[PATCH] x86: Deinline cpuid_eax and friends

[Denys Vlasenko]

cpuid_e{a,b,c,d}x() functions compile to 44 bytes of machine code each.
On x86 allyesconfig build they have 48 callsites.
Deinlining all four of them shrinks kernel by about 1k:

   text      data      bss       dec     hex filename
82434909 22255384 20627456 125317749 7783275 vmlinux.before
82433898 22255384 20627456 125316738 7782e82 vmlinux

Speed impact: CPUID instruction takes from 50 to 350+ cycles,
call overhead is negligible in comparison.

Signed-off-by: Denys Vlasenko <dvlasenk@xxxxxxxxxx>
CC: Steven Rostedt <rostedt@xxxxxxxxxxx>
CC: Ingo Molnar <mingo@xxxxxxxxxx>
CC: Borislav Petkov <bp@xxxxxxxxx>
CC: "H. Peter Anvin" <hpa@xxxxxxxxx>
CC: Andy Lutomirski <luto@xxxxxxxxxxxxxx>
CC: Frederic Weisbecker <fweisbec@xxxxxxxxx>
CC: Alexei Starovoitov <ast@xxxxxxxxxxxx>
CC: Will Drewry <wad@xxxxxxxxxxxx>
CC: Kees Cook <keescook@xxxxxxxxxxxx>
CC: x86@xxxxxxxxxx
CC: linux-kernel@xxxxxxxxxxxxxxx
---
 arch/x86/include/asm/processor.h | 39 ++++--------------------------------
 arch/x86/kernel/cpu/common.c     | 43 ++++++++++++++++++++++++++++++++++++++++
 2 files changed, 47 insertions(+), 35 deletions(-)

diff --git a/arch/x86/include/asm/processor.h b/arch/x86/include/asm/processor.h
index ec1c935..67e1974 100644
--- a/arch/x86/include/asm/processor.h
+++ b/arch/x86/include/asm/processor.h
@@ -616,41 +616,10 @@ static inline void cpuid_count(unsigned int op, int count,
 /*
  * CPUID functions returning a single datum
  */
-static inline unsigned int cpuid_eax(unsigned int op)
-{
-	unsigned int eax, ebx, ecx, edx;
-
-	cpuid(op, &eax, &ebx, &ecx, &edx);
-
-	return eax;
-}
-
-static inline unsigned int cpuid_ebx(unsigned int op)
-{
-	unsigned int eax, ebx, ecx, edx;
-
-	cpuid(op, &eax, &ebx, &ecx, &edx);
-
-	return ebx;
-}
-
-static inline unsigned int cpuid_ecx(unsigned int op)
-{
-	unsigned int eax, ebx, ecx, edx;
-
-	cpuid(op, &eax, &ebx, &ecx, &edx);
-
-	return ecx;
-}
-
-static inline unsigned int cpuid_edx(unsigned int op)
-{
-	unsigned int eax, ebx, ecx, edx;
-
-	cpuid(op, &eax, &ebx, &ecx, &edx);
-
-	return edx;
-}
+unsigned int cpuid_eax(unsigned int op);
+unsigned int cpuid_ebx(unsigned int op);
+unsigned int cpuid_ecx(unsigned int op);
+unsigned int cpuid_edx(unsigned int op);
 
 /* REP NOP (PAUSE) is a good thing to insert into busy-wait loops. */
 static inline void rep_nop(void)
diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c
index 2346c95..1d2e270 100644
--- a/arch/x86/kernel/cpu/common.c
+++ b/arch/x86/kernel/cpu/common.c
@@ -307,6 +307,49 @@ static __always_inline void setup_smap(struct cpuinfo_x86 *c)
 }
 
 /*
+ * CPUID functions returning a single datum
+ */
+unsigned int cpuid_eax(unsigned int op)
+{
+	unsigned int eax, ebx, ecx, edx;
+
+	cpuid(op, &eax, &ebx, &ecx, &edx);
+
+	return eax;
+}
+EXPORT_SYMBOL(cpuid_eax);
+
+unsigned int cpuid_ebx(unsigned int op)
+{
+	unsigned int eax, ebx, ecx, edx;
+
+	cpuid(op, &eax, &ebx, &ecx, &edx);
+
+	return ebx;
+}
+EXPORT_SYMBOL(cpuid_ebx);
+
+unsigned int cpuid_ecx(unsigned int op)
+{
+	unsigned int eax, ebx, ecx, edx;
+
+	cpuid(op, &eax, &ebx, &ecx, &edx);
+
+	return ecx;
+}
+EXPORT_SYMBOL(cpuid_ecx);
+
+unsigned int cpuid_edx(unsigned int op)
+{
+	unsigned int eax, ebx, ecx, edx;
+
+	cpuid(op, &eax, &ebx, &ecx, &edx);
+
+	return edx;
+}
+EXPORT_SYMBOL(cpuid_edx);
+
+/*
  * Some CPU features depend on higher CPUID levels, which may not always
  * be available due to CPUID level capping or broken virtualization
  * software.  Add those features to this table to auto-disable them.
-- 
1.8.1.4

--
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/