linux.git/scripts/decode_stacktrace.sh, branch v6.18.21

scripts/decode_stacktrace.sh: fix build ID and PC source parsing

2025-11-10T05:19:45+00:00

Support for parsing PC source info in stacktraces (e.g.  '(P)') was added
in commit 2bff77c665ed ("scripts/decode_stacktrace.sh: fix decoding of
lines with an additional info").  However, this logic was placed after the
build ID processing.  This incorrect order fails to parse lines containing
both elements, e.g.:

  drm_gem_mmap_obj+0x114/0x200 [drm 03d0564e0529947d67bb2008c3548be77279fd27] (P)

This patch fixes the problem by extracting the PC source info first and
then processing the module build ID.  With this change, the line above is
now properly parsed as such:

  drm_gem_mmap_obj (./include/linux/mmap_lock.h:212 ./include/linux/mm.h:811 drivers/gpu/drm/drm_gem.c:1177) drm (P)

While here, also add a brief explanation the build ID section.

Link: https://lkml.kernel.org/r/20251030010347.2731925-1-cmllamas@google.com
Fixes: 2bff77c665ed ("scripts/decode_stacktrace.sh: fix decoding of lines with an additional info")
Signed-off-by: Carlos Llamas 
Reviewed-by: Matthieu Baerts (NGI0) 
Reviewed-by: Luca Ceresoli 
Cc: Breno Leitao 
Cc: Catalin Marinas 
Cc: Marc Rutland 
Cc: Mark Brown 
Cc: Matthieu Baerts 
Cc: Miroslav Benes 
Cc: Puranjay Mohan 
Cc: 
Signed-off-by: Andrew Morton

scripts/decode_stacktrace.sh: code: preserve alignment

2025-09-21T21:22:28+00:00

With lines having a code to decode, the alignment was not preserved for
the first line.

With this sample ...

  [   52.238089][   T55] RIP: 0010:__ip_queue_xmit+0x127c/0x1820
  [   52.238401][   T55] Code: c1 83 e0 07 48 c1 e9 03 83 c0 03 (...)

... the script was producing the following output:

  [   52.238089][   T55] RIP: 0010:__ip_queue_xmit (...)
  [ 52.238401][ T55] Code: c1 83 e0 07 48 c1 e9 03 83 c0 03 (...)

That's because scripts/decodecode doesn't preserve the alignment.  No need
to modify it, it is enough to give only the "Code: (...)" part to this
script, and print the prefix without modifications.

With the same sample, we now have:

  [   52.238089][   T55] RIP: 0010:__ip_queue_xmit (...)
  [   52.238401][   T55] Code: c1 83 e0 07 48 c1 e9 03 83 c0 03 (...)

Link: https://lkml.kernel.org/r/20250908-decode_strace_indent-v1-3-28e5e4758080@kernel.org
Signed-off-by: Matthieu Baerts (NGI0) 
Tested-by: Carlos Llamas 
Cc: Breno Leitao 
Cc: Elliot Berman 
Cc: Luca Ceresoli 
Cc: Stephen Boyd 
Signed-off-by: Andrew Morton

scripts/decode_stacktrace.sh: symbol: preserve alignment

2025-09-21T21:22:27+00:00

With lines having a symbol to decode, the script was only trying to
preserve the alignment for the timestamps, but not the rest, nor when the
caller was set (CONFIG_PRINTK_CALLER=y).

With this sample ...

  [   52.080924] Call Trace:
  [   52.080926]  
  [   52.080931]  dump_stack_lvl+0x6f/0xb0

... the script was producing the following output:

  [   52.080924] Call Trace:
  [   52.080926]  
  [   52.080931] dump_stack_lvl (arch/x86/include/asm/irqflags.h:19)

  (dump_stack_lvl is no longer aligned with : one missing space)

With this other sample ...

  [   52.080924][   T48] Call Trace:
  [   52.080926][   T48]  
  [   52.080931][   T48]  dump_stack_lvl+0x6f/0xb0

... the script was producing the following output:

  [   52.080924][   T48] Call Trace:
  [   52.080926][   T48]  
  [ 52.080931][ T48] dump_stack_lvl (arch/x86/include/asm/irqflags.h:19)

  (the misalignment is clearer here)

That's because the script had a workaround for CONFIG_PRINTK_TIME=y only,
see the previous comment called "Format timestamps with tabs".

