Merge tag 'perf-core-for-mingo-4.12-20170424' of...

Merge tag 'perf-core-for-mingo-4.12-20170424' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/core

Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo:

User visible changes:

 - Fix display of data source snoop indication in 'perf mem' (Andi Kleen)

 - Fix the code to strip command name from /proc/PID/stat (Jiri Olsa)

Infrastructure changes:

 - Continue the disentanglement of headers, specially util.h (Arnaldo Carvalho de Melo)

 - Synchronize some header files with the kernel (Arnaldo Carvalho de Melo)
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>

arch/um/include/shared/os.h

@@ -302,8 +302,8 @@ extern int ignore_sigio_fd(int fd);
 extern void maybe_sigio_broken(int fd, int read);
 extern void sigio_broken(int fd, int read);
 
-/* sys-x86_64/prctl.c */
-extern int os_arch_prctl(int pid, int code, unsigned long *addr);
+/* prctl.c */
+extern int os_arch_prctl(int pid, int option, unsigned long *arg2);
 
 /* tty.c */
 extern int get_pty(void);

arch/x86/entry/syscalls/syscall_32.tbl

@@ -390,3 +390,4 @@
 381	i386	pkey_alloc		sys_pkey_alloc
 382	i386	pkey_free		sys_pkey_free
 383	i386	statx			sys_statx
+384	i386	arch_prctl		sys_arch_prctl			compat_sys_arch_prctl

arch/x86/include/asm/cpufeatures.h

@@ -187,6 +187,7 @@
  * Reuse free bits when adding new feature flags!
  */
 #define X86_FEATURE_RING3MWAIT	( 7*32+ 0) /* Ring 3 MONITOR/MWAIT */
+#define X86_FEATURE_CPUID_FAULT ( 7*32+ 1) /* Intel CPUID faulting */
 #define X86_FEATURE_CPB		( 7*32+ 2) /* AMD Core Performance Boost */
 #define X86_FEATURE_EPB		( 7*32+ 3) /* IA32_ENERGY_PERF_BIAS support */
 #define X86_FEATURE_CAT_L3	( 7*32+ 4) /* Cache Allocation Technology L3 */

arch/x86/include/asm/msr-index.h

@@ -45,6 +45,8 @@
 #define MSR_IA32_PERFCTR1		0x000000c2
 #define MSR_FSB_FREQ			0x000000cd
 #define MSR_PLATFORM_INFO		0x000000ce
+#define MSR_PLATFORM_INFO_CPUID_FAULT_BIT	31
+#define MSR_PLATFORM_INFO_CPUID_FAULT		BIT_ULL(MSR_PLATFORM_INFO_CPUID_FAULT_BIT)
 
 #define MSR_PKG_CST_CONFIG_CONTROL	0x000000e2
 #define NHM_C3_AUTO_DEMOTE		(1UL << 25)
@@ -127,6 +129,7 @@
 
 /* DEBUGCTLMSR bits (others vary by model): */
 #define DEBUGCTLMSR_LBR			(1UL <<  0) /* last branch recording */
+#define DEBUGCTLMSR_BTF_SHIFT		1
 #define DEBUGCTLMSR_BTF			(1UL <<  1) /* single-step on branches */
 #define DEBUGCTLMSR_TR			(1UL <<  6)
 #define DEBUGCTLMSR_BTS			(1UL <<  7)
@@ -552,10 +555,12 @@
 #define MSR_IA32_MISC_ENABLE_IP_PREF_DISABLE_BIT	39
 #define MSR_IA32_MISC_ENABLE_IP_PREF_DISABLE		(1ULL << MSR_IA32_MISC_ENABLE_IP_PREF_DISABLE_BIT)
 
-/* MISC_FEATURE_ENABLES non-architectural features */
-#define MSR_MISC_FEATURE_ENABLES	0x00000140
+/* MISC_FEATURES_ENABLES non-architectural features */
+#define MSR_MISC_FEATURES_ENABLES	0x00000140
 
-#define MSR_MISC_FEATURE_ENABLES_RING3MWAIT_BIT		1
+#define MSR_MISC_FEATURES_ENABLES_CPUID_FAULT_BIT	0
+#define MSR_MISC_FEATURES_ENABLES_CPUID_FAULT		BIT_ULL(MSR_MISC_FEATURES_ENABLES_CPUID_FAULT_BIT)
+#define MSR_MISC_FEATURES_ENABLES_RING3MWAIT_BIT	1
 
 #define MSR_IA32_TSC_DEADLINE		0x000006E0
 

arch/x86/include/asm/processor.h

@@ -884,6 +884,8 @@ extern void start_thread(struct pt_regs *regs, unsigned long new_ip,
 extern int get_tsc_mode(unsigned long adr);
 extern int set_tsc_mode(unsigned int val);
 
+DECLARE_PER_CPU(u64, msr_misc_features_shadow);
+
 /* Register/unregister a process' MPX related resource */
 #define MPX_ENABLE_MANAGEMENT()	mpx_enable_management()
 #define MPX_DISABLE_MANAGEMENT()	mpx_disable_management()

arch/x86/include/asm/proto.h

@@ -9,6 +9,7 @@ void syscall_init(void);
 
 #ifdef CONFIG_X86_64
 void entry_SYSCALL_64(void);
+long do_arch_prctl_64(struct task_struct *task, int option, unsigned long arg2);
 #endif
 
 #ifdef CONFIG_X86_32
@@ -30,6 +31,7 @@ void x86_report_nx(void);
 
 extern int reboot_force;
 
-long do_arch_prctl(struct task_struct *task, int code, unsigned long addr);
+long do_arch_prctl_common(struct task_struct *task, int option,
+			  unsigned long cpuid_enabled);
 
 #endif /* _ASM_X86_PROTO_H */

arch/x86/include/asm/thread_info.h

@@ -87,6 +87,7 @@ struct thread_info {
 #define TIF_SECCOMP		8	/* secure computing */
 #define TIF_USER_RETURN_NOTIFY	11	/* notify kernel of userspace return */
 #define TIF_UPROBE		12	/* breakpointed or singlestepping */
+#define TIF_NOCPUID		15	/* CPUID is not accessible in userland */
 #define TIF_NOTSC		16	/* TSC is not accessible in userland */
 #define TIF_IA32		17	/* IA32 compatibility process */
 #define TIF_NOHZ		19	/* in adaptive nohz mode */
@@ -110,6 +111,7 @@ struct thread_info {
 #define _TIF_SECCOMP		(1 << TIF_SECCOMP)
 #define _TIF_USER_RETURN_NOTIFY	(1 << TIF_USER_RETURN_NOTIFY)
 #define _TIF_UPROBE		(1 << TIF_UPROBE)
+#define _TIF_NOCPUID		(1 << TIF_NOCPUID)
 #define _TIF_NOTSC		(1 << TIF_NOTSC)
 #define _TIF_IA32		(1 << TIF_IA32)
 #define _TIF_NOHZ		(1 << TIF_NOHZ)
@@ -138,7 +140,7 @@ struct thread_info {
 
 /* flags to check in __switch_to() */
 #define _TIF_WORK_CTXSW							\
-	(_TIF_IO_BITMAP|_TIF_NOTSC|_TIF_BLOCKSTEP)
+	(_TIF_IO_BITMAP|_TIF_NOCPUID|_TIF_NOTSC|_TIF_BLOCKSTEP)
 
 #define _TIF_WORK_CTXSW_PREV (_TIF_WORK_CTXSW|_TIF_USER_RETURN_NOTIFY)
 #define _TIF_WORK_CTXSW_NEXT (_TIF_WORK_CTXSW)
@@ -239,6 +241,8 @@ static inline int arch_within_stack_frames(const void * const stack,
 extern void arch_task_cache_init(void);
 extern int arch_dup_task_struct(struct task_struct *dst, struct task_struct *src);
 extern void arch_release_task_struct(struct task_struct *tsk);
+extern void arch_setup_new_exec(void);
+#define arch_setup_new_exec arch_setup_new_exec
 #endif	/* !__ASSEMBLY__ */
 
 #endif /* _ASM_X86_THREAD_INFO_H */

arch/x86/include/asm/tlbflush.h

@@ -110,6 +110,16 @@ static inline void cr4_clear_bits(unsigned long mask)
 	}
 }
 
