I've been using NVMe M.2 Samsung Pro 960 for 18 months and never had a problem.
Ubuntu 16.04.6 LTS, Kernel 4.14.114 LTS, Skylake 6700HQ, nVidia 970m
UEFI, GPT, AHCI (Intel Raid off), Secure Boot off
I've never had a single fsck error ever. Granted the `grub` boot option `fastboot` means `fsck` is not run on boot but I can check once FS is mounted RW with:
$ sudo fsck -n /dev/nvme0n1p6
fsck from util-linux 2.27.1
e2fsck 1.42.13 (17-May-2015)
Warning! /dev/nvme0n1p6 is mounted.
Warning: skipping journal recovery because doing a read-only filesystem check.
New_Ubuntu_16.04: clean, 712096/2953920 files, 5733245/11829504 blocks
Assuming your `/etc/fstab` is the same, the two important `grub` boot parameters are: `acpiphp.disable=1 pcie_aspm=force`. If memory serves me correct though these were setup for suspend/resume reasons though.
I hope this helps those effected by bug a little but more importantly that people realize the vast majority of NVMe installations work fine in Linux.
I've been using NVMe M.2 Samsung Pro 960 for 18 months and never had a problem.
Ubuntu 16.04.6 LTS, Kernel 4.14.114 LTS, Skylake 6700HQ, nVidia 970m
UEFI, GPT, AHCI (Intel Raid off), Secure Boot off
`/etc/fstab`:
UUID=b40b3925- 70ef-447f- 923e-1b05467c00 e7 / ext4 errors=remount-ro 0 1 0ec8-4b17- 9edd-88db0f0313 32 none swap sw 0 0
UUID=D656-F2A8 /boot/efi vfat umask=0077 0 1
UUID=b4512bc6-
`/etc/default/ grub`: LINUX_DEFAULT= "noplymouth fastboot acpiphp.disable=1 pcie_aspm=force vt.handoff=7 i915.fastboot=1 nopti nospectre_v2 nospec"
GRUB_CMDLINE_
I've never had a single fsck error ever. Granted the `grub` boot option `fastboot` means `fsck` is not run on boot but I can check once FS is mounted RW with:
$ sudo fsck -n /dev/nvme0n1p6
fsck from util-linux 2.27.1
e2fsck 1.42.13 (17-May-2015)
Warning! /dev/nvme0n1p6 is mounted.
Warning: skipping journal recovery because doing a read-only filesystem check.
New_Ubuntu_16.04: clean, 712096/2953920 files, 5733245/11829504 blocks
Assuming your `/etc/fstab` is the same, the two important `grub` boot parameters are: `acpiphp.disable=1 pcie_aspm=force`. If memory serves me correct though these were setup for suspend/resume reasons though.
I hope this helps those effected by bug a little but more importantly that people realize the vast majority of NVMe installations work fine in Linux.