17 files changed, 708 insertions, 38 deletions

diff --git a/tools/testing/nvdimm/Kbuild b/tools/testing/nvdimm/Kbuild
index 405212be044a..d870520da68b 100644
--- a/tools/testing/nvdimm/Kbuild
+++ b/tools/testing/nvdimm/Kbuild
@@ -28,7 +28,10 @@ obj-$(CONFIG_ND_BTT) += nd_btt.o
 obj-$(CONFIG_ND_BLK) += nd_blk.o
 obj-$(CONFIG_X86_PMEM_LEGACY) += nd_e820.o
 obj-$(CONFIG_ACPI_NFIT) += nfit.o
-obj-$(CONFIG_DEV_DAX) += dax.o
+ifeq ($(CONFIG_DAX),m)
+obj-$(CONFIG_DAX) += dax.o
+endif
+obj-$(CONFIG_DEV_DAX) += device_dax.o
 obj-$(CONFIG_DEV_DAX_PMEM) += dax_pmem.o
 
 nfit-y := $(ACPI_SRC)/core.o
@@ -48,9 +51,13 @@ nd_blk-y += config_check.o
 nd_e820-y := $(NVDIMM_SRC)/e820.o
 nd_e820-y += config_check.o
 
-dax-y := $(DAX_SRC)/dax.o
+dax-y := $(DAX_SRC)/super.o
 dax-y += config_check.o
 
+device_dax-y := $(DAX_SRC)/device.o
+device_dax-y += dax-dev.o
+device_dax-y += config_check.o
+
 dax_pmem-y := $(DAX_SRC)/pmem.o
 dax_pmem-y += config_check.o
 
diff --git a/tools/testing/nvdimm/dax-dev.c b/tools/testing/nvdimm/dax-dev.c
new file mode 100644
index 000000000000..36ee3d8797c3
--- /dev/null
+++ b/tools/testing/nvdimm/dax-dev.c
@@ -0,0 +1,49 @@
+/*
+ * Copyright (c) 2016, Intel Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ */
+#include "test/nfit_test.h"
+#include <linux/mm.h>
+#include "../../../drivers/dax/dax-private.h"
+
+phys_addr_t dax_pgoff_to_phys(struct dev_dax *dev_dax, pgoff_t pgoff,
+		unsigned long size)
+{
+	struct resource *res;
+	phys_addr_t addr;
+	int i;
+
+	for (i = 0; i < dev_dax->num_resources; i++) {
+		res = &dev_dax->res[i];
+		addr = pgoff * PAGE_SIZE + res->start;
+		if (addr >= res->start && addr <= res->end)
+			break;
+		pgoff -= PHYS_PFN(resource_size(res));
+	}
+
+	if (i < dev_dax->num_resources) {
+		res = &dev_dax->res[i];
+		if (addr + size - 1 <= res->end) {
+			if (get_nfit_res(addr)) {
+				struct page *page;
+
+				if (dev_dax->region->align > PAGE_SIZE)
+					return -1;
+
+				page = vmalloc_to_page((void *)addr);
+				return PFN_PHYS(page_to_pfn(page));
+			} else
+				return addr;
+		}
+	}
+
+	return -1;
+}
diff --git a/tools/testing/nvdimm/pmem-dax.c b/tools/testing/nvdimm/pmem-dax.c
index c9b8c48f85fc..b53596ad601b 100644
--- a/tools/testing/nvdimm/pmem-dax.c
+++ b/tools/testing/nvdimm/pmem-dax.c
@@ -15,13 +15,13 @@
 #include <pmem.h>
 #include <nd.h>
 
-long pmem_direct_access(struct block_device *bdev, sector_t sector,
-		void **kaddr, pfn_t *pfn, long size)
+long __pmem_direct_access(struct pmem_device *pmem, pgoff_t pgoff,
+		long nr_pages, void **kaddr, pfn_t *pfn)
 {
-	struct pmem_device *pmem = bdev->bd_queue->queuedata;
-	resource_size_t offset = sector * 512 + pmem->data_offset;
+	resource_size_t offset = PFN_PHYS(pgoff) + pmem->data_offset;
 
-	if (unlikely(is_bad_pmem(&pmem->bb, sector, size)))
+	if (unlikely(is_bad_pmem(&pmem->bb, PFN_PHYS(pgoff) / 512,
+					PFN_PHYS(nr_pages))))
 		return -EIO;
 
 	/*
@@ -34,11 +34,10 @@ long pmem_direct_access(struct block_device *bdev, sector_t sector,
 		*kaddr = pmem->virt_addr + offset;
 		page = vmalloc_to_page(pmem->virt_addr + offset);
 		*pfn = page_to_pfn_t(page);
-		dev_dbg_ratelimited(disk_to_dev(bdev->bd_disk)->parent,
-				"%s: sector: %#llx pfn: %#lx\n", __func__,
-				(unsigned long long) sector, page_to_pfn(page));
+		pr_debug_ratelimited("%s: pmem: %p pgoff: %#lx pfn: %#lx\n",
+				__func__, pmem, pgoff, page_to_pfn(page));
 
