diff mbox series

[Virtual,ppce500] virtio_gpu virtio0: swiotlb buffer is full

Message ID 3162da18-462c-72b4-f8f0-eef896c6b162@xenosoft.de (mailing list archive)
State Handled Elsewhere
Headers show
Series [Virtual,ppce500] virtio_gpu virtio0: swiotlb buffer is full | expand

Checks

Context Check Description
snowpatch_ozlabs/apply_patch warning Failed to apply on branch powerpc/merge (f04b169e953c4db1a3a3c1d23eea09c726f01ee5)
snowpatch_ozlabs/apply_patch warning Failed to apply on branch powerpc/next (a7aaa2f26bfd932a654706b19859e7adf802bee2)
snowpatch_ozlabs/apply_patch warning Failed to apply on branch linus/master (fb893de323e2d39f7a1f6df425703a2edbdf56ea)
snowpatch_ozlabs/apply_patch warning Failed to apply on branch powerpc/fixes (6553fb799f601497ca0703682e2aff131197dc5c)
snowpatch_ozlabs/apply_patch warning Failed to apply on branch linux-next (bc09acc9f224c0923794b69534c9bfa999c27b64)
snowpatch_ozlabs/apply_patch fail Failed to apply to any branch

Commit Message

Christian Zigotzky Aug. 12, 2020, 1:09 p.m. UTC 
Hello Daniel,

The VirtIO-GPU doesn't work anymore with the latest Git kernel in a 
virtual e5500 PPC64 QEMU machine [1,2] after the commit "drm/virtio: 
Call the right shmem helpers". [3]
The kernel 5.8 works with the VirtIO-GPU in this virtual machine.

I bisected today [4].

Result: drm/virtio: Call the right shmem helpers ( 
d323bb44e4d23802eb25d13de1f93f2335bd60d0) [3] is the first bad commit.

I was able to revert the first bad commit. [5] After that I compiled a 
new kernel again. Then I was able to boot Linux with this kernel in a 
virtual e5500 PPC64 QEMU machine with the VirtIO-GPU.

I created a patch. [6] With this patch I can use the VirtIO-GPU again.

Could you please check the first bad commit?

Thanks,
Christian

[1] QEMU command: qemu-system-ppc64 -M ppce500 -cpu e5500 -enable-kvm -m 
1024 -kernel uImage -drive 
format=raw,file=fienix-soar_3.0-2020608-net.img,index=0,if=virtio -nic 
user,model=e1000 -append "rw root=/dev/vda2" -device virtio-vga -device 
virtio-mouse-pci -device virtio-keyboard-pci -device pci-ohci,id=newusb 
-device usb-audio,bus=newusb.0 -smp 4

[2] Error messages:

