Re: [PATCH] nvdimm: Fix devs leaks in scan_labels()

["Zhijian Li (Fujitsu)" <lizhijian@fujitsu.com>]

On 07/06/2024 00:49, Ira Weiny wrote:
> Li Zhijian wrote:
>> Don't allocate devs again when it's valid pointer which has pionted to
>> the memory allocated above with size (count + 2 * sizeof(dev)).
>>
>> A kmemleak reports:
>> unreferenced object 0xffff88800dda1980 (size 16):
>>    comm "kworker/u10:5", pid 69, jiffies 4294671781
>>    hex dump (first 16 bytes):
>>      00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
>>    backtrace (crc 0):
>>      [<00000000c5dea560>] __kmalloc+0x32c/0x470
>>      [<000000009ed43c83>] nd_region_register_namespaces+0x6fb/0x1120 [libnvdimm]
>>      [<000000000e07a65c>] nd_region_probe+0xfe/0x210 [libnvdimm]
>>      [<000000007b79ce5f>] nvdimm_bus_probe+0x7a/0x1e0 [libnvdimm]
>>      [<00000000a5f3da2e>] really_probe+0xc6/0x390
>>      [<00000000129e2a69>] __driver_probe_device+0x78/0x150
>>      [<000000002dfed28b>] driver_probe_device+0x1e/0x90
>>      [<00000000e7048de2>] __device_attach_driver+0x85/0x110
>>      [<0000000032dca295>] bus_for_each_drv+0x85/0xe0
>>      [<00000000391c5a7d>] __device_attach+0xbe/0x1e0
>>      [<0000000026dabec0>] bus_probe_device+0x94/0xb0
>>      [<00000000c590d936>] device_add+0x656/0x870
>>      [<000000003d69bfaa>] nd_async_device_register+0xe/0x50 [libnvdimm]
>>      [<000000003f4c52a4>] async_run_entry_fn+0x2e/0x110
>>      [<00000000e201f4b0>] process_one_work+0x1ee/0x600
>>      [<000000006d90d5a9>] worker_thread+0x183/0x350
>>
>> Fixes: 1b40e09a1232 ("libnvdimm: blk labels and namespace instantiation")
>> Signed-off-by: Li Zhijian <lizhijian@fujitsu.com>
>> ---
>>   drivers/nvdimm/namespace_devs.c | 4 +++-
>>   1 file changed, 3 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/nvdimm/namespace_devs.c b/drivers/nvdimm/namespace_devs.c
>> index d6d558f94d6b..56b016dbe307 100644
>> --- a/drivers/nvdimm/namespace_devs.c
>> +++ b/drivers/nvdimm/namespace_devs.c
>> @@ -1994,7 +1994,9 @@ static struct device **scan_labels(struct nd_region *nd_region)
>>   		/* Publish a zero-sized namespace for userspace to configure. */
>>   		nd_mapping_free_labels(nd_mapping);
>>   
>> -		devs = kcalloc(2, sizeof(dev), GFP_KERNEL);
>> +		/* devs probably has been allocated */
> 
> I don't think this is where the bug is.  The loop above is processing the
> known labels and should exit with a count > 0 if devs is not NULL.
> 
>  From what I can tell create_namespace_pmem() must be returning EAGAIN
> which leaves devs allocated but fails to increment count.  Thus there are
> no valid labels but devs was not free'ed.

Per the piece of the code, return EAGAIN and ENODEV could cause this issue in theory.

1980                 dev = create_namespace_pmem(nd_region, nd_mapping, nd_label);
1981                 if (IS_ERR(dev)) {
1982                         switch (PTR_ERR(dev)) {
1983                         case -EAGAIN:
1984                                 /* skip invalid labels */
1985                                 continue;
1986                         case -ENODEV:
1987                                 /* fallthrough to seed creation */
1988                                 break;
1989                         default:
1990                                 goto err;
1991                         }
1992                 } else
1993                         devs[count++] = dev;


> 
> Can you trace the error you are seeing a bit more to see if this is the
> case?


I just tried, but I cannot reproduce this leaking again.
When it happened(100% reproduce at that time), I probably had a corrupted LSA(I see empty
output with command 'ndctl list'). It seemed the QEMU emulated Nvdimm device was broken
for some reasons.


Thanks
Zhijian


> 
> Thanks,
> Ira
> 
>> +		if (!devs)
>> +			devs = kcalloc(2, sizeof(dev), GFP_KERNEL);
>>   		if (!devs)
>>   			goto err;
>>   
>> -- 
>> 2.29.2
>>
> 
>