-		return PAGE_SIZE;
+		return 1;
 	}
 
 	*kaddr = pmem->virt_addr + offset;
@@ -49,6 +48,6 @@ long pmem_direct_access(struct block_device *bdev, sector_t sector,
 	 * requested range.
 	 */
 	if (unlikely(pmem->bb.count))
-		return size;
-	return pmem->size - pmem->pfn_pad - offset;
+		return nr_pages;
+	return PHYS_PFN(pmem->size - pmem->pfn_pad - offset);
 }
diff --git a/tools/testing/nvdimm/test/nfit.c b/tools/testing/nvdimm/test/nfit.c
index 798f17655433..c2187178fb13 100644
--- a/tools/testing/nvdimm/test/nfit.c
+++ b/tools/testing/nvdimm/test/nfit.c
@@ -132,6 +132,7 @@ static u32 handle[] = {
 	[3] = NFIT_DIMM_HANDLE(0, 0, 1, 0, 1),
 	[4] = NFIT_DIMM_HANDLE(0, 1, 0, 0, 0),
 	[5] = NFIT_DIMM_HANDLE(1, 0, 0, 0, 0),
+	[6] = NFIT_DIMM_HANDLE(1, 0, 0, 0, 1),
 };
 
 static unsigned long dimm_fail_cmd_flags[NUM_DCR];
@@ -728,8 +729,8 @@ static int nfit_test0_alloc(struct nfit_test *t)
 static int nfit_test1_alloc(struct nfit_test *t)
 {
 	size_t nfit_size = sizeof(struct acpi_nfit_system_address) * 2
-		+ sizeof(struct acpi_nfit_memory_map)
-		+ offsetof(struct acpi_nfit_control_region, window_size);
+		+ sizeof(struct acpi_nfit_memory_map) * 2
+		+ offsetof(struct acpi_nfit_control_region, window_size) * 2;
 	int i;
 
 	t->nfit_buf = test_alloc(t, nfit_size, &t->nfit_dma);
@@ -906,6 +907,7 @@ static void nfit_test0_setup(struct nfit_test *t)
 	memdev->address = 0;
 	memdev->interleave_index = 0;
 	memdev->interleave_ways = 2;
+	memdev->flags = ACPI_NFIT_MEM_HEALTH_ENABLED;
 
 	/* mem-region2 (spa1, dimm0) */
 	memdev = nfit_buf + offset + sizeof(struct acpi_nfit_memory_map) * 2;
@@ -921,6 +923,7 @@ static void nfit_test0_setup(struct nfit_test *t)
 	memdev->address = SPA0_SIZE/2;
 	memdev->interleave_index = 0;
 	memdev->interleave_ways = 4;
+	memdev->flags = ACPI_NFIT_MEM_HEALTH_ENABLED;
 
 	/* mem-region3 (spa1, dimm1) */
 	memdev = nfit_buf + offset + sizeof(struct acpi_nfit_memory_map) * 3;
@@ -951,6 +954,7 @@ static void nfit_test0_setup(struct nfit_test *t)
 	memdev->address = SPA0_SIZE/2;
 	memdev->interleave_index = 0;
 	memdev->interleave_ways = 4;
+	memdev->flags = ACPI_NFIT_MEM_HEALTH_ENABLED;
 
 	/* mem-region5 (spa1, dimm3) */
 	memdev = nfit_buf + offset + sizeof(struct acpi_nfit_memory_map) * 5;
@@ -1086,6 +1090,7 @@ static void nfit_test0_setup(struct nfit_test *t)
 	memdev->address = 0;
 	memdev->interleave_index = 0;
 	memdev->interleave_ways = 1;
+	memdev->flags = ACPI_NFIT_MEM_HEALTH_ENABLED;
 
 	offset = offset + sizeof(struct acpi_nfit_memory_map) * 14;
 	/* dcr-descriptor0: blk */
@@ -1384,6 +1389,7 @@ static void nfit_test0_setup(struct nfit_test *t)
 		memdev->address = 0;
 		memdev->interleave_index = 0;
 		memdev->interleave_ways = 1;
+		memdev->flags = ACPI_NFIT_MEM_HEALTH_ENABLED;
 
 		/* mem-region16 (spa/bdw4, dimm4) */
 		memdev = nfit_buf + offset +
@@ -1486,6 +1492,34 @@ static void nfit_test1_setup(struct nfit_test *t)
 	dcr->code = NFIT_FIC_BYTE;
 	dcr->windows = 0;
 
