[tip: x86/percpu] x86/percpu: Clean up <asm/percpu.h> a bit

From: tip-bot2 for Ingo Molnar
Date: Tue May 21 2024 - 04:21:44 EST


The following commit has been merged into the x86/percpu branch of tip:

Commit-ID: 61d73e4f7d538f3907d954a531169e8164aef56b
Gitweb: https://git.kernel.org/tip/61d73e4f7d538f3907d954a531169e8164aef56b
Author: Ingo Molnar <mingo@xxxxxxxxxx>
AuthorDate: Mon, 20 May 2024 10:22:39 +02:00
Committer: Ingo Molnar <mingo@xxxxxxxxxx>
CommitterDate: Mon, 20 May 2024 10:56:23 +02:00

x86/percpu: Clean up <asm/percpu.h> a bit

- Fix misc typos

- There's 4 variants of the same spelling right now:

'per-CPU', 'per CPU', 'percpu' and 'per-cpu'

Standardize on 'per-CPU' only.

- s/makes gcc load
/makes the compiler load

- Instead of:

#ifdef CONFIG_XXXX
#define YYYY FOO
#else
#define YYYY BAR
#endif

Use the slightly more readable form of:

#ifdef CONFIG_XXXX
# define YYYY FOO
#else
# define YYYY BAR
#endif

- Standardize & expand '#else' and '#endif' comments

- Fix comment style

- Capitalize x86 instruction names in comments

No change in code.

Signed-off-by: Ingo Molnar <mingo@xxxxxxxxxx>
Cc: Uros Bizjak <ubizjak@xxxxxxxxx>
Cc: Andy Lutomirski <luto@xxxxxxxxxx>
Cc: Josh Poimboeuf <jpoimboe@xxxxxxxxxx>
Cc: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
Cc: linux-kernel@xxxxxxxxxxxxxxx
---
arch/x86/include/asm/percpu.h | 91 ++++++++++++++++++----------------
1 file changed, 50 insertions(+), 41 deletions(-)

diff --git a/arch/x86/include/asm/percpu.h b/arch/x86/include/asm/percpu.h
index 0f0d897..b424cb1 100644
--- a/arch/x86/include/asm/percpu.h
+++ b/arch/x86/include/asm/percpu.h
@@ -3,30 +3,30 @@
#define _ASM_X86_PERCPU_H

#ifdef CONFIG_X86_64
-#define __percpu_seg gs
-#define __percpu_rel (%rip)
+# define __percpu_seg gs
+# define __percpu_rel (%rip)
#else
-#define __percpu_seg fs
-#define __percpu_rel
+# define __percpu_seg fs
+# define __percpu_rel
#endif

#ifdef __ASSEMBLY__

#ifdef CONFIG_SMP
-#define __percpu %__percpu_seg:
+# define __percpu %__percpu_seg:
#else
-#define __percpu
+# define __percpu
#endif

#define PER_CPU_VAR(var) __percpu(var)__percpu_rel

#ifdef CONFIG_X86_64_SMP
-#define INIT_PER_CPU_VAR(var) init_per_cpu__##var
+# define INIT_PER_CPU_VAR(var) init_per_cpu__##var
#else
-#define INIT_PER_CPU_VAR(var) var
+# define INIT_PER_CPU_VAR(var) var
#endif

-#else /* ...!ASSEMBLY */
+#else /* !__ASSEMBLY__: */

#include <linux/build_bug.h>
#include <linux/stringify.h>
@@ -37,19 +37,19 @@
#ifdef CONFIG_CC_HAS_NAMED_AS

#ifdef __CHECKER__
-#define __seg_gs __attribute__((address_space(__seg_gs)))
-#define __seg_fs __attribute__((address_space(__seg_fs)))
+# define __seg_gs __attribute__((address_space(__seg_gs)))
+# define __seg_fs __attribute__((address_space(__seg_fs)))
#endif

#ifdef CONFIG_X86_64
-#define __percpu_seg_override __seg_gs
+# define __percpu_seg_override __seg_gs
#else
-#define __percpu_seg_override __seg_fs
+# define __percpu_seg_override __seg_fs
#endif

#define __percpu_prefix ""

