const is ignored on the function return type. A warning is reported with
-Wignored-qualifers. Remove the ignored const.
Signed-off-by: Ryan McClelland <ryanmcclelland@meta.com>
Decoder was casting uint8_t pointers to uint64_t pointers which could
result in double word instruction which does not support unaligned
access on Cortex-M. Issue was revealed when -O3 optimization was
used instead of -Os. In size optimized version, compiler was
using word load and store instructions which support unaligned
access and issue was not visible.
Signed-off-by: Krzysztof Chruściński <krzysztof.chruscinski@nordicsemi.no>
This adds the bits to call into architecture code to dump
the privileged stack for user threads.
The weak implementation is simply there as a stub until
all architectures have implemented the associated function.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Add missing braces to comply with MISRA C:2012 Rule 15.6 and
also following Zephyr's style guideline.
Signed-off-by: Pisit Sawangvonganan <pisit@ndrsolution.com>
Implement shell commands to access the symbol table.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
Signed-off-by: Yong Cong Sin <yongcong.sin@gmail.com>
This adds the bits to display privileged stack usage for
architectures that support obtaining this information.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
The per-thread output after the first stack usage line is not
aligned which makes it harder to decipher the content.
So align them for easier reading.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Having `CONFIG_EXCEPTION_STACK_TRACE_SYMTAB` to select the
`CONFIG_SYMTAB` or to explicitly not print the symbol name
during exception stack unwind seems unnecessary, as the extra
code to print the symbol name is negligible when compared with
the symbol table, so just remove it.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
Signed-off-by: Yong Cong Sin <yongcong.sin@gmail.com>
Currently it supports `esf` based unwinding only.
Then, update the exception stack unwinding to use
`arch_stack_walk()`, and update the Kconfigs & testcase
accordingly.
Also, `EXCEPTION_STACK_TRACE_MAX_FRAMES` is unused and
made redundant after this change, so remove it.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
Signed-off-by: Yong Cong Sin <yongcong.sin@gmail.com>
Currently it supports `esf` based unwinding only.
Then, update the exception stack unwinding to use
`arch_stack_walk()`, and update the Kconfigs & testcase
accordingly.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
Signed-off-by: Yong Cong Sin <yongcong.sin@gmail.com>
This commit introduces a new ARCH_STACKWALK Kconfig which
determines if the `arch_stack_walk()` is available should the
arch supports it.
Starting from RISCV, this will be able to converge the exception
stack trace implementation & stack walking features. Existing
exception stack trace implementation will be updated later.
Eventually we will end up with the following:
1. If an arch implements `arch_stack_walk()`
`ARCH_HAS_STACKWALK` should be selected.
2. If the above is enabled, `ARCH_SUPPORTS_STACKWALK` indicates
if the dependencies are met for arch to enable stack walking.
This Kconfig replaces `<arch>_EXCEPTION_STACK_TRACE`
2. If the above is enabled, then, `ARCH_STACKWALK` determines
if `arch_stack_walk()` should be compiled.
3. `EXCEPTION_STACK_TRACE` should build on top of the
`ARCH_STACKWALK`, stack traces will be printed when it
is enabled.
4. `ARCH_STACKWALK_MAX_FRAMES` will be removed as it is
replaced by `ARCH_STACKWALK_MAX_FRAMES`
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
Signed-off-by: Yong Cong Sin <yongcong.sin@gmail.com>
Add missing braces to comply with MISRA C:2012 Rule 15.6 and
also following Zephyr's style guideline.
Signed-off-by: Pisit Sawangvonganan <pisit@ndrsolution.com>
Add new config, ARCH_SUPPORTS_COREDUMP_THREADS, and
only enable it for ARM CORTEX M where the gdb server
can support it.
Signed-off-by: Mark Holden <mholden@meta.com>
Add z_test which uses new configs to capture multiple
threads in a core dump and with all of the context
necessary to debug the threads.
Signed-off-by: Mark Holden <mholden@meta.com>
Update core dump file format to support a new section which contains
metadata about threads necessary for debugging.
Define configs to capture that metadata and include it in the dumps
when enabled.
Update documentation to reflect the changes.
Signed-off-by: Mark Holden <mholden@meta.com>
With CONFIG_DEBUG_COREDUMP_MEMORY_DUMP_LINKER_RAM, buffer_output() is
called on the entire RAM memory area. This includes the stack for the
coredump thread, which is where tmp_buf is stored. Eventually, it will copy
(parts of) tmp_buf into tmp_buf itself, which invokes Undefined Behaviour
in memcpy():
> The memory areas must not overlap. Use memmove(3) if the memory areas do
> overlap.
- memcpy(3)
With picolibc, this is detected in __memcpy_chk() and causes a fault in
__chk_fail().
Signed-off-by: Armin Brauns <armin.brauns@embedded-solutions.at>
Fix the second k_thread_foreach_unlocked argument, it's supposed to be
an instance of ta_cb_user_data as that's what's it casted back to in
thread_analyze_cb. Current code results in an exception and crash for
single core applications.
This is a regression introduced in 1b6e0f6479.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
In thread_analyzer_auto(), it casts one function argument
directly into unsigned int. However, on 64-bit platforms,
the compiler complains about casting from pointer of
different size (-Wpointer-to-int-cast). So cast it first to
uintptr_t before casting it into unsigned int.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Add implementation to analyze threads on each cpu separately. This
feature can be enabled with THREAD_ANALYZER_AUTO_SEPARATE_CORES Kconfig
option. If enabled, an analyzer thread is started for each cpu, and
the threads will only analyze thread on the cpu its running on.