+	offset += dcr->header.length;
+	memdev = nfit_buf + offset;
+	memdev->header.type = ACPI_NFIT_TYPE_MEMORY_MAP;
+	memdev->header.length = sizeof(*memdev);
+	memdev->device_handle = handle[6];
+	memdev->physical_id = 0;
+	memdev->region_id = 0;
+	memdev->range_index = 0;
+	memdev->region_index = 0+2;
+	memdev->region_size = SPA2_SIZE;
+	memdev->region_offset = 0;
+	memdev->address = 0;
+	memdev->interleave_index = 0;
+	memdev->interleave_ways = 1;
+	memdev->flags = ACPI_NFIT_MEM_MAP_FAILED;
+
+	/* dcr-descriptor1 */
+	offset += sizeof(*memdev);
+	dcr = nfit_buf + offset;
+	dcr->header.type = ACPI_NFIT_TYPE_CONTROL_REGION;
+	dcr->header.length = offsetof(struct acpi_nfit_control_region,
+			window_size);
+	dcr->region_index = 0+2;
+	dcr_common_init(dcr);
+	dcr->serial_number = ~handle[6];
+	dcr->code = NFIT_FIC_BYTE;
+	dcr->windows = 0;
+
 	post_ars_status(&t->ars_state, t->spa_set_dma[0], SPA2_SIZE);
 
 	acpi_desc = &t->acpi_desc;
@@ -1817,6 +1851,10 @@ static int nfit_test_probe(struct platform_device *pdev)
 	if (rc)
 		return rc;
 
+	rc = devm_add_action_or_reset(&pdev->dev, acpi_nfit_shutdown, acpi_desc);
+	if (rc)
+		return rc;
+
 	if (nfit_test->setup != nfit_test0_setup)
 		return 0;
 
@@ -1907,7 +1945,7 @@ static __init int nfit_test_init(void)
 		case 1:
 			nfit_test->num_pm = 1;
 			nfit_test->dcr_idx = NUM_DCR;
-			nfit_test->num_dcr = 1;
+			nfit_test->num_dcr = 2;
 			nfit_test->alloc = nfit_test1_alloc;
 			nfit_test->setup = nfit_test1_setup;
 			break;
@@ -1924,6 +1962,7 @@ static __init int nfit_test_init(void)
 			put_device(&pdev->dev);
 			goto err_register;
 		}
+		get_device(&pdev->dev);
 
 		rc = dma_coerce_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(64));
 		if (rc)
@@ -1942,6 +1981,10 @@ static __init int nfit_test_init(void)
 		if (instances[i])
 			platform_device_unregister(&instances[i]->pdev);
 	nfit_test_teardown();
+	for (i = 0; i < NUM_NFITS; i++)
+		if (instances[i])
+			put_device(&instances[i]->pdev.dev);
+
 	return rc;
 }
 
@@ -1949,10 +1992,13 @@ static __exit void nfit_test_exit(void)
 {
 	int i;
 
-	platform_driver_unregister(&nfit_test_driver);
 	for (i = 0; i < NUM_NFITS; i++)
 		platform_device_unregister(&instances[i]->pdev);
+	platform_driver_unregister(&nfit_test_driver);
 	nfit_test_teardown();
+
+	for (i = 0; i < NUM_NFITS; i++)
+		put_device(&instances[i]->pdev.dev);
 	class_destroy(nfit_test_dimm);
 }
 
diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile
index d8d94b9bd76c..91edd0566237 100644
--- a/tools/testing/selftests/bpf/Makefile
+++ b/tools/testing/selftests/bpf/Makefile
@@ -13,7 +13,7 @@ LDLIBS += -lcap -lelf
 
 TEST_GEN_PROGS = test_verifier test_tag test_maps test_lru_map test_lpm_map test_progs
 
-TEST_GEN_FILES = test_pkt_access.o test_xdp.o test_l4lb.o
+TEST_GEN_FILES = test_pkt_access.o test_xdp.o test_l4lb.o test_tcp_estats.o
 
 TEST_PROGS := test_kmod.sh
 
@@ -34,6 +34,6 @@ $(BPFOBJ): force
 CLANG ?= clang
 
 %.o: %.c
-	$(CLANG) -I../../../include/uapi -I../../../../samples/bpf/ \
-		-D__x86_64__ -Wno-compare-distinct-pointer-types \
+	$(CLANG) -I. -I../../../include/uapi -I../../../../samples/bpf/ \
+		-Wno-compare-distinct-pointer-types \
 		-O2 -target bpf -c $< -o $@
diff --git a/tools/testing/selftests/bpf/gnu/stubs.h b/tools/testing/selftests/bpf/gnu/stubs.h
new file mode 100644
index 000000000000..719225b16626
--- /dev/null
+++ b/tools/testing/selftests/bpf/gnu/stubs.h
@@ -0,0 +1 @@
+/* dummy .h to trick /usr/include/features.h to work with 'clang -target bpf' */
diff --git a/tools/testing/selftests/bpf/test_progs.c b/tools/testing/selftests/bpf/test_progs.c
index 4ed049a0b14b..b59f5ed4ae40 100644
--- a/tools/testing/selftests/bpf/test_progs.c
+++ b/tools/testing/selftests/bpf/test_progs.c
@@ -268,6 +268,21 @@ out:
 	bpf_object__close(obj);
 }
 
