[PATCH v2 1/4] x86/percpu: Introduce PER_CPU_ARG and use it in cmpxchg{8,16}b_emu.S

[Uros Bizjak <ubizjak@xxxxxxxxx>]

PER_CPU_VAR macro is intended to be applied to a symbol and should not
be used with general operands. Introduce new PER_CPU_ARG macro and
use it in cmpxchg{8,16}b_emu.S instead.

PER_CPU_VAR macro will be repurposed for %rip-relative addressing.

Also add a missing function comment to this_cpu_cmpxchg8b_emu.

No functional changes intended.

Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
Cc: Ingo Molnar <mingo@xxxxxxxxxx>
Cc: Borislav Petkov <bp@xxxxxxxxx>
Cc: Dave Hansen <dave.hansen@xxxxxxxxxxxxxxx>
Cc: "H. Peter Anvin" <hpa@xxxxxxxxx>
Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Signed-off-by: Uros Bizjak <ubizjak@xxxxxxxxx>
--
v2: Introduce PER_CPU_ARG macro to conditionally enable
    segment registers in cmpxchg{8,16}b_emu.S for CONFIG_SMP.
---
 arch/x86/include/asm/percpu.h |  2 ++
 arch/x86/lib/cmpxchg16b_emu.S | 12 ++++++------
 arch/x86/lib/cmpxchg8b_emu.S  | 24 ++++++++++++++++++------
 3 files changed, 26 insertions(+), 12 deletions(-)

diff --git a/arch/x86/include/asm/percpu.h b/arch/x86/include/asm/percpu.h
index 34734d730463..83e6a4bcea38 100644
--- a/arch/x86/include/asm/percpu.h
+++ b/arch/x86/include/asm/percpu.h
@@ -11,8 +11,10 @@
 #ifdef __ASSEMBLY__
 
 #ifdef CONFIG_SMP
+#define PER_CPU_ARG(arg)       %__percpu_seg:arg
 #define PER_CPU_VAR(var)       %__percpu_seg:var
 #else /* ! SMP */
+#define PER_CPU_ARG(arg)       arg
 #define PER_CPU_VAR(var)       var
 #endif /* SMP */
 
diff --git a/arch/x86/lib/cmpxchg16b_emu.S b/arch/x86/lib/cmpxchg16b_emu.S
index 6962df315793..b6b942d07a00 100644
--- a/arch/x86/lib/cmpxchg16b_emu.S
+++ b/arch/x86/lib/cmpxchg16b_emu.S
@@ -23,14 +23,14 @@ SYM_FUNC_START(this_cpu_cmpxchg16b_emu)
        cli
 
        /* if (*ptr == old) */
-       cmpq    PER_CPU_VAR(0(%rsi)), %rax
+       cmpq    PER_CPU_ARG(0(%rsi)), %rax
        jne     .Lnot_same
-       cmpq    PER_CPU_VAR(8(%rsi)), %rdx
+       cmpq    PER_CPU_ARG(8(%rsi)), %rdx
        jne     .Lnot_same
 
        /* *ptr = new */
-       movq    %rbx, PER_CPU_VAR(0(%rsi))
-       movq    %rcx, PER_CPU_VAR(8(%rsi))
+       movq    %rbx, PER_CPU_ARG(0(%rsi))
+       movq    %rcx, PER_CPU_ARG(8(%rsi))
 
        /* set ZF in EFLAGS to indicate success */
        orl     $X86_EFLAGS_ZF, (%rsp)
@@ -42,8 +42,8 @@ SYM_FUNC_START(this_cpu_cmpxchg16b_emu)
        /* *ptr != old */
 
        /* old = *ptr */
-       movq    PER_CPU_VAR(0(%rsi)), %rax
-       movq    PER_CPU_VAR(8(%rsi)), %rdx
+       movq    PER_CPU_ARG(0(%rsi)), %rax
+       movq    PER_CPU_ARG(8(%rsi)), %rdx
 
        /* clear ZF in EFLAGS to indicate failure */
        andl    $(~X86_EFLAGS_ZF), (%rsp)
diff --git a/arch/x86/lib/cmpxchg8b_emu.S b/arch/x86/lib/cmpxchg8b_emu.S
index 49805257b125..9a0a7feeaf7c 100644
--- a/arch/x86/lib/cmpxchg8b_emu.S
+++ b/arch/x86/lib/cmpxchg8b_emu.S
@@ -53,18 +53,30 @@ EXPORT_SYMBOL(cmpxchg8b_emu)
 
 #ifndef CONFIG_UML
 
+/*
+ * Emulate 'cmpxchg8b %fs:(%rsi)'
+ *
+ * Inputs:
+ * %esi : memory location to compare
+ * %eax : low 32 bits of old value
+ * %edx : high 32 bits of old value
+ * %ebx : low 32 bits of new value
+ * %ecx : high 32 bits of new value
+ *
+ * Notably this is not LOCK prefixed and is not safe against NMIs
+ */
 SYM_FUNC_START(this_cpu_cmpxchg8b_emu)
 
        pushfl
        cli
 
-       cmpl    PER_CPU_VAR(0(%esi)), %eax
+       cmpl    PER_CPU_ARG(0(%esi)), %eax
        jne     .Lnot_same2
-       cmpl    PER_CPU_VAR(4(%esi)), %edx
+       cmpl    PER_CPU_ARG(4(%esi)), %edx
        jne     .Lnot_same2
 
-       movl    %ebx, PER_CPU_VAR(0(%esi))
-       movl    %ecx, PER_CPU_VAR(4(%esi))
+       movl    %ebx, PER_CPU_ARG(0(%esi))
+       movl    %ecx, PER_CPU_ARG(4(%esi))
 
        orl     $X86_EFLAGS_ZF, (%esp)
 
@@ -72,8 +84,8 @@ SYM_FUNC_START(this_cpu_cmpxchg8b_emu)
        RET
 
 .Lnot_same2:
-       movl    PER_CPU_VAR(0(%esi)), %eax
-       movl    PER_CPU_VAR(4(%esi)), %edx
+       movl    PER_CPU_ARG(0(%esi)), %eax
+       movl    PER_CPU_ARG(4(%esi)), %edx
 
        andl    $(~X86_EFLAGS_ZF), (%esp)
 
-- 
2.41.0