objtool: Shrink instruction::{type,visited}

Since we don't have that many types in enum insn_type, force it into a
u8 and re-arrange member to get rid of the holes, saves another 8
bytes.

 struct instruction {
 	struct list_head           list;                 /*     0    16 */
 	struct hlist_node          hash;                 /*    16    16 */
 	struct list_head           call_node;            /*    32    16 */
 	struct section *           sec;                  /*    48     8 */
 	long unsigned int          offset;               /*    56     8 */
 	/* --- cacheline 1 boundary (64 bytes) --- */
-	unsigned int               len;                  /*    64     4 */
-	enum insn_type             type;                 /*    68     4 */
-	long unsigned int          immediate;            /*    72     8 */
-	u16                        dead_end:1;           /*    80: 0  2 */
-	u16                        ignore:1;             /*    80: 1  2 */
-	u16                        ignore_alts:1;        /*    80: 2  2 */
-	u16                        hint:1;               /*    80: 3  2 */
-	u16                        save:1;               /*    80: 4  2 */
-	u16                        restore:1;            /*    80: 5  2 */
-	u16                        retpoline_safe:1;     /*    80: 6  2 */
-	u16                        noendbr:1;            /*    80: 7  2 */
-	u16                        entry:1;              /*    80: 8  2 */
+	long unsigned int          immediate;            /*    64     8 */
+	unsigned int               len;                  /*    72     4 */
+	u8                         type;                 /*    76     1 */

-	/* XXX 7 bits hole, try to pack */
+	/* Bitfield combined with previous fields */

-	s8                         instr;                /*    82     1 */
-	u8                         visited;              /*    83     1 */
+	u16                        dead_end:1;           /*    76: 8  2 */
+	u16                        ignore:1;             /*    76: 9  2 */
+	u16                        ignore_alts:1;        /*    76:10  2 */
+	u16                        hint:1;               /*    76:11  2 */
+	u16                        save:1;               /*    76:12  2 */
+	u16                        restore:1;            /*    76:13  2 */
+	u16                        retpoline_safe:1;     /*    76:14  2 */
+	u16                        noendbr:1;            /*    76:15  2 */
+	u16                        entry:1;              /*    78: 0  2 */
+	u16                        visited:4;            /*    78: 1  2 */

-	/* XXX 4 bytes hole, try to pack */
+	/* XXX 3 bits hole, try to pack */
+	/* Bitfield combined with next fields */

-	struct alt_group *         alt_group;            /*    88     8 */
-	struct symbol *            call_dest;            /*    96     8 */
-	struct instruction *       jump_dest;            /*   104     8 */
-	struct instruction *       first_jump_src;       /*   112     8 */
-	struct reloc *             jump_table;           /*   120     8 */
+	s8                         instr;                /*    79     1 */
+	struct alt_group *         alt_group;            /*    80     8 */
+	struct symbol *            call_dest;            /*    88     8 */
+	struct instruction *       jump_dest;            /*    96     8 */
+	struct instruction *       first_jump_src;       /*   104     8 */
+	struct reloc *             jump_table;           /*   112     8 */
+	struct reloc *             reloc;                /*   120     8 */
 	/* --- cacheline 2 boundary (128 bytes) --- */
-	struct reloc *             reloc;                /*   128     8 */
-	struct alternative *       alts;                 /*   136     8 */
-	struct symbol *            sym;                  /*   144     8 */
-	struct stack_op *          stack_ops;            /*   152     8 */
-	struct cfi_state *         cfi;                  /*   160     8 */
+	struct alternative *       alts;                 /*   128     8 */
+	struct symbol *            sym;                  /*   136     8 */
+	struct stack_op *          stack_ops;            /*   144     8 */
+	struct cfi_state *         cfi;                  /*   152     8 */

-	/* size: 168, cachelines: 3, members: 29 */
-	/* sum members: 162, holes: 1, sum holes: 4 */
-	/* sum bitfield members: 9 bits, bit holes: 1, sum bit holes: 7 bits */
-	/* last cacheline: 40 bytes */
+	/* size: 160, cachelines: 3, members: 29 */
+	/* sum members: 158 */
+	/* sum bitfield members: 13 bits, bit holes: 1, sum bit holes: 3 bits */
+	/* last cacheline: 32 bytes */
 };

pre:	5:48.86 real,   220.30 user,    128.34 sys,     24834672 mem
post:	5:48.89 real,   220.96 user,    127.55 sys,     24834672 mem

Signed-off-by: Peter Zijlstra (Intel) <[email protected]>
Signed-off-by: Ingo Molnar <[email protected]>
Acked-by: Josh Poimboeuf <[email protected]>
Tested-by: Nathan Chancellor <[email protected]> # build only
Tested-by: Thomas Weißschuh <[email protected]> # compile and run
Link: https://lore.kernel.org/r/[email protected]

Author: Peter Zijlstra

tools/objtool/include/objtool/check.h

@@ -42,9 +42,9 @@ struct instruction {
 	struct list_head call_node;
 	struct section *sec;
 	unsigned long offset;
-	unsigned int len;
-	enum insn_type type;
 	unsigned long immediate;
+	unsigned int len;
+	u8 type;
 
 	u16 dead_end		: 1,
 	   ignore		: 1,
@@ -54,11 +54,11 @@ struct instruction {
 	   restore		: 1,
 	   retpoline_safe	: 1,
 	   noendbr		: 1,
-	   entry		: 1;
-		/* 7 bit hole */
+	   entry		: 1,
+	   visited		: 4;
+		/* 3 bit hole */
 
 	s8 instr;
-	u8 visited;
 
 	struct alt_group *alt_group;
 	struct symbol *call_dest;