+static void test_tcp_estats(void)
+{
+	const char *file = "./test_tcp_estats.o";
+	int err, prog_fd;
+	struct bpf_object *obj;
+	__u32 duration = 0;
+
+	err = bpf_prog_load(file, BPF_PROG_TYPE_TRACEPOINT, &obj, &prog_fd);
+	CHECK(err, "", "err %d errno %d\n", err, errno);
+	if (err)
+		return;
+
+	bpf_object__close(obj);
+}
+
 int main(void)
 {
 	struct rlimit rinf = { RLIM_INFINITY, RLIM_INFINITY };
@@ -277,6 +292,7 @@ int main(void)
 	test_pkt_access();
 	test_xdp();
 	test_l4lb();
+	test_tcp_estats();
 
 	printf("Summary: %d PASSED, %d FAILED\n", pass_cnt, error_cnt);
 	return 0;
diff --git a/tools/testing/selftests/bpf/test_tcp_estats.c b/tools/testing/selftests/bpf/test_tcp_estats.c
new file mode 100644
index 000000000000..bee3bbecc0c4
--- /dev/null
+++ b/tools/testing/selftests/bpf/test_tcp_estats.c
@@ -0,0 +1,258 @@
+/* Copyright (c) 2017 Facebook
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of version 2 of the GNU General Public
+ * License as published by the Free Software Foundation.
+ */
+
+/* This program shows clang/llvm is able to generate code pattern
+ * like:
+ *   _tcp_send_active_reset:
+ *      0:       bf 16 00 00 00 00 00 00         r6 = r1
+ *    ......
+ *    335:       b7 01 00 00 0f 00 00 00         r1 = 15
+ *    336:       05 00 48 00 00 00 00 00         goto 72
+ *
+ *   LBB0_3:
+ *    337:       b7 01 00 00 01 00 00 00         r1 = 1
+ *    338:       63 1a d0 ff 00 00 00 00         *(u32 *)(r10 - 48) = r1
+ *    408:       b7 01 00 00 03 00 00 00         r1 = 3
+ *
+ *   LBB0_4:
+ *    409:       71 a2 fe ff 00 00 00 00         r2 = *(u8 *)(r10 - 2)
+ *    410:       bf a7 00 00 00 00 00 00         r7 = r10
+ *    411:       07 07 00 00 b8 ff ff ff         r7 += -72
+ *    412:       bf 73 00 00 00 00 00 00         r3 = r7
+ *    413:       0f 13 00 00 00 00 00 00         r3 += r1
+ *    414:       73 23 2d 00 00 00 00 00         *(u8 *)(r3 + 45) = r2
+ *
+ * From the above code snippet, the code generated by the compiler
+ * is reasonable. The "r1" is assigned to different values in basic
+ * blocks "_tcp_send_active_reset" and "LBB0_3", and used in "LBB0_4".
+ * The verifier should be able to handle such code patterns.
+ */
+#include <string.h>
+#include <linux/bpf.h>
+#include <linux/ipv6.h>
+#include <linux/version.h>
+#include <sys/socket.h>
+#include "bpf_helpers.h"
+
+#define _(P) ({typeof(P) val = 0; bpf_probe_read(&val, sizeof(val), &P); val;})
+#define TCP_ESTATS_MAGIC 0xBAADBEEF
+
+/* This test case needs "sock" and "pt_regs" data structure.
+ * Recursively, "sock" needs "sock_common" and "inet_sock".
+ * However, this is a unit test case only for
+ * verifier purpose without bpf program execution.
+ * We can safely mock much simpler data structures, basically
+ * only taking the necessary fields from kernel headers.
+ */
+typedef __u32 __bitwise __portpair;
