linux-toradex.git/include/linux/module.h, branch v5.12-rc5

module: remove never implemented MODULE_SUPPORTED_DEVICE

2021-03-17T20:16:18+00:00

MODULE_SUPPORTED_DEVICE was added in pre-git era and never was
implemented. We can safely remove it, because the kernel has grown
to have many more reliable mechanisms to determine if device is
supported or not.

Signed-off-by: Leon Romanovsky 
Signed-off-by: Linus Torvalds

module: remove EXPORT_UNUSED_SYMBOL*

2021-02-08T11:28:07+00:00

EXPORT_UNUSED_SYMBOL* is not actually used anywhere.  Remove the
unused functionality as we generally just remove unused code anyway.

Reviewed-by: Miroslav Benes 
Reviewed-by: Emil Velikov 
Signed-off-by: Christoph Hellwig 
Signed-off-by: Jessica Yu

module: remove EXPORT_SYMBOL_GPL_FUTURE

2021-02-08T11:28:02+00:00

As far as I can tell this has never been used at all, and certainly
not any time recently.

Reviewed-by: Miroslav Benes 
Reviewed-by: Emil Velikov 
Signed-off-by: Christoph Hellwig 
Signed-off-by: Jessica Yu

module: move struct symsearch to module.c

2021-02-08T11:27:43+00:00

struct symsearch is only used inside of module.h, so move the definition
out of module.h.

Reviewed-by: Miroslav Benes 
Reviewed-by: Emil Velikov 
Signed-off-by: Christoph Hellwig 
Signed-off-by: Jessica Yu

module: mark module_mutex static

2021-02-08T11:24:26+00:00

Except for two lockdep asserts module_mutex is only used in module.c.
Remove the two asserts given that the functions they are in are not
exported and just called from the module code, and mark module_mutex
static.

Reviewed-by: Miroslav Benes 
Signed-off-by: Christoph Hellwig 
Signed-off-by: Jessica Yu

kallsyms: only build {,module_}kallsyms_on_each_symbol when required

2021-02-08T11:24:04+00:00

kallsyms_on_each_symbol and module_kallsyms_on_each_symbol are only used
by the livepatching code, so don't build them if livepatching is not
enabled.

Reviewed-by: Miroslav Benes 
Signed-off-by: Christoph Hellwig 
Signed-off-by: Jessica Yu

module: use RCU to synchronize find_module

2021-02-08T11:21:40+00:00

Allow for a RCU-sched critical section around find_module, following
the lower level find_module_all helper, and switch the two callers
outside of module.c to use such a RCU-sched critical section instead
of module_mutex.

Reviewed-by: Petr Mladek 
Acked-by: Miroslav Benes 
Signed-off-by: Christoph Hellwig 
Signed-off-by: Jessica Yu

Merge tag 'modules-for-v5.11' of git://git.kernel.org/pub/scm/linux/kernel/git/jeyu/linux

2020-12-17T21:01:31+00:00

Pull modules updates from Jessica Yu:
 "Summary of modules changes for the 5.11 merge window:

   - Fix a race condition between systemd/udev and the module loader.

     The module loader was sending a uevent before the module was fully
     initialized (i.e., before its init function has been called). This
     means udev can start processing the module uevent before the module
     has finished initializing, and some udev rules expect that the
     module has initialized already upon receiving the uevent.

     This resulted in some systemd mount units failing if udev processes
     the event faster than the module can finish init. This is fixed by
     delaying the uevent until after the module has called its init
     routine.

   - Make the linker array sections for kernel params and module version
     attributes more robust by switching to use the alignment of the
     type in question.

     Namely, linker section arrays will be constructed using the
     alignment required by the struct (using __alignof__()) as opposed
     to a specific value such as sizeof(void *) or sizeof(long). This is
     less likely to cause breakages should the size of the type ever
     change (Johan Hovold)

   - Fix module state inconsistency by setting it back to GOING when a
     module fails to load and is on its way out (Miroslav Benes)

   - Some comment and code cleanups (Sergey Shtylyov)"

* tag 'modules-for-v5.11' of git://git.kernel.org/pub/scm/linux/kernel/git/jeyu/linux:
  module: delay kobject uevent until after module init call
  module: drop semicolon from version macro
  init: use type alignment for kernel parameters
  params: clean up module-param macros
  params: use type alignment for kernel parameters
  params: drop redundant "unused" attributes
  module: simplify version-attribute handling
  module: drop version-attribute alignment
  module: fix comment style
  module: add more 'kernel-doc' comments
  module: fix up 'kernel-doc' comments
  module: only handle errors with the *switch* statement in module_sig_check()
  module: avoid *goto*s in module_sig_check()
  module: merge repetitive strings in module_sig_check()
  module: set MODULE_STATE_GOING state when a module fails to load

module: drop semicolon from version macro

2020-12-07T12:00:52+00:00

Drop the trailing semicolon from the MODULE_VERSION() macro definition
which was left when removing the array-of-pointer indirection.

Signed-off-by: Johan Hovold 
Signed-off-by: Jessica Yu

module: simplify version-attribute handling

2020-11-25T14:44:46+00:00

Instead of using the array-of-pointers trick to avoid having gcc mess up
the built-in module-version array stride, specify type alignment when
declaring entries to prevent gcc from increasing alignment.

This is essentially an alternative (one-line) fix to the problem
addressed by commit b4bc842802db ("module: deal with alignment issues in
built-in module versions").

gcc can increase the alignment of larger objects with static extent as
an optimisation, but this can be suppressed by using the aligned
attribute when declaring variables.

Note that we have been relying on this behaviour for kernel parameters
for 16 years and it indeed hasn't changed since the introduction of the
aligned attribute in gcc-3.1.

Link: https://lore.kernel.org/lkml/20201103175711.10731-1-johan@kernel.org
Signed-off-by: Johan Hovold 
Signed-off-by: Jessica Yu