perf/arch: Remove perf_sample_data::regs_user_copy

struct perf_sample_data lives on-stack, we should be careful about it's
size. Furthermore, the pt_regs copy in there is only because x86_64 is a
trainwreck, solve it differently.
Reported-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: NSteven Rostedt <rostedt@goodmis.org>
Link: https://lkml.kernel.org/r/20201030151955.258178461@infradead.org

arch/arm/kernel/perf_regs.c

@@ -32,8 +32,7 @@ u64 perf_reg_abi(struct task_struct *task)
 }
 
 void perf_get_regs_user(struct perf_regs *regs_user,
-			struct pt_regs *regs,
-			struct pt_regs *regs_user_copy)
+			struct pt_regs *regs)
 {
 	regs_user->regs = task_pt_regs(current);
 	regs_user->abi = perf_reg_abi(current);

arch/arm64/kernel/perf_regs.c

@@ -73,8 +73,7 @@ u64 perf_reg_abi(struct task_struct *task)
 }
 
 void perf_get_regs_user(struct perf_regs *regs_user,
-			struct pt_regs *regs,
-			struct pt_regs *regs_user_copy)
+			struct pt_regs *regs)
 {
 	regs_user->regs = task_pt_regs(current);
 	regs_user->abi = perf_reg_abi(current);

arch/csky/kernel/perf_regs.c

@@ -32,8 +32,7 @@ u64 perf_reg_abi(struct task_struct *task)
 }
 
 void perf_get_regs_user(struct perf_regs *regs_user,
-			struct pt_regs *regs,
-			struct pt_regs *regs_user_copy)
+			struct pt_regs *regs)
 {
 	regs_user->regs = task_pt_regs(current);
 	regs_user->abi = perf_reg_abi(current);

arch/powerpc/perf/perf_regs.c

@@ -144,8 +144,7 @@ u64 perf_reg_abi(struct task_struct *task)
 }
 
 void perf_get_regs_user(struct perf_regs *regs_user,
-			struct pt_regs *regs,
-			struct pt_regs *regs_user_copy)
+			struct pt_regs *regs)
 {
 	regs_user->regs = task_pt_regs(current);
 	regs_user->abi = (regs_user->regs) ? perf_reg_abi(current) :

arch/riscv/kernel/perf_regs.c

@@ -36,8 +36,7 @@ u64 perf_reg_abi(struct task_struct *task)
 }
 
 void perf_get_regs_user(struct perf_regs *regs_user,
-			struct pt_regs *regs,
-			struct pt_regs *regs_user_copy)
+			struct pt_regs *regs)
 {
 	regs_user->regs = task_pt_regs(current);
 	regs_user->abi = perf_reg_abi(current);

arch/s390/kernel/perf_regs.c

@@ -53,8 +53,7 @@ u64 perf_reg_abi(struct task_struct *task)
 }
 
 void perf_get_regs_user(struct perf_regs *regs_user,
-			struct pt_regs *regs,
-			struct pt_regs *regs_user_copy)
+			struct pt_regs *regs)
 {
 	/*
 	 * Use the regs from the first interruption and let

arch/x86/kernel/perf_regs.c

@@ -101,8 +101,7 @@ u64 perf_reg_abi(struct task_struct *task)
 }
 
 void perf_get_regs_user(struct perf_regs *regs_user,
-			struct pt_regs *regs,
-			struct pt_regs *regs_user_copy)
+			struct pt_regs *regs)
 {
 	regs_user->regs = task_pt_regs(current);
 	regs_user->abi = perf_reg_abi(current);
@@ -129,12 +128,20 @@ u64 perf_reg_abi(struct task_struct *task)
 		return PERF_SAMPLE_REGS_ABI_64;
 }
 
+static DEFINE_PER_CPU(struct pt_regs, nmi_user_regs);
+
 void perf_get_regs_user(struct perf_regs *regs_user,
-			struct pt_regs *regs,
-			struct pt_regs *regs_user_copy)
+			struct pt_regs *regs)
 {
+	struct pt_regs *regs_user_copy = this_cpu_ptr(&nmi_user_regs);
 	struct pt_regs *user_regs = task_pt_regs(current);
 
+	if (!in_nmi()) {
+		regs_user->regs = user_regs;
+		regs_user->abi = perf_reg_abi(current);
+		return;
+	}
+
 	/*
 	 * If we're in an NMI that interrupted task_pt_regs setup, then
 	 * we can't sample user regs at all.  This check isn't really

include/linux/perf_event.h

@@ -1022,13 +1022,7 @@ struct perf_sample_data {
 	struct perf_callchain_entry	*callchain;
 	u64				aux_size;
 
-	/*
-	 * regs_user may point to task_pt_regs or to regs_user_copy, depending
-	 * on arch details.
-	 */
 	struct perf_regs		regs_user;
-	struct pt_regs			regs_user_copy;
-
 	struct perf_regs		regs_intr;
 	u64				stack_user_size;
 

include/linux/perf_regs.h

@@ -20,8 +20,7 @@ u64 perf_reg_value(struct pt_regs *regs, int idx);
 int perf_reg_validate(u64 mask);
 u64 perf_reg_abi(struct task_struct *task);
 void perf_get_regs_user(struct perf_regs *regs_user,
-			struct pt_regs *regs,
-			struct pt_regs *regs_user_copy);
+			struct pt_regs *regs);
 #else
 
 #define PERF_REG_EXTENDED_MASK	0
@@ -42,8 +41,7 @@ static inline u64 perf_reg_abi(struct task_struct *task)
 }
 
 static inline void perf_get_regs_user(struct perf_regs *regs_user,
-				      struct pt_regs *regs,
-				      struct pt_regs *regs_user_copy)
+				      struct pt_regs *regs)
 {
 	regs_user->regs = task_pt_regs(current);
 	regs_user->abi = perf_reg_abi(current);

kernel/events/core.c

@@ -6374,14 +6374,13 @@ perf_output_sample_regs(struct perf_output_handle *handle,
 }
 
 static void perf_sample_regs_user(struct perf_regs *regs_user,
-				  struct pt_regs *regs,
-				  struct pt_regs *regs_user_copy)
+				  struct pt_regs *regs)
 {
 	if (user_mode(regs)) {
 		regs_user->abi = perf_reg_abi(current);
 		regs_user->regs = regs;
 	} else if (!(current->flags & PF_KTHREAD)) {
-		perf_get_regs_user(regs_user, regs, regs_user_copy);
+		perf_get_regs_user(regs_user, regs);
 	} else {
 		regs_user->abi = PERF_SAMPLE_REGS_ABI_NONE;
 		regs_user->regs = NULL;
@@ -7083,8 +7082,7 @@ void perf_prepare_sample(struct perf_event_header *header,
 	}
 
 	if (sample_type & (PERF_SAMPLE_REGS_USER | PERF_SAMPLE_STACK_USER))
-		perf_sample_regs_user(&data->regs_user, regs,
-				      &data->regs_user_copy);
+		perf_sample_regs_user(&data->regs_user, regs);
 
 	if (sample_type & PERF_SAMPLE_REGS_USER) {
 		/* regs dump ABI info */