+static inline void cr4_toggle_bits(unsigned long mask)
+{
+	unsigned long cr4;
+
+	cr4 = this_cpu_read(cpu_tlbstate.cr4);
+	cr4 ^= mask;
+	this_cpu_write(cpu_tlbstate.cr4, cr4);
+	__write_cr4(cr4);
+}
+
 /* Read the CR4 shadow. */
 static inline unsigned long cr4_read_shadow(void)
 {

arch/x86/include/uapi/asm/prctl.h

 #ifndef _ASM_X86_PRCTL_H
 #define _ASM_X86_PRCTL_H
 
-#define ARCH_SET_GS 0x1001
-#define ARCH_SET_FS 0x1002
-#define ARCH_GET_FS 0x1003
-#define ARCH_GET_GS 0x1004
+#define ARCH_SET_GS		0x1001
+#define ARCH_SET_FS		0x1002
+#define ARCH_GET_FS		0x1003
+#define ARCH_GET_GS		0x1004
+
+#define ARCH_GET_CPUID		0x1011
+#define ARCH_SET_CPUID		0x1012
 
 #define ARCH_MAP_VDSO_X32	0x2001
 #define ARCH_MAP_VDSO_32	0x2002

arch/x86/kernel/cpu/intel.c

@@ -90,16 +90,12 @@ static void probe_xeon_phi_r3mwait(struct cpuinfo_x86 *c)
 		return;
 	}
 
-	if (ring3mwait_disabled) {
-		msr_clear_bit(MSR_MISC_FEATURE_ENABLES,
-			      MSR_MISC_FEATURE_ENABLES_RING3MWAIT_BIT);
+	if (ring3mwait_disabled)
 		return;
-	}
-
-	msr_set_bit(MSR_MISC_FEATURE_ENABLES,
-		    MSR_MISC_FEATURE_ENABLES_RING3MWAIT_BIT);
 
 	set_cpu_cap(c, X86_FEATURE_RING3MWAIT);
+	this_cpu_or(msr_misc_features_shadow,
+		    1UL << MSR_MISC_FEATURES_ENABLES_RING3MWAIT_BIT);
 
 	if (c == &boot_cpu_data)
 		ELF_HWCAP2 |= HWCAP2_RING3MWAIT;
@@ -488,6 +484,34 @@ static void intel_bsp_resume(struct cpuinfo_x86 *c)
 	init_intel_energy_perf(c);
 }
 
+static void init_cpuid_fault(struct cpuinfo_x86 *c)
+{
+	u64 msr;
+
+	if (!rdmsrl_safe(MSR_PLATFORM_INFO, &msr)) {
+		if (msr & MSR_PLATFORM_INFO_CPUID_FAULT)
+			set_cpu_cap(c, X86_FEATURE_CPUID_FAULT);
+	}
+}
+
+static void init_intel_misc_features(struct cpuinfo_x86 *c)
+{
+	u64 msr;
+
+	if (rdmsrl_safe(MSR_MISC_FEATURES_ENABLES, &msr))
+		return;
+
+	/* Clear all MISC features */
+	this_cpu_write(msr_misc_features_shadow, 0);
+
+	/* Check features and update capabilities and shadow control bits */
+	init_cpuid_fault(c);
+	probe_xeon_phi_r3mwait(c);
+
+	msr = this_cpu_read(msr_misc_features_shadow);
+	wrmsrl(MSR_MISC_FEATURES_ENABLES, msr);
+}
+
 static void init_intel(struct cpuinfo_x86 *c)
 {
 	unsigned int l2 = 0;
@@ -602,7 +626,7 @@ static void init_intel(struct cpuinfo_x86 *c)
 
 	init_intel_energy_perf(c);
 
-	probe_xeon_phi_r3mwait(c);
+	init_intel_misc_features(c);
 }
 
 #ifdef CONFIG_X86_32

arch/x86/kernel/process.c

@@ -37,6 +37,7 @@
 #include <asm/vm86.h>
 #include <asm/switch_to.h>
 #include <asm/desc.h>
+#include <asm/prctl.h>
 
 /*
  * per-CPU TSS segments. Threads are completely 'soft' on Linux,
@@ -124,11 +125,6 @@ void flush_thread(void)
 	fpu__clear(&tsk->thread.fpu);
 }
 
-static void hard_disable_TSC(void)
-{
-	cr4_set_bits(X86_CR4_TSD);
-}
-
 void disable_TSC(void)
 {
 	preempt_disable();
@@ -137,15 +133,10 @@ void disable_TSC(void)
 		 * Must flip the CPU state synchronously with
 		 * TIF_NOTSC in the current running context.
 		 */
-		hard_disable_TSC();
+		cr4_set_bits(X86_CR4_TSD);
 	preempt_enable();
 }
 
-static void hard_enable_TSC(void)
-{
-	cr4_clear_bits(X86_CR4_TSD);
-}
-
 static void enable_TSC(void)
 {
 	preempt_disable();
@@ -154,7 +145,7 @@ static void enable_TSC(void)
 		 * Must flip the CPU state synchronously with
 		 * TIF_NOTSC in the current running context.
 		 */
-		hard_enable_TSC();
+		cr4_clear_bits(X86_CR4_TSD);
 	preempt_enable();
 }
 
@@ -182,54 +173,129 @@ int set_tsc_mode(unsigned int val)
 	return 0;
 }
 
-void __switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p,
-		      struct tss_struct *tss)
-{
-	struct thread_struct *prev, *next;
-
-	prev = &prev_p->thread;
-	next = &next_p->thread;
+DEFINE_PER_CPU(u64, msr_misc_features_shadow);
 
-	if (test_tsk_thread_flag(prev_p, TIF_BLOCKSTEP) ^
-	    test_tsk_thread_flag(next_p, TIF_BLOCKSTEP)) {
-		unsigned long debugctl = get_debugctlmsr();
+static void set_cpuid_faulting(bool on)
+{
+	u64 msrval;
 
-		debugctl &= ~DEBUGCTLMSR_BTF;
-		if (test_tsk_thread_flag(next_p, TIF_BLOCKSTEP))
-			debugctl |= DEBUGCTLMSR_BTF;
+	msrval = this_cpu_read(msr_misc_features_shadow);
+	msrval &= ~MSR_MISC_FEATURES_ENABLES_CPUID_FAULT;
+	msrval |= (on << MSR_MISC_FEATURES_ENABLES_CPUID_FAULT_BIT);
+	this_cpu_write(msr_misc_features_shadow, msrval);
+	wrmsrl(MSR_MISC_FEATURES_ENABLES, msrval);
+}
 
-		update_debugctlmsr(debugctl);
+static void disable_cpuid(void)
+{
+	preempt_disable();
+	if (!test_and_set_thread_flag(TIF_NOCPUID)) {
+		/*
+		 * Must flip the CPU state synchronously with
+		 * TIF_NOCPUID in the current running context.
+		 */
+		set_cpuid_faulting(true);
 	}
+	preempt_enable();
+}
 
-	if (test_tsk_thread_flag(prev_p, TIF_NOTSC) ^
-	    test_tsk_thread_flag(next_p, TIF_NOTSC)) {
-		/* prev and next are different */
-		if (test_tsk_thread_flag(next_p, TIF_NOTSC))
-			hard_disable_TSC();
-		else
-			hard_enable_TSC();
+static void enable_cpuid(void)
+{
+	preempt_disable();
+	if (test_and_clear_thread_flag(TIF_NOCPUID)) {
+		/*
+		 * Must flip the CPU state synchronously with
+		 * TIF_NOCPUID in the current running context.
+		 */
+		set_cpuid_faulting(false);
 	}
+	preempt_enable();
+}
+
+static int get_cpuid_mode(void)
+{
+	return !test_thread_flag(TIF_NOCPUID);
+}
+
+static int set_cpuid_mode(struct task_struct *task, unsigned long cpuid_enabled)
+{
+	if (!static_cpu_has(X86_FEATURE_CPUID_FAULT))
+		return -ENODEV;
+
+	if (cpuid_enabled)
+		enable_cpuid();
+	else
+		disable_cpuid();
+
+	return 0;
+}
+
+/*
+ * Called immediately after a successful exec.
+ */
+void arch_setup_new_exec(void)
+{
+	/* If cpuid was previously disabled for this task, re-enable it. */
+	if (test_thread_flag(TIF_NOCPUID))
+		enable_cpuid();
+}
 
