Add a function to modify inplace only a portion of a property..
This is especially useful when the property is an array of values, and you
want to update one of them without changing the DT size.
Acked-by: Simon Glass <sjg@chromium.org>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
[dwg: Remove unnecessary unsigned qualifier, correct a comment]
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Add a function to retrieve the highest phandle in a given device tree.
Acked-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Stefan Agner <stefan@agner.ch>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Implement a macro based on fdt_first_property_offset and
fdt_next_property_offset that provides a convenience to iterate over all
the properties of a given node.
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
Acked-by: Simon Glass <sjg@chromium.org>
[dwg: Removed a stray trailing blank line]
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
The fdt_for_each_subnode() iterator macro provided by this patch can be
used to iterate over a device tree node's subnodes. At each iteration a
loop variable will be set to the next subnode.
Signed-off-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
This patch allows dtc to accept multiple /dts-v1/ tags (provided they're
all at the beginning of the input), rather than giving a syntax error.
This makes it more convenient to include one .dts file from another without
having to be careful that the /dts-v1/ tag is in exactly one of them.
We a couple of existing testcases to take advantage of this, which
simplifies them slightly.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
ePAPR 1.1 section 2.2.1.1 "Node Name Requirements" specifies that any
node that has a reg property must include a unit address in its name
with value matching the first entry in its reg property. Conversely, if
a node does not have a reg property, the node name must not include a
unit address. Also allow ranges property as it is deemed valid, but ePAPR
is not clear about it.
Implement a check for this. The code doesn't validate the format of the
unit address; ePAPR implies this may vary from (containing bus) binding
to binding, so doing so would be much more complex.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
[robh: also allow non-empty ranges]
Signed-off-by: Rob Herring <robh@kernel.org>
[moved new test in check_table]
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
We have a couple of checks of the form:
if (offset+size > totalsize)
die();
We need to check that offset+size doesn't overflow, otherwise the check
will pass, and we may access past totalsize.
Found with AFL.
Signed-off-by: Anton Blanchard <anton@samba.org>
[Added a testcase]
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
1937095 "Prevent crash on division by zero" fixed a crash when attempting
a division by zero using the / operator in a dts. However, it missed the
precisely equivalent crash with the % (modulus) operator. This patch fixes
the oversight.
Reported-by: Anton Blanchard <anton@samba.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
In order to work with preprocessed dts files more easily, dts will parse
line number information in the form emitted by cpp.
Anton Blanchard (using a fuzzer) reported that including a line number
directive with a nul character (a literal nul in the input file, not a \0
sequence) would cause dtc to SEGV. I spotted several more problems on
examining the code:
* It modified yytext in place which seems to work, but is ugly and I'm
not sure if it's safe on all lex/flex versions
* The regexp used in the lexer to recognize line number information
accepts strings with escape characters, but it won't process these
escapes.
- GNU cpp at least, will generate \ escapes in line number
information, at least with files containing " or \ in the name
This patch reworks the handling of line number information to address
these problems. \ escapes should now be handled directly. nuls in file
names (either with a literal nul in the input file, or with a \0 escape
sequence) are still not permitted, but will now result in a lexical error
rather than a SEGV.
Reported-by: Anton Blanchard <anton@samba.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
If a dts file contains a string with \ followed by a nul byte - an actual
nul in the input file, not the \\0 escape - then the assert() in
get_escape_char() will trip, crashing dtc.
As far as I can tell, there isn't any valid need for this assert(), so just
remove it.
Reported-by: Anton Blanchard <anton@samba.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
The code handling integer literals in dtc-lexer.l assumes that the flex
regexp means that strtoull() can't fail to interpret the string as a valid
integer (either decimal, octal, or hexadecimal). This is not true for
octals. For example '09' is accepted as a literal by the regexp,
strtoull() attempts to handle it as octal, but it has a bad digit.
This changes the code to give a more useful error in this case.