This feature is needed for Intel ADSP platform, where cpu specific
caches are not synchronized between the cpu. It is also probably
needed by other platforms having CONFIG_KERNEL_COHERENCE=y, so default
to THREAD_ANALYZER_AUTO_SEPARATE_CORES=y for those platform.
Signed-off-by: Jyri Sarha <jyri.sarha@linux.intel.com>
Set the UNALIGNED_ACCESS_SUPPORTED only for MCU with cortex M
that are neither M0 nor M0plus
Cortex M0 or M0plus mcus do not support un-aligned address access
Signed-off-by: Francois Ramu <francois.ramu@st.com>
An architecture can indicate that it has an implementation for
the `arch_stack_walk()` function by selecting
`ARCH_HAS_STACKWALK`.
Set the default value of `EXCEPTION_STACK_TRACE_MAX_FRAMES` to
`ARCH_STACKWALK_MAX_FRAMES` if the latter is available.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
Make `struct arch_esf` compulsory for all architectures by
declaring it in the `arch_interface.h` header.
After this commit, the named struct `z_arch_esf_t` is only used
internally to generate offsets, and is slated to be removed
from the `arch_interface.h` header in the future.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
Replace usage of flash_area_erase with flash_area_flatten to
allow code to work with devices that do not provide erase
callback.
Signed-off-by: Dominik Ermel <dominik.ermel@nordicsemi.no>
Fix the dependencies of `CONFIG_EXCEPTION_STACK_TRACE`:
- Architecture-specific Kconfig, i.e.
`X86_EXCEPTION_STACK_TRACE`, will be enabled automatically
when all the dependencies are met.
- `EXCEPTION_STACK_TRACE` depends on architecture-specific
Kconfig to be enabled.
- The stack trace implementations should be compiled only if
user enables `CONFIG_EXCEPTION_STACK_TRACE`.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
The `DEBUG_INFO` in the `EXCEPTION_STACK_TRACE` is only
required by x86. Move that to `X86_EXCEPTION_STACK_TRACE`
instead.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
The `PRINTK` was required for `EXCEPTION_STACK_TRACE` because
it's initial implementation for x86 in #6653 uses `printk()`.
We are using `LOG_ERR()` now, so this is not required anymore.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
Employs the same linkonce magic of sw_isr_table to fix the
multiple definition of the symtab variables issue that I
get in my application build that doesn't use `west`.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
`start_addr` is the address of the first symbol, rename it to
`first_addr` instead as it seems more intuitive and relatable
to the comments.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
The `symtab_find_symbol_name()` is using an adapted binary
search function to get the entry between 2 addresses, we need
to add a dummy entry at the end so that the search function
can remain simple and straightforward without doing
out-of-bound checks:
20 \
|
|
50 x
|
|
90 x
. |
. |
. |
dummy /
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
This allows a customized coredump_backend_other API to re-use the
COREDUMP_*_STR without re-defining the same strings.
Signed-off-by: Younghyun Park <younghyunpark@google.com>
Current on x86 & risc-v that implement stack trace, the
maximum depth of the stack trace is defined by a macro.
Introduce a new Kconfig:EXCEPTION_STACK_TRACE_MAX_FRAMES
so that this is configurable in software.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
Selecting `CONFIG_EXCEPTION_STACK_TRACE_SYMTAB` will
enable the symtab generation which will be used in the
stack trace to print the function name of the return
address.
Updated the `stack_unwind` test to test the symbols in a
stack trace.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
Use pyelftools to extract the symbol table from the
link stage executable. Then, filter out the function names
and sort them based on their offsets before writing into the
`symtab.c`, this is similar to how the `isr_tables` works.
To access the structure, simply include the new header:
```c
#include <zephyr/debug/symtab.h>
```
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
Add `gdb_q_packet` function to handle proccessing of query packets. Also
add `gdb_qsupported` function to serve as a single place to process and
respond to `qSupported` packets.
Signed-off-by: Robert Zieba <robertzieba@google.com>
Make the current gdb loop enum public under the name `gdb_loop_state`.
This will allow for extending the current stub with further
functionality in the future as new features will require control over
the loop state.
Signed-off-by: Robert Zieba <robertzieba@google.com>
Simple rename to align the kernel naming scheme. This is being
used throughout the tree, especially in the architecture code.
As this is not a private API internal to kernel, prefix it
appropriately with K_.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
The current implementation can raise a warning as the compiler sees
an attempt to index an array with a size of zero. This can be fixed by
giving `gdb_mem_num_regions` a default value of zero, but this gets
flagged by CI checks. Disable the warning around the array access.
Signed-off-by: Robert Zieba <robertzieba@google.com>
By default `GDBSTUB_SERIAL_BACKEND` will be selected as it's the only
option here. This can cause problems for code that wants to provide its
own custom backend. Add a choice for a custom backend.
Signed-off-by: Robert Zieba <robertzieba@google.com>
Add GDBSTUB_TRACE config option to extend GDB backend debug logging
for remote commands received and to debug the GDB stub itself.
Signed-off-by: Dmitrii Golovanov <dmitrii.golovanov@intel.com>
Do not enable subsystem/driver shell modules by default and stop abusing
CONFIG_SHELL_MINIMAL, which is internal to the shell subsystem, to decide
when to enable a driver shell.
The list of shell modules has grown considerably through the
years. Enabling CONFIG_SHELL for doing e.g. an interactive debug session
leads to a large number of shell modules also being enabled unless
explicitly disabled, which again leads to non-negligible increases in
RAM/ROM usage.
This commit attempts to establish a policy of subsystem/driver shell
modules being disabled by default, requiring the user/application to
explicitly enable only those needed.
Signed-off-by: Henrik Brix Andersen <hebad@vestas.com>
Ryan McClelland