-	if (test_tsk_thread_flag(next_p, TIF_IO_BITMAP)) {
+static inline void switch_to_bitmap(struct tss_struct *tss,
+				    struct thread_struct *prev,
+				    struct thread_struct *next,
+				    unsigned long tifp, unsigned long tifn)
+{
+	if (tifn & _TIF_IO_BITMAP) {
 		/*
 		 * Copy the relevant range of the IO bitmap.
 		 * Normally this is 128 bytes or less:
 		 */
 		memcpy(tss->io_bitmap, next->io_bitmap_ptr,
 		       max(prev->io_bitmap_max, next->io_bitmap_max));
-
 		/*
 		 * Make sure that the TSS limit is correct for the CPU
 		 * to notice the IO bitmap.
 		 */
 		refresh_tss_limit();
-	} else if (test_tsk_thread_flag(prev_p, TIF_IO_BITMAP)) {
+	} else if (tifp & _TIF_IO_BITMAP) {
 		/*
 		 * Clear any possible leftover bits:
 		 */
 		memset(tss->io_bitmap, 0xff, prev->io_bitmap_max);
 	}
+}
+
+void __switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p,
+		      struct tss_struct *tss)
+{
+	struct thread_struct *prev, *next;
+	unsigned long tifp, tifn;
+
+	prev = &prev_p->thread;
+	next = &next_p->thread;
+
+	tifn = READ_ONCE(task_thread_info(next_p)->flags);
+	tifp = READ_ONCE(task_thread_info(prev_p)->flags);
+	switch_to_bitmap(tss, prev, next, tifp, tifn);
+
 	propagate_user_return_notify(prev_p, next_p);
+
+	if ((tifp & _TIF_BLOCKSTEP || tifn & _TIF_BLOCKSTEP) &&
+	    arch_has_block_step()) {
+		unsigned long debugctl, msk;
+
+		rdmsrl(MSR_IA32_DEBUGCTLMSR, debugctl);
+		debugctl &= ~DEBUGCTLMSR_BTF;
+		msk = tifn & _TIF_BLOCKSTEP;
+		debugctl |= (msk >> TIF_BLOCKSTEP) << DEBUGCTLMSR_BTF_SHIFT;
+		wrmsrl(MSR_IA32_DEBUGCTLMSR, debugctl);
+	}
+
+	if ((tifp ^ tifn) & _TIF_NOTSC)
+		cr4_toggle_bits(X86_CR4_TSD);
+
+	if ((tifp ^ tifn) & _TIF_NOCPUID)
+		set_cpuid_faulting(!!(tifn & _TIF_NOCPUID));
 }
 
 /*
@@ -550,3 +616,16 @@ unsigned long get_wchan(struct task_struct *p)
 	put_task_stack(p);
 	return ret;
 }
+
+long do_arch_prctl_common(struct task_struct *task, int option,
+			  unsigned long cpuid_enabled)
+{
+	switch (option) {
+	case ARCH_GET_CPUID:
+		return get_cpuid_mode();
+	case ARCH_SET_CPUID:
+		return set_cpuid_mode(task, cpuid_enabled);
+	}
+
+	return -EINVAL;
+}

arch/x86/kernel/process_32.c

@@ -37,6 +37,7 @@
 #include <linux/uaccess.h>
 #include <linux/io.h>
 #include <linux/kdebug.h>
+#include <linux/syscalls.h>
 
 #include <asm/pgtable.h>
 #include <asm/ldt.h>
@@ -56,6 +57,7 @@
 #include <asm/switch_to.h>
 #include <asm/vm86.h>
 #include <asm/intel_rdt.h>
+#include <asm/proto.h>
 
 void __show_regs(struct pt_regs *regs, int all)
 {
@@ -304,3 +306,8 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p)
 
 	return prev_p;
 }
+
+SYSCALL_DEFINE2(arch_prctl, int, option, unsigned long, arg2)
+{
+	return do_arch_prctl_common(current, option, arg2);
+}

arch/x86/kernel/process_64.c

@@ -37,6 +37,7 @@
 #include <linux/uaccess.h>
 #include <linux/io.h>
 #include <linux/ftrace.h>
+#include <linux/syscalls.h>
 
 #include <asm/pgtable.h>
 #include <asm/processor.h>
@@ -204,7 +205,7 @@ int copy_thread_tls(unsigned long clone_flags, unsigned long sp,
 				(struct user_desc __user *)tls, 0);
 		else
 #endif
-			err = do_arch_prctl(p, ARCH_SET_FS, tls);
+			err = do_arch_prctl_64(p, ARCH_SET_FS, tls);
 		if (err)
 			goto out;
 	}
@@ -547,70 +548,72 @@ static long prctl_map_vdso(const struct vdso_image *image, unsigned long addr)
 }
 #endif
 
-long do_arch_prctl(struct task_struct *task, int code, unsigned long addr)
+long do_arch_prctl_64(struct task_struct *task, int option, unsigned long arg2)
 {
 	int ret = 0;
 	int doit = task == current;
 	int cpu;
 
-	switch (code) {
+	switch (option) {
 	case ARCH_SET_GS:
-		if (addr >= TASK_SIZE_MAX)
+		if (arg2 >= TASK_SIZE_MAX)
 			return -EPERM;
 		cpu = get_cpu();
 		task->thread.gsindex = 0;
-		task->thread.gsbase = addr;
+		task->thread.gsbase = arg2;
 		if (doit) {
 			load_gs_index(0);
-			ret = wrmsrl_safe(MSR_KERNEL_GS_BASE, addr);
+			ret = wrmsrl_safe(MSR_KERNEL_GS_BASE, arg2);
 		}
 		put_cpu();
 		break;
 	case ARCH_SET_FS:
 		/* Not strictly needed for fs, but do it for symmetry
 		   with gs */
-		if (addr >= TASK_SIZE_MAX)
+		if (arg2 >= TASK_SIZE_MAX)
 			return -EPERM;
 		cpu = get_cpu();
 		task->thread.fsindex = 0;
-		task->thread.fsbase = addr;
+		task->thread.fsbase = arg2;
 		if (doit) {
 			/* set the selector to 0 to not confuse __switch_to */
 			loadsegment(fs, 0);
-			ret = wrmsrl_safe(MSR_FS_BASE, addr);
+			ret = wrmsrl_safe(MSR_FS_BASE, arg2);
 		}
 		put_cpu();
 		break;
 	case ARCH_GET_FS: {
 		unsigned long base;
+
 		if (doit)
 			rdmsrl(MSR_FS_BASE, base);
 		else
 			base = task->thread.fsbase;
-		ret = put_user(base, (unsigned long __user *)addr);
+		ret = put_user(base, (unsigned long __user *)arg2);
 		break;
 	}
 	case ARCH_GET_GS: {
 		unsigned long base;
+
 		if (doit)
 			rdmsrl(MSR_KERNEL_GS_BASE, base);
 		else
 			base = task->thread.gsbase;
-		ret = put_user(base, (unsigned long __user *)addr);
+		ret = put_user(base, (unsigned long __user *)arg2);
 		break;
 	}
 
 #ifdef CONFIG_CHECKPOINT_RESTORE
 # ifdef CONFIG_X86_X32_ABI
 	case ARCH_MAP_VDSO_X32:
-		return prctl_map_vdso(&vdso_image_x32, addr);
+		return prctl_map_vdso(&vdso_image_x32, arg2);
 # endif
 # if defined CONFIG_X86_32 || defined CONFIG_IA32_EMULATION
 	case ARCH_MAP_VDSO_32:
-		return prctl_map_vdso(&vdso_image_32, addr);
+		return prctl_map_vdso(&vdso_image_32, arg2);
 # endif
 	case ARCH_MAP_VDSO_64:
-		return prctl_map_vdso(&vdso_image_64, addr);
+		return prctl_map_vdso(&vdso_image_64, arg2);
 #endif
 
 	default:
@@ -621,10 +624,23 @@ long do_arch_prctl(struct task_struct *task, int code, unsigned long addr)
 	return ret;
 }
 