Reported-by: Anton Blanchard <anton@samba.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Currently, attempting to divide by zero in an integer expression in a dts
file will cause dtc to crash with a division by zero (SIGFPE).
This patch corrects this to properly detect this case and raise an error.
Reported-by: Anton Blanchard <anton@samba.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Given a device tree node, a property name and an index, the new function
fdt_stringlist_get() will return a pointer to the index'th string in the
property's value and return its length (or an error code on failure) in
an output argument.
Signed-off-by: Thierry Reding <treding@nvidia.com>
[Fix some -Wshadow warnings --dwg]
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
The new fdt_stringlist_search() function will look up a given string in
the list contained in the value of a named property of a given device
tree node and return its index.
Signed-off-by: Thierry Reding <treding@nvidia.com>
[Fix some -Wshadow warnings --dwg]
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Given a device tree node and a property name, the fdt_stringlist_count()
function counts the number of strings found in the property value.
This also adds a new error code, FDT_ERR_BADVALUE, that the function
returns when it encounters a non-NUL-terminated string list.
Signed-off-by: Thierry Reding <treding@nvidia.com>
[Changed testcase name --dwg]
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
On systems where 'char' is signed, fdtdump will currently print the wrong
thing on properties containing bytestring values with "negative" bytes
(that is with values from 0x80..0xff). The fdtdump testcase is extended
to cover this case too.
This corrects the problem by forcing use of unsigned char - although this
is perhaps another indication that fdtdump is a buggy hack and if you want
to do real work you should use dtc -O dts.
Reported-by: Igor Prusov <Igor.V.Prusov@mcst.ru>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
The overall test runner script, for the fdtdump tests invokes the helper
script fdtdump-runtest.sh. It then includes directly some code very
similar to fdtdump-runtest.sh, which is never reached due to a "return".
Remove the never-reached test code.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
The test script includes several specific tests for the handling of aliases
in fdt_path_offset(). These are primarily tests of the fdt_path_offset()
libfdt function itself, although dtc is used to generate a test file for
convenience.
Move these from the dtc tests section to the libfdt tests section
accordingly.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
The -n option is not standard in POSIX, so convert to printf which should
work the same in every shell.
Signed-off-by: Mike Frysinger <vapier@chromium.org>
If you have a parent block with #size-cells improperly set to 0, and
then subsequently try to include a regs property in the child, dtc will
crash with SIGFPE while validating it. This patch fixes that crash,
instead printing the same invalid length warning that was causing it.
Test included.
Signed-off-by: Jack Miller <jack@codezen.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Paths with multiple '/' characters in a row (e.g. //somenode//somsubnode),
or trailing '/' characters (e.g. '/somenode/somesubnode/') should be
handled by fdt_path_offset(), and treated as equivalent to
/somenode/somesubnode.
Our current path_offset testcase doesn't check for these cases, so extend
it so it does.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
This introduces a check_path_offset() helper function into the path_offset
testcase to simplify it. This will also make extending the test case with
tests for path_offset_namelen() and some edge cases easier.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
This patch changes the dtc grammar to allow following syntax
i2cexp: &i2c2 {
...
};
Current device tree compiler allows to define multiple labels when defining
the device node the first time. Typically device nodes are defined in
DTSI files. Now these nodes can be overwritten for updating some of the
properties. Typically, device nodes are overridden in DTS files.
When working with adapter boards, most of the time adapter board can fit to
multiple base boards. But depending on which base board it is connected to,
the devices on the adapter board would be children of different devices.
e.g. On dra7-evm.dts, i2c2 is exported for expansion connector whereas
on dra72-evm.dts, i2c5 is exported for expansion connector.
This causes a problem when writing a generic device tree file for
the adapter board. Because, you cannot know whether all the devices on
adapter board are present on i2c or i2c5.
The problem can be solved by adding a common label (e.g. i2cexp) in both
of the DTS files when overriding the device nodes for i2c2 or i2c5.
