[PATCH] lkdtm/core: fix resource leaks on module init error path

[PATCH] lkdtm/core: fix resource leaks on module init error path

[Jiangshan Yi]

When lkdtm_register_cpoint() fails during lkdtm_module_init(), the code
jumps to out_err:, which only calls debugfs_remove_recursive() and
returns the error code. Since module_init() returned an error, the
kernel does not load the module and never invokes module_exit(), so
lkdtm_module_exit() is not called.

As a result, everything allocated earlier in the init path is leaked:

  - the lkdtm_kernel_info string allocated by kasprintf(),
  - the four kmem_caches created by lkdtm_usercopy_init() and
    lkdtm_heap_init().

The corresponding cleanup — lkdtm_heap_exit(), lkdtm_usercopy_exit()
and kfree(lkdtm_kernel_info) — lives only in lkdtm_module_exit() and
is therefore never executed on this failure path.

Free these resources on the error path before returning, mirroring the
cleanup done in lkdtm_module_exit().

Signed-off-by: Jiangshan Yi <yijiangshan@kylinos.cn>
---
 drivers/misc/lkdtm/core.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/misc/lkdtm/core.c b/drivers/misc/lkdtm/core.c
index ededa32d6744..d741c6051408 100644
--- a/drivers/misc/lkdtm/core.c
+++ b/drivers/misc/lkdtm/core.c
@@ -468,6 +468,9 @@ static int __init lkdtm_module_init(void)
 
 out_err:
 	debugfs_remove_recursive(lkdtm_debugfs_root);
+	lkdtm_heap_exit();
+	lkdtm_usercopy_exit();
+	kfree(lkdtm_kernel_info);
 	return ret;
 }
 
-- 
2.25.1

Re: [PATCH] lkdtm/core: fix resource leaks on module init error path

[kernel test robot]

Hi Jiangshan,

kernel test robot noticed the following build warnings:

[auto build test WARNING on char-misc/char-misc-testing]
[also build test WARNING on char-misc/char-misc-next char-misc/char-misc-linus soc/for-next linus/master v7.2-rc6 next-20260804]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Jiangshan-Yi/lkdtm-core-fix-resource-leaks-on-module-init-error-path/20260804-225548
base:   char-misc/char-misc-testing
patch link:    https://lore.kernel.org/r/20260701151438.3772375-1-yijiangshan%40kylinos.cn
patch subject: [PATCH] lkdtm/core: fix resource leaks on module init error path
config: i386-allmodconfig (https://download.01.org/0day-ci/archive/20260805/202608050915.CRrKhB4T-lkp@intel.com/config)
compiler: gcc-14 (Debian 14.2.0-19) 14.2.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20260805/202608050915.CRrKhB4T-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202608050915.CRrKhB4T-lkp@intel.com/

All warnings (new ones prefixed by >>, old ones prefixed by <<):

>> WARNING: modpost: drivers/misc/lkdtm/lkdtm: section mismatch in reference: lkdtm_module_init+0x221 (section: .init.text) -> lkdtm_heap_exit (section: .exit.text)
>> WARNING: modpost: drivers/misc/lkdtm/lkdtm: section mismatch in reference: lkdtm_module_init+0x226 (section: .init.text) -> lkdtm_usercopy_exit (section: .exit.text)

--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki

Re: Re: [PATCH] lkdtm/core: fix resource leaks on module init error path

[Jiangshan Yi]

Hi,

Thanks for the build test report and for catching these section mismatch
warnings:

> > > WARNING: modpost: drivers/misc/lkdtm/lkdtm: section mismatch in reference: lkdtm_module_init+0x221 (section: .init.text) -> lkdtm_heap_exit (section: .exit.text)
> > > WARNING: modpost: drivers/misc/lkdtm/lkdtm: section mismatch in reference: lkdtm_module_init+0x226 (section: .init.text) -> lkdtm_usercopy_exit (section: .exit.text)

These warnings have already been addressed in the v2 of this patch,
posted at:

https://lore.kernel.org/all/20260730154911.3164527-1-yijiangshan@kylinos.cn/

In v2, the error-handling path in lkdtm_module_init() no longer
references the .exit.text cleanup helpers (lkdtm_heap_exit,
lkdtm_usercopy_exit), which resolves the section mismatch reported
by modpost.

Since the fix is part of a new version of the same patch rather than a
separate commit, I haven't added the Reported-by / Closes tags to
that version. However, your report was very helpful in confirming the
issue — thank you again for the testing.

