Public bug reported:
Sorry that this is a little vague, I ran into it and didn't run it down
to ground; this was on the cosmic daily server installer build.
hardware: 4 HDD;s, 1 NVMe SSD.
configuration: 1x 512MB ESP per drive, then for the HDDs the second partition
for an mdraid raid5 set.
layered on that, bcache with the SSD as cache, and md0 as backing
layered on that, lvm with one big vg and a lv for root
mkfs.ext4 on that lv would break, as a TRIM command it issued would error in
bcache.
The cache was constructed like so: make-bcache -o 7741440 -w 2k -B
/dev/md127 -C /dev/nvme0n1p2
I'm not familiar enough with TRIM/discard to know whether commands
/layers are meant to not issue them unless supported, or whether they
should be issued and discarded; either way something breaks here -
bcache may be incorrectly advertising that bcache0 supports trim, or lvm
may be checking in the wrong way, or mkfs.ext4 etc etc.
Though I can rule mkfs.ext4 out - lvremove also issues a discard and
fails in the same way.
** Affects: linux (Ubuntu)
Importance: Undecided
Status: New
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1810276
Title:
bcache TRIM support buggy - mkfs.ext4 error discarding blocks
Status in linux package in Ubuntu:
New
Bug description:
Sorry that this is a little vague, I ran into it and didn't run it
down to ground; this was on the cosmic daily server installer build.
hardware: 4 HDD;s, 1 NVMe SSD.
configuration: 1x 512MB ESP per drive, then for the HDDs the second partition
for an mdraid raid5 set.
layered on that, bcache with the SSD as cache, and md0 as backing
layered on that, lvm with one big vg and a lv for root
mkfs.ext4 on that lv would break, as a TRIM command it issued would error in
bcache.
The cache was constructed like so: make-bcache -o 7741440 -w 2k -B
/dev/md127 -C /dev/nvme0n1p2
I'm not familiar enough with TRIM/discard to know whether commands
/layers are meant to not issue them unless supported, or whether they
should be issued and discarded; either way something breaks here -
bcache may be incorrectly advertising that bcache0 supports trim, or
lvm may be checking in the wrong way, or mkfs.ext4 etc etc.
Though I can rule mkfs.ext4 out - lvremove also issues a discard and
fails in the same way.
To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1810276/+subscriptions
--
Mailing list: https://launchpad.net/~kernel-packages
Post to : kernel-packages@lists.launchpad.net
Unsubscribe : https://launchpad.net/~kernel-packages
More help : https://help.launchpad.net/ListHelp
