Re: [PATCH 2/3] x86: Pass in pt_regs pointer for syscalls that need it

From: "H. Peter Anvin" <hpa@zytor.com>
To: Brian Gerst <brgerst@gmail.com>
Cc: "H. Peter Anvin" <hpa@kernel.org>, Tejun Heo <tj@kernel.org>,
	Ingo Molnar <mingo@elte.hu>,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH 2/3] x86: Pass in pt_regs pointer for syscalls that need it
Date: Wed, 11 Feb 2009 12:00:17 -0800	[thread overview]
Message-ID: <49932E51.8040009@zytor.com> (raw)
In-Reply-To: <73c1f2160902111157g202dbdd2h1bc75b0a8cb242b3@mail.gmail.com>

Brian Gerst wrote:
> 
> IMHO, copying the 4th-6th args to a new stack frame is the only way to
> guarantee that gcc won't trash any part of pt_regs.  The question is
> whether to do it unconditionally, or try to be clever and only copy
> them for the syscalls that actually need them.
> 

My guess is that the conditionalization would actually cost more than 
doing it unconditionally.  We're talking a small fraction of a cache 
line, and a set of stores to RAM, which can be buffered.

It's in many ways easier than reorganizing the struct pt_regs.  I'm just 
hypersensitive to adding system call overhead in any way.

	-hpa