linux-mtd.lists.infradead.org archive mirror
 help / color / mirror / Atom feed
From: "Uwe Kleine-König" <u.kleine-koenig@pengutronix.de>
To: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
Cc: Richard Weinberger <richard@nod.at>,
	linux-mtd@lists.infradead.org, entwicklung@pengutronix.de
Subject: [PATCH mtd-utils RFC] Add an ubifs mount helper
Date: Tue,  6 Oct 2020 11:19:13 +0200	[thread overview]
Message-ID: <20201006091912.1139071-1-u.kleine-koenig@pengutronix.de> (raw)

This abstracts away attaching of the right ubi and then selecting the right
ubi device and volume to mount.

As described in the comment at the top this allows to mount ubifs volumes
directly from /etc/fstab without having to use hardcoded numbers (which
depend on mount order and so are unreliable) and extra magic to care for
attaching.

Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
---
Hello,

this is a nice script that helps attaching ubis and mounting ubifs'. I
already discussed this a bit with Richard on irc and the idea was
welcome in principle.

A concern was that the entry in /etc/fstab is too magic and there should
be an indication that this relies on the mount helper. Something like
using ubifs.autoattach as file system type (fs_vfstype) or requiring
"autoattach" in the fourth field (fs_mntops).

Having thought a bit I'm not convinced we need this indication. Yes
there is magic but nothing out of the ordinary as mount helpers are a
known concept and the fs_spec already looks interesting enough to make
it obvious there is something going on. And once people are used to the
magic it is annoying to have to always write this indication in your
fstab. Additionally it complicates the implementation. 

Also "autoattach" doesn't completely cover the script as it does a
translation from mtd name to mtd number, too. "autoattach,mtdtranslate"
seems a bit excessive.

Best regards
Uwe

 ubifs-utils/Makemodule.am |   2 +
 ubifs-utils/mount.ubifs   | 101 ++++++++++++++++++++++++++++++++++++++
 2 files changed, 103 insertions(+)
 create mode 100755 ubifs-utils/mount.ubifs

diff --git a/ubifs-utils/Makemodule.am b/ubifs-utils/Makemodule.am
index 59109ccd613c..5c5d99f7572b 100644
--- a/ubifs-utils/Makemodule.am
+++ b/ubifs-utils/Makemodule.am
@@ -47,4 +47,6 @@ UBIFS_EXTRA = \
 
 EXTRA_DIST += $(UBIFS_HEADER) $(UBIFS_EXTRA)
 
+dist_sbin_SCRIPTS = ubifs-utils/mount.ubifs
+
 sbin_PROGRAMS += $(UBIFS_BINS)
diff --git a/ubifs-utils/mount.ubifs b/ubifs-utils/mount.ubifs
new file mode 100755
index 000000000000..b94ddc5649f4
--- /dev/null
+++ b/ubifs-utils/mount.ubifs
@@ -0,0 +1,101 @@
+#!/bin/sh
+
+# This script should be installed as /sbin/mount.ubifs. The benefit is that an
+# fstab entry like:
+#
+# 	mtd=mtddev:home     /home        ubifs   defaults 0 0
+#
+# results in the ubi contained in the mtd named "mtddev" to be attached (if not
+# already done) and then the volume named "home" being mounted to /home.
+
+# This is called by mount with the following options:
+# /sbin/mount.ubifs spec dir [-sfnv] [-N namespace] [-o options] [-t type.subtype]
+
+spec="$1"
+shift
+
+mtdname2num() {
+	local name
+
+	name="$1"
+
+	for d in $(find /sys/class/mtd/ -regex '.*/mtd[0-9]*'); do
+		case "$d" in
+			*ro)
+				continue
+				;;
+		esac
+
+		if test "$name" = "$(cat "$d/name")"; then
+			local dev mtdnum
+
+			dev="$(basename "$d")"
+			mtdnum="${dev#mtd}"
+			echo "$mtdnum"
+			return
+		fi
+	done
+
+	return 1
+}
+
+mtdnum2ubi() {
+	local mtdnum
+
+	mtdnum="$1"
+
+	for d in $(find /sys/class/ubi/ -regex '.*/ubi[0-9]*'); do
+		case "$d" in
+			*_[0-9]*)
+				continue
+				;;
+		esac
+
+		if test "$mtdnum" = "$(cat "$d/mtd_num")"; then
+			local ubi
+
+			ubi="$(basename "$d")"
+			echo "$ubi"
+			return;
+		fi
+	done
+
+	return 1
+}
+
+mtdnum2ubi_autoattach() {
+	local mtdnum ubi
+
+	mtdnum="$1"
+
+	ubi="$(mtdnum2ubi "$mtdnum")" && { echo "$ubi"; return; }
+
+	# ubiattach might fail with "mtdX is already attached to ubiY" if there
+	# is more than one mount to do in the same mtd partition. So ignore errors.
+	ubiattach -m "$mtdnum" >&2 || true
+
+	mtdnum2ubi "$mtdnum"
+}
+
+case "$spec" in
+	mtd=*:*)
+		spec="${spec#mtd=}"
+		mtd="${spec%:*}"
+		rspec="${spec#*:}"
+
+		mtdnum="$(mtdname2num "$mtd")" || {
+			echo "Failed to find mtdnum for mtd \"$mtd\""
+			exit 1
+		}
+
+		ubi="$(mtdnum2ubi_autoattach "$mtdnum")" || {
+			echo "Failed to find ubi for mtd \"$mtd\""
+			exit 1
+		}
+
+		spec="$ubi:$rspec"
+
+		;;
+esac
+
+/bin/mount -i -t ubifs "$spec" "$@"
-- 
2.28.0


______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/

             reply	other threads:[~2020-10-06  9:20 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-10-06  9:19 Uwe Kleine-König [this message]
2020-11-16  9:39 ` [PATCH mtd-utils RFC] Add an ubifs mount helper Uwe Kleine-König
2020-11-18 14:43   ` David Oberhollenzer

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20201006091912.1139071-1-u.kleine-koenig@pengutronix.de \
    --to=u.kleine-koenig@pengutronix.de \
    --cc=david.oberhollenzer@sigma-star.at \
    --cc=entwicklung@pengutronix.de \
    --cc=linux-mtd@lists.infradead.org \
    --cc=richard@nod.at \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).