Re: [GIT PULL] x86/topology changes for v5.3

From: Kees Cook <keescook@chromium.org>
To: Thomas Gleixner <tglx@linutronix.de>
Cc: Linus Torvalds <torvalds@linux-foundation.org>,
	Ingo Molnar <mingo@kernel.org>,
	Linux List Kernel Mailing <linux-kernel@vger.kernel.org>,
	Borislav Petkov <bp@alien8.de>, Len Brown <lenb@kernel.org>,
	Peter Zijlstra <a.p.zijlstra@chello.nl>,
	Andrew Morton <akpm@linux-foundation.org>,
	"Rafael J. Wysocki" <rafael.j.wysocki@intel.com>,
	Tony Luck <tony.luck@intel.com>, Jiri Kosina <jkosina@suse.cz>,
	Bob Moore <robert.moore@intel.com>,
	Erik Schmauss <erik.schmauss@intel.com>
Subject: Re: [GIT PULL] x86/topology changes for v5.3
Date: Tue, 9 Jul 2019 17:31:07 -0700	[thread overview]
Message-ID: <201907091727.91CC6C72D8@keescook> (raw)
In-Reply-To: <alpine.DEB.2.21.1907100115220.1758@nanos.tec.linutronix.de>

On Wed, Jul 10, 2019 at 01:17:11AM +0200, Thomas Gleixner wrote:
> On Wed, 10 Jul 2019, Thomas Gleixner wrote:
> > 
> > That still does not explain the cr4/0 issue you have. Can you send me your
> > .config please?
> 
> Does your machine have UMIP support? None of my test boxes has. So that'd
> be the difference of bits enforced in CR4. Should not matter because it's
> User mode instruction prevention, but who knows.

Ew. Yeah, I don't have i9 nor i7 for testing this. I did try everything
else I had (and hibernation). Is only Linus able to reproduce this so far?

To rule out (in?) UMIP, this would remove UMIP from the pinning:

diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c
index 309b6b9b49d4..f3beedb6da8a 100644
--- a/arch/x86/kernel/cpu/common.c
+++ b/arch/x86/kernel/cpu/common.c
@@ -380,7 +380,7 @@ static void __init setup_cr_pinning(void)
 {
 	unsigned long mask;
 
-	mask = (X86_CR4_SMEP | X86_CR4_SMAP | X86_CR4_UMIP);
+	mask = (X86_CR4_SMEP | X86_CR4_SMAP);
 	cr4_pinned_bits = this_cpu_read(cpu_tlbstate.cr4) & mask;
 	static_key_enable(&cr_pinning.key);
 }


-- 
Kees Cook