+typedef __u64 __bitwise __addrpair;
+
+struct sock_common {
+	unsigned short		skc_family;
+	union {
+		__addrpair	skc_addrpair;
+		struct {
+			__be32	skc_daddr;
+			__be32	skc_rcv_saddr;
+		};
+	};
+	union {
+		__portpair	skc_portpair;
+		struct {
+			__be16	skc_dport;
+			__u16	skc_num;
+		};
+	};
+	struct in6_addr		skc_v6_daddr;
+	struct in6_addr		skc_v6_rcv_saddr;
+};
+
+struct sock {
+	struct sock_common	__sk_common;
+#define sk_family		__sk_common.skc_family
+#define sk_v6_daddr		__sk_common.skc_v6_daddr
+#define sk_v6_rcv_saddr		__sk_common.skc_v6_rcv_saddr
+};
+
+struct inet_sock {
+	struct sock		sk;
+#define inet_daddr		sk.__sk_common.skc_daddr
+#define inet_dport		sk.__sk_common.skc_dport
+	__be32			inet_saddr;
+	__be16			inet_sport;
+};
+
+struct pt_regs {
+	long di;
+};
+
+static inline struct inet_sock *inet_sk(const struct sock *sk)
+{
+	return (struct inet_sock *)sk;
+}
+
+/* Define various data structures for state recording.
+ * Some fields are not used due to test simplification.
+ */
+enum tcp_estats_addrtype {
+	TCP_ESTATS_ADDRTYPE_IPV4 = 1,
+	TCP_ESTATS_ADDRTYPE_IPV6 = 2
+};
+
+enum tcp_estats_event_type {
+	TCP_ESTATS_ESTABLISH,
+	TCP_ESTATS_PERIODIC,
+	TCP_ESTATS_TIMEOUT,
+	TCP_ESTATS_RETRANSMIT_TIMEOUT,
+	TCP_ESTATS_RETRANSMIT_OTHER,
+	TCP_ESTATS_SYN_RETRANSMIT,
+	TCP_ESTATS_SYNACK_RETRANSMIT,
+	TCP_ESTATS_TERM,
+	TCP_ESTATS_TX_RESET,
+	TCP_ESTATS_RX_RESET,
+	TCP_ESTATS_WRITE_TIMEOUT,
+	TCP_ESTATS_CONN_TIMEOUT,
+	TCP_ESTATS_ACK_LATENCY,
+	TCP_ESTATS_NEVENTS,
+};
+
+struct tcp_estats_event {
+	int pid;
+	int cpu;
+	unsigned long ts;
+	unsigned int magic;
+	enum tcp_estats_event_type event_type;
+};
+
+/* The below data structure is packed in order for
+ * llvm compiler to generate expected code.
+ */
+struct tcp_estats_conn_id {
+	unsigned int localaddressType;
+	struct {
+		unsigned char data[16];
+	} localaddress;
+	struct {
+		unsigned char data[16];
+	} remaddress;
+	unsigned short    localport;
+	unsigned short    remport;
+} __attribute__((__packed__));
+
+struct tcp_estats_basic_event {
+	struct tcp_estats_event event;
+	struct tcp_estats_conn_id conn_id;
+};
+
+struct bpf_map_def SEC("maps") ev_record_map = {
+	.type = BPF_MAP_TYPE_HASH,
+	.key_size = sizeof(__u32),
+	.value_size = sizeof(struct tcp_estats_basic_event),
+	.max_entries = 1024,
+};
+
+struct dummy_tracepoint_args {
+	unsigned long long pad;
+	struct sock *sock;
+};
+
+static __always_inline void tcp_estats_ev_init(struct tcp_estats_event *event,
+					       enum tcp_estats_event_type type)
+{
+	event->magic = TCP_ESTATS_MAGIC;
+	event->ts = bpf_ktime_get_ns();
+	event->event_type = type;
+}
+
+static __always_inline void unaligned_u32_set(unsigned char *to, __u8 *from)
+{
+	to[0] = _(from[0]);
+	to[1] = _(from[1]);
+	to[2] = _(from[2]);
+	to[3] = _(from[3]);
+}
+
