[PATCH v4 15/20] binman: Support simple templates
Simon Glass
sjg at chromium.org
Tue Jul 11 16:59:17 CEST 2023
Collections can used to collect the contents of other entries into a
single entry, but they result in a single entry, with the original entries
'left behind' in their old place.
It is useful to be able to specific a set of entries ones and have it used
in multiple images, or parts of an image.
Implement this mechanism.
Signed-off-by: Simon Glass <sjg at chromium.org>
---
Changes in v4:
- Copy over properties from the top-level template node
Changes in v2:
- Correct ordering of template nodes
- Fix 'preseverd' and 'inserter' typos
tools/binman/binman.rst | 87 ++++++++++++++++++++++++++++++
tools/binman/control.py | 26 +++++++++
tools/binman/etype/section.py | 3 +-
tools/binman/ftest.py | 8 +++
tools/binman/test/286_template.dts | 42 +++++++++++++++
5 files changed, 165 insertions(+), 1 deletion(-)
create mode 100644 tools/binman/test/286_template.dts
diff --git a/tools/binman/binman.rst b/tools/binman/binman.rst
index a4b31fe5397b..8f57b6cfc76f 100644
--- a/tools/binman/binman.rst
+++ b/tools/binman/binman.rst
@@ -727,6 +727,13 @@ optional:
Note that missing, optional blobs do not produce a non-zero exit code from
binman, although it does show a warning about the missing external blob.
+insert-template:
+ This is not strictly speaking an entry property, since it is processed early
+ in Binman before the entries are read. It is a list of phandles of nodes to
+ include in the current (target) node. For each node, its subnodes and their
+ properties are brought into the target node. See Templates_ below for
+ more information.
+
The attributes supported for images and sections are described below. Several
are similar to those for entries.
@@ -1172,6 +1179,86 @@ If you are having trouble figuring out what is going on, you can use
arch/arm/dts/u-boot.dtsi ... found: "arch/arm/dts/juno-r2-u-boot.dtsi"
+Templates
+=========
+
+Sometimes multiple images need to be created which have all have a common
+part. For example, a board may generate SPI and eMMC images which both include
+a FIT. Since the FIT includes many entries, it is tedious to repeat them twice
+in the image description.
+
+Templates provide a simple way to handle this::
+
+ binman {
+ multiple-images;
+ common_part: template-1 {
+ some-property;
+ fit {
+ ... lots of entries in here
+ };
+
+ text {
+ text = "base image";
+ };
+ };
+
+ spi-image {
+ filename = "image-spi.bin";
+ insert-template = <&fit>;
+
+ /* things specific to SPI follow */
+ footer {
+ ];
+
+ text {
+ text = "SPI image";
+ };
+ };
+
+ mmc-image {
+ filename = "image-mmc.bin";
+ insert-template = <&fit>;
+
+ /* things specific to MMC follow */
+ footer {
+ ];
+
+ text {
+ text = "MMC image";
+ };
+ };
+ };
+
+The template node name must start with 'template', so it is not considered to be
+an image itself.
+
+The mechanism is very simple. For each phandle in the 'insert-templates'
+property, the source node is looked up. Then the subnodes of that source node
+are copied into the target node, i.e. the one containing the `insert-template`
+property.
+
+If the target node has a node with the same name as a template, its properties
+override corresponding properties in the template. This allows the template to
+be uses as a base, with the node providing updates to the properties as needed.
+The overriding happens recursively.
+
+Template nodes appear first in each node that they are inserted into and
+ordering of template nodes is preserved. Other nodes come afterwards. If a
+template node also appears in the target node, then the template node sets the
+order. Thus the template can be used to set the ordering, even if the target
+node provides all the properties. In the above example, `fit` and `text` appear
+first in the `spi-image` and `mmc-image` images, followed by `footer`.
+
+Where there are multiple template nodes, they are inserted in that order. so
+the first template node appears first, then the second.
+
+Properties in the template node are inserted into the destination node if they
+do not exist there. In the example above, `some-property` is added to each of
+`spi-image` and `mmc-image`.
+
+Note that template nodes are not removed from the binman description at present.
