[PATCH v4] virtio-mmio: improve virtio-mmio get_dev_path alog

schspa posted 1 patch 5 months, 1 week ago
Test checkpatch passed
Patches applied successfully (tree, apply log)
git fetch https://github.com/patchew-project/qemu tags/patchew/20210225053606.385884-1-schspa@gmail.com
Maintainers: "Michael S. Tsirkin" <mst@redhat.com>
hw/virtio/virtio-mmio.c | 13 +++++++------
1 file changed, 7 insertions(+), 6 deletions(-)

[PATCH v4] virtio-mmio: improve virtio-mmio get_dev_path alog

Posted by schspa 5 months, 1 week ago
At the moment the following QEMU command line triggers an assertion
failure On xlnx-versal SOC:
  qemu-system-aarch64 \
      -machine xlnx-versal-virt -nographic -smp 2 -m 128 \
      -fsdev local,id=shareid,path=${HOME}/work,security_model=none \
      -device virtio-9p-device,fsdev=shareid,mount_tag=share \
      -fsdev local,id=shareid1,path=${HOME}/Music,security_model=none \
      -device virtio-9p-device,fsdev=shareid1,mount_tag=share1

  qemu-system-aarch64: ../migration/savevm.c:860:
  vmstate_register_with_alias_id:
  Assertion `!se->compat || se->instance_id == 0' failed.

This problem was fixed on arm virt platform in commit f58b39d2d5b
("virtio-mmio: format transport base address in BusClass.get_dev_path")

It works perfectly on arm virt platform. but there is still there on
xlnx-versal SOC.

The main difference between arm virt and xlnx-versal is they use
different way to create virtio-mmio qdev. on arm virt, it calls
sysbus_create_simple("virtio-mmio", base, pic[irq]); which will call
sysbus_mmio_map internally and assign base address to subsys device
mmio correctly. but xlnx-versal's implements won't do this.

However, xlnx-versal can't switch to sysbus_create_simple() to create
virtio-mmio device. It's because xlnx-versal's cpu use
VersalVirt.soc.fpd.apu.mr as it's memory. which is subregion of
system_memory. sysbus_create_simple will add virtio to system_memory,
which can't be accessed by cpu.

Besides, xlnx-versal can't add sysbus_mmio_map api call too, because
this will add memory region to system_memory, and it can't be added
to VersalVirt.soc.fpd.apu.mr again.

We can solve this by assign correct base address offset on dev_path.

This path was test on aarch64 virt & xlnx-versal platform.

Signed-off-by: schspa <schspa@gmail.com>
---
 hw/virtio/virtio-mmio.c | 13 +++++++------
 1 file changed, 7 insertions(+), 6 deletions(-)

diff --git a/hw/virtio/virtio-mmio.c b/hw/virtio/virtio-mmio.c
index 610661d6a5..6990b9879c 100644
--- a/hw/virtio/virtio-mmio.c
+++ b/hw/virtio/virtio-mmio.c
@@ -737,8 +737,8 @@ static char *virtio_mmio_bus_get_dev_path(DeviceState *dev)
     BusState *virtio_mmio_bus;
     VirtIOMMIOProxy *virtio_mmio_proxy;
     char *proxy_path;
-    SysBusDevice *proxy_sbd;
     char *path;
+    MemoryRegionSection section;
 
     virtio_mmio_bus = qdev_get_parent_bus(dev);
     virtio_mmio_proxy = VIRTIO_MMIO(virtio_mmio_bus->parent);
@@ -757,17 +757,18 @@ static char *virtio_mmio_bus_get_dev_path(DeviceState *dev)
     }
 
     /* Otherwise, we append the base address of the transport. */
-    proxy_sbd = SYS_BUS_DEVICE(virtio_mmio_proxy);
-    assert(proxy_sbd->num_mmio == 1);
-    assert(proxy_sbd->mmio[0].memory == &virtio_mmio_proxy->iomem);
+    section = memory_region_find(&virtio_mmio_proxy->iomem, 0, 0x200);
+    assert(section.mr);
 
     if (proxy_path) {
         path = g_strdup_printf("%s/virtio-mmio@" TARGET_FMT_plx, proxy_path,
-                               proxy_sbd->mmio[0].addr);
+                               section.offset_within_address_space);
     } else {
         path = g_strdup_printf("virtio-mmio@" TARGET_FMT_plx,
-                               proxy_sbd->mmio[0].addr);
+                               section.offset_within_address_space);
     }
+    memory_region_unref(section.mr);
+
     g_free(proxy_path);
     return path;
 }
-- 
2.30.1