+static __always_inline void conn_id_ipv4_init(struct tcp_estats_conn_id *conn_id,
+					      __be32 *saddr, __be32 *daddr)
+{
+	conn_id->localaddressType = TCP_ESTATS_ADDRTYPE_IPV4;
+
+	unaligned_u32_set(conn_id->localaddress.data, (__u8 *)saddr);
+	unaligned_u32_set(conn_id->remaddress.data, (__u8 *)daddr);
+}
+
+static __always_inline void conn_id_ipv6_init(struct tcp_estats_conn_id *conn_id,
+					      __be32 *saddr, __be32 *daddr)
+{
+	conn_id->localaddressType = TCP_ESTATS_ADDRTYPE_IPV6;
+
+	unaligned_u32_set(conn_id->localaddress.data, (__u8 *)saddr);
+	unaligned_u32_set(conn_id->localaddress.data + sizeof(__u32),
+			  (__u8 *)(saddr + 1));
+	unaligned_u32_set(conn_id->localaddress.data + sizeof(__u32) * 2,
+			  (__u8 *)(saddr + 2));
+	unaligned_u32_set(conn_id->localaddress.data + sizeof(__u32) * 3,
+			  (__u8 *)(saddr + 3));
+
+	unaligned_u32_set(conn_id->remaddress.data,
+			  (__u8 *)(daddr));
+	unaligned_u32_set(conn_id->remaddress.data + sizeof(__u32),
+			  (__u8 *)(daddr + 1));
+	unaligned_u32_set(conn_id->remaddress.data + sizeof(__u32) * 2,
+			  (__u8 *)(daddr + 2));
+	unaligned_u32_set(conn_id->remaddress.data + sizeof(__u32) * 3,
+			  (__u8 *)(daddr + 3));
+}
+
+static __always_inline void tcp_estats_conn_id_init(struct tcp_estats_conn_id *conn_id,
+						    struct sock *sk)
+{
+	conn_id->localport = _(inet_sk(sk)->inet_sport);
+	conn_id->remport = _(inet_sk(sk)->inet_dport);
+
+	if (_(sk->sk_family) == AF_INET6)
+		conn_id_ipv6_init(conn_id,
+				  sk->sk_v6_rcv_saddr.s6_addr32,
+				  sk->sk_v6_daddr.s6_addr32);
+	else
+		conn_id_ipv4_init(conn_id,
+				  &inet_sk(sk)->inet_saddr,
+				  &inet_sk(sk)->inet_daddr);
+}
+
+static __always_inline void tcp_estats_init(struct sock *sk,
+					    struct tcp_estats_event *event,
+					    struct tcp_estats_conn_id *conn_id,
+					    enum tcp_estats_event_type type)
+{
+	tcp_estats_ev_init(event, type);
+	tcp_estats_conn_id_init(conn_id, sk);
+}
+
+static __always_inline void send_basic_event(struct sock *sk,
+					     enum tcp_estats_event_type type)
+{
+	struct tcp_estats_basic_event ev;
+	__u32 key = bpf_get_prandom_u32();
+
+	memset(&ev, 0, sizeof(ev));
+	tcp_estats_init(sk, &ev.event, &ev.conn_id, type);
+	bpf_map_update_elem(&ev_record_map, &key, &ev, BPF_ANY);
+}
+
+SEC("dummy_tracepoint")
+int _dummy_tracepoint(struct dummy_tracepoint_args *arg)
+{
+	if (!arg->sock)
+		return 0;
+
+	send_basic_event(arg->sock, TCP_ESTATS_TX_RESET);
+	return 0;
+}
+
+char _license[] SEC("license") = "GPL";
+__u32 _version SEC("version") = 1; /* ignored by tracepoints, required by libbpf.a */
diff --git a/tools/testing/selftests/powerpc/Makefile b/tools/testing/selftests/powerpc/Makefile
index 2132ff8eb4e7..03e1617367d3 100644
--- a/tools/testing/selftests/powerpc/Makefile
+++ b/tools/testing/selftests/powerpc/Makefile
@@ -14,6 +14,7 @@ export CFLAGS
 