To always preserve spaces, they need to be recorded along the words.  That
is what is now done with the new 'spaces' array.

Some notes:

- 'extglob' is needed only for this operation, and that's why it is set
  in a dedicated subshell.

- 'read' is used with '-r' not to treat a  character in any
  special way, e.g. when followed by a space.

- When a word is removed from the 'words' array, the corresponding space
  needs to be removed from the 'spaces' array as well.

With the last sample, we now have:

  [   52.080924][   T48] Call Trace:
  [   52.080926][   T48]  
  [   52.080931][   T48]  dump_stack_lvl (arch/x86/include/asm/irqflags.h:19)

  (the alignment is preserved)

Link: https://lkml.kernel.org/r/20250908-decode_strace_indent-v1-2-28e5e4758080@kernel.org
Signed-off-by: Matthieu Baerts (NGI0) 
Tested-by: Carlos Llamas 
Cc: Breno Leitao 
Cc: Elliot Berman 
Cc: Luca Ceresoli 
Cc: Stephen Boyd 
Signed-off-by: Andrew Morton

scripts/decode_stacktrace.sh: symbol: avoid trailing whitespaces

2025-09-21T21:22:27+00:00

A few patches slightly improving the output generated by
decode_stacktrace.sh.


This patch (of 3):

Lines having a symbol to decode might not always have info after this
symbol.  It means ${info_str} might not be set, but it will always be
printed after a space, causing trailing whitespaces.

That's a detail, but when the output is opened with an editor marking
these trailing whitespaces, that's a bit disturbing.  It is easy to remove
them by printing this variable with a space only if it is set.

While at it, do the same with ${module} and print everything in one line.

Link: https://lkml.kernel.org/r/20250908-decode_strace_indent-v1-0-28e5e4758080@kernel.org
Link: https://lkml.kernel.org/r/20250908-decode_strace_indent-v1-1-28e5e4758080@kernel.org
Signed-off-by: Matthieu Baerts (NGI0) 
Reviewed-by: Carlos Llamas 
Reviewed-by: Breno Leitao 
Reviewed-by: Luca Ceresoli 
Cc: Carlos Llamas 
Cc: Elliot Berman 
Cc: Stephen Boyd 
Signed-off-by: Andrew Morton

scripts/decode_stacktrace.sh: fix decoding of lines with an additional info

2025-01-13T03:03:34+00:00

Since commit bdf8eafbf7f5 ("arm64: stacktrace: report source of unwind
data") a stack trace line can contain an additional info field that was not
present before, in the form of one or more letters in parentheses. E.g.:

  [  504.517915]  led_sysfs_enable+0x54/0x80 (P)
                                             ^^^

When this is present, decode_stacktrace decodes the line incorrectly:

  [  504.517915] led_sysfs_enable+0x54/0x80 P

Extend parsing to decode it correctly:

  [  504.517915] led_sysfs_enable (drivers/leds/led-core.c:455 (discriminator 7)) (P)

The regex to match such lines assumes the info can be extended in the
future to other uppercase characters, and will need to be extended in case
other characters will be used. Using a much more generic regex might incur
in false positives, so this looked like a good tradeoff.

Link: https://lkml.kernel.org/r/20241230-decode_stacktrace-fix-info-v1-1-984910659173@bootlin.com
Fixes: bdf8eafbf7f5 ("arm64: stacktrace: report source of unwind data")
Signed-off-by: Luca Ceresoli 
Cc: Catalin Marinas 
Cc: Mark Brown 
Cc: Mark Rutland 
Cc: Miroslav Benes 
Cc: Puranjay Mohan 
Cc: Thomas Petazzoni 
Signed-off-by: Andrew Morton

scripts/decode_stacktrace.sh: remove trailing space

2024-11-06T01:12:32+00:00

decode_stacktrace.sh adds a trailing space at the end of the decoded stack
if the module is not set (in most of the lines), which makes the some
lines of the stack having trailing space and some others not.

Do not add an extra space at the end of the line if module is not set,
adding consistency in output formatting.

Link: https://lkml.kernel.org/r/20241014100213.1873611-1-leitao@debian.org
Signed-off-by: Breno Leitao 
Reviewed-by: Elliot Berman 
Reviewed-by: Carlos Llamas 
Reviewed-by: Stephen Boyd 
Cc: Bjorn Andersson 
Cc: Luca Ceresoli 
Cc: Xiong Nandi 
Signed-off-by: Andrew Morton

scripts/decode_stacktrace.sh: add '-h' flag

2024-09-02T03:43:41+00:00

When no parameters are passed, the usage instructions are presented only
when debuginfod-find is not found.  This makes sense because with
debuginfod none of the positional parameters are needed.  However it means
that users having debuginfod-find installed will have no chance of reading
the usage text without opening the file.