virtio_gpu virtio0: swiotlb buffer is full (sz: 4096 bytes), total 0 
(slots), used 0 (slots)
BUG: Kernel NULL pointer dereference on read at 0x00000010
Faulting instruction address: 0xc0000000000c7324
Oops: Kernel access of bad area, sig: 11 [#1]
BE PAGE_SIZE=4K PREEMPT SMP NR_CPUS=4 QEMU e500
Modules linked in:
CPU: 2 PID: 1678 Comm: kworker/2:2 Not tainted 
5.9-a3_A-EON_X5000-11735-g06a81c1c7db9-dirty #1
Workqueue: events .virtio_gpu_dequeue_ctrl_func
NIP:  c0000000000c7324 LR: c0000000000c72e4 CTR: c000000000462930
REGS: c00000003dba75e0 TRAP: 0300   Not tainted 
(5.9-a3_A-EON_X5000-11735-g06a81c1c7db9-dirty)
MSR:  0000000090029000 <CE,EE,ME>  CR: 24002288  XER: 00000000
DEAR: 0000000000000010 ESR: 0000000000000000 IRQMASK: 0
GPR00: c0000000000c6188 c00000003dba7870 c0000000017f2300 c00000003d893010
GPR04: 0000000000000000 0000000000000001 0000000000000000 0000000000000000
GPR08: 0000000000000000 0000000000000000 0000000000000000 7f7f7f7f7f7f7f7f
GPR12: 0000000024002284 c00000003fff9200 c00000000008c3a0 c0000000061566c0
GPR16: 0000000000000000 0000000000000000 0000000000000000 0000000000000000
GPR20: 0000000000000000 0000000000000000 0000000000000000 0000000000000000
GPR24: 0000000000000001 0000000000110000 0000000000000000 0000000000000000
GPR28: c00000003d893010 0000000000000000 0000000000000000 c00000003d893010
NIP [c0000000000c7324] .dma_direct_unmap_sg+0x4c/0xd8
LR [c0000000000c72e4] .dma_direct_unmap_sg+0xc/0xd8
Call Trace:
[c00000003dba7870] [c00000003dba7950] 0xc00000003dba7950 (unreliable)
[c00000003dba7920] [c0000000000c6188] .dma_unmap_sg_attrs+0x5c/0x98
[c00000003dba79d0] [c0000000005cd438] .drm_gem_shmem_free_object+0x98/0xcc
[c00000003dba7a50] [c0000000006af5b4] .virtio_gpu_cleanup_object+0xc8/0xd4
[c00000003dba7ad0] [c0000000006ad3bc] .virtio_gpu_cmd_unref_cb+0x1c/0x30
[c00000003dba7b40] [c0000000006adab8] 
.virtio_gpu_dequeue_ctrl_func+0x208/0x28c
[c00000003dba7c10] [c000000000086b70] .process_one_work+0x1a4/0x258
[c00000003dba7cb0] [c0000000000870f4] .worker_thread+0x214/0x284
[c00000003dba7d70] [c00000000008c4f0] .kthread+0x150/0x158
[c00000003dba7e20] [c00000000000082c] .ret_from_kernel_thread+0x58/0x60
Instruction dump:
f821ff51 7cb82b78 7cdb3378 4e000000 7cfa3b78 3bc00000 7f9ec000 41fc0014
382100b0 81810008 7d808120 48bc1ba8 <e93d0010> ebfc0248 833d0018 7fff4850
---[ end trace f28d194d9f0955a8 ]---

virtio_gpu virtio0: swiotlb buffer is full (sz: 4096 bytes), total 0 
(slots), used 0 (slots)
virtio_gpu virtio0: swiotlb buffer is full (sz: 16384 bytes), total 0 
(slots), used 0 (slots)

---

[3] 
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=d323bb44e4d23802eb25d13de1f93f2335bd60d0

[4] https://forum.hyperion-entertainment.com/viewtopic.php?p=51377#p51377

[5] git revert d323bb44e4d23802eb25d13de1f93f2335bd60d0 //Output: 
[master 966950f724e4] Revert "drm/virtio: Call the right shmem helpers" 
1 file changed, 1 insertion(+), 1 deletion(-)

[6]
          return -EINVAL;
---

Comments

Christian Zigotzky Aug. 17, 2020, 9:19 a.m. UTC | #1 
Hello

I compiled the RC1 of kernel 5.9 today. Unfortunately the issue with the 
VirtIO-GPU (see below) still exists. Therefore we still need the patch 
(see below) for using the VirtIO-GPU in a virtual e5500 PPC64 QEMU machine.

Could you please check the first bad commit?

