On 04/12/2017 03:51 AM, Michał Kępień wrote: > $ tail -20 qemu.log > > == ide_cancel_dma_sync == > > ATA Registers: > cmd 0x06 > feature 0x01 > error 0x00 > nsector 0x00000001 > sector 0x00 > lcyl 0x00 > hcyl 0x00 > hob_feature 0x00 > hob_nsector 0x00 > hob_sector 0x00 > hob_lcyl 0x00 > hob_hcyl 0x00 > select 0x60 > status 0x58 > lba48 0x00000000 > qemu-system-x86_64: hw/ide/core.c:704: ide_cancel_dma_sync: Assertion > `s->bus->dma->aiocb == NULL' failed. > $ grep ^cmd qemu.log | sort | uniq -c > 128 cmd 0x06 > 151854 cmd 0xc8 > 217496 cmd 0xca > > I am happy to help if any further debugging is required. >
Whoops, I misunderstood exactly how often cancel would be invoked here, sorry about that. It looks like when DMA is finished and the guest signals that it's over, we cancel any outstanding DMA just to be safe, and that'd explain the nearly 400,000 calls in your logs. However, this looks like it might legitimately be trying to cancel a TRIM command (I don't know why ...) but we don't clean up after those properly. Let's try and see if this doesn't fix your problem: https://github.com/jnsnow/qemu/commit/57bf2ccdfe8dd35838c1e6642bf9bd76dc9ad1a9 Optionally, you can delete the printf from the last patch if you want. I'm still a little concerned that your guest is trying to cancel in-flight commands which I didn't think would happen under normal circumstances unless some other problem arose, but I think this will clear up the assert for us. Thanks, -John -- You received this bug notification because you are a member of qemu- devel-ml, which is subscribed to QEMU. https://bugs.launchpad.net/bugs/1681439 Title: qemu-system-x86_64: hw/ide/core.c:685: ide_cancel_dma_sync: Assertion `s->bus->dma->aiocb == NULL' failed. Status in QEMU: New Bug description: Since upgrading to QEMU 2.8.0, my Windows 7 64-bit virtual machines started crashing due to the assertion quoted in the summary failing. The assertion in question was added by commit 9972354856 ("block: add BDS field to count in-flight requests"). My tests show that setting discard=unmap is needed to reproduce the issue. Speaking of reproduction, it is a bit flaky, because I have been unable to come up with specific instructions that would allow the issue to be triggered outside of my environment, but I do have a semi-sane way of testing that appears to depend on a specific initial state of data on the underlying storage volume, actions taken within the VM and waiting for about 20 minutes. Here is the shortest QEMU command line that I managed to reproduce the bug with: qemu-system-x86_64 \ -machine pc-i440fx-2.7,accel=kvm \ -m 3072 \ -drive file=/dev/lvm/qemu,format=raw,if=ide,discard=unmap \ -netdev tap,id=hostnet0,ifname=tap0,script=no,downscript=no,vhost=on \ -device virtio-net-pci,netdev=hostnet0 \ -vnc :0 The underlying storage (/dev/lvm/qemu) is a thin LVM snapshot. QEMU was compiled using: ./configure --python=/usr/bin/python2.7 --target-list=x86_64-softmmu make -j3 My virtualization environment is not really a critical one and reproduction is not that much of a hassle, so if you need me to gather further diagnostic information or test patches, I will be happy to help. To manage notifications about this bug go to: https://bugs.launchpad.net/qemu/+bug/1681439/+subscriptions