linux-kernel - Re: [PATCH -tip] x86: move vmware to hypervisor

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <1238001853.32497.22.camel@alok-dev1>
Date:	Wed, 25 Mar 2009 10:24:13 -0700
From:	Alok Kataria <akataria@...are.com>
To:	Jaswinder Singh Rajput <jaswinder@...nel.org>
Cc:	Ingo Molnar <mingo@...e.hu>, "H. Peter Anvin" <hpa@...or.com>,
	x86 maintainers <x86@...nel.org>,
	LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH -tip] x86: move vmware to hypervisor

On Wed, 2009-03-25 at 10:07 -0700, Jaswinder Singh Rajput wrote:
> On Wed, 2009-03-25 at 09:52 -0700, Alok Kataria wrote:
> > > 
> > > vmware can be considered a CPU here, so i think making the disabling 
> > > also depend on PROCESSOR_SELECT.
> > 
> > Ingo, this code is not just to be used by VMware, the reason we did this
> > generically was so that a guest could run unaltered on *any* fully
> > virtualized hypervisor.
> > And give that this code is just a boot setup thing, the only thing this
> > patch saves over here is not running the detection code on native
> > systems. All the rest of the code is guarded by the
> > "boot_cpu_data.x86_hyper_vendor" checks anyways.  
> > 
> > I don't really see the point of adding one more config option just for
> > this. 
> > 
> 
> Can you please explain what is the point of adding this support all the
> time if this is useless for 99.9% of cases. IMHO, it should be optional.

First of all, I don't know how did you get to the 99.9% number, though I
think its not a point worth debating, just like to share some info with
you. More and more people are adopting virtualization now a days and
give the trend i don't see just 0.1% people running Linux on virtualized
hardware. So though its not a common case there is still a large user
base. 
I am not saying we should not hide this behind a config at all. The
point is there is nothing that we save by adding a new config, so what's
the point at all. If you can give me a solid reason like, say, you save
1% code space with this config option, or 'n' sec in the boottime,  I am
all ears for such an argument.

Thanks,
Alok

> 
> --
> JSR
> 
> 
> 
> 
> 

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/