-long sys_arch_prctl(int code, unsigned long addr)
+SYSCALL_DEFINE2(arch_prctl, int, option, unsigned long, arg2)
+{
+	long ret;
+
+	ret = do_arch_prctl_64(current, option, arg2);
+	if (ret == -EINVAL)
+		ret = do_arch_prctl_common(current, option, arg2);
+
+	return ret;
+}
+
+#ifdef CONFIG_IA32_EMULATION
+COMPAT_SYSCALL_DEFINE2(arch_prctl, int, option, unsigned long, arg2)
 {
-	return do_arch_prctl(current, code, addr);
+	return do_arch_prctl_common(current, option, arg2);
 }
+#endif
 
 unsigned long KSTK_ESP(struct task_struct *task)
 {

arch/x86/kernel/ptrace.c

@@ -396,12 +396,12 @@ static int putreg(struct task_struct *child,
 		if (value >= TASK_SIZE_MAX)
 			return -EIO;
 		/*
-		 * When changing the segment base, use do_arch_prctl
+		 * When changing the segment base, use do_arch_prctl_64
 		 * to set either thread.fs or thread.fsindex and the
 		 * corresponding GDT slot.
 		 */
 		if (child->thread.fsbase != value)
-			return do_arch_prctl(child, ARCH_SET_FS, value);
+			return do_arch_prctl_64(child, ARCH_SET_FS, value);
 		return 0;
 	case offsetof(struct user_regs_struct,gs_base):
 		/*
@@ -410,7 +410,7 @@ static int putreg(struct task_struct *child,
 		if (value >= TASK_SIZE_MAX)
 			return -EIO;
 		if (child->thread.gsbase != value)
-			return do_arch_prctl(child, ARCH_SET_GS, value);
+			return do_arch_prctl_64(child, ARCH_SET_GS, value);
 		return 0;
 #endif
 	}
@@ -869,7 +869,7 @@ long arch_ptrace(struct task_struct *child, long request,
 		   Works just like arch_prctl, except that the arguments
 		   are reversed. */
 	case PTRACE_ARCH_PRCTL:
-		ret = do_arch_prctl(child, data, addr);
+		ret = do_arch_prctl_64(child, data, addr);
 		break;
 #endif
 

arch/x86/um/Makefile

@@ -16,7 +16,7 @@ obj-y = bug.o bugs_$(BITS).o delay.o fault.o ldt.o \
 
 ifeq ($(CONFIG_X86_32),y)
 
-obj-y += checksum_32.o
+obj-y += checksum_32.o syscalls_32.o
 obj-$(CONFIG_ELF_CORE) += elfcore.o
 
 subarch-y = ../lib/string_32.o ../lib/atomic64_32.o ../lib/atomic64_cx8_32.o

arch/x86/um/asm/ptrace.h

@@ -78,7 +78,7 @@ static inline int ptrace_set_thread_area(struct task_struct *child, int idx,
         return -ENOSYS;
 }
 
-extern long arch_prctl(struct task_struct *task, int code,
+extern long arch_prctl(struct task_struct *task, int option,
 		       unsigned long __user *addr);
 
 #endif

arch/x86/um/os-Linux/prctl.c

@@ -6,7 +6,7 @@
 #include <sys/ptrace.h>
 #include <asm/ptrace.h>
 
-int os_arch_prctl(int pid, int code, unsigned long *addr)
+int os_arch_prctl(int pid, int option, unsigned long *arg2)
 {
-        return ptrace(PTRACE_ARCH_PRCTL, pid, (unsigned long) addr, code);
+	return ptrace(PTRACE_ARCH_PRCTL, pid, (unsigned long) arg2, option);
 }

arch/x86/um/syscalls_32.c

+#include <linux/syscalls.h>
+#include <os.h>
+
+SYSCALL_DEFINE2(arch_prctl, int, option, unsigned long, arg2)
+{
+	return -EINVAL;
+}

arch/x86/um/syscalls_64.c

@@ -7,13 +7,15 @@
 
 #include <linux/sched.h>
 #include <linux/sched/mm.h>
+#include <linux/syscalls.h>
 #include <linux/uaccess.h>
 #include <asm/prctl.h> /* XXX This should get the constants from libc */
 #include <os.h>
 
-long arch_prctl(struct task_struct *task, int code, unsigned long __user *addr)
+long arch_prctl(struct task_struct *task, int option)
+		unsigned long __user *arg2)
 {
-	unsigned long *ptr = addr, tmp;
+	unsigned long *ptr = arg2, tmp;
 	long ret;
 	int pid = task->mm->context.id.u.pid;
 
@@ -30,7 +32,7 @@ long arch_prctl(struct task_struct *task, int code, unsigned long __user *addr)
 	 * arch_prctl is run on the host, then the registers are read
 	 * back.
 	 */
-	switch (code) {
+	switch (option) {
 	case ARCH_SET_FS:
 	case ARCH_SET_GS:
 		ret = restore_registers(pid, &current->thread.regs.regs);
@@ -50,11 +52,11 @@ long arch_prctl(struct task_struct *task, int code, unsigned long __user *addr)
 		ptr = &tmp;
 	}
 
-	ret = os_arch_prctl(pid, code, ptr);
+	ret = os_arch_prctl(pid, option, ptr);
 	if (ret)
 		return ret;
 
-	switch (code) {
+	switch (option) {
 	case ARCH_SET_FS:
 		current->thread.arch.fs = (unsigned long) ptr;
 		ret = save_registers(pid, &current->thread.regs.regs);
@@ -63,19 +65,19 @@ long arch_prctl(struct task_struct *task, int code, unsigned long __user *addr)
 		ret = save_registers(pid, &current->thread.regs.regs);
 		break;
 	case ARCH_GET_FS:
-		ret = put_user(tmp, addr);
+		ret = put_user(tmp, arg2);
 		break;
 	case ARCH_GET_GS:
-		ret = put_user(tmp, addr);
+		ret = put_user(tmp, arg2);
 		break;
 	}
 
 	return ret;
 }
 
-long sys_arch_prctl(int code, unsigned long addr)
+SYSCALL_DEFINE2(arch_prctl, int, option, unsigned long, arg2)
 {
-	return arch_prctl(current, code, (unsigned long __user *) addr);
+	return arch_prctl(current, option, (unsigned long __user *) arg2);
 }
 
 void arch_switch_to(struct task_struct *to)

fs/exec.c

@@ -1320,6 +1320,7 @@ void setup_new_exec(struct linux_binprm * bprm)
 	else
 		set_dumpable(current->mm, suid_dumpable);
 
+	arch_setup_new_exec();
 	perf_event_exec();
 	__set_task_comm(current, kbasename(bprm->filename), true);
 

include/linux/compat.h

@@ -723,6 +723,8 @@ asmlinkage long compat_sys_sched_rr_get_interval(compat_pid_t pid,
 asmlinkage long compat_sys_fanotify_mark(int, unsigned int, __u32, __u32,
 					    int, const char __user *);
 
+asmlinkage long compat_sys_arch_prctl(int option, unsigned long arg2);
+
 /*
  * For most but not all architectures, "am I in a compat syscall?" and
  * "am I a compat task?" are the same question.  For architectures on which

include/linux/thread_info.h

@@ -101,6 +101,10 @@ static inline void check_object_size(const void *ptr, unsigned long n,
 { }
 #endif /* CONFIG_HARDENED_USERCOPY */
 
+#ifndef arch_setup_new_exec
+static inline void arch_setup_new_exec(void) { }
+#endif
+
 #endif	/* __KERNEL__ */
 
 #endif /* _LINUX_THREAD_INFO_H */

tools/arch/x86/include/asm/cpufeatures.h

@@ -187,6 +187,7 @@
  * Reuse free bits when adding new feature flags!
  */
 #define X86_FEATURE_RING3MWAIT	( 7*32+ 0) /* Ring 3 MONITOR/MWAIT */
+#define X86_FEATURE_CPUID_FAULT ( 7*32+ 1) /* Intel CPUID faulting */
 #define X86_FEATURE_CPB		( 7*32+ 2) /* AMD Core Performance Boost */
 #define X86_FEATURE_EPB		( 7*32+ 3) /* IA32_ENERGY_PERF_BIAS support */
 #define X86_FEATURE_CAT_L3	( 7*32+ 4) /* Cache Allocation Technology L3 */

tools/arch/x86/lib/memcpy_64.S

@@ -286,7 +286,7 @@ ENDPROC(memcpy_mcsafe_unrolled)
 	_ASM_EXTABLE_FAULT(.L_copy_leading_bytes, .L_memcpy_mcsafe_fail)
 	_ASM_EXTABLE_FAULT(.L_cache_w0, .L_memcpy_mcsafe_fail)
 	_ASM_EXTABLE_FAULT(.L_cache_w1, .L_memcpy_mcsafe_fail)
-	_ASM_EXTABLE_FAULT(.L_cache_w3, .L_memcpy_mcsafe_fail)
+	_ASM_EXTABLE_FAULT(.L_cache_w2, .L_memcpy_mcsafe_fail)
 	_ASM_EXTABLE_FAULT(.L_cache_w3, .L_memcpy_mcsafe_fail)
 	_ASM_EXTABLE_FAULT(.L_cache_w4, .L_memcpy_mcsafe_fail)
 	_ASM_EXTABLE_FAULT(.L_cache_w5, .L_memcpy_mcsafe_fail)

tools/include/linux/bug.h

+#ifndef _TOOLS_PERF_LINUX_BUG_H
+#define _TOOLS_PERF_LINUX_BUG_H
+
+/* Force a compilation error if condition is true, but also produce a
+   result (of value 0 and type size_t), so the expression can be used
+   e.g. in a structure initializer (or where-ever else comma expressions
+   aren't permitted). */
+#define BUILD_BUG_ON_ZERO(e) (sizeof(struct { int:-!!(e); }))
+
+#endif	/* _TOOLS_PERF_LINUX_BUG_H */

tools/include/linux/compiler-gcc.h

@@ -16,3 +16,6 @@
 #if GCC_VERSION >= 40300
 # define __compiletime_error(message) __attribute__((error(message)))
 #endif /* GCC_VERSION >= 40300 */
+
+/* &a[0] degrades to a pointer: a different type from an array */
+#define __must_be_array(a)	BUILD_BUG_ON_ZERO(__same_type((a), &(a)[0]))

tools/include/linux/compiler.h

@@ -17,6 +17,11 @@
 # define __always_inline	inline __attribute__((always_inline))
 #endif
 