This way, generic adapter board file would override the i2cexp. And
depending on which base board you use the adapter board, all the devices
are automatically added for correct device nodes.
Signed-off-by: Nikhil Devshatwar <nikhil.nd@ti.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
For example:
reserved-names="res1\0res2\0res3";
Where \0 is an actual embedded NUL in the source instead of a string
escape. To achieve this, use the len given by the lexer instead of
strlen.
Without this patch dtc will mangle the output and possibly hang on
realloc.
This has been there for ages, but the assertion makes no sense in the
context of the test immediately preceding it. This caused an abort()
when in -I dts -O dts mode with the right sort of internal labels in a
string property value.
Add a testcase for this and another candidate edge case (though this one
we already get right).
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
We can test fdtdump by comparing its output with the source file that was
compiled by dtc. Add a simple test that should at least catch regressions
in basic functionality.
Signed-off-by: Simon Glass <sjg@chromium.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
This patch makes a small start on libfdt functions which actually help to
parse the contents of device trees, rather than purely manipulating the
tree's structure.
We add simple helpers to read and sanity check the #address-cells and
#size-cells values for a given node.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
At present, the lexer token for references to a path doesn't permit a
reference to the root node &{/}. Fixing the lexer exposes another bug
handling this case.
This patch fixes both bugs and adds testcases.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
The sw_tree1 testcase has accumulated some valgrind errors, at least in
the "realloc" mode.
* It had both a realloc_fdt() and explicit xmalloc() for the initial
allocation which was redundant and caused errors.
* It doesn't make sense to call fdt_resize() until after we've created
the initial stub tree
* Alignment gaps inserted into the tree contain uninitialized data, which
trips an error when we write it out. We could zero the buffer, but that
would make it easier to miss real bugs, so we add suppressions for the
valgrind warnings instead.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
This is a debugging convenience option, which makes it much easier to find
the failing tests and fix them one by one.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
At present, when using sequential write mode, there's no straightforward
means of resizing the buffer the fdt is being built into. This patch
adds an fdt_resize() function for this purpose.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
ePAPR 1.1 section 2.2.1.1 "Node Name Requirements" specifies that any
node that has a reg property must include a unit address in its name
with value matching the first entry in its reg property. Conversely, if
a node does not have a reg property, the node name must not include a
unit address.
Adjust all the dtc test-cases to conform to this rule.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Previously, the #line parsing regex ended with ({WS}+[0-9]+)?. The {WS}
could match line-break characters. If the #line directive did not contain
the optional flags field at the end, this could cause any integer data on
the next line to be consumed as part of the #line directive parsing. This
could cause syntax errors (i.e. #line parsing consuming the leading 0
from a hex literal 0x1234, leaving x1234 to be parsed as cell data,
which is a syntax error), or invalid compilation results (i.e. simply
consuming literal 1234 as part of the #line processing, thus removing it
from the cell data).
Fix this by replacing {WS} with [ \t] so that it can't match line-breaks.
Convert all instances of {WS}, even though the other instances should be
irrelevant for any well-formed #line directive. This is done for
consistency and ultimate safety.
Reported-by: Ian Campbell <Ian.Campbell@citrix.com>
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
If you try to insert a new node or extend a property with large value,
using fdtput you will notice that it always fails.
example:
fdtput -v -p -ts ./tst.dtb "/node-1" "property-1" "value-1
Error at 'node-1': FDT_ERR_NOSPACE
or
fdtput -v -c ./tst.dtb "/node-1"
Error at 'node-1': FDT_ERR_NOSPACE
or
fdtput -v -ts ./tst.dtb "/node" "property" "very big value"
Decoding value:
string: 'very big value'
Value size 15
Error at 'property': FDT_ERR_NOSPACE
All these error are returned from libfdt, as the size of the fdt passed
has no space to accomdate these new properties.
