linux-kernel - [PATCH] x86/cpuid: Fix up "virtual" IBRS/IBPB/STIBP feature bits on Intel

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-next>] [day] [month] [year] [list]

Message-Id: <1517269773-16750-1-git-send-email-dwmw@amazon.co.uk>
Date:   Mon, 29 Jan 2018 23:49:33 +0000
From:   David Woodhouse <dwmw@...zon.co.uk>
To:     arjan@...ux.intel.com, tglx@...utronix.de, karahmed@...zon.de,
        x86@...nel.org, linux-kernel@...r.kernel.org,
        tim.c.chen@...ux.intel.com, bp@...en8.de, peterz@...radead.org,
        pbonzini@...hat.com, ak@...ux.intel.com,
        torvalds@...ux-foundation.org, gregkh@...ux-foundation.org
Subject: [PATCH] x86/cpuid: Fix up "virtual" IBRS/IBPB/STIBP feature bits on Intel

Despite the fact that all the other code there seems to be doing it,
just using set_cpu_cap() in early_intel_init() doesn't actually work.

When the CPU is queried again in identify_boot_cpu(), it all gets
overwritten again. Do it in init_scattered_cpuid_features() instead.

Turning the bits off for bad microcode can use setup_clear_cpu_cap()
to force them off for all CPUs; I was less keen on forcing the feature
bits *on* that way.

Signed-off-by: David Woodhouse <dwmw@...zon.co.uk>
Fixes: 2961298e ("x86/cpufeatures: Clean up Spectre v2 related CPUID flags")
---
I feel I must be missing something. Is the rest of early_init_intel() broken too?

 arch/x86/kernel/cpu/intel.c     | 27 ++++++++-------------------
 arch/x86/kernel/cpu/scattered.c | 13 +++++++++++++
 2 files changed, 21 insertions(+), 19 deletions(-)

diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c
index 0c8b916..4cf4f8c 100644
--- a/arch/x86/kernel/cpu/intel.c
+++ b/arch/x86/kernel/cpu/intel.c
@@ -175,28 +175,17 @@ static void early_init_intel(struct cpuinfo_x86 *c)
 	if (c->x86 >= 6 && !cpu_has(c, X86_FEATURE_IA64))
 		c->microcode = intel_get_microcode_revision();
 
-	/*
-	 * The Intel SPEC_CTRL CPUID bit implies IBRS and IBPB support,
-	 * and they also have a different bit for STIBP support. Also,
-	 * a hypervisor might have set the individual AMD bits even on
-	 * Intel CPUs, for finer-grained selection of what's available.
-	 */
-	if (cpu_has(c, X86_FEATURE_SPEC_CTRL)) {
-		set_cpu_cap(c, X86_FEATURE_IBRS);
-		set_cpu_cap(c, X86_FEATURE_IBPB);
-	}
-	if (cpu_has(c, X86_FEATURE_INTEL_STIBP))
-		set_cpu_cap(c, X86_FEATURE_STIBP);
-
 	/* Now if any of them are set, check the blacklist and clear the lot */
-	if ((cpu_has(c, X86_FEATURE_IBRS) || cpu_has(c, X86_FEATURE_IBPB) ||
+	if ((cpu_has(c, X86_FEATURE_SPEC_CTRL) ||
+	     cpu_has(c, X86_FEATURE_INTEL_STIBP) ||
+	     cpu_has(c, X86_FEATURE_IBRS) || cpu_has(c, X86_FEATURE_IBPB) ||
 	     cpu_has(c, X86_FEATURE_STIBP)) && bad_spectre_microcode(c)) {
 		pr_warn("Intel Spectre v2 broken microcode detected; disabling Speculation Control\n");
-		clear_cpu_cap(c, X86_FEATURE_IBRS);
-		clear_cpu_cap(c, X86_FEATURE_IBPB);
-		clear_cpu_cap(c, X86_FEATURE_STIBP);
-		clear_cpu_cap(c, X86_FEATURE_SPEC_CTRL);
-		clear_cpu_cap(c, X86_FEATURE_INTEL_STIBP);
+		setup_clear_cpu_cap(X86_FEATURE_IBRS);
+		setup_clear_cpu_cap(X86_FEATURE_IBPB);
+		setup_clear_cpu_cap(X86_FEATURE_STIBP);
+		setup_clear_cpu_cap(X86_FEATURE_SPEC_CTRL);
+		setup_clear_cpu_cap(X86_FEATURE_INTEL_STIBP);
 	}
 
 	/*
diff --git a/arch/x86/kernel/cpu/scattered.c b/arch/x86/kernel/cpu/scattered.c
index df11f5d..121b91e 100644
--- a/arch/x86/kernel/cpu/scattered.c
+++ b/arch/x86/kernel/cpu/scattered.c
@@ -53,6 +53,19 @@ void init_scattered_cpuid_features(struct cpuinfo_x86 *c)
 		if (regs[cb->reg] & (1 << cb->bit))
 			set_cpu_cap(c, cb->feature);
 	}
+
+	/*
+	 * The Intel SPEC_CTRL CPUID bit implies IBRS and IBPB support,
+	 * and they also have a different bit for STIBP support. Also,
+	 * a hypervisor might have set the individual AMD bits even on
+	 * Intel CPUs, for finer-grained selection of what's available.
+	 */
+	if (cpu_has(c, X86_FEATURE_SPEC_CTRL)) {
+		set_cpu_cap(c, X86_FEATURE_IBRS);
+		set_cpu_cap(c, X86_FEATURE_IBPB);
+	}
+	if (cpu_has(c, X86_FEATURE_INTEL_STIBP))
+		set_cpu_cap(c, X86_FEATURE_STIBP);
 }
 
 u32 get_scattered_cpuid_leaf(unsigned int level, unsigned int sub_leaf,
-- 
2.7.4