diff --git a/arch/s390/kernel/perf_cpum_sf.c b/arch/s390/kernel/perf_cpum_sf.c
index a8e8321..92619cc 100644
--- a/arch/s390/kernel/perf_cpum_sf.c
+++ b/arch/s390/kernel/perf_cpum_sf.c
@@ -979,12 +979,15 @@
 	struct pt_regs regs;
 	struct perf_sf_sde_regs *sde_regs;
 	struct perf_sample_data data;
-	struct perf_raw_record raw;
+	struct perf_raw_record raw = {
+		.frag = {
+			.size = sfr->size,
+			.data = sfr,
+		},
+	};
 
 	/* Setup perf sample */
 	perf_sample_data_init(&data, 0, event->hw.last_period);
-	raw.size = sfr->size;
-	raw.data = sfr;
 	data.raw = &raw;
 
 	/* Setup pt_regs to look like an CPU-measurement external interrupt
diff --git a/arch/x86/events/amd/ibs.c b/arch/x86/events/amd/ibs.c
index feb90f6..72dea2f 100644
--- a/arch/x86/events/amd/ibs.c
+++ b/arch/x86/events/amd/ibs.c
@@ -655,8 +655,12 @@
 	}
 
 	if (event->attr.sample_type & PERF_SAMPLE_RAW) {
-		raw.size = sizeof(u32) + ibs_data.size;
-		raw.data = ibs_data.data;
+		raw = (struct perf_raw_record){
+			.frag = {
+				.size = sizeof(u32) + ibs_data.size,
+				.data = ibs_data.data,
+			},
+		};
 		data.raw = &raw;
 	}
 
diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
index 1a827ce..e79e6c6 100644
--- a/include/linux/perf_event.h
+++ b/include/linux/perf_event.h
@@ -69,9 +69,22 @@
 	bool			    contexts_maxed;
 };
 
-struct perf_raw_record {
-	u32				size;
+typedef unsigned long (*perf_copy_f)(void *dst, const void *src,
+				     unsigned long len);
+
+struct perf_raw_frag {
+	union {
+		struct perf_raw_frag	*next;
+		unsigned long		pad;
+	};
+	perf_copy_f			copy;
 	void				*data;
+	u32				size;
+} __packed;
+
+struct perf_raw_record {
+	struct perf_raw_frag		frag;
+	u32				size;
 };
 
 /*
@@ -1283,6 +1296,11 @@
 static inline void perf_restore_debug_store(void)			{ }
 #endif
 
+static __always_inline bool perf_raw_frag_last(const struct perf_raw_frag *frag)
+{
+	return frag->pad < sizeof(u64);
+}
+
 #define perf_output_put(handle, x) perf_output_copy((handle), &(x), sizeof(x))
 
 /*
diff --git a/kernel/events/core.c b/kernel/events/core.c
index 9c51ec3..b1891b6 100644
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -5553,16 +5553,26 @@
 	}
 
 	if (sample_type & PERF_SAMPLE_RAW) {
-		if (data->raw) {
-			u32 raw_size = data->raw->size;
-			u32 real_size = round_up(raw_size + sizeof(u32),
-						 sizeof(u64)) - sizeof(u32);
-			u64 zero = 0;
+		struct perf_raw_record *raw = data->raw;
 
-			perf_output_put(handle, real_size);
-			__output_copy(handle, data->raw->data, raw_size);
-			if (real_size - raw_size)
-				__output_copy(handle, &zero, real_size - raw_size);
+		if (raw) {
+			struct perf_raw_frag *frag = &raw->frag;
+
+			perf_output_put(handle, raw->size);
+			do {
+				if (frag->copy) {
+					__output_custom(handle, frag->copy,
+							frag->data, frag->size);
+				} else {
+					__output_copy(handle, frag->data,
+						      frag->size);
+				}
+				if (perf_raw_frag_last(frag))
+					break;
+				frag = frag->next;
+			} while (1);
+			if (frag->pad)
+				__output_skip(handle, NULL, frag->pad);
 		} else {
 			struct {
 				u32	size;
@@ -5687,14 +5697,28 @@
 	}
 
 	if (sample_type & PERF_SAMPLE_RAW) {
-		int size = sizeof(u32);
+		struct perf_raw_record *raw = data->raw;
+		int size;
 
-		if (data->raw)
-			size += data->raw->size;
-		else
-			size += sizeof(u32);
+		if (raw) {
+			struct perf_raw_frag *frag = &raw->frag;
+			u32 sum = 0;
 
-		header->size += round_up(size, sizeof(u64));
+			do {
+				sum += frag->size;
+				if (perf_raw_frag_last(frag))
+					break;
+				frag = frag->next;
+			} while (1);
+
+			size = round_up(sum + sizeof(u32), sizeof(u64));
+			raw->size = size - sizeof(u32);
+			frag->pad = raw->size - sum;
+		} else {
+			size = sizeof(u64);
+		}
+
+		header->size += size;
 	}
 
 	if (sample_type & PERF_SAMPLE_BRANCH_STACK) {
@@ -7331,7 +7355,7 @@
 static int perf_tp_filter_match(struct perf_event *event,
 				struct perf_sample_data *data)
 {
-	void *record = data->raw->data;
+	void *record = data->raw->frag.data;
 
 	/* only top level events have filters set */
 	if (event->parent)
@@ -7387,8 +7411,10 @@
 	struct perf_event *event;
 
 	struct perf_raw_record raw = {
-		.size = entry_size,
-		.data = record,
+		.frag = {
+			.size = entry_size,
+			.data = record,
+		},
 	};
 
 	perf_sample_data_init(&data, 0, 0);
diff --git a/kernel/events/internal.h b/kernel/events/internal.h
index 05f9f6d..2417eb5 100644
--- a/kernel/events/internal.h
+++ b/kernel/events/internal.h
@@ -123,10 +123,7 @@
 	return rb->aux_nr_pages << PAGE_SHIFT;
 }
 
-#define DEFINE_OUTPUT_COPY(func_name, memcpy_func)			\
-static inline unsigned long						\
-func_name(struct perf_output_handle *handle,				\
-	  const void *buf, unsigned long len)				\
+#define __DEFINE_OUTPUT_COPY_BODY(memcpy_func)				\
 {									\
 	unsigned long size, written;					\
 									\
@@ -152,6 +149,17 @@
 	return len;							\
 }
 
+#define DEFINE_OUTPUT_COPY(func_name, memcpy_func)			\
+static inline unsigned long						\
+func_name(struct perf_output_handle *handle,				\
+	  const void *buf, unsigned long len)				\
+__DEFINE_OUTPUT_COPY_BODY(memcpy_func)
+
+static inline unsigned long
+__output_custom(struct perf_output_handle *handle, perf_copy_f copy_func,
+		const void *buf, unsigned long len)
+__DEFINE_OUTPUT_COPY_BODY(copy_func)
+
 static inline unsigned long
 memcpy_common(void *dst, const void *src, unsigned long n)
 {
diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c
index 094c716..35ab1b2 100644
--- a/kernel/trace/bpf_trace.c
+++ b/kernel/trace/bpf_trace.c
@@ -245,8 +245,10 @@
 	struct bpf_event_entry *ee;
 	struct perf_event *event;
 	struct perf_raw_record raw = {
-		.size = size,
-		.data = data,
+		.frag = {
+			.size = size,
+			.data = data,
+		},
 	};
 
 	if (unlikely(flags & ~(BPF_F_INDEX_MASK)))