btrfs: Kernel warning when using/mount RAID 5/6

[Diederik de Haas]

Hi,

Since https://bugs.debian.org/863290 (2017) the Debian kernel has had a
patch to warn about the use of RAID 5/6 with BTRFS.
That bug mentions "It looks like there's a consensus that such a warning
should live in the kernel rather than userland"

Via [1] and [2] it seems userland did get a warning. It is mentioned in
the kernel documentation (``Documentation/btrfs-man5.rst``) (and [3]
and [4]), but AFAICT users are still not warned by the kernel when
using RAID 5/6.

I stumbled upon this issue when I was (locally) rebasing the Debian kernel
patch for kernel 6.8. I attached my rebased version of the original patch.
(I may have done it completely wrong as I know very little about BTRFS.)

But more importantly, IMO such a warning shouldn't be in a downstream
kernel (Debian), but in the upstream kernel source?

Cheers,
  Diederik

[1] https://lore.kernel.org/linux-btrfs/20161208153004.GA31795@xxxxxxxxxx/
[2] https://lore.kernel.org/linux-btrfs/bf9594ea55ce40af80548888070427ad97daf78a.1598374255.git.josef@xxxxxxxxxxxxxx/
[3] https://btrfs.readthedocs.io/en/latest/btrfs-man5.html#raid56-status-and-recommended-practices
[4] https://lore.kernel.org/linux-btrfs/dbf47c42-932c-9cf0-0e50-75f1d779d024@xxxxxxxxxxxxx/From: Adam Borowski <kilobyte@xxxxxxxxxx>
Date: Tue, 28 Mar 2017 16:55:05 +0200
Subject: btrfs: warn about RAID5/6 being experimental at mount time
Bug-Debian: https://bugs.debian.org/863290
Origin: https://bugs.debian.org/863290#5

Too many people come complaining about losing their data -- and indeed,
there's no warning outside a wiki and the mailing list tribal knowledge.
Message severity chosen for consistency with XFS -- "alert" makes dmesg
produce nice red background which should get the point across.

Signed-off-by: Adam Borowski <kilobyte@xxxxxxxxxx>
[bwh: Also add_taint() so this is flagged in bug reports]
[2023-01-10: still accurate according to btrfs-progs own manpage:
https://git.kernel.org/pub/scm/linux/kernel/git/kdave/btrfs-progs.git/commit/?id=922797e15590b836e377d6dc47b828356cafc2a9]
[2024-03-17: still accurate; manpage is now in Documentation/btrfs-man5.rst
implementation went from disk-io.c to super.c]
---
 fs/btrfs/super.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index 101f786963d4..2c409bce1bf5 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -731,6 +731,18 @@ static void set_device_specific_options(struct btrfs_fs_info *fs_info)
 	    !fs_info->fs_devices->rotating)
 		btrfs_set_opt(fs_info->mount_opt, SSD);
 
+	/*
+	 * Warn about RAID5/6 being experimental at mount time
+	 */
+	if ((fs_info->avail_data_alloc_bits |
+	     fs_info->avail_metadata_alloc_bits |
+	     fs_info->avail_system_alloc_bits) &
+	    BTRFS_BLOCK_GROUP_RAID56_MASK) {
+		btrfs_alert(fs_info,
+		"btrfs RAID5/6 is EXPERIMENTAL and has known data-loss bugs");
+		add_taint(TAINT_AUX, LOCKDEP_STILL_OK);
+	}
+
 	/*
 	 * For devices supporting discard turn on discard=async automatically,
 	 * unless it's already set or disabled. This could be turned off by

Attachment: signature.asc
Description: This is a digitally signed message part.