Re: Regression in v5.0-rc1: Panic at boot

[Logan Gunthorpe]

On 2019-01-08 6:19 a.m., Christoph Hellwig wrote:
> On Mon, Jan 07, 2019 at 12:41:06PM -0700, Logan Gunthorpe wrote:
>> Hey,
>>
>> I found a regression in v5.0-rc1 this morning. My system panics on boot.
>> I've attached a log of the panic.
>>
>> I bisected to find the problematic commit is:
>>
>> Fixes: 9d037ad707ed ("block: remove req->timeout_list")
>>
>> But it makes no sense to me why this commit would cause a problem like
>> this. I've attached a bisect log. I've also tested v5.0-rc1 with this
>> commit reverted and that boots fine.
>>
>> The traceback seems to indicate the problem is on the bip_get_seed()
>> line in t10_pi_complete(). Which suggests that bio_integrity() is
>> returning NULL.
> 
> Very odd.  Can you try an experiment?  Can you add padding the size
> of struct timer_list to struct request to check if that makes the
> problem go away?  Then move the padding from the where the field
> was to the end and see if that still "helps"?

Ok I tried these things and they all boot without panic:

1) Add two void pointers to where 'timer_list' was
2) Add two void pointer to the end of the struct
3) Add one void pointer to the end of the struct

So it seems to be a struct size issue...

Logan