Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 23 Nov 2014 19:02:06 +0100
From:      Tijl Coosemans <tijl@FreeBSD.org>
To:        Scott Long <scott4long@yahoo.com>
Cc:        svn-src-head@freebsd.org, svn-src-all@freebsd.org, src-committers@freebsd.org, Scott Long <scottl@FreeBSD.org>, Rui Paulo <rpaulo@me.com>
Subject:   Re: svn commit: r274489 - in head/sys/amd64: amd64 include
Message-ID:  <20141123190206.37942760@kalimero.tijl.coosemans.org>
In-Reply-To: <13EC3116-6146-42FC-8941-2C7C009224B3@yahoo.com>
References:  <201411132211.sADMBjP3009246@svn.freebsd.org> <35E5EAD8-99C1-43C0-8D01-B3B5B86ECA25@me.com> <13EC3116-6146-42FC-8941-2C7C009224B3@yahoo.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, 21 Nov 2014 16:26:47 -0700 Scott Long <scott4long@yahoo.com> wrote:
> On Nov 20, 2014, at 11:33 PM, Rui Paulo <rpaulo@me.com> wrote:
>> On Nov 13, 2014, at 14:11, Scott Long <scottl@FreeBSD.org> wrote:
>>>=20
>>> Author: scottl
>>> Date: Thu Nov 13 22:11:44 2014
>>> New Revision: 274489
>>> URL: https://svnweb.freebsd.org/changeset/base/274489
>>>=20
>>> Log:
>>> Extend earlier addition of stack frames to most of support.S.  This mak=
es
>>> stack traces in KDB, HWPMC, and DTrace much more reliable and useful.
>>=20
>> No performance differences?  The kernel enables/disables the compiler
>> option to omit the frame pointer based on the kernel config file.  If
>> DDB, DTrace, or HWPMC is enabled, the frame pointer is always saved in
>> C functions.
>>=20
>> Some of these functions are in the hot path, so if you didn't see any
>> performance problem, I wonder if we should disable -fomit-frame-pointer
>> always.
>=20
> That=E2=80=99s a good question to look further into.  I didn=E2=80=99t se=
e any measurable
> differences with this change.  I think that the cost of the function call
> itself masks the cost of a few extra instructions, but I didn=E2=80=99t t=
est with
> switching it on/off for the entire kernel.  That said, I purposely
> implemented this as macros so it could be easily changed in the future.
> If someone finds that this measurably impacts a certain workload, I
> wouldn=E2=80=99t object to making it conditional, though it does complica=
te any
> hand-written ASM code that tries to access the stack via %esp offsets.
> We don=E2=80=99t have anything like that now, but Kip was concerned enoug=
h about
> it in the future that I left it enabled unconditionally.

An alternative is to annotate the functions with .cfi directives.
Stack unwinding doesn't need frame pointers then.
http://www.logix.cz/michal/devel/gas-cfi/
https://sourceware.org/binutils/docs-2.24/as/CFI-directives.html



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20141123190206.37942760>