Thanks
Christian


On 12 August 2020 at 3:09 pm, Christian Zigotzky wrote:
> Hello Daniel,
>
> The VirtIO-GPU doesn't work anymore with the latest Git kernel in a 
> virtual e5500 PPC64 QEMU machine [1,2] after the commit "drm/virtio: 
> Call the right shmem helpers". [3]
> The kernel 5.8 works with the VirtIO-GPU in this virtual machine.
>
> I bisected today [4].
>
> Result: drm/virtio: Call the right shmem helpers ( 
> d323bb44e4d23802eb25d13de1f93f2335bd60d0) [3] is the first bad commit.
>
> I was able to revert the first bad commit. [5] After that I compiled a 
> new kernel again. Then I was able to boot Linux with this kernel in a 
> virtual e5500 PPC64 QEMU machine with the VirtIO-GPU.
>
> I created a patch. [6] With this patch I can use the VirtIO-GPU again.
>
> Could you please check the first bad commit?
>
> Thanks,
> Christian
>
> [1] QEMU command: qemu-system-ppc64 -M ppce500 -cpu e5500 -enable-kvm 
> -m 1024 -kernel uImage -drive 
> format=raw,file=fienix-soar_3.0-2020608-net.img,index=0,if=virtio -nic 
> user,model=e1000 -append "rw root=/dev/vda2" -device virtio-vga 
> -device virtio-mouse-pci -device virtio-keyboard-pci -device 
> pci-ohci,id=newusb -device usb-audio,bus=newusb.0 -smp 4
>
> [2] Error messages:
>
> virtio_gpu virtio0: swiotlb buffer is full (sz: 4096 bytes), total 0 
> (slots), used 0 (slots)
> BUG: Kernel NULL pointer dereference on read at 0x00000010
> Faulting instruction address: 0xc0000000000c7324
> Oops: Kernel access of bad area, sig: 11 [#1]
> BE PAGE_SIZE=4K PREEMPT SMP NR_CPUS=4 QEMU e500
> Modules linked in:
> CPU: 2 PID: 1678 Comm: kworker/2:2 Not tainted 
> 5.9-a3_A-EON_X5000-11735-g06a81c1c7db9-dirty #1
> Workqueue: events .virtio_gpu_dequeue_ctrl_func
> NIP:  c0000000000c7324 LR: c0000000000c72e4 CTR: c000000000462930
> REGS: c00000003dba75e0 TRAP: 0300   Not tainted 
> (5.9-a3_A-EON_X5000-11735-g06a81c1c7db9-dirty)
> MSR:  0000000090029000 <CE,EE,ME>  CR: 24002288  XER: 00000000
> DEAR: 0000000000000010 ESR: 0000000000000000 IRQMASK: 0
> GPR00: c0000000000c6188 c00000003dba7870 c0000000017f2300 
> c00000003d893010
> GPR04: 0000000000000000 0000000000000001 0000000000000000 
> 0000000000000000
> GPR08: 0000000000000000 0000000000000000 0000000000000000 
> 7f7f7f7f7f7f7f7f
> GPR12: 0000000024002284 c00000003fff9200 c00000000008c3a0 
> c0000000061566c0
> GPR16: 0000000000000000 0000000000000000 0000000000000000 
> 0000000000000000
> GPR20: 0000000000000000 0000000000000000 0000000000000000 
> 0000000000000000
> GPR24: 0000000000000001 0000000000110000 0000000000000000 
> 0000000000000000
> GPR28: c00000003d893010 0000000000000000 0000000000000000 
> c00000003d893010
> NIP [c0000000000c7324] .dma_direct_unmap_sg+0x4c/0xd8
> LR [c0000000000c72e4] .dma_direct_unmap_sg+0xc/0xd8
> Call Trace:
> [c00000003dba7870] [c00000003dba7950] 0xc00000003dba7950 (unreliable)
> [c00000003dba7920] [c0000000000c6188] .dma_unmap_sg_attrs+0x5c/0x98
> [c00000003dba79d0] [c0000000005cd438] 
> .drm_gem_shmem_free_object+0x98/0xcc
> [c00000003dba7a50] [c0000000006af5b4] 
> .virtio_gpu_cleanup_object+0xc8/0xd4
> [c00000003dba7ad0] [c0000000006ad3bc] .virtio_gpu_cmd_unref_cb+0x1c/0x30
> [c00000003dba7b40] [c0000000006adab8] 
> .virtio_gpu_dequeue_ctrl_func+0x208/0x28c
> [c00000003dba7c10] [c000000000086b70] .process_one_work+0x1a4/0x258
> [c00000003dba7cb0] [c0000000000870f4] .worker_thread+0x214/0x284
> [c00000003dba7d70] [c00000000008c4f0] .kthread+0x150/0x158
> [c00000003dba7e20] [c00000000000082c] .ret_from_kernel_thread+0x58/0x60
> Instruction dump:
> f821ff51 7cb82b78 7cdb3378 4e000000 7cfa3b78 3bc00000 7f9ec000 41fc0014
> 382100b0 81810008 7d808120 48bc1ba8 <e93d0010> ebfc0248 833d0018 7fff4850
> ---[ end trace f28d194d9f0955a8 ]---
>
> virtio_gpu virtio0: swiotlb buffer is full (sz: 4096 bytes), total 0 
> (slots), used 0 (slots)
> virtio_gpu virtio0: swiotlb buffer is full (sz: 16384 bytes), total 0 
> (slots), used 0 (slots)
>
> ---
>
> [3] 
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=d323bb44e4d23802eb25d13de1f93f2335bd60d0
>
> [4] https://forum.hyperion-entertainment.com/viewtopic.php?p=51377#p51377
>
> [5] git revert d323bb44e4d23802eb25d13de1f93f2335bd60d0 //Output: 
> [master 966950f724e4] Revert "drm/virtio: Call the right shmem 
> helpers" 1 file changed, 1 insertion(+), 1 deletion(-)
>
> [6]
> diff --git a/drivers/gpu/drm/virtio/virtgpu_object.c 
> b/drivers/gpu/drm/virtio/virtgpu_object.c
> index 6ccbd01cd888..346cef5ce251 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_object.c
> +++ b/drivers/gpu/drm/virtio/virtgpu_object.c
> @@ -150,7 +150,7 @@ static int virtio_gpu_object_shmem_init(struct 
> virtio_gpu_device *vgdev,
>      if (ret < 0)
>          return -EINVAL;
>
> -    shmem->pages = drm_gem_shmem_get_pages_sgt(&bo->base.base);
> +    shmem->pages = drm_gem_shmem_get_sg_table(&bo->base.base);
>      if (!shmem->pages) {
>          drm_gem_shmem_unpin(&bo->base.base);
>          return -EINVAL;
> ---
Gerd Hoffmann Aug. 18, 2020, 8:18 a.m. UTC | #2 
On Mon, Aug 17, 2020 at 11:19:58AM +0200, Christian Zigotzky wrote:
> Hello
> 
> I compiled the RC1 of kernel 5.9 today. Unfortunately the issue with the
> VirtIO-GPU (see below) still exists. Therefore we still need the patch (see
> below) for using the VirtIO-GPU in a virtual e5500 PPC64 QEMU machine.

