cEwnw@mail.gmail.com \
    --to=alistair23@gmail.com \
    --cc=Alistair.Francis@wdc.com \
    --cc=ani@anisinha.ca \
    --cc=anthony.perard@citrix.com \
    --cc=armbru@redhat.com \
    --cc=berrange@redhat.com \
    --cc=damien.hedde@greensocs.com \
    --cc=david@redhat.com \
    --cc=dgilbert@redhat.com \
    --cc=eblake@redhat.com \
    --cc=edgari@xilinx.com \
    --cc=ehabkost@redhat.com \
    --cc=eric.auger@redhat.com \
    --cc=imammedo@redhat.com \
    --cc=kraxel@redhat.com \
    --cc=marcandre.lureau@redhat.com \
    --cc=mark.burton@greensocs.com \
    --cc=mirela.grujic@greensocs.com \
    --cc=mst@redhat.com \
    --cc=paul@xen.org \
    --cc=pbonzini@redhat.com \
    --cc=peter.maydell@linaro.org \
    --cc=peterx@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=qemu-riscv@nongnu.org \
    --cc=sstabellini@kernel.org \
    --cc=xen-devel@lists.xenproject.org \
    /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.