linux.git/kernel/module.c, branch v2.6.26-rc2

module: don't ignore vermagic string if module doesn't have modversions

2008-05-09T14:45:18+00:00

Linus found a logic bug: we ignore the version number in a module's
vermagic string if we have CONFIG_MODVERSIONS set, but modversions
also lets through a module with no __versions section for modprobe
--force (with tainting, but still).

We should only ignore the start of the vermagic string if the module
actually *has* crcs to check.  Rather than (say) having an
entertaining hissy fit and creating a config option to work around the
buggy code.

Signed-off-by: Rusty Russell 
Signed-off-by: Linus Torvalds

module: be more picky about allowing missing module versions

2008-05-09T14:45:18+00:00

We allow missing __versions sections, because modprobe --force strips
it.  It makes less sense to allow sections where there's no version
for a specific symbol the module uses, so disallow that.

Signed-off-by: Rusty Russell 
Signed-off-by: Linus Torvalds

Make forced module loading optional

2008-05-05T00:04:16+00:00

The kernel module loader used to be much too happy to allow loading of
modules for the wrong kernel version by default.  For example, if you
had MODVERSIONS enabled, but tried to load a module with no version
info, it would happily load it and taint the kernel - whether it was
likely to actually work or not!

Generally, such forced module loading should be considered a really
really bad idea, so make it conditional on a new config option
(MODULE_FORCE_LOAD), and make it default to off.

If somebody really wants to force module loads, that's their problem,
but we should not encourage it.  Especially as it happened to me by
mistake (ie regular unversioned Fedora modules getting loaded) causing
lots of strange behavior.

Signed-off-by: Linus Torvalds

module: add MODULE_STATE_GOING notifier call

2008-05-01T11:15:01+00:00

Provide module unload callback. Required by the gcov profiling
infrastructure to keep track of profiling data structures.

Signed-off-by: Peter Oberparleiter 
Signed-off-by: Rusty Russell

module: Enhance verify_export_symbols

2008-05-01T11:15:00+00:00

Make verify_export_symbols check the modules unused, unused_gpl and
gpl_future syms.

Inspired by Jan Beulich's fix, but table-driven.

Signed-off-by: Rusty Russell

module: set unused_gpl_crcs instead of overwriting unused_crcs

2008-05-01T11:15:00+00:00

Obvious typo, but I don't know of any modules with unused GPL exports,
and then it would take someone noticing that the version shouldn't
have matched in a dependent module.

Signed-off-by: Rusty Russell

module: neaten __find_symbol, rename to find_symbol

2008-05-01T11:15:00+00:00

__find_symbol() has grown over time: there are now 5 different arrays
of symbols it traverses.  It also shouldn't print out a warning on
some calls (ie. verify_symbol which simply checks for name clashes,
and __symbol_put which checks for bugs).

1) Rename to find_symbol: no need for underscores.
2) Use bool and add "warn" parameter to suppress warnings.
3) Make table-driven rather than open coded.

Signed-off-by: Rusty Russell

module: reduce module image and resident size

2008-05-01T11:14:59+00:00

Resulting reduction (x86-64, gcc 4.1.2) with my (special purpose, i.e.
much reduced) configurations:
- 16k kernel resident size
- 180k module resident size
- 10k module image size

Signed-off-by: Jan Beulich 
Signed-off-by: Rusty Russell

module: make module_sect_attrs private to kernel/module.c

2008-05-01T11:14:59+00:00

No-one else is using these afaics.

Signed-off-by: Jan Beulich 
Signed-off-by: Rusty Russell

kernel: Remove unnecessary inclusions of asm/semaphore.h

2008-04-19T02:17:04+00:00

None of these files use any of the functionality promised by
asm/semaphore.h.

Signed-off-by: Matthew Wilcox