[RFC PATCH v1] powerpc: Enable KFENCE for PPC32

[Christophe Leroy]

Le 04/03/2021 à 12:31, Marco Elver a écrit :
> On Thu, 4 Mar 2021 at 12:23, Christophe Leroy
> <christophe.leroy at csgroup.eu> wrote:
>> Le 03/03/2021 à 11:56, Marco Elver a écrit :
>>>
>>> Somewhat tangentially, I also note that e.g. show_regs(regs) (which
>>> was printed along the KFENCE report above) didn't include the top
>>> frame in the "Call Trace", so this assumption is definitely not
>>> isolated to KFENCE.
>>>
>>
>> Now, I have tested PPC64 (with the patch I sent yesterday to modify save_stack_trace_regs()
>> applied), and I get many failures. Any idea ?
>>
>> [   17.653751][   T58] ==================================================================
>> [   17.654379][   T58] BUG: KFENCE: invalid free in .kfence_guarded_free+0x2e4/0x530
>> [   17.654379][   T58]
>> [   17.654831][   T58] Invalid free of 0xc00000003c9c0000 (in kfence-#77):
>> [   17.655358][   T58]  .kfence_guarded_free+0x2e4/0x530
>> [   17.655775][   T58]  .__slab_free+0x320/0x5a0
>> [   17.656039][   T58]  .test_double_free+0xe0/0x198
>> [   17.656308][   T58]  .kunit_try_run_case+0x80/0x110
>> [   17.656523][   T58]  .kunit_generic_run_threadfn_adapter+0x38/0x50
>> [   17.657161][   T58]  .kthread+0x18c/0x1a0
>> [   17.659148][   T58]  .ret_from_kernel_thread+0x58/0x70
>> [   17.659869][   T58]
>> [   17.663954][   T58] kfence-#77 [0xc00000003c9c0000-0xc00000003c9c001f, size=32, cache=kmalloc-32]
>> allocated by task 58:
>> [   17.666113][   T58]  .__kfence_alloc+0x1bc/0x510
>> [   17.667069][   T58]  .__kmalloc+0x280/0x4f0
>> [   17.667452][   T58]  .test_alloc+0x19c/0x430
>> [   17.667732][   T58]  .test_double_free+0x88/0x198
>> [   17.667971][   T58]  .kunit_try_run_case+0x80/0x110
>> [   17.668283][   T58]  .kunit_generic_run_threadfn_adapter+0x38/0x50
>> [   17.668553][   T58]  .kthread+0x18c/0x1a0
>> [   17.669315][   T58]  .ret_from_kernel_thread+0x58/0x70
>> [   17.669711][   T58]
>> [   17.669711][   T58] freed by task 58:
>> [   17.670116][   T58]  .kfence_guarded_free+0x3d0/0x530
>> [   17.670421][   T58]  .__slab_free+0x320/0x5a0
>> [   17.670603][   T58]  .test_double_free+0xb4/0x198
>> [   17.670827][   T58]  .kunit_try_run_case+0x80/0x110
>> [   17.671073][   T58]  .kunit_generic_run_threadfn_adapter+0x38/0x50
>> [   17.671410][   T58]  .kthread+0x18c/0x1a0
>> [   17.671618][   T58]  .ret_from_kernel_thread+0x58/0x70
>> [   17.671972][   T58]
>> [   17.672638][   T58] CPU: 0 PID: 58 Comm: kunit_try_catch Tainted: G    B
>> 5.12.0-rc1-01540-g0783285cc1b8-dirty #4685
>> [   17.673768][   T58] ==================================================================
>> [   17.677031][   T58]     # test_double_free: EXPECTATION FAILED at mm/kfence/kfence_test.c:380
>> [   17.677031][   T58]     Expected report_matches(&expect) to be true, but is false
>> [   17.684397][    T1]     not ok 7 - test_double_free
>> [   17.686463][   T59]     # test_double_free-memcache: setup_test_cache: size=32, ctor=0x0
>> [   17.688403][   T59]     # test_double_free-memcache: test_alloc: size=32, gfp=cc0, policy=any,
>> cache=1
> 
> Looks like something is prepending '.' to function names. We expect
> the function name to appear as-is, e.g. "kfence_guarded_free",
> "test_double_free", etc.
> 
> Is there something special on ppc64, where the '.' is some convention?
> 

I think so, see https://refspecs.linuxfoundation.org/ELF/ppc64/PPC-elf64abi.html#FUNC-DES

Also see commit https://github.com/linuxppc/linux/commit/02424d896

Christophe