-#else /* CONFIG_CC_HAS_NAMED_AS */
+#else /* !CONFIG_CC_HAS_NAMED_AS: */

#define __percpu_seg_override
#define __percpu_prefix "%%"__stringify(__percpu_seg)":"
@@ -80,7 +80,8 @@

#define PER_CPU_VAR(var) %__percpu_seg:(var)__percpu_rel

-#else /* CONFIG_SMP */
+#else /* !CONFIG_SMP: */
+
#define __percpu_seg_override
#define __percpu_prefix ""
#define __force_percpu_prefix ""
@@ -96,7 +97,7 @@
#define __force_percpu_arg(x) __force_percpu_prefix "%" #x

/*
- * Initialized pointers to per-cpu variables needed for the boot
+ * Initialized pointers to per-CPU variables needed for the boot
* processor need to use these macros to get the proper address
* offset from __per_cpu_load on SMP.
*
@@ -106,13 +107,15 @@
extern typeof(var) init_per_cpu_var(var)

#ifdef CONFIG_X86_64_SMP
-#define init_per_cpu_var(var) init_per_cpu__##var
+# define init_per_cpu_var(var) init_per_cpu__##var
#else
-#define init_per_cpu_var(var) var
+# define init_per_cpu_var(var) var
#endif

-/* For arch-specific code, we can use direct single-insn ops (they
- * don't give an lvalue though). */
+/*
+ * For arch-specific code, we can use direct single-insn ops (they
+ * don't give an lvalue though).
+ */

#define __pcpu_type_1 u8
#define __pcpu_type_2 u16
@@ -158,7 +161,7 @@ do { \

#define __raw_cpu_read_const(pcp) __raw_cpu_read(, , pcp)

-#else /* CONFIG_USE_X86_SEG_SUPPORT */
+#else /* !CONFIG_USE_X86_SEG_SUPPORT: */

#define __raw_cpu_read(size, qual, _var) \
({ \
@@ -183,7 +186,7 @@ do { \
} while (0)

/*
- * The generic per-cpu infrastrucutre is not suitable for
+ * The generic per-CPU infrastrucutre is not suitable for
* reading const-qualified variables.
*/
#define __raw_cpu_read_const(pcp) ({ BUILD_BUG(); (typeof(pcp))0; })
@@ -219,7 +222,7 @@ do { \
} while (0)

/*
- * Generate a percpu add to memory instruction and optimize code
+ * Generate a per-CPU add to memory instruction and optimize code
* if one is added or subtracted.
*/
#define percpu_add_op(size, qual, var, val) \
@@ -266,9 +269,9 @@ do { \
})

/*
- * this_cpu_xchg() is implemented using cmpxchg without a lock prefix.
- * xchg is expensive due to the implied lock prefix. The processor
- * cannot prefetch cachelines if xchg is used.
+ * this_cpu_xchg() is implemented using CMPXCHG without a LOCK prefix.
+ * XCHG is expensive due to the implied LOCK prefix. The processor
+ * cannot prefetch cachelines if XCHG is used.
*/
#define this_percpu_xchg_op(_var, _nval) \
({ \
@@ -278,8 +281,8 @@ do { \
})

/*
- * cmpxchg has no such implied lock semantics as a result it is much
- * more efficient for cpu local operations.
+ * CMPXCHG has no such implied lock semantics as a result it is much
+ * more efficient for CPU-local operations.
*/
#define percpu_cmpxchg_op(size, qual, _var, _oval, _nval) \
({ \
@@ -314,6 +317,7 @@ do { \
})

#if defined(CONFIG_X86_32) && !defined(CONFIG_UML)
+
#define percpu_cmpxchg64_op(size, qual, _var, _oval, _nval) \
({ \
union { \
@@ -374,7 +378,8 @@ do { \

#define raw_cpu_try_cmpxchg64(pcp, ovalp, nval) percpu_try_cmpxchg64_op(8, , pcp, ovalp, nval)
#define this_cpu_try_cmpxchg64(pcp, ovalp, nval) percpu_try_cmpxchg64_op(8, volatile, pcp, ovalp, nval)
-#endif
+
+#endif /* defined(CONFIG_X86_32) && !defined(CONFIG_UML) */

