cloud-utils growroot doesn't work with newer sfdisk >= 2.26 : fixed patch solution for newer and older sfdisk
Hi Scott and Juerg,
Someone has filed this bug in Debian:
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=783826
Unfortunately, the patch here:
http://pkgs.fedoraproject.org/cgit/cloud-utils.git/plain/0002-Support-new-sfdisk-version-2.26.patch
doesn't work with the older sfdisk, currently in Sid, but works with
version 2.26 only.
To make a smooth transition, best would be to make a conditional test,
and check if we're running with util-linux 2.25, or >= 2.26. This would
also allow newer cloud-utils to be backported to Jessie.
This is what I've done on the attached patch. Scott, could you please
use that upstream from now on?
Also, what about the do-not-use-EXIT-when-trapping.patch patch which I
had to add in Debian? Could you please apply them both?
Cheers,
Thomas Goirand (zigo)
Description: Adds support for sfdisk 2.26 and higher
The usage of sfdisk has changed in util-linux >= 2.26, so this patch makes it
possible to use both old and newer versions.
.
This patch has been made adapting this patch from Fedora:
http://pkgs.fedoraproject.org/cgit/cloud-utils.git/plain/0002-Support-new-sfdisk-version-2.26.patch
which was attached to the below bug:
https://bugzilla.redhat.com/show_bug.cgi?id=1197894
Author: Thomas Goirand <zigo@debian.org>
Bug-Debian: https://bugs.debian.org/783826
Forwarded: no
Last-Update: 2015-05-04
--- cloud-utils-0.27.orig/bin/growpart
+++ cloud-utils-0.27/bin/growpart
@@ -132,8 +132,46 @@ bad_Usage() {
exit 2
}
+is_sfdisk_higher_than_226 () {
+ local SFDISK_VERSION MAJOR MID
+ # Do convenient caching of the result
+ # to avoid forking of sfdisk
+ if [ -n "${SFDISK_HIGHER_THAN_226}" ] ; then
+ return ${SFDISK_HIGHER_THAN_226}
+ fi
+ SFDISK_VERSION=`sfdisk -v | awk '{print $4}'`
+ # Using dpkg --compare-version if dpkg is present
+ if [ -x /usr/bin/dpkg ] ; then
+ dpkg --compare-versions ${SFDISK_VERSION} ge 2.26
+ return $?
+ # Compatibility layer for non-debian platforms
+ else
+ MAJOR=`echo $SFDISK_VERSION | cut -d. -f1`
+ MID=`echo $SFDISK_VERSION | cut -d. -f2`
+ if [ ${MAJOR} -ge 2 ] ; then
+ if [ ${MAJOR} -gt 2 ] ; then
+ SFDISK_HIGHER_THAN_226=0
+ else
+ if [ ${MID} -ge 26 ] ; then
+ SFDISK_HIGHER_THAN_226=0
+ else
+ SFDISK_HIGHER_THAN_226=1
+ fi
+ fi
+ else
+ SFDISK_HIGHER_THAN_226=1
+ fi
+ fi
+ return ${SFDISK_HIGHER_THAN_226}
+}
+
mbr_restore() {
- sfdisk --no-reread "${DISK}" ${MBR_CHS} -I "${MBR_BACKUP}"
+ if is_sfdisk_higher_than_226 ; then
+ dd if="${MBR_BACKUP}-${DISK#/dev/}-0x00000000.bak" of="${DISK}" bs=1 \
+ conv=notrunc
+ else
+ sfdisk --no-reread "${DISK}" ${MBR_CHS} -I "${MBR_BACKUP}"
+ fi
}
sfdisk_worked_but_blkrrpart_failed() {
@@ -148,7 +186,7 @@ sfdisk_worked_but_blkrrpart_failed() {
mbr_resize() {
RESTORE_HUMAN="${TEMP_D}/recovery"
- MBR_BACKUP="${TEMP_D}/orig.save"
+ MBR_BACKUP="${TEMP_D}/backup"
local change_out=${TEMP_D}/change.out
local dump_out=${TEMP_D}/dump.out
@@ -160,24 +198,39 @@ mbr_resize() {
local _devc cyl _w1 heads _w2 sectors _w3 tot dpart
