Date: Sun, 11 Jan 2009 22:54:22 +0000 (GMT) From: Robert Watson <rwatson@FreeBSD.org> To: Jason DiCioccio <jd@ods.org> Cc: freebsd-net@freebsd.org Subject: Re: Still trying to debug mbuf leak.. Anyone familiar with uma_*? Message-ID: <alpine.BSF.2.00.0901112252470.16794@fledge.watson.org> In-Reply-To: <496A5800.5080506@ods.org> References: <496A5800.5080506@ods.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, 11 Jan 2009, Jason DiCioccio wrote:
> So I've got various vmcores here from kmem_malloc failures.. I already know
> that it's an excess of mbufs that's causing the crash, I'm just trying to
> figure out what the mbufs contain so that I can try and figure out where
> they came from..
>
> Currently I'm able to get at (uma_zone_t)zone_mbuf, but I'm unsure how to
> get to the actual mbuf data from there. I've tried looking the the uma_keg,
> etc. No luck so far. Is there anyone familiar with these data structures
> that could point me in the right direction?
It might require tweaking to work as things have changed a bit, but last time
I was futzing with the mbuf allocator I put a UMA state dumper into
src/tools/tools/umastat. It inspects /dev/kmem (or a core, I suppose) to
print out the layout of cached data across per-cpu caches, etc. I'm not sure
it would be all that useful for debugging a leak, but might prove informative
for getting a sense of how it fits together, and the source code might be
useful similarly.
Robert N M Watson
Computer Laboratory
University of Cambridge
>
> (kgdb) p *zone_mbuf
> $66 = {uz_name = 0xc076d6d1 "mbuf", uz_lock = 0xc1048108, uz_keg =
> 0xc1048100,
> uz_link = {le_next = 0xc1047180, le_prev = 0xc104812c}, uz_full_bucket = {
> lh_first = 0x0}, uz_free_bucket = {lh_first = 0x0},
> uz_ctor = 0xc05068f0 <mb_ctor_mbuf>, uz_dtor = 0xc0507040 <mb_dtor_mbuf>,
> uz_init = 0, uz_fini = 0, uz_allocs = 137806761, uz_frees = 136018292,
> uz_fails = 136, uz_fills = 4, uz_count = 128, uz_cpu = {{
> uc_freebucket = 0xc73a8c48, uc_allocbucket = 0x0, uc_allocs = 0,
> uc_frees = 0}}}
>
> Thanks!
> -JD-
> _______________________________________________
> freebsd-net@freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-net
> To unsubscribe, send any mail to "freebsd-net-unsubscribe@freebsd.org"
>
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?alpine.BSF.2.00.0901112252470.16794>