+/* Are two types/vars the same type (ignoring qualifiers)? */
+#ifndef __same_type
+# define __same_type(a, b) __builtin_types_compatible_p(typeof(a), typeof(b))
+#endif
+
 #ifdef __ANDROID__
 /*
  * FIXME: Big hammer to get rid of tons of:

tools/include/linux/hashtable.h

@@ -13,10 +13,6 @@
 #include <linux/hash.h>
 #include <linux/log2.h>
 
-#ifndef ARRAY_SIZE
-#define ARRAY_SIZE(x) (sizeof(x) / sizeof((x)[0]))
-#endif
-
 #define DEFINE_HASHTABLE(name, bits)						\
 	struct hlist_head name[1 << (bits)] =					\
 			{ [0 ... ((1 << (bits)) - 1)] = HLIST_HEAD_INIT }

tools/include/linux/kernel.h

@@ -4,6 +4,7 @@
 #include <stdarg.h>
 #include <stddef.h>
 #include <assert.h>
+#include <linux/compiler.h>
 
 #ifndef UINT_MAX
 #define UINT_MAX	(~0U)
@@ -76,6 +77,8 @@
 int vscnprintf(char *buf, size_t size, const char *fmt, va_list args);
 int scnprintf(char * buf, size_t size, const char * fmt, ...);
 
+#define ARRAY_SIZE(arr) (sizeof(arr) / sizeof((arr)[0]) + __must_be_array(arr))
+
 /*
  * This looks more complex than it should be. But we need to
  * get the type for the ~ right in round_down (it needs to be

tools/include/linux/log2.h

@@ -12,6 +12,9 @@
 #ifndef _TOOLS_LINUX_LOG2_H
 #define _TOOLS_LINUX_LOG2_H
 
+#include <linux/bitops.h>
+#include <linux/types.h>
+
 /*
  * non-constant log of base 2 calculators
  * - the arch may override these in asm/bitops.h if they can be implemented

tools/include/uapi/linux/stat.h

@@ -114,7 +114,7 @@ struct statx {
 	__u64	stx_ino;	/* Inode number */
 	__u64	stx_size;	/* File size */
 	__u64	stx_blocks;	/* Number of 512-byte blocks allocated */
-	__u64	__spare1[1];
+	__u64	stx_attributes_mask; /* Mask to show what's supported in stx_attributes */
 	/* 0x40 */
 	struct statx_timestamp	stx_atime;	/* Last access time */
 	struct statx_timestamp	stx_btime;	/* File creation time */
@@ -152,9 +152,10 @@ struct statx {
 #define STATX_BASIC_STATS	0x000007ffU	/* The stuff in the normal stat struct */
 #define STATX_BTIME		0x00000800U	/* Want/got stx_btime */
 #define STATX_ALL		0x00000fffU	/* All currently supported flags */
+#define STATX__RESERVED		0x80000000U	/* Reserved for future struct statx expansion */
 
 /*
- * Attributes to be found in stx_attributes
+ * Attributes to be found in stx_attributes and masked in stx_attributes_mask.
  *
  * These give information about the features or the state of a file that might
  * be of use to ordinary userspace programs such as GUIs or ls rather than

tools/lib/subcmd/help.h

@@ -2,6 +2,7 @@
 #define __SUBCMD_HELP_H
 
 #include <sys/types.h>
+#include <stdio.h>
 
 struct cmdnames {
 	size_t alloc;

tools/lib/symbol/kallsyms.c

+#include <ctype.h>
 #include "symbol/kallsyms.h"
 #include <stdio.h>
 #include <stdlib.h>

tools/objtool/builtin-check.c

@@ -36,8 +36,7 @@
 #include "warn.h"
 
 #include <linux/hashtable.h>
-
-#define ARRAY_SIZE(x) (sizeof(x) / sizeof((x)[0]))
+#include <linux/kernel.h>
 
 #define STATE_FP_SAVED		0x1
 #define STATE_FP_SETUP		0x2

tools/objtool/objtool.c

@@ -31,11 +31,10 @@
 #include <stdlib.h>
 #include <subcmd/exec-cmd.h>
 #include <subcmd/pager.h>
+#include <linux/kernel.h>
 
 #include "builtin.h"
 
-#define ARRAY_SIZE(x) (sizeof(x)/sizeof(x[0]))
-
 struct cmd_struct {
 	const char *name;
 	int (*fn)(int, const char **);

tools/perf/MANIFEST

@@ -64,6 +64,7 @@ tools/include/linux/bitops.h
 tools/include/linux/compiler.h
 tools/include/linux/compiler-gcc.h
 tools/include/linux/coresight-pmu.h
+tools/include/linux/bug.h
 tools/include/linux/filter.h
 tools/include/linux/hash.h
 tools/include/linux/kernel.h

tools/perf/arch/arm/util/cs-etm.c

@@ -33,6 +33,7 @@
 #include "../../util/cs-etm.h"
 
 #include <stdlib.h>
+#include <sys/stat.h>
 
 #define ENABLE_SINK_MAX	128
 #define CS_BUS_DEVICE_PATH "/bus/coresight/devices/"

tools/perf/arch/arm/util/dwarf-regs.c

@@ -9,6 +9,7 @@
  */
 
 #include <stddef.h>
+#include <linux/stringify.h>
 #include <dwarf-regs.h>
 
 struct pt_regs_dwarfnum {
@@ -16,10 +17,9 @@ struct pt_regs_dwarfnum {
 	unsigned int dwarfnum;
 };
 
-#define STR(s) #s
 #define REG_DWARFNUM_NAME(r, num) {.name = r, .dwarfnum = num}
 #define GPR_DWARFNUM_NAME(num) \
-	{.name = STR(%r##num), .dwarfnum = num}
+	{.name = __stringify(%r##num), .dwarfnum = num}
 #define REG_DWARFNUM_END {.name = NULL, .dwarfnum = 0}
 
 /*

tools/perf/arch/arm/util/unwind-libdw.c

 #include <elfutils/libdwfl.h>
 #include "../../util/unwind-libdw.h"
 #include "../../util/perf_regs.h"
+#include "../../util/event.h"
 
 bool libdw__arch_set_initial_registers(Dwfl_Thread *thread, void *arg)
 {

tools/perf/arch/arm64/util/dwarf-regs.c

@@ -8,9 +8,12 @@
  * published by the Free Software Foundation.
  */
 
+#include <errno.h>
 #include <stddef.h>
+#include <string.h>
 #include <dwarf-regs.h>
 #include <linux/ptrace.h> /* for struct user_pt_regs */
+#include <linux/stringify.h>
 #include "util.h"
 
 struct pt_regs_dwarfnum {
@@ -20,7 +23,7 @@ struct pt_regs_dwarfnum {
 
 #define REG_DWARFNUM_NAME(r, num) {.name = r, .dwarfnum = num}
 #define GPR_DWARFNUM_NAME(num) \
-	{.name = STR(%x##num), .dwarfnum = num}
+	{.name = __stringify(%x##num), .dwarfnum = num}
 #define REG_DWARFNUM_END {.name = NULL, .dwarfnum = 0}
 #define DWARFNUM2OFFSET(index) \
 	(index * sizeof((struct user_pt_regs *)0)->regs[0])

tools/perf/arch/arm64/util/unwind-libunwind.c

+#include <errno.h>
 
 #ifndef REMOTE_UNWIND_LIBUNWIND
-#include <errno.h>
 #include <libunwind.h>
 #include "perf_regs.h"
 #include "../../util/unwind.h"

tools/perf/arch/common.c

@@ -4,6 +4,8 @@
 #include "../util/util.h"
 #include "../util/debug.h"
 