+
+
Updating an ELF file
====================
diff --git a/tools/binman/control.py b/tools/binman/control.py
index 7e2dd3541b96..e9c4a65a75a1 100644
--- a/tools/binman/control.py
+++ b/tools/binman/control.py
@@ -22,6 +22,7 @@ from binman import bintool
from binman import cbfs_util
from binman import elf
from binman import entry
+from dtoc import fdt_util
from u_boot_pylib import command
from u_boot_pylib import tools
from u_boot_pylib import tout
@@ -478,6 +479,29 @@ def SignEntries(image_fname, input_fname, privatekey_fname, algo, entry_paths,
AfterReplace(image, allow_resize=True, write_map=write_map)
+def _ProcessTemplates(parent):
+ """Handle any templates in the binman description
+
+ Args:
+ parent: Binman node to process (typically /binman)
+
+ Search though each target node looking for those with an 'insert-template'
+ property. Use that as a list of references to template nodes to use to
+ adjust the target node.
+
+ Processing involves copying each subnode of the template node into the
+ target node.
+
+ For now this is not done recursively, so templates must be at the top level
+ of the binman image.
+
+ See 'Templates' in the Binman documnentation for details.
+ """
+ for node in parent.subnodes:
+ tmpl = fdt_util.GetPhandleList(node, 'insert-template')
+ if tmpl:
+ node.copy_subnodes_from_phandles(tmpl)
+
def PrepareImagesAndDtbs(dtb_fname, select_images, update_fdt, use_expanded):
"""Prepare the images to be processed and select the device tree
@@ -520,6 +544,8 @@ def PrepareImagesAndDtbs(dtb_fname, select_images, update_fdt, use_expanded):
raise ValueError("Device tree '%s' does not have a 'binman' "
"node" % dtb_fname)
+ _ProcessTemplates(node)
+
images = _ReadImageDesc(node, use_expanded)
if select_images:
diff --git a/tools/binman/etype/section.py b/tools/binman/etype/section.py
index d9b9e428024a..7c4d312c16c0 100644
--- a/tools/binman/etype/section.py
+++ b/tools/binman/etype/section.py
@@ -179,7 +179,8 @@ class Entry_section(Entry):
Returns:
bool: True if the node is a special one, else False
"""
- return node.name.startswith('hash') or node.name.startswith('signature')
+ start_list = ('hash', 'signature', 'template')
+ return any(node.name.startswith(name) for name in start_list)
def ReadNode(self):
"""Read properties from the section node"""
diff --git a/tools/binman/ftest.py b/tools/binman/ftest.py
index dc9a95d341e5..fc5d8a839e6b 100644
--- a/tools/binman/ftest.py
+++ b/tools/binman/ftest.py
@@ -6785,6 +6785,14 @@ fdt fdtmap Extract the devicetree blob from the fdtmap
data = self._DoReadFileDtb('285_spl_expand.dts',
use_expanded=True, entry_args=entry_args)[0]
+ def testTemplate(self):
+ """Test using a template"""
+ TestFunctional._MakeInputFile('vga2.bin', b'#' + VGA_DATA)
+ data = self._DoReadFile('286_template.dts')
+ first = U_BOOT_DATA + VGA_DATA + U_BOOT_DTB_DATA
+ second = U_BOOT_DATA + b'#' + VGA_DATA + U_BOOT_DTB_DATA
+ self.assertEqual(U_BOOT_IMG_DATA + first + second, data)
+
if __name__ == "__main__":
unittest.main()
diff --git a/tools/binman/test/286_template.dts b/tools/binman/test/286_template.dts
new file mode 100644
index 000000000000..6980dbfafcc6
--- /dev/null
+++ b/tools/binman/test/286_template.dts
@@ -0,0 +1,42 @@
+// SPDX-License-Identifier: GPL-2.0+
+
+/dts-v1/;
+
+/ {
+ #address-cells = <1>;
+ #size-cells = <1>;
+
+ binman {
+ u-boot-img {
+ };
+
+ common_part: template {
+ u-boot {
+ };
+
+ intel-vga {
+ filename = "vga.bin";
+ };
+ };
+
+ first {
+ type = "section";
+ insert-template = <&common_part>;
+
+ u-boot-dtb {
+ };
+ };
+
+ second {
+ type = "section";
+ insert-template = <&common_part>;
+
+ u-boot-dtb {
+ };
+
+ intel-vga {
+ filename = "vga2.bin";
+ };
+ };
+ };
+};
--
2.41.0.390.g38632f3daf-goog
More information about the U-Boot
mailing list