[PATCH v2 1/2] x86/msr: Carry on after a non-"safe" MSR access fails without !panic_on_oops

[Andy Lutomirski]

This demotes an OOPS and likely panic due to a failed non-"safe" MSR
access to a WARN_ON_ONCE and a return of zero (in the RDMSR case).
We still write a pr_info entry unconditionally for debugging.

To be clear, this type of failure should *not* happen.  This patch
exists to minimize the chance of nasty undebuggable failures due on
systems that used to work due to a now-fixed CONFIG_PARAVIRT=y bug.

Signed-off-by: Andy Lutomirski <luto@xxxxxxxxxx>
---
 arch/x86/kernel/traps.c | 55 +++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 55 insertions(+)

diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c
index 346eec73f7db..f82987643e32 100644
--- a/arch/x86/kernel/traps.c
+++ b/arch/x86/kernel/traps.c
@@ -437,6 +437,58 @@ exit_trap:
 	do_trap(X86_TRAP_BR, SIGSEGV, "bounds", regs, error_code, NULL);
 }
 
+static bool paper_over_kernel_gpf(struct pt_regs *regs)
+{
+	/*
+	 * Try to decode the opcode that failed.  So far, we only care
+	 * about boring two-byte unprefixed opcodes, so we don't need
+	 * the full instruction decoder machinery.
+	 */
+	u16 opcode;
+
+	if (probe_kernel_read(&opcode, (const void *)regs->ip, sizeof(opcode)))
+		return false;
+
+	if (opcode == 0x320f) {
+		/* RDMSR */
+		pr_info("bad kernel RDMSR from non-existent MSR 0x%x",
+			(unsigned int)regs->cx);
+		if (!panic_on_oops) {
+			WARN_ON_ONCE(true);
+
+			/*
+			 * Pretend that RDMSR worked and returned zero.  We
+			 * chose zero because zero seems less likely to
+			 * cause further malfunctions than any other value.
+			 */
+			regs->ax = 0;
+			regs->dx = 0;
+			regs->ip += 2;
+			return true;
+		} else {
+			/* Don't fix it up. */
+			return false;
+		}
+	} else if (opcode == 0x300f) {
+		/* WRMSR */
+		pr_info("bad kernel WRMSR writing 0x%08x%08x to MSR 0x%x",
+			(unsigned int)regs->dx, (unsigned int)regs->ax,
+			(unsigned int)regs->cx);
+		if (!panic_on_oops) {
+			WARN_ON_ONCE(true);
+
+			/* Pretend it worked and carry on. */
+			regs->ip += 2;
+			return true;
+		} else {
+			/* Don't fix it up. */
+			return false;
+		}
+	}
+
+	return false;
+}
+
 dotraplinkage void
 do_general_protection(struct pt_regs *regs, long error_code)
 {
@@ -456,6 +508,9 @@ do_general_protection(struct pt_regs *regs, long error_code)
 		if (fixup_exception(regs))
 			return;
 
+		if (paper_over_kernel_gpf(regs))
+			return;
+
 		tsk->thread.error_code = error_code;
 		tsk->thread.trap_nr = X86_TRAP_GP;
 		if (notify_die(DIE_GPF, "general protection fault", regs, error_code,
-- 
2.4.3

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