+#include "sane_ctype.h"
+
 const char *const arm_triplets[] = {
 	"arm-eabi-",
 	"arm-linux-androideabi-",

tools/perf/arch/powerpc/util/dwarf-regs.c

@@ -15,6 +15,7 @@
 #include <dwarf-regs.h>
 #include <linux/ptrace.h>
 #include <linux/kernel.h>
+#include <linux/stringify.h>
 #include "util.h"
 
 struct pt_regs_dwarfnum {
@@ -24,10 +25,10 @@ struct pt_regs_dwarfnum {
 };
 
 #define REG_DWARFNUM_NAME(r, num)					\
-		{.name = STR(%)STR(r), .dwarfnum = num,			\
+		{.name = __stringify(%)__stringify(r), .dwarfnum = num,			\
 		.ptregs_offset = offsetof(struct pt_regs, r)}
 #define GPR_DWARFNUM_NAME(num)						\
-		{.name = STR(%gpr##num), .dwarfnum = num,		\
+		{.name = __stringify(%gpr##num), .dwarfnum = num,		\
 		.ptregs_offset = offsetof(struct pt_regs, gpr[num])}
 #define REG_DWARFNUM_END {.name = NULL, .dwarfnum = 0, .ptregs_offset = 0}
 

tools/perf/arch/powerpc/util/kvm-stat.c

+#include <errno.h>
 #include "util/kvm-stat.h"
 #include "util/parse-events.h"
 #include "util/debug.h"

tools/perf/arch/powerpc/util/perf_regs.c

+#include <errno.h>
 #include <string.h>
 #include <regex.h>
 

tools/perf/arch/s390/util/kvm-stat.c

@@ -9,6 +9,7 @@
  * as published by the Free Software Foundation.
  */
 
+#include <errno.h>
 #include "../../util/kvm-stat.h"
 #include <asm/sie.h>
 

tools/perf/arch/x86/tests/intel-cqm.c

@@ -6,7 +6,10 @@
 #include "evsel.h"
 #include "arch-tests.h"
 
+#include <signal.h>
 #include <sys/mman.h>
+#include <sys/wait.h>
+#include <errno.h>
 #include <string.h>
 
 static pid_t spawn(void)

tools/perf/arch/x86/tests/perf-time-to-tsc.c

+#include <errno.h>
+#include <inttypes.h>
 #include <stdio.h>
 #include <unistd.h>
 #include <linux/types.h>

tools/perf/arch/x86/util/auxtrace.c

@@ -13,6 +13,7 @@
  *
  */
 
+#include <errno.h>
 #include <stdbool.h>
 
 #include "../../util/header.h"

tools/perf/arch/x86/util/intel-bts.c

@@ -13,6 +13,7 @@
  *
  */
 
+#include <errno.h>
 #include <linux/kernel.h>
 #include <linux/types.h>
 #include <linux/bitops.h>

tools/perf/arch/x86/util/intel-pt.c

@@ -13,6 +13,7 @@
  *
  */
 
+#include <errno.h>
 #include <stdbool.h>
 #include <linux/kernel.h>
 #include <linux/types.h>

tools/perf/arch/x86/util/kvm-stat.c

+#include <errno.h>
 #include "../../util/kvm-stat.h"
 #include <asm/svm.h>
 #include <asm/vmx.h>

tools/perf/arch/x86/util/perf_regs.c

+#include <errno.h>
 #include <string.h>
 #include <regex.h>
 

tools/perf/arch/x86/util/unwind-libdw.c

 #include <elfutils/libdwfl.h>
 #include "../../util/unwind-libdw.h"
 #include "../../util/perf_regs.h"
+#include "../../util/event.h"
 
 bool libdw__arch_set_initial_registers(Dwfl_Thread *thread, void *arg)
 {

tools/perf/bench/mem-functions.c

@@ -12,6 +12,7 @@
 #include <subcmd/parse-options.h>
 #include "../util/header.h"
 #include "../util/cloexec.h"
+#include "../util/string2.h"
 #include "bench.h"
 #include "mem-memcpy-arch.h"
 #include "mem-memset-arch.h"

tools/perf/bench/numa.c

@@ -4,6 +4,7 @@
  * numa: Simulate NUMA-sensitive workload and measure their NUMA performance
  */
 
+#include <inttypes.h>
 /* For the CLR_() macros */
 #include <pthread.h>
 
@@ -30,6 +31,7 @@
 #include <sys/wait.h>
 #include <sys/prctl.h>
 #include <sys/types.h>
+#include <linux/kernel.h>
 #include <linux/time64.h>
 
 #include <numa.h>

tools/perf/builtin-annotate.c

@@ -33,6 +33,7 @@
 #include "util/block-range.h"
 
 #include <dlfcn.h>
+#include <errno.h>
 #include <linux/bitmap.h>
 
 struct perf_annotate {

tools/perf/builtin-buildid-cache.c

@@ -10,6 +10,7 @@
 #include <sys/time.h>
 #include <time.h>
 #include <dirent.h>
+#include <errno.h>
 #include <unistd.h>
 #include "builtin.h"
 #include "perf.h"
@@ -21,6 +22,7 @@
 #include "util/build-id.h"
 #include "util/session.h"
 #include "util/symbol.h"
+#include "util/time-utils.h"
 
 static int build_id_cache__kcore_buildid(const char *proc_dir, char *sbuildid)
 {

tools/perf/builtin-buildid-list.c

@@ -16,6 +16,7 @@
 #include "util/session.h"
 #include "util/symbol.h"
 #include "util/data.h"
+#include <errno.h>
 
 static int sysfs__fprintf_build_id(FILE *fp)
 {

tools/perf/builtin-c2c.c

@@ -9,10 +9,13 @@
  *   Dick Fowles <fowles@inreach.com>
  *   Joe Mario <jmario@redhat.com>
  */
+#include <errno.h>
+#include <inttypes.h>
 #include <linux/compiler.h>
 #include <linux/kernel.h>
 #include <linux/stringify.h>
 #include <asm/bug.h>
+#include <sys/param.h>
 #include "util.h"
 #include "debug.h"
 #include "builtin.h"
@@ -29,6 +32,7 @@
 #include <asm/bug.h>
 #include "ui/browsers/hists.h"
 #include "evlist.h"
+#include "thread.h"
 
 struct c2c_hists {
 	struct hists		hists;

tools/perf/builtin-diff.c

@@ -19,6 +19,8 @@
 #include "util/data.h"
 #include "util/config.h"
 
+#include <errno.h>
+#include <inttypes.h>
 #include <stdlib.h>
 #include <math.h>
 

tools/perf/builtin-ftrace.c

@@ -9,12 +9,15 @@
 #include "builtin.h"
 #include "perf.h"
 
+#include <errno.h>
 #include <unistd.h>
 #include <signal.h>
 #include <fcntl.h>
+#include <poll.h>
 
 #include "debug.h"
 #include <subcmd/parse-options.h>
+#include <api/fs/tracing_path.h>
 #include "evlist.h"
 #include "target.h"
 #include "cpumap.h"

tools/perf/builtin-help.c

@@ -12,16 +12,22 @@
 #include <subcmd/run-command.h>
 #include <subcmd/help.h>
 #include "util/debug.h"
+#include <linux/kernel.h>
+#include <errno.h>
+#include <stdio.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <unistd.h>
 
 static struct man_viewer_list {
 	struct man_viewer_list *next;
-	char name[FLEX_ARRAY];
+	char name[0];
 } *man_viewer_list;
 
 static struct man_viewer_info_list {
 	struct man_viewer_info_list *next;
 	const char *info;
-	char name[FLEX_ARRAY];
+	char name[0];
 } *man_viewer_info_list;
 
 enum help_format {

tools/perf/builtin-inject.c

@@ -18,10 +18,13 @@
 #include "util/data.h"
 #include "util/auxtrace.h"
 #include "util/jit.h"
+#include "util/thread.h"
 
 #include <subcmd/parse-options.h>
 
 #include <linux/list.h>
+#include <errno.h>
+#include <signal.h>
 
 struct perf_inject {
 	struct perf_tool	tool;

tools/perf/builtin-kallsyms.c

@@ -7,6 +7,7 @@
  *
  * Released under the GPL v2. (and only v2, not any later version)
  */
+#include <inttypes.h>
 #include "builtin.h"
 #include <linux/compiler.h>
 #include <subcmd/parse-options.h>

tools/perf/builtin-kmem.c

@@ -20,11 +20,16 @@
 
 #include "util/debug.h"
 
+#include <linux/kernel.h>
 #include <linux/rbtree.h>
 #include <linux/string.h>
+#include <errno.h>
+#include <inttypes.h>
 #include <locale.h>
 #include <regex.h>
 
