Re: [tip:x86/asm] x86/umip: Add emulation code for UMIP instructions

From: "H. Peter Anvin" <hpa@zytor.com>
To: Linus Torvalds <torvalds@linux-foundation.org>,
	Denys Vlasenko <dvlasenk@redhat.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>, Borislav Petkov <bp@suse.de>,
	Vlastimil Babka <vbabka@suse.cz>, Tony Luck <tony.luck@intel.com>,
	Fenghua Yu <fenghua.yu@intel.com>,
	Josh Poimboeuf <jpoimboe@redhat.com>,
	Borislav Petkov <bp@alien8.de>, Brian Gerst <brgerst@gmail.com>,
	Jonathan Corbet <corbet@lwn.net>,
	"Ravi V. Shankar" <ravi.v.shankar@intel.com>,
	ricardo.neri-calderon@linux.intel.com,
	Chris Metcalf <cmetcalf@mellanox.com>,
	Andrew Morton <akpm@linux-foundation.org>,
	Paul Gortmaker <paul.gortmaker@windriver.com>,
	Masami Hiramatsu <mhiramat@kernel.org>,
	Andrew Lutomirski <luto@kernel.org>,
	Dave Hansen <dave.hansen@linux.intel.com>,
	ray.huang@amd.com, "Michael S. Tsirkin" <mst@redhat.com>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	Peter Zijlstra <peterz@infradead.org>,
	Chen Yucong <slaoub@gmail.com>, Ingo Molnar <mingo@kernel.org>,
	Jiri Slaby <jslaby@suse.cz>, Shuah Khan <shuah@kernel.org>,
	Thomas Gleixner <tglx@linutronix.de>,
	"linux-tip-commits@vger.kernel.org"@zytor.com
Subject: Re: [tip:x86/asm] x86/umip: Add emulation code for UMIP instructions
Date: Wed, 08 Nov 2017 08:38:18 -0800	[thread overview]
Message-ID: <201711081638.vA8GcSMV026939@mail.zytor.com> (raw)
In-Reply-To: <CA+55aFyzb5D6DfM_1mxcMbuYOaxZWFxqfv7WZS6zdkqnN7+p0A@mail.gmail.com>

<linux-tip-commits@vger.kernel.org>
From: hpa@zytor.com
Message-ID: <67A3DB9F-E58D-42B3-BF16-AC59D2FCECC2@zytor.com>

On November 8, 2017 8:34:45 AM PST, Linus Torvalds <torvalds@linux-foundation.org> wrote:
>On Wed, Nov 8, 2017 at 8:14 AM, Denys Vlasenko <dvlasenk@redhat.com>
>wrote:
>>
>> Can we avoid maintain emulation of these isns, by asking Wine to
>remove
>> their use instead?
>
>If we ask the Wine people to remove the instruction use, that may mean
>that we can avoid the emulation in four or five _years_ once everybody
>has updated.
>
>But it wouldn't mean that we could avoid it today.
>
>We *do* probably want to print a warning about these instructions,
>whether emulated or not. If emulated, we want to print a warning about
>the emulation being expensive (as a nudge to the Wine people to not do
>that), and if _not_ emulated, we'd want to print a warning about why
>people suddenly got a SIGSEGV in case there are programs that use
>those instructions that people didn't even think of.
>
>But make it ratelimit the warning (perhaps at most once every few
>minutes or whatever), so that people won't flood message logs just
>because they have old stupid programs that use those things.
>
>             Linus

The real problem is that 16-bit software uses the instructions mostly for stupid reasons, but smsw in particular is used exactly because it is a virtualization hole.
-- 
Sent from my Android device with K-9 Mail. Please excuse my brevity.