Best regards,
Jiangshan Yi

Re: [PATCH] lkdtm/core: fix resource leaks on module init error path

[kernel test robot]

Hi Jiangshan,

kernel test robot noticed the following build warnings:

[auto build test WARNING on char-misc/char-misc-testing]
[also build test WARNING on char-misc/char-misc-next char-misc/char-misc-linus soc/for-next linus/master v7.2-rc6 next-20260804]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Jiangshan-Yi/lkdtm-core-fix-resource-leaks-on-module-init-error-path/20260804-225548
base:   char-misc/char-misc-testing
patch link:    https://lore.kernel.org/r/20260701151438.3772375-1-yijiangshan%40kylinos.cn
patch subject: [PATCH] lkdtm/core: fix resource leaks on module init error path
config: alpha-allyesconfig (https://download.01.org/0day-ci/archive/20260805/202608051446.cKZTkVXV-lkp@intel.com/config)
compiler: alpha-linux-gcc (GCC) 16.1.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20260805/202608051446.cKZTkVXV-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202608051446.cKZTkVXV-lkp@intel.com/

All warnings (new ones prefixed by >>, old ones prefixed by <<):

WARNING: modpost: "saved_config" [vmlinux] is COMMON symbol
>> WARNING: modpost: vmlinux: section mismatch in reference: lkdtm_module_init+0x1ec (section: .init.text) -> lkdtm_heap_exit (section: .exit.text)
WARNING: modpost: vmlinux: section mismatch in reference: lkdtm_module_init+0x1f0 (section: .init.text) -> lkdtm_heap_exit (section: .exit.text)
>> WARNING: modpost: vmlinux: section mismatch in reference: lkdtm_module_init+0x1fc (section: .init.text) -> lkdtm_usercopy_exit (section: .exit.text)
WARNING: modpost: vmlinux: section mismatch in reference: lkdtm_module_init+0x200 (section: .init.text) -> lkdtm_usercopy_exit (section: .exit.text)

--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki

Re: Re: [PATCH] lkdtm/core: fix resource leaks on module init error path

[Jiangshan Yi]

Hi,

Thanks for the build test report and for catching these section mismatch
warnings:

> All warnings (new ones prefixed by >>, old ones prefixed by <<):
>
> WARNING: modpost: "saved_config" [vmlinux] is COMMON symbol
> >> WARNING: modpost: vmlinux: section mismatch in reference: lkdtm_module_init+0x1ec (section: .init.text) -> lkdtm_heap_exit (section: .exit.text)
> WARNING: modpost: vmlinux: section mismatch in reference: lkdtm_module_init+0x1f0 (section: .init.text) -> lkdtm_heap_exit (section: .exit.text)
> >> WARNING: modpost: vmlinux: section mismatch in reference: lkdtm_module_init+0x1fc (section: .init.text) -> lkdtm_usercopy_exit (section: .exit.text)
> WARNING: modpost: vmlinux: section mismatch in reference: lkdtm_module_init+0x200 (section: .init.text) -> lkdtm_usercopy_exit (section: .exit.text)
>
> --
> 0-DAY CI Kernel Test Service
> https://github.com/intel/lkp-tests/wiki

These warnings have already been addressed in the v2 of this patch,
posted at:

https://lore.kernel.org/all/20260730154911.3164527-1-yijiangshan@kylinos.cn/

In v2, the error-handling path in lkdtm_module_init() no longer
references the .exit.text cleanup helpers (lkdtm_heap_exit,
lkdtm_usercopy_exit), which resolves the section mismatch reported
by modpost.

Since the fix is part of a new version of the same patch rather than a
separate commit, I haven't added the Reported-by / Closes tags to
that version. However, your report was very helpful in confirming the
issue — thank you again for the testing.

Best regards,
Jiangshan Yi

Re: Re: [PATCH] lkdtm/core: fix resource leaks on module init error path

[Philip Li]

On Mon, Aug 10, 2026 at 09:48:15AM +0800, Jiangshan Yi wrote:
> Hi,
> 
> Thanks for the build test report and for catching these section mismatch
> warnings:
> 
> > All warnings (new ones prefixed by >>, old ones prefixed by <<):
> >
> > WARNING: modpost: "saved_config" [vmlinux] is COMMON symbol
> > >> WARNING: modpost: vmlinux: section mismatch in reference: lkdtm_module_init+0x1ec (section: .init.text) -> lkdtm_heap_exit (section: .exit.text)
> > WARNING: modpost: vmlinux: section mismatch in reference: lkdtm_module_init+0x1f0 (section: .init.text) -> lkdtm_heap_exit (section: .exit.text)
> > >> WARNING: modpost: vmlinux: section mismatch in reference: lkdtm_module_init+0x1fc (section: .init.text) -> lkdtm_usercopy_exit (section: .exit.text)
> > WARNING: modpost: vmlinux: section mismatch in reference: lkdtm_module_init+0x200 (section: .init.text) -> lkdtm_usercopy_exit (section: .exit.text)
> >
> > --
> > 0-DAY CI Kernel Test Service
> > https://github.com/intel/lkp-tests/wiki
> 
> These warnings have already been addressed in the v2 of this patch,
> posted at:
> 
> https://lore.kernel.org/all/20260730154911.3164527-1-yijiangshan@kylinos.cn/

Sorry about the noise, the bot doesn't recognize the v2 correctly, we will
fix it to avoid false positive.

> 
> In v2, the error-handling path in lkdtm_module_init() no longer
> references the .exit.text cleanup helpers (lkdtm_heap_exit,
> lkdtm_usercopy_exit), which resolves the section mismatch reported
> by modpost.
> 
> Since the fix is part of a new version of the same patch rather than a
> separate commit, I haven't added the Reported-by / Closes tags to
> that version. However, your report was very helpful in confirming the
> issue — thank you again for the testing.
> 
> Best regards,
> Jiangshan Yi
>