| Message ID | 20240819121045.1533891-1-Quirin.Gylstorff@siemens.com (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
| Series | [isar-cip-core,RFC] fix: Check return code value of e2fsck | expand |
On 19.08.24 14:09, Quirin Gylstorff wrote: > From: Quirin Gylstorff <quirin.gylstorff@siemens.com> > > This fixes the overlay script in case of a recoverable file system error. > In that case we execute the erronously the recouvery script and mke2fs > ask if the filesystem should be overwritten. > > The issue was introduced with: > > 8644fb1 initramfs-overlay-hook: Check file system of INITRAMFS_OVERLAY_STORAGE_DEVICE > > Signed-off-by: Quirin Gylstorff <quirin.gylstorff@siemens.com> > --- > This fixes the issue but we need to address the problem in case of a > return code of 2 which requires a reboot of the system. Should we: > 1. panic > 2. silently reboot > 3. execute the user defined recovery action (add the return value as a > additional argument) > 4. do nothing > In my testing i used 4. Why does e2fsck require a reboot in some cases? > > Also should we force mke2fs? I would say so. Who knows in which state that target partition was left. > > .../initramfs-overlay-hook/files/overlay.script.tmpl | 10 +++++++++- > 1 file changed, 9 insertions(+), 1 deletion(-) > > diff --git a/recipes-initramfs/initramfs-overlay-hook/files/overlay.script.tmpl b/recipes-initramfs/initramfs-overlay-hook/files/overlay.script.tmpl > index c655a4f..02df53d 100644 > --- a/recipes-initramfs/initramfs-overlay-hook/files/overlay.script.tmpl > +++ b/recipes-initramfs/initramfs-overlay-hook/files/overlay.script.tmpl > @@ -40,7 +40,15 @@ partition_fstype=$(get_fstype "${ovl_partition_device}") > if ! mountpoint -q "${rootmnt}${storage_mount_point}"; then > case $partition_fstype in > ext*) > - if ! e2fsck -p -f "$ovl_partition_device" && [ -x "$ovl_recovery_script" ]; then > + e2fsck -p -f "$ovl_partition_device" > + fsck_ret="$?" > + # e2fsck returns 1 in case ofrepairing the file system or a 2 in case > + # a reboot is required. > + # https://man7.org/linux/man-pages/man8/e2fsck.8.html#EXIT_CODE > + if [ "$fsck_ret" -gt "1" ]; then > + # TODO should we force are reboot or customer action > + panic "fsck requires a reboot" > + elif [ "$fsck_ret" -gt "2" ] && [ -x "$ovl_recovery_script" ]; then > "$ovl_recovery_script" "$partition_fstype" "$ovl_partition_device" > fi > ;; Please typo check everything. Does this fix https://gitlab.com/cip-project/cip-core/isar-cip-core/-/issues/113? Jan
On 8/19/24 2:16 PM, Jan Kiszka wrote: > On 19.08.24 14:09, Quirin Gylstorff wrote: >> From: Quirin Gylstorff <quirin.gylstorff@siemens.com> >> >> This fixes the overlay script in case of a recoverable file system error. >> In that case we execute the erronously the recouvery script and mke2fs >> ask if the filesystem should be overwritten. >> >> The issue was introduced with: >> >> 8644fb1 initramfs-overlay-hook: Check file system of INITRAMFS_OVERLAY_STORAGE_DEVICE >> >> Signed-off-by: Quirin Gylstorff <quirin.gylstorff@siemens.com> >> --- >> This fixes the issue but we need to address the problem in case of a >> return code of 2 which requires a reboot of the system. Should we: >> 1. panic >> 2. silently reboot >> 3. execute the user defined recovery action (add the return value as a >> additional argument) >> 4. do nothing >> In my testing i used 4. > > Why does e2fsck require a reboot in some cases? After looking into it: In case of a ext2 system an the check file system is root. https://github.com/tytso/e2fsprogs/blob/950a0d69c82b585aba30118f01bf80151deffe8c/e2fsck/unix.c#L2133 https://github.com/tytso/e2fsprogs/blob/950a0d69c82b585aba30118f01bf80151deffe8c/e2fsck/util.c#L70 I will remove that check and say that greater than 1 needs the recovery script in v2. > >> >> Also should we force mke2fs? > > I would say so. Who knows in which state that target partition was left. I will add it in v2. Quirin > >> >> .../initramfs-overlay-hook/files/overlay.script.tmpl | 10 +++++++++- >> 1 file changed, 9 insertions(+), 1 deletion(-) >> >> diff --git a/recipes-initramfs/initramfs-overlay-hook/files/overlay.script.tmpl b/recipes-initramfs/initramfs-overlay-hook/files/overlay.script.tmpl >> index c655a4f..02df53d 100644 >> --- a/recipes-initramfs/initramfs-overlay-hook/files/overlay.script.tmpl >> +++ b/recipes-initramfs/initramfs-overlay-hook/files/overlay.script.tmpl >> @@ -40,7 +40,15 @@ partition_fstype=$(get_fstype "${ovl_partition_device}") >> if ! mountpoint -q "${rootmnt}${storage_mount_point}"; then >> case $partition_fstype in >> ext*) >> - if ! e2fsck -p -f "$ovl_partition_device" && [ -x "$ovl_recovery_script" ]; then >> + e2fsck -p -f "$ovl_partition_device" >> + fsck_ret="$?" >> + # e2fsck returns 1 in case ofrepairing the file system or a 2 in case >> + # a reboot is required. >> + # https://man7.org/linux/man-pages/man8/e2fsck.8.html#EXIT_CODE >> + if [ "$fsck_ret" -gt "1" ]; then >> + # TODO should we force are reboot or customer action >> + panic "fsck requires a reboot" >> + elif [ "$fsck_ret" -gt "2" ] && [ -x "$ovl_recovery_script" ]; then >> "$ovl_recovery_script" "$partition_fstype" "$ovl_partition_device" >> fi >> ;; > > Please typo check everything. > > Does this fix > https://gitlab.com/cip-project/cip-core/isar-cip-core/-/issues/113? > > Jan >
diff --git a/recipes-initramfs/initramfs-overlay-hook/files/overlay.script.tmpl b/recipes-initramfs/initramfs-overlay-hook/files/overlay.script.tmpl index c655a4f..02df53d 100644 --- a/recipes-initramfs/initramfs-overlay-hook/files/overlay.script.tmpl +++ b/recipes-initramfs/initramfs-overlay-hook/files/overlay.script.tmpl @@ -40,7 +40,15 @@ partition_fstype=$(get_fstype "${ovl_partition_device}") if ! mountpoint -q "${rootmnt}${storage_mount_point}"; then case $partition_fstype in ext*) - if ! e2fsck -p -f "$ovl_partition_device" && [ -x "$ovl_recovery_script" ]; then + e2fsck -p -f "$ovl_partition_device" + fsck_ret="$?" + # e2fsck returns 1 in case ofrepairing the file system or a 2 in case + # a reboot is required. + # https://man7.org/linux/man-pages/man8/e2fsck.8.html#EXIT_CODE + if [ "$fsck_ret" -gt "1" ]; then + # TODO should we force are reboot or customer action + panic "fsck requires a reboot" + elif [ "$fsck_ret" -gt "2" ] && [ -x "$ovl_recovery_script" ]; then "$ovl_recovery_script" "$partition_fstype" "$ovl_partition_device" fi ;;