It is fixed in drm-misc-next (commit 51c3b0cc32d2e17581fce5b487ee95bbe9e8270a).

Will cherry-pick into drm-misc-fixes once the branch is 5.9-based, which
in turn should bring it to 5.9-rc2 or -rc3.

take care,
  Gerd
Christian Zigotzky Aug. 18, 2020, 11:08 a.m. UTC | #3 
On 18 August 2020 at 10:18 am, Gerd Hoffmann wrote:
> On Mon, Aug 17, 2020 at 11:19:58AM +0200, Christian Zigotzky wrote:
>> Hello
>>
>> I compiled the RC1 of kernel 5.9 today. Unfortunately the issue with the
>> VirtIO-GPU (see below) still exists. Therefore we still need the patch (see
>> below) for using the VirtIO-GPU in a virtual e5500 PPC64 QEMU machine.
> It is fixed in drm-misc-next (commit 51c3b0cc32d2e17581fce5b487ee95bbe9e8270a).
>
> Will cherry-pick into drm-misc-fixes once the branch is 5.9-based, which
> in turn should bring it to 5.9-rc2 or -rc3.
>
> take care,
>    Gerd
>
Thank you!
Christian Zigotzky Aug. 18, 2020, 2:41 p.m. UTC | #4 
On 18 August 2020 at 10:18 am, Gerd Hoffmann wrote:
> On Mon, Aug 17, 2020 at 11:19:58AM +0200, Christian Zigotzky wrote:
>> Hello
>>
>> I compiled the RC1 of kernel 5.9 today. Unfortunately the issue with the
>> VirtIO-GPU (see below) still exists. Therefore we still need the patch (see
>> below) for using the VirtIO-GPU in a virtual e5500 PPC64 QEMU machine.
> It is fixed in drm-misc-next (commit 51c3b0cc32d2e17581fce5b487ee95bbe9e8270a).
>
> Will cherry-pick into drm-misc-fixes once the branch is 5.9-based, which
> in turn should bring it to 5.9-rc2 or -rc3.
>
> take care,
>    Gerd
>
Hello Gerd,

I compiled a new kernel with the latest DRM misc updates today. The 
patch is included in these updates.