Re: [PATCH v4] virtio-mmio: improve virtio-mmio get_dev_path alog

Posted by Peter Maydell 4 months, 4 weeks ago
On Thu, 25 Feb 2021 at 05:36, schspa <schspa@gmail.com> wrote:
>
> At the moment the following QEMU command line triggers an assertion
> failure On xlnx-versal SOC:
>   qemu-system-aarch64 \
>       -machine xlnx-versal-virt -nographic -smp 2 -m 128 \
>       -fsdev local,id=shareid,path=${HOME}/work,security_model=none \
>       -device virtio-9p-device,fsdev=shareid,mount_tag=share \
>       -fsdev local,id=shareid1,path=${HOME}/Music,security_model=none \
>       -device virtio-9p-device,fsdev=shareid1,mount_tag=share1
>
>   qemu-system-aarch64: ../migration/savevm.c:860:
>   vmstate_register_with_alias_id:
>   Assertion `!se->compat || se->instance_id == 0' failed.
>
> This problem was fixed on arm virt platform in commit f58b39d2d5b
> ("virtio-mmio: format transport base address in BusClass.get_dev_path")
>
> It works perfectly on arm virt platform. but there is still there on
> xlnx-versal SOC.
>
> The main difference between arm virt and xlnx-versal is they use
> different way to create virtio-mmio qdev. on arm virt, it calls
> sysbus_create_simple("virtio-mmio", base, pic[irq]); which will call
> sysbus_mmio_map internally and assign base address to subsys device
> mmio correctly. but xlnx-versal's implements won't do this.
>
> However, xlnx-versal can't switch to sysbus_create_simple() to create
> virtio-mmio device. It's because xlnx-versal's cpu use
> VersalVirt.soc.fpd.apu.mr as it's memory. which is subregion of
> system_memory. sysbus_create_simple will add virtio to system_memory,
> which can't be accessed by cpu.
>
> Besides, xlnx-versal can't add sysbus_mmio_map api call too, because
> this will add memory region to system_memory, and it can't be added
> to VersalVirt.soc.fpd.apu.mr again.
>
> We can solve this by assign correct base address offset on dev_path.
>
> This path was test on aarch64 virt & xlnx-versal platform.
>
> Signed-off-by: schspa <schspa@gmail.com>
> ---

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>

Applied to target-arm.next, thanks (unless MST would rather it
go in via another route).

-- PMM

Re: [PATCH v4] virtio-mmio: improve virtio-mmio get_dev_path alog

Posted by Shi Schspa 4 months, 4 weeks ago
Thank you very much.

Best regards.


Peter Maydell <peter.maydell@linaro.org> 于2021年3月5日周五 下午7:57写道:

> On Thu, 25 Feb 2021 at 05:36, schspa <schspa@gmail.com> wrote:
> >
> > At the moment the following QEMU command line triggers an assertion
> > failure On xlnx-versal SOC:
> >   qemu-system-aarch64 \
> >       -machine xlnx-versal-virt -nographic -smp 2 -m 128 \
> >       -fsdev local,id=shareid,path=${HOME}/work,security_model=none \
> >       -device virtio-9p-device,fsdev=shareid,mount_tag=share \
> >       -fsdev local,id=shareid1,path=${HOME}/Music,security_model=none \
> >       -device virtio-9p-device,fsdev=shareid1,mount_tag=share1
> >
> >   qemu-system-aarch64: ../migration/savevm.c:860:
> >   vmstate_register_with_alias_id:
> >   Assertion `!se->compat || se->instance_id == 0' failed.
> >
> > This problem was fixed on arm virt platform in commit f58b39d2d5b
> > ("virtio-mmio: format transport base address in BusClass.get_dev_path")
> >
> > It works perfectly on arm virt platform. but there is still there on
> > xlnx-versal SOC.
> >
> > The main difference between arm virt and xlnx-versal is they use
> > different way to create virtio-mmio qdev. on arm virt, it calls
> > sysbus_create_simple("virtio-mmio", base, pic[irq]); which will call
> > sysbus_mmio_map internally and assign base address to subsys device
> > mmio correctly. but xlnx-versal's implements won't do this.
> >
> > However, xlnx-versal can't switch to sysbus_create_simple() to create
> > virtio-mmio device. It's because xlnx-versal's cpu use
> > VersalVirt.soc.fpd.apu.mr as it's memory. which is subregion of
> > system_memory. sysbus_create_simple will add virtio to system_memory,
> > which can't be accessed by cpu.
> >
> > Besides, xlnx-versal can't add sysbus_mmio_map api call too, because
> > this will add memory region to system_memory, and it can't be added
> > to VersalVirt.soc.fpd.apu.mr again.
> >
> > We can solve this by assign correct base address offset on dev_path.
> >
> > This path was test on aarch64 virt & xlnx-versal platform.
> >
> > Signed-off-by: schspa <schspa@gmail.com>
> > ---
>
> Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
>
> Applied to target-arm.next, thanks (unless MST would rather it
> go in via another route).
>
> -- PMM
>