#ifdef CONFIG_X86_64
#define raw_cpu_cmpxchg64(pcp, oval, nval) percpu_cmpxchg_op(8, , pcp, oval, nval);
@@ -443,7 +448,8 @@ do { \

#define raw_cpu_try_cmpxchg128(pcp, ovalp, nval) percpu_try_cmpxchg128_op(16, , pcp, ovalp, nval)
#define this_cpu_try_cmpxchg128(pcp, ovalp, nval) percpu_try_cmpxchg128_op(16, volatile, pcp, ovalp, nval)
-#endif
+
+#endif /* CONFIG_X86_64 */

#define raw_cpu_read_1(pcp) __raw_cpu_read(1, , pcp)
#define raw_cpu_read_2(pcp) __raw_cpu_read(2, , pcp)
@@ -510,8 +516,8 @@ do { \
#define this_cpu_try_cmpxchg_4(pcp, ovalp, nval) percpu_try_cmpxchg_op(4, volatile, pcp, ovalp, nval)

/*
- * Per cpu atomic 64 bit operations are only available under 64 bit.
- * 32 bit must fall back to generic operations.
+ * Per-CPU atomic 64-bit operations are only available under 64-bit kernels.
+ * 32-bit kernels must fall back to generic operations.
*/
#ifdef CONFIG_X86_64
#define raw_cpu_read_8(pcp) __raw_cpu_read(8, , pcp)
@@ -539,20 +545,23 @@ do { \
#define this_cpu_try_cmpxchg_8(pcp, ovalp, nval) percpu_try_cmpxchg_op(8, volatile, pcp, ovalp, nval)

#define raw_cpu_read_long(pcp) raw_cpu_read_8(pcp)
-#else
-/* There is no generic 64 bit read stable operation for 32 bit targets. */
+
+#else /* !CONFIG_X86_64: */
+
+/* There is no generic 64-bit read stable operation for 32-bit targets. */
#define this_cpu_read_stable_8(pcp) ({ BUILD_BUG(); (typeof(pcp))0; })

#define raw_cpu_read_long(pcp) raw_cpu_read_4(pcp)
-#endif
+
+#endif /* CONFIG_X86_64 */

#define this_cpu_read_const(pcp) __raw_cpu_read_const(pcp)

/*
- * this_cpu_read() makes gcc load the percpu variable every time it is
- * accessed while this_cpu_read_stable() allows the value to be cached.
+ * this_cpu_read() makes the compiler load the per-CPU variable every time
+ * it is accessed while this_cpu_read_stable() allows the value to be cached.
* this_cpu_read_stable() is more efficient and can be used if its value
- * is guaranteed to be valid across cpus. The current users include
+ * is guaranteed to be valid across CPUs. The current users include
* pcpu_hot.current_task and pcpu_hot.top_of_stack, both of which are
* actually per-thread variables implemented as per-CPU variables and
* thus stable for the duration of the respective task.
@@ -626,12 +635,12 @@ DECLARE_PER_CPU_READ_MOSTLY(unsigned long, this_cpu_off);

#define early_per_cpu_ptr(_name) (_name##_early_ptr)
#define early_per_cpu_map(_name, _idx) (_name##_early_map[_idx])
-#define early_per_cpu(_name, _cpu) \
+#define early_per_cpu(_name, _cpu) \
*(early_per_cpu_ptr(_name) ? \
&early_per_cpu_ptr(_name)[_cpu] : \
&per_cpu(_name, _cpu))

-#else /* !CONFIG_SMP */
+#else /* !CONFIG_SMP: */
#define DEFINE_EARLY_PER_CPU(_type, _name, _initvalue) \
DEFINE_PER_CPU(_type, _name) = _initvalue

@@ -651,6 +660,6 @@ DECLARE_PER_CPU_READ_MOSTLY(unsigned long, this_cpu_off);
#define early_per_cpu_ptr(_name) NULL
/* no early_per_cpu_map() */

-#endif /* !CONFIG_SMP */
+#endif /* CONFIG_SMP */

#endif /* _ASM_X86_PERCPU_H */