Date: Wed, 24 Oct 2012 21:37:47 +0200 From: Andre Oppermann <andre@freebsd.org> To: attilio@FreeBSD.org Cc: mdf@freebsd.org, src-committers@freebsd.org, Bruce Evans <brde@optusnet.com.au>, John Baldwin <jhb@freebsd.org>, svn-src-user@freebsd.org Subject: Re: svn commit: r241889 - in user/andre/tcp_workqueue/sys: arm/arm cddl/compat/opensolaris/kern cddl/contrib/opensolaris/uts/common/dtrace cddl/contrib/opensolaris/uts/common/fs/zfs ddb dev/acpica dev/... Message-ID: <5088438B.90204@freebsd.org> In-Reply-To: <CAJ-FndBENEuyaH%2B2Q%2Bigj39tdGmsHh=3arL-Cb2GP3i9WSr_hQ@mail.gmail.com> References: <201210221418.q9MEINkr026751@svn.freebsd.org> <CAMBSHm-NYtqL71_5zvKj5RNO%2B-ryc6av760eSwBncgP=1d7%2BmQ@mail.gmail.com> <50872624.6060901@freebsd.org> <201210241005.38977.jhb@freebsd.org> <CAJ-FndBENEuyaH%2B2Q%2Bigj39tdGmsHh=3arL-Cb2GP3i9WSr_hQ@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 24.10.2012 16:34, Attilio Rao wrote:
> On Wed, Oct 24, 2012 at 3:05 PM, John Baldwin <jhb@freebsd.org> wrote:
>> On Tuesday, October 23, 2012 7:20:04 pm Andre Oppermann wrote:
>>> Is there a way to do the cache line alignment in a sane way without
>>> littering __aligned(CACHE_LINE_SIZE) all over the place?
>>
>> I was hoping to do something with an anonymous union or some such like:
>>
>> union mtx_aligned {
>> struct mtx;
>> char[roundup2(sizeof(struct mtx), CACHE_LINE_SIZE)];
>> }
>>
>> I don't know if there is a useful way to define an 'aligned mutex' type
>> that will transparently map to a 'struct mtx', e.g.:
>>
>> typedef struct mtx __aligned(CACHE_LINE_SIZE) aligned_mtx_t;
>
> Unfortunately that doesn't work as I've verified with alc@ few months ago.
> The __aligned() attribute only works with structures definition, not
> objects declaration.
There is a huge amount of false cache line sharing in the .bss
section:
ffffffff81515980 B accept_mtx
ffffffff815159a0 B maxsockets
ffffffff815159a8 B socket_zone
ffffffff815159b0 B so_gencnt
ffffffff815159b8 B sysctl__kern_ipc_children
accept_mtx, socket_zone and so_gencnt are frequently accessed but
not related to each other and on different code path's. There is
a lot of pointless cache dirtying going on.
Example randomly picked from "nm -n kernel.debug".
--
Andre
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?5088438B.90204>