Dakota Williams
Dakota Williams
Still working our way toward kernel 5.15 compat - no packages yet, it won't build.
It going to depend on the page cache. If it's synchronized, then the mapping should be null. If for some reason, there's a race, then it should have the old...
@nickchen-cpu You're correct in that we wouldn't need to worry about the deleted data on a separate disk, but we'd run into a similar problem as my second limitation in...
Duplicate of https://github.com/datto/dattobd/issues/144. Mount hooks can't be found because the kernel restructured how the mount/umount syscalls are organized. Kernel mount and userspace mount don't follow the same trace. It would...
I really don't have a good estimate for you at this time.
We already have some significant work toward kernel 5.9 compatibility. The issue is that 5.9 turned off many of the interfaces used for single-queue i/o that was dominant in the...
The .ko should output in `src/`, and I cannot reproduce on my deb9 vm. Do you have a build log?
Just an update/haven't forgot about this, I'm looking into this to find a root cause. At first I thought it might have to do with REQ_OP_DRV_{IN,OUT} conflicting with our DATTOBD_PASSTHROUGH...
0.10.4 is a considerably old version. Does this happen with the latest release (0.10.11)?
What are your reproduction steps? Also, are there more dattobd log entries in dmesg before the crash happened?