This kernel works with the VirtIO-GPU in a virtual e5500 QEMU/KVM HV 
machine on my X5000.

Unfortunately I can only use the VirtIO-GPU (Monitor: Red Hat, Inc. 8") 
with a resolution of 640x480. If I set a higher resolution then the 
guest disables the monitor.
I can use higher resolutions with the stable kernel 5.8 and the VirtIO-GPU.

Please check the latest DRM updates.

Thanks,
Christian
Gerd Hoffmann Aug. 19, 2020, 4:35 a.m. UTC | #5 
On Tue, Aug 18, 2020 at 04:41:38PM +0200, Christian Zigotzky wrote:
> Hello Gerd,
> 
> I compiled a new kernel with the latest DRM misc updates today. The patch is
> included in these updates.
> 
> This kernel works with the VirtIO-GPU in a virtual e5500 QEMU/KVM HV machine
> on my X5000.
> 
> Unfortunately I can only use the VirtIO-GPU (Monitor: Red Hat, Inc. 8") with
> a resolution of 640x480. If I set a higher resolution then the guest
> disables the monitor.
> I can use higher resolutions with the stable kernel 5.8 and the VirtIO-GPU.
> 
> Please check the latest DRM updates.

https://patchwork.freedesktop.org/patch/385980/

(tests & reviews & acks are welcome)

HTH,
  Gerd
Christian Zigotzky Aug. 19, 2020, 10:22 a.m. UTC | #6 
On 19 August 2020 at 06:35 am, Gerd Hoffmann wrote:
> On Tue, Aug 18, 2020 at 04:41:38PM +0200, Christian Zigotzky wrote:
>> Hello Gerd,
>>
>> I compiled a new kernel with the latest DRM misc updates today. The patch is
>> included in these updates.
>>
>> This kernel works with the VirtIO-GPU in a virtual e5500 QEMU/KVM HV machine
>> on my X5000.
>>
>> Unfortunately I can only use the VirtIO-GPU (Monitor: Red Hat, Inc. 8") with
>> a resolution of 640x480. If I set a higher resolution then the guest
>> disables the monitor.
>> I can use higher resolutions with the stable kernel 5.8 and the VirtIO-GPU.
>>
>> Please check the latest DRM updates.
> https://patchwork.freedesktop.org/patch/385980/
>
> (tests & reviews & acks are welcome)
>
> HTH,
>    Gerd
>
Hello Gerd,

I compiled a new RC1 with our patches today. With these patches, the 
VirtIO-GPU works without any problems. I can use higher resolutions again.

Screenshot of the RC1-3 with the VirtIO-GPU in a virtual e5500 QEMU/KVM 
HV machine on my X5000: 
https://i.pinimg.com/originals/4f/b0/14/4fb01476edd7abe6be1e1203a8e7e152.png

Thanks a lot for your help!

Cheers,
Christian
diff mbox series

Patch

diff --git a/drivers/gpu/drm/virtio/virtgpu_object.c 
b/drivers/gpu/drm/virtio/virtgpu_object.c
index 6ccbd01cd888..346cef5ce251 100644
--- a/drivers/gpu/drm/virtio/virtgpu_object.c
+++ b/drivers/gpu/drm/virtio/virtgpu_object.c
@@ -150,7 +150,7 @@  static int virtio_gpu_object_shmem_init(struct 
virtio_gpu_device *vgdev,
      if (ret < 0)
          return -EINVAL;

-    shmem->pages = drm_gem_shmem_get_pages_sgt(&bo->base.base);
+    shmem->pages = drm_gem_shmem_get_sg_table(&bo->base.base);
      if (!shmem->pages) {
          drm_gem_shmem_unpin(&bo->base.base);