[PATCH v4 4/4] x86/asm: Rewrite sync_core() to use IRET-to-self

From: Andy Lutomirski
Date: Fri Dec 09 2016 - 13:24:32 EST


Aside from being excessively slow, CPUID is problematic: Linux runs
on a handful of CPUs that don't have CPUID. Use IRET-to-self
instead. IRET-to-self works everywhere, so it makes testing easy.

For reference, On my laptop, IRET-to-self is ~110ns,
CPUID(eax=1, ecx=0) is ~83ns on native and very very slow under KVM,
and MOV-to-CR2 is ~42ns.

While we're at it: sync_core() serves a very specific purpose.
Document it.

Cc: "H. Peter Anvin" <hpa@xxxxxxxxx>
Signed-off-by: Andy Lutomirski <luto@xxxxxxxxxx>
---
arch/x86/include/asm/processor.h | 80 +++++++++++++++++++++++++++++-----------
1 file changed, 58 insertions(+), 22 deletions(-)

diff --git a/arch/x86/include/asm/processor.h b/arch/x86/include/asm/processor.h
index 64fbc937d586..ceb1f4d3f3fa 100644
--- a/arch/x86/include/asm/processor.h
+++ b/arch/x86/include/asm/processor.h
@@ -590,33 +590,69 @@ static __always_inline void cpu_relax(void)

#define cpu_relax_lowlatency() cpu_relax()

-/* Stop speculative execution and prefetching of modified code. */
+/*
+ * This function forces the icache and prefetched instruction stream to
+ * catch up with reality in two very specific cases:
+ *
+ * a) Text was modified using one virtual address and is about to be executed
+ * from the same physical page at a different virtual address.
+ *
+ * b) Text was modified on a different CPU, may subsequently be
+ * executed on this CPU, and you want to make sure the new version
+ * gets executed. This generally means you're calling this in a IPI.
+ *
+ * If you're calling this for a different reason, you're probably doing
+ * it wrong.
+ */
static inline void sync_core(void)
{
- int tmp;
-
-#ifdef CONFIG_X86_32
/*
- * Do a CPUID if available, otherwise do a jump. The jump
- * can conveniently enough be the jump around CPUID.
+ * There are quite a few ways to do this. IRET-to-self is nice
+ * because it works on every CPU, at any CPL (so it's compatible
+ * with paravirtualization), and it never exits to a hypervisor.
+ * The only down sides are that it's a bit slow (it seems to be
+ * a bit more than 2x slower than the fastest options) and that
+ * it unmasks NMIs. The "push %cs" is needed because, in
+ * paravirtual environments, __KERNEL_CS may not be a valid CS
+ * value when we do IRET directly.
+ *
+ * In case NMI unmasking or performance ever becomes a problem,
+ * the next best option appears to be MOV-to-CR2 and an
+ * unconditional jump. That sequence also works on all CPUs,
+ * but it will fault at CPL3 (i.e. Xen PV and lguest).
+ *
+ * CPUID is the conventional way, but it's nasty: it doesn't
+ * exist on some 486-like CPUs, and it usually exits to a
+ * hypervisor.
+ *
+ * Like all of Linux's memory ordering operations, this is a
+ * compiler barrier as well.
*/
- asm volatile("cmpl %2,%1\n\t"
- "jl 1f\n\t"
- "cpuid\n"
- "1:"
- : "=a" (tmp)
- : "rm" (boot_cpu_data.cpuid_level), "ri" (0), "0" (1)
- : "ebx", "ecx", "edx", "memory");
+ register void *__sp asm(_ASM_SP);
+
+#ifdef CONFIG_X86_32
+ asm volatile (
+ "pushfl\n\t"
+ "pushl %%cs\n\t"
+ "pushl $1f\n\t"
+ "iret\n\t"
+ "1:"
+ : "+r" (__sp) : : "memory");
#else
- /*
- * CPUID is a barrier to speculative execution.
- * Prefetched instructions are automatically
- * invalidated when modified.
- */
- asm volatile("cpuid"
- : "=a" (tmp)
- : "0" (1)
- : "ebx", "ecx", "edx", "memory");
+ unsigned int tmp;
+
+ asm volatile (
+ "mov %%ss, %0\n\t"
+ "pushq %q0\n\t"
+ "pushq %%rsp\n\t"
+ "addq $8, (%%rsp)\n\t"
+ "pushfq\n\t"
+ "mov %%cs, %0\n\t"
+ "pushq %q0\n\t"
+ "pushq $1f\n\t"
+ "iretq\n\t"
+ "1:"
+ : "=&r" (tmp), "+r" (__sp) : : "cc", "memory");
#endif
}

--
2.9.3