 SUB_DIRS = alignment		\
 	   benchmarks		\
+	   cache_shape		\
 	   copyloops		\
 	   context_switch	\
 	   dscr			\
diff --git a/tools/testing/selftests/powerpc/cache_shape/.gitignore b/tools/testing/selftests/powerpc/cache_shape/.gitignore
new file mode 100644
index 000000000000..ec1848434be5
--- /dev/null
+++ b/tools/testing/selftests/powerpc/cache_shape/.gitignore
@@ -0,0 +1 @@
+cache_shape
diff --git a/tools/testing/selftests/powerpc/cache_shape/Makefile b/tools/testing/selftests/powerpc/cache_shape/Makefile
new file mode 100644
index 000000000000..b24485ab30e2
--- /dev/null
+++ b/tools/testing/selftests/powerpc/cache_shape/Makefile
@@ -0,0 +1,10 @@
+TEST_PROGS := cache_shape
+
+all: $(TEST_PROGS)
+
+$(TEST_PROGS): ../harness.c ../utils.c
+
+include ../../lib.mk
+
+clean:
+	rm -f $(TEST_PROGS) *.o
diff --git a/tools/testing/selftests/powerpc/cache_shape/cache_shape.c b/tools/testing/selftests/powerpc/cache_shape/cache_shape.c
new file mode 100644
index 000000000000..29ec07eba7f9
--- /dev/null
+++ b/tools/testing/selftests/powerpc/cache_shape/cache_shape.c
@@ -0,0 +1,125 @@
+/*
+ * Copyright 2017, Michael Ellerman, IBM Corp.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version
+ * 2 of the License, or (at your option) any later version.
+ */
+
+#include <elf.h>
+#include <errno.h>
+#include <fcntl.h>
+#include <link.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <sys/stat.h>
+#include <sys/types.h>
+#include <sys/wait.h>
+#include <unistd.h>
+
+#include "utils.h"
+
+#ifndef AT_L1I_CACHESIZE
+#define AT_L1I_CACHESIZE	40
+#define AT_L1I_CACHEGEOMETRY	41
+#define AT_L1D_CACHESIZE	42
+#define AT_L1D_CACHEGEOMETRY	43
+#define AT_L2_CACHESIZE		44
+#define AT_L2_CACHEGEOMETRY	45
+#define AT_L3_CACHESIZE		46
+#define AT_L3_CACHEGEOMETRY	47
+#endif
+
+static void print_size(const char *label, uint32_t val)
+{
+	printf("%s cache size: %#10x %10dB %10dK\n", label, val, val, val / 1024);
+}
+
+static void print_geo(const char *label, uint32_t val)
+{
+	uint16_t assoc;
+
+	printf("%s line size:  %#10x       ", label, val & 0xFFFF);
+
+	assoc = val >> 16;
+	if (assoc)
+		printf("%u-way", assoc);
+	else
+		printf("fully");
+
+	printf(" associative\n");
+}
+
+static int test_cache_shape()
+{
+	static char buffer[4096];
+	ElfW(auxv_t) *p;
+	int found;
+
+	FAIL_IF(read_auxv(buffer, sizeof(buffer)));
+
+	found = 0;
+
+	p = find_auxv_entry(AT_L1I_CACHESIZE, buffer);
+	if (p) {
+		found++;
+		print_size("L1I ", (uint32_t)p->a_un.a_val);
+	}
+
+	p = find_auxv_entry(AT_L1I_CACHEGEOMETRY, buffer);
+	if (p) {
+		found++;
+		print_geo("L1I ", (uint32_t)p->a_un.a_val);
+	}
+
+	p = find_auxv_entry(AT_L1D_CACHESIZE, buffer);
+	if (p) {
+		found++;
+		print_size("L1D ", (uint32_t)p->a_un.a_val);
+	}
+
+	p = find_auxv_entry(AT_L1D_CACHEGEOMETRY, buffer);
+	if (p) {
+		found++;
+		print_geo("L1D ", (uint32_t)p->a_un.a_val);
+	}
+
+	p = find_auxv_entry(AT_L2_CACHESIZE, buffer);
+	if (p) {
+		found++;
+		print_size("L2  ", (uint32_t)p->a_un.a_val);
+	}
+
+	p = find_auxv_entry(AT_L2_CACHEGEOMETRY, buffer);
+	if (p) {
+		found++;
+		print_geo("L2  ", (uint32_t)p->a_un.a_val);
+	}
+
+	p = find_auxv_entry(AT_L3_CACHESIZE, buffer);
+	if (p) {
+		found++;
+		print_size("L3  ", (uint32_t)p->a_un.a_val);
+	}
+
+	p = find_auxv_entry(AT_L3_CACHEGEOMETRY, buffer);
+	if (p) {
+		found++;
+		print_geo("L3  ", (uint32_t)p->a_un.a_val);
+	}
+
+	/* If we found none we're probably on a system where they don't exist */
+	SKIP_IF(found == 0);
+
+	/* But if we found any, we expect to find them all */
+	FAIL_IF(found != 8);
+
+	return 0;
+}
+
+int main(void)
+{
+	return test_harness(test_cache_shape, "cache_shape");
+}
diff --git a/tools/testing/selftests/powerpc/include/utils.h b/tools/testing/selftests/powerpc/include/utils.h
index 53405e8a52ab..735815b3ad7f 100644
--- a/tools/testing/selftests/powerpc/include/utils.h
+++ b/tools/testing/selftests/powerpc/include/utils.h
@@ -24,7 +24,11 @@ typedef uint8_t u8;
 
