mirror of
https://git.proxmox.com/git/mirror_zfs.git
synced 2024-11-17 10:01:01 +03:00
df554c148e
This change fixes a kernel panic which would occur when resizing a dataset which was not open. The objset_t stored in the zvol_state_t will be set to NULL when the block device is closed. To avoid this issue we pass the correct objset_t as the third arg. The code has also been updated to correctly notify the kernel when the block device capacity changes. For 2.6.28 and newer kernels the capacity change will be immediately detected. For earlier kernels the capacity change will be detected when the device is next opened. This is a known limitation of older kernels. Online ext3 resize test case passes on 2.6.28+ kernels: $ dd if=/dev/zero of=/tmp/zvol bs=1M count=1 seek=1023 $ zpool create tank /tmp/zvol $ zfs create -V 500M tank/zd0 $ mkfs.ext3 /dev/zd0 $ mkdir /mnt/zd0 $ mount /dev/zd0 /mnt/zd0 $ df -h /mnt/zd0 $ zfs set volsize=800M tank/zd0 $ resize2fs /dev/zd0 $ df -h /mnt/zd0 Original-patch-by: Fajar A. Nugraha <github@fajar.net> Closes #68 Closes #84 |
||
|---|---|---|
| .. | ||
| dm0-raid0.sh | ||
| file-raid0.sh | ||
| file-raid10.sh | ||
| file-raidz2.sh | ||
| file-raidz.sh | ||
| hda-raid0.sh | ||
| lo-faulty-raid0.sh | ||
| lo-faulty-raid10.sh | ||
| lo-faulty-raidz2.sh | ||
| lo-faulty-raidz3.sh | ||
| lo-faulty-raidz.sh | ||
| lo-raid0.sh | ||
| lo-raid10.sh | ||
| lo-raidz2.sh | ||
| lo-raidz.sh | ||
| Makefile.am | ||
| Makefile.in | ||
| md0-raid5.sh | ||
| md0-raid10.sh | ||
| ram0-raid0.sh | ||
| scsi_debug-noraid.sh | ||
| scsi_debug-raid0.sh | ||
| scsi_debug-raid10.sh | ||
| scsi_debug-raidz2.sh | ||
| scsi_debug-raidz3.sh | ||
| scsi_debug-raidz.sh | ||
| sda-raid0.sh | ||
| zpool-raid0.sh | ||
| zpool-raid10.sh | ||
| zpool-raidz.sh | ||