svn commit: r281451 - head/sys/vm

[Scott Long]

> On Apr 25, 2015, at 1:59 PM, Dmitry Morozovsky <marck at rinet.ru> wrote:
> 
> On Sat, 25 Apr 2015, Scott Long via svn-src-all wrote:
> 
>>>> True, it's not actually odd, it's just surprising the first time
>>>> one comes across it.
>>>> 
>>>> Also, I goofed in the text:
>>>> 
>>>>>> With the flexible array, (sizeof(struct uma_cache)) is going to be
>>>>>> 32 bytes smaller than without it.
>>>> 
>>>> It's `struct uma_zone` that shrinks by (potentially) more than one
>>>> would expect.  (The uma_cache struct is itself 32 bytes and is not
>>>> changed, so the args.size value should be the same -- implying that
>>>> some OTHER "sizeof(struct uma_zone)" is where things are going wrong.)
>>>> 
>>> vm_mem_init() called before uma init. unfortunately, we have not seen
>>> a proper info about panic from scottl
> 
>> Too early in boot to get a crashdump.
> 
> ... and even console not set up enough to have a backtrace/some debugging 
> printfs displayed?

Maybe you didn’t see the backtrace that was printed and included?  What debugging printfs should I have predicted that you would want?

Scott