+#include "sane_ctype.h"
+
 static int	kmem_slab;
 static int	kmem_page;
 

tools/perf/builtin-kvm.c

@@ -3,6 +3,7 @@
 
 #include "util/evsel.h"
 #include "util/evlist.h"
+#include "util/term.h"
 #include "util/util.h"
 #include "util/cache.h"
 #include "util/symbol.h"
@@ -23,13 +24,33 @@
 #ifdef HAVE_TIMERFD_SUPPORT
 #include <sys/timerfd.h>
 #endif
+#include <sys/time.h>
 
+#include <linux/kernel.h>
 #include <linux/time64.h>
+#include <errno.h>
+#include <inttypes.h>
+#include <poll.h>
 #include <termios.h>
 #include <semaphore.h>
+#include <signal.h>
 #include <pthread.h>
 #include <math.h>
 
+static const char *get_filename_for_perf_kvm(void)
+{
+	const char *filename;
+
+	if (perf_host && !perf_guest)
+		filename = strdup("perf.data.host");
+	else if (!perf_host && perf_guest)
+		filename = strdup("perf.data.guest");
+	else
+		filename = strdup("perf.data.kvm");
+
+	return filename;
+}
+
 #ifdef HAVE_KVM_STAT_SUPPORT
 #include "util/kvm-stat.h"
 

tools/perf/builtin-lock.c

+#include <errno.h>
+#include <inttypes.h>
 #include "builtin.h"
 #include "perf.h"
 
@@ -26,6 +28,7 @@
 
 #include <linux/list.h>
 #include <linux/hash.h>
+#include <linux/kernel.h>
 
 static struct perf_session *session;
 

tools/perf/builtin-mem.c

+#include <inttypes.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <unistd.h>
 #include "builtin.h"
 #include "perf.h"
 
@@ -8,6 +12,7 @@
 #include "util/data.h"
 #include "util/mem-events.h"
 #include "util/debug.h"
+#include "util/symbol.h"
 
 #define MEM_OPERATION_LOAD	0x1
 #define MEM_OPERATION_STORE	0x2

tools/perf/builtin-record.c

@@ -38,11 +38,18 @@
 #include "util/bpf-loader.h"
 #include "util/trigger.h"
 #include "util/perf-hooks.h"
+#include "util/time-utils.h"
+#include "util/units.h"
 #include "asm/bug.h"
 
+#include <errno.h>
+#include <inttypes.h>
+#include <poll.h>
 #include <unistd.h>
 #include <sched.h>
+#include <signal.h>
 #include <sys/mman.h>
+#include <sys/wait.h>
 #include <asm/bug.h>
 #include <linux/time64.h>
 

tools/perf/builtin-report.c

@@ -16,7 +16,6 @@
 #include <linux/rbtree.h>
 #include "util/symbol.h"
 #include "util/callchain.h"
-#include "util/strlist.h"
 #include "util/values.h"
 
 #include "perf.h"
@@ -38,10 +37,18 @@
 #include "arch/common.h"
 #include "util/time-utils.h"
 #include "util/auxtrace.h"
+#include "util/units.h"
 
 #include <dlfcn.h>
+#include <errno.h>
+#include <inttypes.h>
+#include <regex.h>
+#include <signal.h>
 #include <linux/bitmap.h>
 #include <linux/stringify.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <unistd.h>
 
 struct report {
 	struct perf_tool	tool;

tools/perf/builtin-sched.c

@@ -22,16 +22,21 @@
 
 #include "util/debug.h"
 
+#include <linux/kernel.h>
 #include <linux/log2.h>
 #include <sys/prctl.h>
 #include <sys/resource.h>
+#include <inttypes.h>
 
+#include <errno.h>
 #include <semaphore.h>
 #include <pthread.h>
 #include <math.h>
 #include <api/fs/fs.h>
 #include <linux/time64.h>
 
+#include "sane_ctype.h"
+
 #define PR_SET_NAME		15               /* Set process name */
 #define MAX_CPUS		4096
 #define COMM_LEN		20

tools/perf/builtin-script.c

@@ -21,14 +21,27 @@
 #include "util/cpumap.h"
 #include "util/thread_map.h"
 #include "util/stat.h"
+#include "util/string2.h"
 #include "util/thread-stack.h"
 #include "util/time-utils.h"
+#include "print_binary.h"
 #include <linux/bitmap.h>
+#include <linux/kernel.h>
 #include <linux/stringify.h>
 #include <linux/time64.h>
 #include "asm/bug.h"
 #include "util/mem-events.h"
 #include "util/dump-insn.h"
+#include <dirent.h>
+#include <errno.h>
+#include <inttypes.h>
+#include <signal.h>
+#include <sys/param.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <unistd.h>
+
+#include "sane_ctype.h"
 
 static char const		*script_name;
 static char const		*generate_script_lang;

tools/perf/builtin-stat.c

@@ -64,14 +64,24 @@
 #include "util/session.h"
 #include "util/tool.h"
 #include "util/group.h"
+#include "util/string2.h"
 #include "asm/bug.h"
 
 #include <linux/time64.h>
 #include <api/fs/fs.h>
+#include <errno.h>
+#include <signal.h>
 #include <stdlib.h>
 #include <sys/prctl.h>
+#include <inttypes.h>
 #include <locale.h>
 #include <math.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <sys/wait.h>
+#include <unistd.h>
+
+#include "sane_ctype.h"
 
 #define DEFAULT_SEPARATOR	" "
 #define CNTR_NOT_SUPPORTED	"<not supported>"

tools/perf/builtin-timechart.c

@@ -12,6 +12,8 @@
  * of the License.
  */
 
+#include <errno.h>
+#include <inttypes.h>
 #include <traceevent/event-parse.h>
 
 #include "builtin.h"
@@ -23,11 +25,12 @@
 #include "util/cache.h"
 #include "util/evlist.h"
 #include "util/evsel.h"
+#include <linux/kernel.h>
 #include <linux/rbtree.h>
 #include <linux/time64.h>
 #include "util/symbol.h"
+#include "util/thread.h"
 #include "util/callchain.h"
-#include "util/strlist.h"
 
 #include "perf.h"
 #include "util/header.h"

tools/perf/builtin-top.c

@@ -40,6 +40,7 @@
 #include "util/cpumap.h"
 #include "util/xyarray.h"
 #include "util/sort.h"
+#include "util/term.h"
 #include "util/intlist.h"
 #include "util/parse-branch-options.h"
 #include "arch/common.h"
@@ -58,6 +59,7 @@
 #include <errno.h>
 #include <time.h>
 #include <sched.h>
+#include <signal.h>
 
 #include <sys/syscall.h>
 #include <sys/ioctl.h>
@@ -72,6 +74,8 @@
 #include <linux/time64.h>
 #include <linux/types.h>
 
+#include "sane_ctype.h"
+
 static volatile int done;
 
 #define HEADER_LINE_NR  5

tools/perf/builtin-trace.c

@@ -24,6 +24,7 @@
 #include "util/evlist.h"
 #include <subcmd/exec-cmd.h>
 #include "util/machine.h"
+#include "util/path.h"
 #include "util/session.h"
 #include "util/thread.h"
 #include <subcmd/parse-options.h>
@@ -36,19 +37,28 @@
 #include "util/parse-events.h"
 #include "util/bpf-loader.h"
 #include "callchain.h"
+#include "print_binary.h"
+#include "string2.h"
 #include "syscalltbl.h"
 #include "rb_resort.h"
 
+#include <errno.h>
+#include <inttypes.h>
 #include <libaudit.h> /* FIXME: Still needed for audit_errno_to_name */
+#include <poll.h>
+#include <signal.h>
 #include <stdlib.h>
 #include <string.h>
 #include <linux/err.h>
 #include <linux/filter.h>
 #include <linux/audit.h>
+#include <linux/kernel.h>
 #include <linux/random.h>
 #include <linux/stringify.h>
 #include <linux/time64.h>
 
