Date: Mon, 8 Feb 2010 15:05:35 -0500 From: John Baldwin <jhb@freebsd.org> To: Kostik Belousov <kostikbel@gmail.com> Cc: Tom McLaughlin <tmclaugh@sdf.lonestar.org>, freebsd-stable@freebsd.org Subject: Re: Recent MFC to 7 causes crash on VMware ESXi Message-ID: <201002081505.35643.jhb@freebsd.org> In-Reply-To: <20100208194651.GP9991@deviant.kiev.zoral.com.ua> References: <4B6B89E7.8030002@sdf.lonestar.org> <201002081442.33579.jhb@freebsd.org> <20100208194651.GP9991@deviant.kiev.zoral.com.ua>
next in thread | previous in thread | raw e-mail | index | archive | help
On Monday 08 February 2010 2:46:51 pm Kostik Belousov wrote: > On Mon, Feb 08, 2010 at 02:42:33PM -0500, John Baldwin wrote: > > On Monday 08 February 2010 1:51:46 pm Kostik Belousov wrote: > > > On Mon, Feb 08, 2010 at 01:15:06PM -0500, John Baldwin wrote: > > > > > > I thought the old condition only happened under VMware? > > > > > > > > > > Reports I got where from XEN. > > > > > > > > Ok. Those would also be covered under the vm_guest test as it is > > > > non-zero for Xen, VMware, Parallels, etc. > > > > > > What I said was suggestion and not objection. Ignore me. > > > > Were there any reports of problems with Intel CPUs that weren't under a > > virtualization system? If so, we should keep the test, but my understanding > > was that the test was only true under specific virtualization environments. > > Forcing Intel CPU to use CLFLUSH, by clearing SS bit, caused > reserved trap on Pentium M at least. My concern is that if Intel > makes some stripped-down CPU with CLFLUSH but without SS logic, > we would be affected. Hmm, I am content to just workaround things that we know are broken for enough. There are an infinite number of possibilities as far as if future products may be broken. I think for that case we should wait and add a workaround to disable it as new parts that are broken become available. -- John Baldwin
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201002081505.35643.jhb>