From: "H. Hartzer" <h@hartzer.sh>
Subject: Re: [PATCH] softraid(4) man page clarifications
To: "Jason McIntyre" <jmc@kerhand.co.uk>, <tech@openbsd.org>
Date: Sat, 12 Apr 2025 14:37:38 +0000

> Jason McIntyre wrote:

Hi Jason,

Thank you for replying!

> this change seems simpler. but you want to start new sentences on new
> lines.

I have a new diff with this below.

> -Stacking disciplines (CRYPTO on top of RAID 1, for example) is not
> +Stacking disciplines (CRYPTO on top of RAID 5, for example) is not

> can you explain thwe reason for this change?

I found this confusing because there is RAID 1C, which is its own
discipline. I figure if RAID 5 is mentioned, then there's no ambiguity
since RAID 1C is okay, but CRYPTO + RAID 1 is not. I would not
have been confused thinking: "I thought RAID 1 + CRYPTO was okay?"

New patch below.

-Henrich

--- share/man/man4/softraid.4.old       Fri Apr 11 23:18:05 2025
+++ share/man/man4/softraid.4   Sat Apr 12 14:31:44 2025
@@ -119,9 +119,9 @@ in the boot storage area of the
 .Nm
 volume.
 All chunks in the volume will then be bootable.
-Boot support is currently limited to the CRYPTO, RAID 1 disciplines
-on the amd64, arm64, i386, riscv64 and sparc64 platforms.
-amd64, arm64, riscv64 and sparc64 also have boot support for the RAID 1C discipline.
+Boot support is currently limited to the CRYPTO, RAID 1, and RAID 1C
+disciplines on the amd64, arm64, riscv64 and sparc64 platforms.
+Boot support on i386 is limited to CRYPTO and RAID 1.
 On sparc64, bootable chunks must be RAID partitions using the letter
 .Sq a .
 At the
@@ -236,6 +236,7 @@ boot> boot sr0a:/bsd
 .Sh SEE ALSO
 .Xr bio 4 ,
 .Xr bioctl 8 ,
+.Xr boot 8 ,
 .Xr boot_sparc64 8 ,
 .Xr disklabel 8 ,
 .Xr fdisk 8 ,
@@ -258,7 +259,7 @@ There is no point in wasting a lot of time syncing ran
 The RAID 5 discipline does not initialize parity upon creation, instead parity
 is only updated upon write.
 .Pp
-Stacking disciplines (CRYPTO on top of RAID 1, for example) is not
+Stacking disciplines (CRYPTO on top of RAID 5, for example) is not
 supported at this time.
 .Pp
 Currently there is no automated mechanism to recover from failed disks.