local pt_start pt_size pt_end max_end new_size change_info
- # --show-pt-geometry outputs something like
- # /dev/sda: 164352 cylinders, 4 heads, 32 sectors/track
- rqe sfd_geom sfdisk "${DISK}" --show-pt-geometry >"${tmp}" &&
- read _devc cyl _w1 heads _w2 sectors _w3 <"${tmp}" &&
- MBR_CHS="-C ${cyl} -H ${heads} -S ${sectors}" ||
- fail "failed to get CHS from ${DISK}"
-
- tot=$((${cyl}*${heads}*${sectors}))
-
- debug 1 "geometry is ${MBR_CHS}. total size=${tot}"
- rqe sfd_dump sfdisk ${MBR_CHS} --unit=S --dump "${DISK}" \
- >"${dump_out}" ||
- fail "failed to dump sfdisk info for ${DISK}"
-
- {
- echo "## sfdisk ${MBR_CHS} --unit=S --dump ${DISK}"
- cat "${dump_out}"
- } >"${RESTORE_HUMAN}"
+ if is_sfdisk_higher_than_226 ; then
+ tot=$(sfdisk --list "${DISK}" | awk '{ print $(NF-1) ; exit }') ||
+ fail "failed to get total number of sectors from ${DISK}"
+ debug 1 "total number of sectors of ${DISK} is ${tot}"
+
+ rqe sfd_dump sfdisk --dump "${DISK}" \
+ >"${dump_out}" ||
+ fail "failed to dump sfdisk info for ${DISK}"
+ {
+ echo "## sfdisk --dump ${DISK}"
+ cat "${dump_out}"
+ } >"${RESTORE_HUMAN}"
+ else
+ # --show-pt-geometry outputs something like
+ # /dev/sda: 164352 cylinders, 4 heads, 32 sectors/track
+ rqe sfd_geom sfdisk "${DISK}" --show-pt-geometry >"${tmp}" &&
+ read _devc cyl _w1 heads _w2 sectors _w3 <"${tmp}" &&
+ MBR_CHS="-C ${cyl} -H ${heads} -S ${sectors}" ||
+ fail "failed to get CHS from ${DISK}"
+
+ tot=$((${cyl}*${heads}*${sectors}))
+ debug 1 "geometry is ${MBR_CHS}. total size=${tot}"
+
+ rqe sfd_dump sfdisk ${MBR_CHS} --unit=S --dump "${DISK}" \
+ >"${dump_out}" ||
+ fail "failed to dump sfdisk info for ${DISK}"
+ {
+ echo "## sfdisk ${MBR_CHS} --unit=S --dump ${DISK}"
+ cat "${dump_out}"
+ } >"${RESTORE_HUMAN}"
+ fi
+
+
[ $? -eq 0 ] || fail "failed to save sfdisk -d output"
debugcat 1 "${RESTORE_HUMAN}"
@@ -237,8 +290,13 @@ mbr_resize() {
exit 0
fi
- LANG=C sfdisk --no-reread "${DISK}" ${MBR_CHS} --force \
- -O "${MBR_BACKUP}" <"${new_out}" >"${change_out}" 2>&1
+ if is_sfdisk_higher_than_226 ; then
+ LANG=C sfdisk --no-reread "${DISK}" --force \
+ -O "${MBR_BACKUP}" <"${new_out}" >"${change_out}" 2>&1
+ else
+ LANG=C sfdisk --no-reread "${DISK}" ${MBR_CHS} --force \
+ -O "${MBR_BACKUP}" <"${new_out}" >"${change_out}" 2>&1
+ fi
ret=$?
[ $ret -eq 0 ] || RESTORE_FUNC="mbr_restore"
Description: Uses 0 instead of EXIT when trapping
Under Ubuntu, this seems harmless, but in Debian, it breaks, so I'm patching.
Author: Thomas Goirand <zigo@debian.org>
Forwarded: not-needed
Last-Update: 2013-06-11
Index: cloud-utils/bin/growpart
===================================================================
--- cloud-utils.orig/bin/growpart
+++ cloud-utils/bin/growpart
@@ -569,7 +569,7 @@ PT_UPDATE=$_RET
debug 1 "update-partition set to $PT_UPDATE"
mktemp_d && TEMP_D="${_RET}" || fail "failed to make temp dir"
-trap cleanup EXIT
+trap cleanup 0
# get the ID of the first partition to determine if it's MBR or GPT
id=$(sfdisk --id --force "${DISK}" 1 2>/dev/null) ||
Reply to: