[PATCH v3 1/8] overlayfs-qa: common functions for overlayfs

[PATCH v3 1/8] overlayfs-qa: common functions for overlayfs

[Vyacheslav Yurkov]

This class provides commom functions for overlayfs and its QA check,
which is performed in ROOTFS_POSTPROCESS_COMMAND

Signed-off-by: Vyacheslav Yurkov <uvv.mail@gmail.com>
---
 meta/classes/overlayfs-qa.bbclass | 14 ++++++++++++++
 1 file changed, 14 insertions(+)
 create mode 100644 meta/classes/overlayfs-qa.bbclass

diff --git a/meta/classes/overlayfs-qa.bbclass b/meta/classes/overlayfs-qa.bbclass
new file mode 100644
index 0000000000..54fa8316a2
--- /dev/null
+++ b/meta/classes/overlayfs-qa.bbclass
@@ -0,0 +1,14 @@
+# This class contains common functions for overlayfs and its QA check,
+# which is performed in ROOTFS_POSTPROCESS_COMMAND
+
+# this function is based on https://github.com/systemd/systemd/blob/main/src/basic/unit-name.c
+def escapeSystemdUnitName(path):
+    escapeMap = {
+        '/': '-',
+        '-': "\\x2d",
+        '\\': "\\x5d"
+    }
+    return "".join([escapeMap.get(c, c) for c in path.strip('/')])
+
+def mountUnitName(unit):
+    return escapeSystemdUnitName(unit) + '.mount'
-- 
2.28.0

[PATCH v3 2/8] overlayfs.bbclass: generate overlayfs mount units

[Vyacheslav Yurkov]

It's often desired in Embedded System design to have a read-only rootfs.
But a lot of different applications might want to have a read-write access
to some parts of a filesystem. It can be especially useful when your update
mechanism overwrites the whole rootfs, but you want your application data
to be preserved between updates. This class provides a way to achieve that
by means of overlayfs and at the same time keeping the base rootfs read-only.

Signed-off-by: Vyacheslav Yurkov <uvv.mail@gmail.com>
---
 meta/classes/overlayfs.bbclass | 133 +++++++++++++++++++++++++++++++++
 1 file changed, 133 insertions(+)
 create mode 100644 meta/classes/overlayfs.bbclass

diff --git a/meta/classes/overlayfs.bbclass b/meta/classes/overlayfs.bbclass
new file mode 100644
index 0000000000..40e6107f0d
--- /dev/null
+++ b/meta/classes/overlayfs.bbclass
@@ -0,0 +1,133 @@
+# Class for generation of overlayfs mount units
+#
+# It's often desired in Embedded System design to have a read-only rootfs.
+# But a lot of different applications might want to have a read-write access to
+# some parts of a filesystem. It can be especially useful when your update mechanism
+# overwrites the whole rootfs, but you want your application data to be preserved
+# between updates. This class provides a way to achieve that by means
+# of overlayfs and at the same time keeping the base rootfs read-only.
+#
+# Usage example.
+#
+# Set a mount point for a partition overlayfs is going to use as upper layer
+# in your machine configuration. Underlying file system can be anything that
+# is supported by overlayfs. This has to be done in your machine configuration.
+# QA check fails to catch file existence if you redefine this variable in your recipe!
+#
+#   OVERLAYFS_MOUNT_POINT[data] ?= "/data"
+#
+# The class assumes you have a data.mount systemd unit defined in your
+# systemd-machine-units recipe and installed to the image.
+#
+# Then you can specify writable directories on a recipe base
+#
+#   OVERLAYFS_WRITABLE_PATHS[data] = "/usr/share/my-custom-application"
+#
+# To support several mount points you can use a different variable flag. Assume we
+# want to have a writable location on the file system, but not interested where the data
+# survive a reboot. Then we could have a mnt-overlay.mount unit for a tmpfs file system:
+#
+#   OVERLAYFS_MOUNT_POINT[mnt-overlay] = "/mnt/overlay"
+#   OVERLAYFS_WRITABLE_PATHS[mnt-overlay] = "/usr/share/another-application"
+#
+# Note: the class does not support /etc directory itself, because systemd depends on it
+
+REQUIRED_DISTRO_FEATURES += "systemd overlayfs"
+
+inherit systemd features_check overlayfs-qa
+
+def strForBash(s):
+    return s.replace('\\', '\\\\')
+
+def unitFileList(d):
+    fileList = []
+    overlayMountPoints = d.getVarFlags("OVERLAYFS_MOUNT_POINT")
+
+    if not overlayMountPoints:
+        bb.fatal("A recipe uses overlayfs class but there is no OVERLAYFS_MOUNT_POINT set in your MACHINE configuration")
+
+    # check that we have required mount points set first
+    requiredMountPoints = d.getVarFlags('OVERLAYFS_WRITABLE_PATHS')
+    for mountPoint in requiredMountPoints:
+        if mountPoint not in overlayMountPoints:
+            bb.fatal("Missing required mount point for OVERLAYFS_MOUNT_POINT[%s] in your MACHINE configuration" % mountPoint)
+
+    for mountPoint in overlayMountPoints:
+        for path in d.getVarFlag('OVERLAYFS_WRITABLE_PATHS', mountPoint).split():
+            fileList.append(mountUnitName(path))
+            fileList.append(helperUnitName(path))
+
+    return fileList
+
+def helperUnitName(unit):
+    return escapeSystemdUnitName(unit) + '-create-upper-dir.service'
+
+python do_create_overlayfs_units() {
+    CreateDirsUnitTemplate = """[Unit]
+Description=Overlayfs directories setup
+Requires={DATA_MOUNT_UNIT}
+After={DATA_MOUNT_UNIT}
+DefaultDependencies=no
+
+[Service]
+Type=oneshot
+ExecStart=mkdir -p {DATA_MOUNT_POINT}/workdir{LOWERDIR} && mkdir -p {DATA_MOUNT_POINT}/upper{LOWERDIR}
+RemainAfterExit=true
+StandardOutput=journal
+
+[Install]
+WantedBy=multi-user.target
+"""
+    MountUnitTemplate = """[Unit]
+Description=Overlayfs mount unit
+Requires={CREATE_DIRS_SERVICE}
+After={CREATE_DIRS_SERVICE}
+
+[Mount]
+What=overlay
+Where={LOWERDIR}
+Type=overlay
+Options=lowerdir={LOWERDIR},upperdir={DATA_MOUNT_POINT}/upper{LOWERDIR},workdir={DATA_MOUNT_POINT}/workdir{LOWERDIR}
+
+[Install]
+WantedBy=multi-user.target
+"""
+
+    def prepareUnits(data, lower):
+        args = {
+            'DATA_MOUNT_POINT': data,
+            'DATA_MOUNT_UNIT': mountUnitName(data),
+            'CREATE_DIRS_SERVICE': helperUnitName(lower),
+            'LOWERDIR': lower,
+        }
+
+        with open(os.path.join(d.getVar('WORKDIR'), mountUnitName(lower)), 'w') as f:
+            f.write(MountUnitTemplate.format(**args))
+
+        with open(os.path.join(d.getVar('WORKDIR'), helperUnitName(lower)), 'w') as f:
+            f.write(CreateDirsUnitTemplate.format(**args))
+
+    overlayMountPoints = d.getVarFlags("OVERLAYFS_MOUNT_POINT")
+    for mountPoint in overlayMountPoints:
+        for lower in d.getVarFlag('OVERLAYFS_WRITABLE_PATHS', mountPoint).split():
+            prepareUnits(d.getVarFlag('OVERLAYFS_MOUNT_POINT', mountPoint), lower)
+}
+
+# we need to generate file names early during parsing stage
+python () {
+    unitList = unitFileList(d)
+    for unit in unitList:
+        d.appendVar('SYSTEMD_SERVICE_' + d.getVar('PN'), ' ' + unit);
+        d.appendVar('FILES_' + d.getVar('PN'), ' ' + strForBash(unit))
+
+    d.setVar('OVERLAYFS_UNIT_LIST', ' '.join([strForBash(s) for s in unitList]))
+}
+
+do_install_append() {
+    install -d ${D}${systemd_system_unitdir}
+    for unit in ${OVERLAYFS_UNIT_LIST}; do
+        install -m 0444 ${WORKDIR}/${unit} ${D}${systemd_system_unitdir}
+    done
+}
+
+addtask create_overlayfs_units before do_install
-- 
2.28.0

[PATCH v3 3/8] maintainers.inc: overlayfs bbclass maintainer

[Vyacheslav Yurkov]

Signed-off-by: Vyacheslav Yurkov <uvv.mail@gmail.com>
---
 meta/conf/distro/include/maintainers.inc | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/meta/conf/distro/include/maintainers.inc b/meta/conf/distro/include/maintainers.inc
index e59f01d66a..90b20c9223 100644
--- a/meta/conf/distro/include/maintainers.inc
+++ b/meta/conf/distro/include/maintainers.inc
@@ -808,3 +808,5 @@ RECIPE_MAINTAINER_pn-xz = "Denys Dmytriyenko <denis@denix.org>"
 RECIPE_MAINTAINER_pn-zip = "Denys Dmytriyenko <denis@denix.org>"
 RECIPE_MAINTAINER_pn-zlib = "Denys Dmytriyenko <denis@denix.org>"
 RECIPE_MAINTAINER_pn-zstd = "Alexander Kanavin <alex.kanavin@gmail.com>"
+BBCLASS_MAINTAINER_overlayfs = "Vyacheslav Yurkov <uvv.mail@gmail.com>"
+BBCLASS_MAINTAINER_overlayfs-qa = "Vyacheslav Yurkov <uvv.mail@gmail.com>"
-- 
2.28.0

[PATCH v3 4/8] rootfs-postcommands: add QA check for overlayfs

[Vyacheslav Yurkov]

The check is conditional and only enabled when overlayfs is set in
DISTRO_FEATURES

Signed-off-by: Vyacheslav Yurkov <uvv.mail@gmail.com>
---
 meta/classes/rootfs-postcommands.bbclass | 24 ++++++++++++++++++++++++
 1 file changed, 24 insertions(+)

diff --git a/meta/classes/rootfs-postcommands.bbclass b/meta/classes/rootfs-postcommands.bbclass
index e66ed5938b..b6ddf5475c 100644
--- a/meta/classes/rootfs-postcommands.bbclass
+++ b/meta/classes/rootfs-postcommands.bbclass
@@ -39,7 +39,10 @@ ROOTFS_POSTPROCESS_COMMAND += '${@bb.utils.contains("DISTRO_FEATURES", "systemd"
 
 ROOTFS_POSTPROCESS_COMMAND += 'empty_var_volatile;'
 
+ROOTFS_POSTPROCESS_COMMAND += '${@bb.utils.contains("DISTRO_FEATURES", "overlayfs", "overlayfs_qa_check;", "", d)}'
+
 inherit image-artifact-names
+inherit ${@bb.utils.contains("DISTRO_FEATURES", "overlayfs", "overlayfs-qa", "", d)}
 
 # Sort the user and group entries in /etc by ID in order to make the content
 # deterministic. Package installs are not deterministic, causing the ordering
@@ -373,3 +376,24 @@ rootfs_reproducible () {
 		fi
 	fi
 }
+
+python overlayfs_qa_check() {
+    # this is a dumb check for unit existence, not its validity
+    overlayMountPoints = d.getVarFlags("OVERLAYFS_MOUNT_POINT")
+    imagepath = d.getVar("IMAGE_ROOTFS")
+    searchpaths = [oe.path.join(imagepath, d.getVar("sysconfdir"), "systemd", "system"),
+                   oe.path.join(imagepath, d.getVar("systemd_system_unitdir"))]
+
+    allUnitExist = True;
+    for mountPoint in overlayMountPoints:
+        path = d.getVarFlag('OVERLAYFS_MOUNT_POINT', mountPoint)
+        unit = mountUnitName(path)
+
+        if not any(os.path.isfile(oe.path.join(dirpath, unit))
+                   for dirpath in searchpaths):
+            bb.warn('Unit name %s not found in systemd unit directories' % unit)
+            allUnitExist = False;
+
+    if not allUnitExist:
+        bb.fatal('Not all mount units are installed by the BSP')
+}
-- 
2.28.0

[PATCH v3 5/8] systemd-machine-units: add bbappend for meta-selftest

[Vyacheslav Yurkov]

Signed-off-by: Vyacheslav Yurkov <uvv.mail@gmail.com>
---
 .../systemd-machine-units/systemd-machine-units_%.bbappend      | 2 ++
 1 file changed, 2 insertions(+)
 create mode 100644 meta-selftest/recipes-test/systemd-machine-units/systemd-machine-units_%.bbappend

diff --git a/meta-selftest/recipes-test/systemd-machine-units/systemd-machine-units_%.bbappend b/meta-selftest/recipes-test/systemd-machine-units/systemd-machine-units_%.bbappend
new file mode 100644
index 0000000000..205720982c
--- /dev/null
+++ b/meta-selftest/recipes-test/systemd-machine-units/systemd-machine-units_%.bbappend
@@ -0,0 +1,2 @@
+# This bbappend is used to alter the recipe using the test_recipe.inc file created by tests.
+include test_recipe.inc
-- 
2.28.0

[PATCH v3 6/8] overlayfs: meta-selftest recipe

[Vyacheslav Yurkov]

The recipe demonstrates example usage of overlayfs bbclass

Signed-off-by: Vyacheslav Yurkov <uvv.mail@gmail.com>
---
 .../overlayfs-user/overlayfs-user.bb            | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)
 create mode 100644 meta-selftest/recipes-test/overlayfs-user/overlayfs-user.bb

diff --git a/meta-selftest/recipes-test/overlayfs-user/overlayfs-user.bb b/meta-selftest/recipes-test/overlayfs-user/overlayfs-user.bb
new file mode 100644
index 0000000000..5c7be7da97
--- /dev/null
+++ b/meta-selftest/recipes-test/overlayfs-user/overlayfs-user.bb
@@ -0,0 +1,17 @@
+SUMMARY = "Overlayfs class unit test"
+DESCRIPTION = "Contains an overlayfs configuration"
+LICENSE = "MIT"
+
+INHIBIT_DEFAULT_DEPS = "1"
+EXCLUDE_FROM_WORLD = "1"
+
+inherit ${@bb.utils.contains("DISTRO_FEATURES", "overlayfs", "overlayfs", "", d)}
+include test_recipe.inc
+
+OVERLAYFS_WRITABLE_PATHS[mnt-overlay] = "/usr/share/my-application"
+
+do_install() {
+    install -d ${D}/usr/share/my-application
+}
+
+FILES_${PN} += "/usr"
-- 
2.28.0

[PATCH v3 7/8] oeqa/selftest: overlayfs unit tests

[Vyacheslav Yurkov]

Unit tests for overlayfs.bbclass

Signed-off-by: Vyacheslav Yurkov <uvv.mail@gmail.com>
---
 meta/lib/oeqa/selftest/cases/overlayfs.py | 171 ++++++++++++++++++++++
 1 file changed, 171 insertions(+)
 create mode 100644 meta/lib/oeqa/selftest/cases/overlayfs.py

diff --git a/meta/lib/oeqa/selftest/cases/overlayfs.py b/meta/lib/oeqa/selftest/cases/overlayfs.py
new file mode 100644
index 0000000000..f679a71182
--- /dev/null
+++ b/meta/lib/oeqa/selftest/cases/overlayfs.py
@@ -0,0 +1,171 @@
+#
+# SPDX-License-Identifier: MIT
+#
+
+from oeqa.selftest.case import OESelftestTestCase
+from oeqa.utils.commands import runCmd, bitbake, get_bb_var, runqemu
+
+class OverlayFSTests(OESelftestTestCase):
+    """Overlayfs class usage tests"""
+
+    def getline(self, res, line):
+        for l in res.output.split('\n'):
+            if line in l:
+                return l
+
+    def add_overlay_conf_to_machine(self):
+        machine_inc = """
+OVERLAYFS_MOUNT_POINT[mnt-overlay] = "/mnt/overlay"
+"""
+        self.set_machine_config(machine_inc)
+
+    def test_distro_features_missing(self):
+        """
+        Summary:   Check that required DISTRO_FEATURES are set
+        Expected:  Fail when either systemd or overlayfs are not in DISTRO_FEATURES
+        Author:    Vyacheslav Yurkov <uvv.mail@gmail.com>
+        """
+
+        config = """
+IMAGE_INSTALL_append = " overlayfs-user"
+"""
+        overlayfs_recipe_append = """
+inherit overlayfs
+"""
+        self.write_config(config)
+        self.add_overlay_conf_to_machine()
+        self.write_recipeinc('overlayfs-user', overlayfs_recipe_append)
+
+        res = bitbake('core-image-minimal', ignore_status=True)
+        line = self.getline(res, "overlayfs-user was skipped: missing required distro features")
+        self.assertTrue("overlayfs" in res.output, msg=res.output)
+        self.assertTrue("systemd" in res.output, msg=res.output)
+        self.assertTrue("ERROR: Required build target 'core-image-minimal' has no buildable providers." in res.output, msg=res.output)
+
+    def test_not_all_units_installed(self):
+        """
+        Summary:   Test QA check that we have required mount units in the image
+        Expected:  Fail because mount unit for overlay partition is not installed
+        Author:    Vyacheslav Yurkov <uvv.mail@gmail.com>
+        """
+
+        config = """
+IMAGE_INSTALL_append = " overlayfs-user"
+DISTRO_FEATURES += "systemd overlayfs"
+"""
+
+        self.write_config(config)
+        self.add_overlay_conf_to_machine()
+
+        res = bitbake('core-image-minimal', ignore_status=True)
+        line = self.getline(res, "Unit name mnt-overlay.mount not found in systemd unit directories")
+        self.assertTrue(line and line.startswith("WARNING:"), msg=res.output)
+        line = self.getline(res, "Not all mount units are installed by the BSP")
+        self.assertTrue(line and line.startswith("ERROR:"), msg=res.output)
+
+    def test_mount_unit_not_set(self):
+        """
+        Summary:   Test whether mount unit was set properly
+        Expected:  Fail because mount unit was not set
+        Author:    Vyacheslav Yurkov <uvv.mail@gmail.com>
+        """
+
+        config = """
+IMAGE_INSTALL_append = " overlayfs-user"
+DISTRO_FEATURES += "systemd overlayfs"
+"""
+
+        self.write_config(config)
+
+        res = bitbake('core-image-minimal', ignore_status=True)
+        line = self.getline(res, "A recipe uses overlayfs class but there is no OVERLAYFS_MOUNT_POINT set in your MACHINE configuration")
+        self.assertTrue(line and line.startswith("Parsing recipes...ERROR:"), msg=res.output)
+
+    def test_wrong_mount_unit_set(self):
+        """
+        Summary:   Test whether mount unit was set properly
+        Expected:  Fail because not the correct flag used for mount unit
+        Author:    Vyacheslav Yurkov <uvv.mail@gmail.com>
+        """
+
+        config = """
+IMAGE_INSTALL_append = " overlayfs-user"
+DISTRO_FEATURES += "systemd overlayfs"
+"""
+
+        wrong_machine_config = """
+OVERLAYFS_MOUNT_POINT[usr-share-overlay] = "/usr/share/overlay"
+"""
+
+        self.write_config(config)
+        self.set_machine_config(wrong_machine_config)
+
+        res = bitbake('core-image-minimal', ignore_status=True)
+        line = self.getline(res, "Missing required mount point for OVERLAYFS_MOUNT_POINT[mnt-overlay] in your MACHINE configuration")
+        self.assertTrue(line and line.startswith("Parsing recipes...ERROR:"), msg=res.output)
+
+    def test_correct_image(self):
+        """
+        Summary:   Check that we can create an image when all parameters are
+                   set correctly
+        Expected:  Image is created successfully
+        Author:    Vyacheslav Yurkov <uvv.mail@gmail.com>
+        """
+
+        config = """
+IMAGE_INSTALL_append = " overlayfs-user systemd-machine-units"
+DISTRO_FEATURES += "systemd overlayfs"
+
+# Use systemd as init manager
+VIRTUAL-RUNTIME_init_manager = "systemd"
+
+# enable overlayfs in the kernel
+KERNEL_EXTRA_FEATURES_append = " features/overlayfs/overlayfs.scc"
+"""
+
+        systemd_machine_unit_append = """
+SYSTEMD_SERVICE_${PN} += " \
+    mnt-overlay.mount \
+"
+
+do_install_append() {
+    install -d ${D}${systemd_system_unitdir}
+    cat <<EOT > ${D}${systemd_system_unitdir}/mnt-overlay.mount
+[Unit]
+Description=Tmpfs directory
+DefaultDependencies=no
+
+[Mount]
+What=tmpfs
+Where=/mnt/overlay
+Type=tmpfs
+Options=mode=1777,strictatime,nosuid,nodev
+
+[Install]
+WantedBy=multi-user.target
+EOT
+}
+
+"""
+
+        self.write_config(config)
+        self.add_overlay_conf_to_machine()
+        self.write_recipeinc('systemd-machine-units', systemd_machine_unit_append)
+
+        bitbake('core-image-minimal')
+
+        def getline_qemu(out, line):
+            for l in out.split('\n'):
+                if line in l:
+                    return l
+
+        with runqemu('core-image-minimal') as qemu:
+            # Check that we have /mnt/overlay fs mounted as tmpfs and
+            # /usr/share/my-application as an overlay (see overlayfs-user recipe)
+            status, output = qemu.run_serial("/bin/mount -t tmpfs,overlay")
+
+            line = getline_qemu(output, "on /mnt/overlay")
+            self.assertTrue(line and line.startswith("tmpfs"), msg=output)
+
+            line = getline_qemu(output, "upperdir=/mnt/overlay/upper/usr/share/my-application")
+            self.assertTrue(line and line.startswith("overlay"), msg=output)
-- 
2.28.0

[PATCH v3 8/8] docs: add overlayfs class

[Vyacheslav Yurkov]

Signed-off-by: Vyacheslav Yurkov <uvv.mail@gmail.com>
---
 documentation/ref-manual/classes.rst | 46 ++++++++++++++++++++++++++++
 1 file changed, 46 insertions(+)

diff --git a/documentation/ref-manual/classes.rst b/documentation/ref-manual/classes.rst
index 09878c480f..6d80e7dc4f 100644
--- a/documentation/ref-manual/classes.rst
+++ b/documentation/ref-manual/classes.rst
@@ -1708,6 +1708,52 @@ one such example. However, being aware of this class can reduce the
 proliferation of different versions of similar classes across multiple
 layers.
 
+.. _ref-classes-overlayfs:
+
+``overlayfs.bbclass``
+=======================
+
+
+It's often desired in Embedded System design to have a read-only rootfs.
+But a lot of different applications might want to have a read-write access to
+some parts of a filesystem. It can be especially useful when your update mechanism
+overwrites the whole rootfs, but you want your application data to be preserved
+between updates. The ``overlayfs`` class provides a way to achieve that by means
+of overlayfs and at the same time keeping the base rootfs read-only.
+
+To use this class, set a mount point for a partition overlayfs is going to use as upper
+layer in your machine configuration. Underlying file system can be anything that
+is supported by overlayfs. This has to be done in your machine configuration::
+
+  OVERLAYFS_MOUNT_POINT[data] = "/data"
+
+.. note::
+
+  QA check fails to catch file existence if you redefine this variable in your recipe!
+
+The class assumes you have a data.mount systemd unit defined elsewhere in your BSP
+(e.g. in systemd-machine-units recipe) and it's installed to the image.
+
+Then you can specify writable directories on a recipe basis::
+
+  OVERLAYFS_WRITABLE_PATHS[data] = "/usr/share/my-custom-application"
+
+To support several mount points you can use a different variable flag. Assume we
+want to have a writable location on the file system, but not interested where the data
+survive a reboot. Then we could have a mnt-overlay.mount unit for a tmpfs file system.
+
+In your machine configuration::
+
+  OVERLAYFS_MOUNT_POINT[mnt-overlay] = "/mnt/overlay"
+
+and then in your recipe::
+
+  OVERLAYFS_WRITABLE_PATHS[mnt-overlay] = "/usr/share/another-application"
+
+.. note::
+
+   The class does not support /etc directory itself, because systemd depends on it
+
 .. _ref-classes-own-mirrors:
 
 ``own-mirrors.bbclass``
-- 
2.28.0

Re: [OE-core] [PATCH v3 1/8] overlayfs-qa: common functions for overlayfs

[Richard Purdie]

Hi,

On Thu, 2021-07-15 at 21:39 +0200, Vyacheslav Yurkov wrote:
> This class provides commom functions for overlayfs and its QA check,
> which is performed in ROOTFS_POSTPROCESS_COMMAND
> 
> Signed-off-by: Vyacheslav Yurkov <uvv.mail@gmail.com>
> ---
>  meta/classes/overlayfs-qa.bbclass | 14 ++++++++++++++
>  1 file changed, 14 insertions(+)
>  create mode 100644 meta/classes/overlayfs-qa.bbclass
> 
> diff --git a/meta/classes/overlayfs-qa.bbclass b/meta/classes/overlayfs-qa.bbclass
> new file mode 100644
> index 0000000000..54fa8316a2
> --- /dev/null
> +++ b/meta/classes/overlayfs-qa.bbclass
> @@ -0,0 +1,14 @@
> +# This class contains common functions for overlayfs and its QA check,
> +# which is performed in ROOTFS_POSTPROCESS_COMMAND
> +
> +# this function is based on https://github.com/systemd/systemd/blob/main/src/basic/unit-name.c
> +def escapeSystemdUnitName(path):
> +    escapeMap = {
> +        '/': '-',
> +        '-': "\\x2d",
> +        '\\': "\\x5d"
> +    }
> +    return "".join([escapeMap.get(c, c) for c in path.strip('/')])
> +
> +def mountUnitName(unit):
> +    return escapeSystemdUnitName(unit) + '.mount'


Thanks for working on this, it is passing our automated tests but I had 
some review comments. I think the above would make much more sense as code
in meta/lib/oe/ somewhere as it is basically just function library code.
There is a general intent to move the more generic python functions there
and try to establish some better APIs so if you could move this there as
new code, that would be great!

Cheers,

Richard

Re: [OE-core] [PATCH v3 3/8] maintainers.inc: overlayfs bbclass maintainer

[Richard Purdie]

On Thu, 2021-07-15 at 21:39 +0200, Vyacheslav Yurkov wrote:
> Signed-off-by: Vyacheslav Yurkov <uvv.mail@gmail.com>
> ---
>  meta/conf/distro/include/maintainers.inc | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/meta/conf/distro/include/maintainers.inc b/meta/conf/distro/include/maintainers.inc
> index e59f01d66a..90b20c9223 100644
> --- a/meta/conf/distro/include/maintainers.inc
> +++ b/meta/conf/distro/include/maintainers.inc
> @@ -808,3 +808,5 @@ RECIPE_MAINTAINER_pn-xz = "Denys Dmytriyenko <denis@denix.org>"
>  RECIPE_MAINTAINER_pn-zip = "Denys Dmytriyenko <denis@denix.org>"
>  RECIPE_MAINTAINER_pn-zlib = "Denys Dmytriyenko <denis@denix.org>"
>  RECIPE_MAINTAINER_pn-zstd = "Alexander Kanavin <alex.kanavin@gmail.com>"
> +BBCLASS_MAINTAINER_overlayfs = "Vyacheslav Yurkov <uvv.mail@gmail.com>"
> +BBCLASS_MAINTAINER_overlayfs-qa = "Vyacheslav Yurkov <uvv.mail@gmail.com>"

This variable doesn't exist. 

I've been trying to figure out how to better establish maintainers and 
there is a MAINTAINERS.md patch queued to create a maintainers file. I 
think this entry could be better off in there? Could you create a follow
on patch for that file instead please?

Cheers,

Richard

Re: [OE-core] [PATCH v3 2/8] overlayfs.bbclass: generate overlayfs mount units

[Richard Purdie]

On Thu, 2021-07-15 at 21:39 +0200, Vyacheslav Yurkov wrote:
> It's often desired in Embedded System design to have a read-only rootfs.
> But a lot of different applications might want to have a read-write access
> to some parts of a filesystem. It can be especially useful when your update
> mechanism overwrites the whole rootfs, but you want your application data
> to be preserved between updates. This class provides a way to achieve that
> by means of overlayfs and at the same time keeping the base rootfs read-only.
> 
> Signed-off-by: Vyacheslav Yurkov <uvv.mail@gmail.com>
> ---
>  meta/classes/overlayfs.bbclass | 133 +++++++++++++++++++++++++++++++++
>  1 file changed, 133 insertions(+)
>  create mode 100644 meta/classes/overlayfs.bbclass
> 
> diff --git a/meta/classes/overlayfs.bbclass b/meta/classes/overlayfs.bbclass
> new file mode 100644
> index 0000000000..40e6107f0d
> --- /dev/null
> +++ b/meta/classes/overlayfs.bbclass
> @@ -0,0 +1,133 @@
> +# Class for generation of overlayfs mount units
> +#
> +# It's often desired in Embedded System design to have a read-only rootfs.
> +# But a lot of different applications might want to have a read-write access to
> +# some parts of a filesystem. It can be especially useful when your update mechanism
> +# overwrites the whole rootfs, but you want your application data to be preserved
> +# between updates. This class provides a way to achieve that by means
> +# of overlayfs and at the same time keeping the base rootfs read-only.
> +#
> +# Usage example.
> +#
> +# Set a mount point for a partition overlayfs is going to use as upper layer
> +# in your machine configuration. Underlying file system can be anything that
> +# is supported by overlayfs. This has to be done in your machine configuration.
> +# QA check fails to catch file existence if you redefine this variable in your recipe!
> +#
> +#   OVERLAYFS_MOUNT_POINT[data] ?= "/data"
> +#
> +# The class assumes you have a data.mount systemd unit defined in your
> +# systemd-machine-units recipe and installed to the image.
> +#
> +# Then you can specify writable directories on a recipe base
> +#
> +#   OVERLAYFS_WRITABLE_PATHS[data] = "/usr/share/my-custom-application"
> +#
> +# To support several mount points you can use a different variable flag. Assume we
> +# want to have a writable location on the file system, but not interested where the data
> +# survive a reboot. Then we could have a mnt-overlay.mount unit for a tmpfs file system:
> +#
> +#   OVERLAYFS_MOUNT_POINT[mnt-overlay] = "/mnt/overlay"
> +#   OVERLAYFS_WRITABLE_PATHS[mnt-overlay] = "/usr/share/another-application"
> +#
> +# Note: the class does not support /etc directory itself, because systemd depends on it
> +
> +REQUIRED_DISTRO_FEATURES += "systemd overlayfs"
> +
> +inherit systemd features_check overlayfs-qa
> +
> +def strForBash(s):
> +    return s.replace('\\', '\\\\')
> +
> +def unitFileList(d):
> +    fileList = []
> +    overlayMountPoints = d.getVarFlags("OVERLAYFS_MOUNT_POINT")
> +
> +    if not overlayMountPoints:
> +        bb.fatal("A recipe uses overlayfs class but there is no OVERLAYFS_MOUNT_POINT set in your MACHINE configuration")
> +
> +    # check that we have required mount points set first
> +    requiredMountPoints = d.getVarFlags('OVERLAYFS_WRITABLE_PATHS')
> +    for mountPoint in requiredMountPoints:
> +        if mountPoint not in overlayMountPoints:
> +            bb.fatal("Missing required mount point for OVERLAYFS_MOUNT_POINT[%s] in your MACHINE configuration" % mountPoint)
> +
> +    for mountPoint in overlayMountPoints:
> +        for path in d.getVarFlag('OVERLAYFS_WRITABLE_PATHS', mountPoint).split():
> +            fileList.append(mountUnitName(path))
> +            fileList.append(helperUnitName(path))
> +
> +    return fileList
> +
> +def helperUnitName(unit):
> +    return escapeSystemdUnitName(unit) + '-create-upper-dir.service'

Again, I'm a little torn on some of the functions. Would some of these
be better as part of a function library in meta/lib/oe/ ?

The downside is that variable dependencies are no longer automatic so
the dependency on OVERLAYFS_WRITABLE_PATHS may be missed, I don't think
it would see OVERLAYFS_MOUNT_POINT as I'm not sure the magic parsing
code understand getVarFlags (compared to getVarFlag which it does).

Cheers,

Richard

Re: [OE-core] [PATCH v3 2/8] overlayfs.bbclass: generate overlayfs mount units

[Vyacheslav Yurkov]

On 22.07.2021 23:13, Richard Purdie wrote:
> On Thu, 2021-07-15 at 21:39 +0200, Vyacheslav Yurkov wrote:
>> It's often desired in Embedded System design to have a read-only rootfs.
>> But a lot of different applications might want to have a read-write access
>> to some parts of a filesystem. It can be especially useful when your update
>> mechanism overwrites the whole rootfs, but you want your application data
>> to be preserved between updates. This class provides a way to achieve that
>> by means of overlayfs and at the same time keeping the base rootfs read-only.
>>
>> Signed-off-by: Vyacheslav Yurkov <uvv.mail@gmail.com>
>> ---
>>   meta/classes/overlayfs.bbclass | 133 +++++++++++++++++++++++++++++++++
>>   1 file changed, 133 insertions(+)
>>   create mode 100644 meta/classes/overlayfs.bbclass
>>
>> diff --git a/meta/classes/overlayfs.bbclass b/meta/classes/overlayfs.bbclass
>> new file mode 100644
>> index 0000000000..40e6107f0d
>> --- /dev/null
>> +++ b/meta/classes/overlayfs.bbclass
>> @@ -0,0 +1,133 @@
>> +# Class for generation of overlayfs mount units
>> +#
>> +# It's often desired in Embedded System design to have a read-only rootfs.
>> +# But a lot of different applications might want to have a read-write access to
>> +# some parts of a filesystem. It can be especially useful when your update mechanism
>> +# overwrites the whole rootfs, but you want your application data to be preserved
>> +# between updates. This class provides a way to achieve that by means
>> +# of overlayfs and at the same time keeping the base rootfs read-only.
>> +#
>> +# Usage example.
>> +#
>> +# Set a mount point for a partition overlayfs is going to use as upper layer
>> +# in your machine configuration. Underlying file system can be anything that
>> +# is supported by overlayfs. This has to be done in your machine configuration.
>> +# QA check fails to catch file existence if you redefine this variable in your recipe!
>> +#
>> +#   OVERLAYFS_MOUNT_POINT[data] ?= "/data"
>> +#
>> +# The class assumes you have a data.mount systemd unit defined in your
>> +# systemd-machine-units recipe and installed to the image.
>> +#
>> +# Then you can specify writable directories on a recipe base
>> +#
>> +#   OVERLAYFS_WRITABLE_PATHS[data] = "/usr/share/my-custom-application"
>> +#
>> +# To support several mount points you can use a different variable flag. Assume we
>> +# want to have a writable location on the file system, but not interested where the data
>> +# survive a reboot. Then we could have a mnt-overlay.mount unit for a tmpfs file system:
>> +#
>> +#   OVERLAYFS_MOUNT_POINT[mnt-overlay] = "/mnt/overlay"
>> +#   OVERLAYFS_WRITABLE_PATHS[mnt-overlay] = "/usr/share/another-application"
>> +#
>> +# Note: the class does not support /etc directory itself, because systemd depends on it
>> +
>> +REQUIRED_DISTRO_FEATURES += "systemd overlayfs"
>> +
>> +inherit systemd features_check overlayfs-qa
>> +
>> +def strForBash(s):
>> +    return s.replace('\\', '\\\\')
>> +
>> +def unitFileList(d):
>> +    fileList = []
>> +    overlayMountPoints = d.getVarFlags("OVERLAYFS_MOUNT_POINT")
>> +
>> +    if not overlayMountPoints:
>> +        bb.fatal("A recipe uses overlayfs class but there is no OVERLAYFS_MOUNT_POINT set in your MACHINE configuration")
>> +
>> +    # check that we have required mount points set first
>> +    requiredMountPoints = d.getVarFlags('OVERLAYFS_WRITABLE_PATHS')
>> +    for mountPoint in requiredMountPoints:
>> +        if mountPoint not in overlayMountPoints:
>> +            bb.fatal("Missing required mount point for OVERLAYFS_MOUNT_POINT[%s] in your MACHINE configuration" % mountPoint)
>> +
>> +    for mountPoint in overlayMountPoints:
>> +        for path in d.getVarFlag('OVERLAYFS_WRITABLE_PATHS', mountPoint).split():
>> +            fileList.append(mountUnitName(path))
>> +            fileList.append(helperUnitName(path))
>> +
>> +    return fileList
>> +
>> +def helperUnitName(unit):
>> +    return escapeSystemdUnitName(unit) + '-create-upper-dir.service'
> Again, I'm a little torn on some of the functions. Would some of these
> be better as part of a function library in meta/lib/oe/ ?
>
> The downside is that variable dependencies are no longer automatic so
> the dependency on OVERLAYFS_WRITABLE_PATHS may be missed, I don't think
> it would see OVERLAYFS_MOUNT_POINT as I'm not sure the magic parsing
> code understand getVarFlags (compared to getVarFlag which it does).
>
> Cheers,
>
> Richard

Sorry for the late response. Do you mean to move only helper functions 
to meta/lib/oe, right?
By variable dependencies you mean that bitbake would not pick up meta 
data changes automatically?

I will try to move common parts to meta/lib/oe then and send a v4

Thanks,
Vyacheslav

Re: [OE-core] [PATCH v3 2/8] overlayfs.bbclass: generate overlayfs mount units

[Richard Purdie]

On Fri, 2021-07-30 at 07:14 +0200, Vyacheslav Yurkov wrote:
> Sorry for the late response. Do you mean to move only helper functions 
> to meta/lib/oe, right?

Correct, yes. The things being defined as "def <function>".

> By variable dependencies you mean that bitbake would not pick up meta 
> data changes automatically?

I mean that if a function is something like:

def function(d):
   d.getVar("VAR")

it will no longer magically depend on the variable VAR.

> I will try to move common parts to meta/lib/oe then and send a v4

Thanks!

Cheers,

Richard