Many programs have a '-h' flag to get the usage, so add such a flag. 
Invoking 'scripts/decode_stacktrace.sh -h' will now show the usage text
and exit.

Link: https://lkml.kernel.org/r/20240823-decode_stacktrace-find_module-improvements-v2-3-d7a57d35558b@bootlin.com
Signed-off-by: Luca Ceresoli 
Reviewed-by: Stephen Boyd 
Cc: Alexis Lothoré (eBPF Foundation) 
Cc: Konstantin Khlebnikov 
Cc: Sasha Levin 
Cc: Thomas Petazzoni 
Signed-off-by: Andrew Morton

scripts/decode_stacktrace.sh: clarify command line

2024-09-02T03:43:41+00:00

The syntax as expressed by usage() is not entirely correct: "" cannot be passed without "|auto".  Additionally human
reading of this syntax can be subject to misunderstanding due the mixture
of '|' and '[]'.

Improve readability in various ways:
 * rewrite using two lines for the two allowed usages
 * add square brackets around "" as it is optional when using
   debuginfod-find
 * move "" to inside the square brackets of the 2nd
   positional parameter
 * use underscores instead of spaces in <...> strings

Link: https://lkml.kernel.org/r/20240823-decode_stacktrace-find_module-improvements-v2-2-d7a57d35558b@bootlin.com
Signed-off-by: Luca Ceresoli 
Reviewed-by: Stephen Boyd 
Cc: Alexis Lothoré (eBPF Foundation) 
Cc: Konstantin Khlebnikov 
Cc: Sasha Levin 
Cc: Thomas Petazzoni 
Signed-off-by: Andrew Morton

scripts/decode_stacktrace.sh: remove find_module recursion and improve error reporting

2024-09-02T03:43:40+00:00

Patch series "scripts/decode_stacktrace.sh: improve error reporting and
usability", v2.

This small series improves usability of scripts/decode_stacktrace.sh by
improving the usage text and correctly reporting when modules are built
without debugging symbols.


This patch (of 3):

The find_module() function can fail for two reasons:

 * the module was not found
 * the module was found but without debugging info

In both cases the user is reported the same error:

   WARNING! Modules path isn't set, but is needed to parse this symbol

This is misleading in case the modules path is set correctly.

find_module() is currently implemented as a recursive function based on
global variables in order to check up to 4 different paths.  This is not
straightforward to read and even less to modify.

Besides, the debuginfo code at the beginning of find_module() is executed
identically every time the function is entered, i.e.  up to 4 times per
each module search due to recursion.

To be able to improve error reporting, first rewrite the find_module()
function to remove recursion.  The new version of the function iterates
over all the same (up to 4) paths as before and for each of them does the
same checks as before.  At the end of the iteration it is now able to
print an appropriate error message, so that has been moved from the caller
into find_module().

Finally, when the module is found but without debugging info, mention the
two Kconfig variables one needs to set in order to have the needed
debugging symbols.

Link: https://lkml.kernel.org/r/20240823-decode_stacktrace-find_module-improvements-v2-0-d7a57d35558b@bootlin.com
Link: https://lkml.kernel.org/r/20240823-decode_stacktrace-find_module-improvements-v2-1-d7a57d35558b@bootlin.com
Signed-off-by: Luca Ceresoli 
Reviewed-by: Stephen Boyd 
Cc: Alexis Lothoré (eBPF Foundation) 
Cc: Konstantin Khlebnikov 
Cc: Luca Ceresoli 
Cc: Sasha Levin 
Cc: Thomas Petazzoni 
Signed-off-by: Andrew Morton

scripts/decode_stacktrace.sh: nix-ify

2024-09-02T03:43:38+00:00

nix only puts /usr/bin/env at the standard location (as required by
posix), so shebangs have to be tweaked.

Link: https://lkml.kernel.org/r/20240817215025.161628-1-kent.overstreet@linux.dev
Signed-off-by: Kent Overstreet 
Cc: Bjorn Andersson 
Cc: Elliot Berman 
Cc: Xiong Nandi 
Signed-off-by: Andrew Morton