Re: Anyone else seeing uma_zalloc_arg panics?

On 23 Nov, Mark Robert Vaughan Murray wrote:
> Hi *
> 
> Anyone else seeing panics like this?
> 
> This is CURRENT, amd64, 2 CPUs, source synced on Thursday 21st Nov evening GMT sometime.
> 
> +panic: uma_zalloc_arg: Returning an empty bucket.
> +cpuid = 1
> +KDB: stack backtrace:
> +db_trace_self_wrapper() at db_trace_self_wrapper+0x2b/frame 0xfffffe003c7dd820
> +kdb_backtrace() at kdb_backtrace+0x39/frame 0xfffffe003c7dd8d0
> +vpanic() at vpanic+0x126/frame 0xfffffe003c7dd910
> +kassert_panic() at kassert_panic+0x136/frame 0xfffffe003c7dd980
> +uma_zalloc_arg() at uma_zalloc_arg+0x49a/frame 0xfffffe003c7dd9f0
> +bucket_alloc() at bucket_alloc+0x8b/frame 0xfffffe003c7dda10
> +uma_zfree_arg() at uma_zfree_arg+0x2a1/frame 0xfffffe003c7dda70
> +bucket_cache_drain() at bucket_cache_drain+0x1b/frame 0xfffffe003c7ddaa0
> +zone_drain_wait() at zone_drain_wait+0xd1/frame 0xfffffe003c7ddb00
> +uma_reclaim() at uma_reclaim+0x42d/frame 0xfffffe003c7ddb30
> +vm_pageout() at vm_pageout+0x3bc/frame 0xfffffe003c7ddbb0
> +fork_exit() at fork_exit+0x84/frame 0xfffffe003c7ddbf0
> +fork_trampoline() at fork_trampoline+0xe/frame 0xfffffe003c7ddbf0
> +--- trap 0, rip = 0, rsp = 0xfffffe003c7ddcb0, rbp = 0 ---
> +KDB: enter: panic
> 
> M

Yes, I got one of these a couple days ago:
<http://docs.freebsd.org/cgi/mid.cgi?201311212234.rALMY0bN009437>

Just curious ... are you running i386 or amd64?  My test machine is
i386, but I suspect that most people running head are running amd64.