+#include "sane_ctype.h"
+
 #ifndef O_CLOEXEC
 # define O_CLOEXEC		02000000
 #endif

tools/perf/builtin-version.c

-#include "util/util.h"
 #include "builtin.h"
 #include "perf.h"
+#include <linux/compiler.h>
+#include <stdio.h>
 
 int cmd_version(int argc __maybe_unused, const char **argv __maybe_unused)
 {

tools/perf/builtin.h

@@ -2,16 +2,12 @@
 #define BUILTIN_H
 
 #include "util/util.h"
-#include "util/strbuf.h"
 
 extern const char perf_usage_string[];
 extern const char perf_more_info_string[];
 
 void list_common_cmds_help(void);
 const char *help_unknown_cmd(const char *cmd);
-void prune_packed_objects(int);
-int read_line_with_nul(char *buf, int size, FILE *file);
-int check_pager_config(const char *cmd);
 
 int cmd_annotate(int argc, const char **argv);
 int cmd_bench(int argc, const char **argv);

tools/perf/perf.c

@@ -19,9 +19,15 @@
 #include "util/debug.h"
 #include <api/fs/fs.h>
 #include <api/fs/tracing_path.h>
+#include <errno.h>
 #include <pthread.h>
+#include <signal.h>
 #include <stdlib.h>
 #include <time.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <unistd.h>
+#include <linux/kernel.h>
 
 const char perf_usage_string[] =
 	"perf [--version] [--help] [OPTIONS] COMMAND [ARGS]";
@@ -88,7 +94,7 @@ static int pager_command_config(const char *var, const char *value, void *data)
 }
 
 /* returns 0 for "no pager", 1 for "use pager", and -1 for "not specified" */
-int check_pager_config(const char *cmd)
+static int check_pager_config(const char *cmd)
 {
 	int err;
 	struct pager_config c;
@@ -327,16 +333,6 @@ static void handle_internal_command(int argc, const char **argv)
 {
 	const char *cmd = argv[0];
 	unsigned int i;
-	static const char ext[] = STRIP_EXTENSION;
-
-	if (sizeof(ext) > 1) {
-		i = strlen(argv[0]) - strlen(ext);
-		if (i > 0 && !strcmp(argv[0] + i, ext)) {
-			char *argv0 = strdup(argv[0]);
-			argv[0] = cmd = argv0;
-			argv0[i] = '\0';
-		}
-	}
 
 	/* Turn "perf cmd --help" into "perf help cmd" */
 	if (argc > 1 && !strcmp(argv[1], "--help")) {

tools/perf/tests/attr.c

@@ -18,10 +18,16 @@
  * permissions. All the event text files are stored there.
  */
 
+#include <errno.h>
+#include <inttypes.h>
 #include <stdlib.h>
 #include <stdio.h>
 #include <linux/types.h>
 #include <linux/kernel.h>
+#include <sys/param.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <unistd.h>
 #include "../perf.h"
 #include "util.h"
 #include <subcmd/exec-cmd.h>

tools/perf/tests/backward-ring-buffer.c

@@ -8,6 +8,7 @@
 #include <sys/prctl.h>
 #include "tests.h"
 #include "debug.h"
+#include <errno.h>
 
 #define NR_ITERS 111
 

tools/perf/tests/bpf.c

+#include <errno.h>
 #include <stdio.h>
 #include <sys/epoll.h>
+#include <sys/types.h>
+#include <sys/stat.h>
 #include <util/util.h>
 #include <util/bpf-loader.h>
 #include <util/evlist.h>
 #include <linux/bpf.h>
 #include <linux/filter.h>
+#include <linux/kernel.h>
 #include <api/fs/fs.h>
 #include <bpf/bpf.h>
 #include "tests.h"

tools/perf/tests/builtin-test.c

@@ -3,8 +3,10 @@
  *
  * Builtin regression testing command: ever growing number of sanity tests
  */
+#include <errno.h>
 #include <unistd.h>
 #include <string.h>
+#include <sys/wait.h>
 #include "builtin.h"
 #include "hist.h"
 #include "intlist.h"
@@ -13,6 +15,7 @@
 #include "color.h"
 #include <subcmd/parse-options.h>
 #include "symbol.h"
+#include <linux/kernel.h>
 
 static bool dont_fork;
 

tools/perf/tests/clang.c

@@ -2,6 +2,7 @@
 #include "debug.h"
 #include "util.h"
 #include "c++/clang-c.h"
+#include <linux/kernel.h>
 
 static struct {
 	int (*func)(void);

tools/perf/tests/code-reading.c

+#include <errno.h>
+#include <linux/kernel.h>
 #include <linux/types.h>
+#include <inttypes.h>
 #include <stdlib.h>
 #include <unistd.h>
 #include <stdio.h>
-#include <ctype.h>
 #include <string.h>
+#include <sys/param.h>
 
 #include "parse-events.h"
 #include "evlist.h"
@@ -16,6 +19,8 @@
 
 #include "tests.h"
 
+#include "sane_ctype.h"
+
 #define BUFSZ	1024
 #define READLEN	128
 

tools/perf/tests/dso-data.c

+#include <dirent.h>
 #include <stdlib.h>
+#include <linux/kernel.h>
 #include <linux/types.h>
 #include <sys/stat.h>
 #include <fcntl.h>

tools/perf/tests/dwarf-unwind.c

 #include <linux/compiler.h>
 #include <linux/types.h>
+#include <inttypes.h>
 #include <unistd.h>
 #include "tests.h"
 #include "debug.h"

tools/perf/tests/event-times.c

 #include <linux/compiler.h>
+#include <errno.h>
+#include <inttypes.h>
 #include <string.h>
+#include <sys/wait.h>
 #include "tests.h"
 #include "evlist.h"
 #include "evsel.h"

tools/perf/tests/evsel-roundtrip-name.c

@@ -3,6 +3,8 @@
 #include "parse-events.h"
 #include "tests.h"
 #include "debug.h"
+#include <errno.h>
+#include <linux/kernel.h>
 
 static int perf_evsel__roundtrip_cache_name_test(void)
 {

tools/perf/tests/hists_common.c

+#include <inttypes.h>
 #include "perf.h"
 #include "util/debug.h"
 #include "util/symbol.h"
@@ -7,6 +8,7 @@
 #include "util/machine.h"
 #include "util/thread.h"
 #include "tests/hists_common.h"
+#include <linux/kernel.h>
 
 static struct {
 	u32 pid;

tools/perf/tests/hists_cumulate.c

@@ -9,6 +9,7 @@
 #include "util/parse-events.h"
 #include "tests/tests.h"
 #include "tests/hists_common.h"
+#include <linux/kernel.h>
 
 struct sample {
 	u32 pid;

tools/perf/tests/hists_filter.c

@@ -9,6 +9,7 @@
 #include "util/parse-events.h"
 #include "tests/tests.h"
 #include "tests/hists_common.h"
+#include <linux/kernel.h>
 
 struct sample {
 	u32 pid;

tools/perf/tests/hists_link.c

@@ -9,6 +9,8 @@
 #include "thread.h"
 #include "parse-events.h"
 #include "hists_common.h"
+#include <errno.h>
+#include <linux/kernel.h>
 
 struct sample {
 	u32 pid;

tools/perf/tests/hists_output.c

@@ -9,6 +9,7 @@
 #include "util/parse-events.h"
 #include "tests/tests.h"
 #include "tests/hists_common.h"
+#include <linux/kernel.h>
 
 struct sample {
 	u32 cpu;

tools/perf/tests/is_printable_array.c

 #include <linux/compiler.h>
+#include <linux/kernel.h>
 #include "tests.h"
 #include "debug.h"
-#include "util.h"
+#include "print_binary.h"
 
 int test__is_printable_array(int subtest __maybe_unused)
 {

tools/perf/tests/mmap-basic.c

+#include <errno.h>
+#include <inttypes.h>
 /* For the CLR_() macros */
 #include <pthread.h>
 
@@ -7,6 +9,7 @@
 #include "cpumap.h"
 #include "tests.h"
 #include <linux/err.h>
+#include <linux/kernel.h>
 
 /*
  * This test will generate random numbers of calls to some getpid syscalls,

tools/perf/tests/mmap-thread-lookup.c

+#include <inttypes.h>
 #include <unistd.h>
 #include <sys/syscall.h>
 #include <sys/types.h>
@@ -11,6 +12,7 @@
 #include "thread_map.h"
 #include "symbol.h"
 #include "thread.h"
+#include "util.h"
 
 #define THREADS 4
 

tools/perf/tests/openat-syscall-all-cpus.c

+#include <errno.h>
+#include <inttypes.h>
 /* For the CPU_* macros */
 #include <pthread.h>
 
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <fcntl.h>
 #include <api/fs/fs.h>
 #include <linux/err.h>
+#include <api/fs/tracing_path.h>
 #include "evsel.h"
 #include "tests.h"
 #include "thread_map.h"

tools/perf/tests/openat-syscall-tp-fields.c

@@ -5,6 +5,7 @@
 #include "thread_map.h"
 #include "tests.h"
 #include "debug.h"
+#include <errno.h>
 
 #ifndef O_DIRECTORY
 #define O_DIRECTORY    00200000