Created attachment 1823016 [details] guest xml Description of problem: Under I/O load, I get intermittent write errors inside the Fedora armv7 guest running on aarch64 host. [ 2063.691215] print_req_error: 6 callbacks suppressed [ 2063.691249] blk_update_request: I/O error, dev vda, sector 8990744 op 0x1:(WRITE) flags 0x100000 phys_seg 60 prio class 0 [ 2063.713844] blk_update_request: I/O error, dev vda, sector 8992744 op 0x1:(WRITE) flags 0x104000 phys_seg 254 prio class 0 [ 2063.722337] blk_update_request: I/O error, dev vda, sector 8994992 op 0x1:(WRITE) flags 0x100000 phys_seg 2 prio class 0 [ 2063.730617] blk_update_request: I/O error, dev vda, sector 8995008 op 0x1:(WRITE) flags 0x104000 phys_seg 191 prio class 0 [ 2063.745903] blk_update_request: I/O error, dev vda, sector 8997568 op 0x1:(WRITE) flags 0x100000 phys_seg 65 prio class 0 [ 2063.760612] blk_update_request: I/O error, dev vda, sector 8998552 op 0x1:(WRITE) flags 0x104000 phys_seg 158 prio class 0 [ 2063.774761] blk_update_request: I/O error, dev vda, sector 9001112 op 0x1:(WRITE) flags 0x100000 phys_seg 98 prio class 0 [ 2063.785913] blk_update_request: I/O error, dev vda, sector 9002712 op 0x1:(WRITE) flags 0x104000 phys_seg 234 prio class 0 [ 2063.798290] blk_update_request: I/O error, dev vda, sector 9005272 op 0x1:(WRITE) flags 0x100000 phys_seg 120 prio class 0 On the host side, no errors are seen, so it doesn't look like this is an actual I/O error. Version-Release number of selected component (if applicable): qemu-6.1.0-7.fc36.aarch64 Host kernel: 5.14.0-60.fc35.aarch64 Guest kernel: 5.14.1-300.fc35.armv7hl+lpae How reproducible: 100% Steps to Reproduce: 1. Do anything in the guest which is write heavy. eg. dnf install, git clone
<disk type='block' device='disk'> <driver name='qemu' type='raw'/> <source dev='/dev/fedora/armv7' index='1'/> <backingStore/> <target dev='vda' bus='virtio'/> <alias name='virtio-disk0'/> <address type='pci' domain='0x0000' bus='0x04' slot='0x00' function='0x0'/> </disk> The guest is backed by a raw-format host device. As mentioned in the bug report, no errors are seen on the host side, and in general the host storage on this Raspberry Pi 4B host has been solid (USB3 UAS SSD).
*** This bug has been marked as a duplicate of bug 2026747 ***