This patch adds realloc functions in fdtput to allocate new space in fdt
when it detects a shortage in space for new value or node. With this
patch, fdtput can insert a new node or property or extend a property
with new value greater than original size. Also it packs the final blob
to clean up any extra padding.
Without this patch fdtput tool complains with FDT_ERR_NOSPACE when we
try to add a node/property or extend the value of a property.
Testcases for the new behaviour added by David Gibson.
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@st.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
There are a couple of fdtput related tests which are rather pointless -
they explicitly test for the presence of an undesirable limitation in
fdtput, which will cause test failures when we fix it. This patch removes
the tests.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
We have certain tests which generate extremely long command lines, which
are shortened in the testsuite output with the 'shorten_echo' function.
Currently that is used in run_fdtput_test and run_wrap_test, this patch
uses it for run_wrap_test as well, allowing more general tests with long
command lines.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Iterating through subnodes with libfdt is a little painful to write as we
need something like this:
for (depth = 0, count = 0,
offset = fdt_next_node(fdt, parent_offset, &depth);
(offset >= 0) && (depth > 0);
offset = fdt_next_node(fdt, offset, &depth)) {
if (depth == 1) {
/* code body */
}
}
Using fdt_next_subnode() we can instead write this, which is shorter and
easier to get right:
for (offset = fdt_first_subnode(fdt, parent_offset);
offset >= 0;
offset = fdt_next_subnode(fdt, offset)) {
/* code body */
}
Also, it doesn't require two levels of indentation for the loop body.
Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
For a follow up commit, we want to be able to scan the buffer that was
returned to us. In order to do that safely, we need to know how big
the buffer actually is, so pass that back if requested.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
This commit which changed the behaviour of this function broke one
of the tests. Also the comment should be updated to reflect its new
behaviour.
Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Not all /bin/echo implementations support the -e option. Instead, use
printf, which appears to be more widely available than /bin/echo -e.
See commit eaec1db "fdtget-runtest.sh: Fix failures when /bin/sh isn't
bash" for history.
I have tested this on Ubuntu 10.04 with /bin/sh pointing to both dash
and bash.
Reported-by: Mike Frysinger <vapier@gentoo.org> # and implemented-by
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
tests will need fdt type definitions provided in a subsequent patch
to libfdt_env.h. Since libfdt.h includes libfdt_env.h in the right
order anyway, just remove the fdt.h include.
Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
The previous definition of for_each_*() would always include the very
first object within the list, irrespective of whether it was marked
deleted, since the deleted flag was not checked on the first object,
but only on any "next" object.
Fix for_each_*() to check the deleted flag in the loop body every
iteration to correct this.
Incidentally, this change is why commit 45013d8 dtc: "Add ability to
delete nodes and properties" only caused two "make checkm" failures;
only two tests actually use multiple labels on the same property or
node. With this current change applied, but commit 317a5d9 "dtc: zero
out new label objects" reverted, "make checkm" fails 29 times; i.e.
for every test that uses any labels at all.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Line control directives of the following formats are supported:
#line LINE "FILE"
# LINE "FILE" [FLAGS]
This allows dtc to consume the output of pre-processors, and to provide
error messages that refer to the original filename, including taking
into account any #include directives that the pre-processor may have
performed.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
The device tree language as currently defined conflicts with the C pre-
processor in one aspect - when a property or node name begins with a #
character, a pre-processor would attempt to interpret it as a directive,
fail, and most likely error out.
This change allows a property/node name to be prefixed with \. This
prevents a pre-processor from seeing # as the first non-whitespace
character on the line, and hence prevents the conflict. \ was previously
an illegal character in property/node names, so this change is
backwards compatible. The \ is stripped from the name during parsing
by dtc.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
dtc currently allows the contents of properties to be changed, and the
contents of nodes to be added to. There are situations where removing
properties or nodes may be useful. This change implements the following
syntax to do that:
/ {
/delete-property/ propname;
/delete-node/ nodename;
};
or:
/delete-node/ &noderef;
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Maxime Ripard