 void test_harness_set_timeout(uint64_t time);
 int test_harness(int (test_function)(void), char *name);
-extern void *get_auxv_entry(int type);
+
+int read_auxv(char *buf, ssize_t buf_size);
+void *find_auxv_entry(int type, char *auxv);
+void *get_auxv_entry(int type);
+
 int pick_online_cpu(void);
 
 static inline bool have_hwcap(unsigned long ftr)
diff --git a/tools/testing/selftests/powerpc/utils.c b/tools/testing/selftests/powerpc/utils.c
index dcf74184bfd0..d46916867a6f 100644
--- a/tools/testing/selftests/powerpc/utils.c
+++ b/tools/testing/selftests/powerpc/utils.c
@@ -19,45 +19,64 @@
 
 static char auxv[4096];
 
-void *get_auxv_entry(int type)
+int read_auxv(char *buf, ssize_t buf_size)
 {
-	ElfW(auxv_t) *p;
-	void *result;
 	ssize_t num;
-	int fd;
+	int rc, fd;
 
 	fd = open("/proc/self/auxv", O_RDONLY);
 	if (fd == -1) {
 		perror("open");
-		return NULL;
+		return -errno;
 	}
 
-	result = NULL;
-
-	num = read(fd, auxv, sizeof(auxv));
+	num = read(fd, buf, buf_size);
 	if (num < 0) {
 		perror("read");
+		rc = -EIO;
 		goto out;
 	}
 
-	if (num > sizeof(auxv)) {
-		printf("Overflowed auxv buffer\n");
+	if (num > buf_size) {
+		printf("overflowed auxv buffer\n");
+		rc = -EOVERFLOW;
 		goto out;
 	}
 
+	rc = 0;
+out:
+	close(fd);
+	return rc;
+}
+
+void *find_auxv_entry(int type, char *auxv)
+{
+	ElfW(auxv_t) *p;
+
 	p = (ElfW(auxv_t) *)auxv;
 
 	while (p->a_type != AT_NULL) {
-		if (p->a_type == type) {
-			result = (void *)p->a_un.a_val;
-			break;
-		}
+		if (p->a_type == type)
+			return p;
 
 		p++;
 	}
-out:
-	close(fd);
-	return result;
+
+	return NULL;
+}
+
+void *get_auxv_entry(int type)
+{
+	ElfW(auxv_t) *p;
+
+	if (read_auxv(auxv, sizeof(auxv)))
+		return NULL;
+
+	p = find_auxv_entry(type, auxv);
+	if (p)
+		return (void *)p->a_un.a_val;
+
+	return NULL;
 }
 
 int pick_online_cpu(void)
diff --git a/tools/testing/selftests/vm/Makefile b/tools/testing/selftests/vm/Makefile
index dba889004ea1..cbb29e41ef2b 100644
--- a/tools/testing/selftests/vm/Makefile
+++ b/tools/testing/selftests/vm/Makefile
@@ -16,6 +16,7 @@ TEST_GEN_FILES += thuge-gen
 TEST_GEN_FILES += transhuge-stress
 TEST_GEN_FILES += userfaultfd
 TEST_GEN_FILES += mlock-random-test
+TEST_GEN_FILES += virtual_address_range
 
 TEST_PROGS := run_vmtests
 
diff --git a/tools/testing/selftests/vm/run_vmtests b/tools/testing/selftests/vm/run_vmtests
index 3214a6456d13..0640923ded7e 100755
--- a/tools/testing/selftests/vm/run_vmtests
+++ b/tools/testing/selftests/vm/run_vmtests
@@ -165,4 +165,15 @@ else
 	echo "[PASS]"
 fi
 
+echo "-----------------------------"
+echo "running virtual_address_range"
+echo "-----------------------------"
+./virtual_address_range
+if [ $? -ne 0 ]; then
+	echo "[FAIL]"
+	exitcode=1
+else
+	echo "[PASS]"
+fi
+
 exit $exitcode
diff --git a/tools/testing/selftests/vm/virtual_address_range.c b/tools/testing/selftests/vm/virtual_address_range.c
new file mode 100644
index 000000000000..3b02aa6eb9da
--- /dev/null
+++ b/tools/testing/selftests/vm/virtual_address_range.c
@@ -0,0 +1,122 @@
+/*
+ * Copyright 2017, Anshuman Khandual, IBM Corp.
+ * Licensed under GPLv2.
+ *
+ * Works on architectures which support 128TB virtual
+ * address range and beyond.
+ */
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <unistd.h>
+#include <errno.h>
+#include <numaif.h>
+#include <sys/mman.h>
+#include <sys/time.h>
+
+/*
+ * Maximum address range mapped with a single mmap()
+ * call is little bit more than 16GB. Hence 16GB is
+ * chosen as the single chunk size for address space
+ * mapping.
+ */
+#define MAP_CHUNK_SIZE   17179869184UL /* 16GB */
+
+/*
+ * Address space till 128TB is mapped without any hint
+ * and is enabled by default. Address space beyond 128TB
+ * till 512TB is obtained by passing hint address as the
+ * first argument into mmap() system call.
+ *
+ * The process heap address space is divided into two
+ * different areas one below 128TB and one above 128TB
+ * till it reaches 512TB. One with size 128TB and the
+ * other being 384TB.
+ */
+#define NR_CHUNKS_128TB   8192UL /* Number of 16GB chunks for 128TB */
+#define NR_CHUNKS_384TB  24576UL /* Number of 16GB chunks for 384TB */
+
+#define ADDR_MARK_128TB  (1UL << 47) /* First address beyond 128TB */
+
+static char *hind_addr(void)
+{
+	int bits = 48 + rand() % 15;
+
+	return (char *) (1UL << bits);
+}
+
+static int validate_addr(char *ptr, int high_addr)
+{
+	unsigned long addr = (unsigned long) ptr;
+
+	if (high_addr) {
+		if (addr < ADDR_MARK_128TB) {
+			printf("Bad address %lx\n", addr);
+			return 1;
+		}
+		return 0;
+	}
+
+	if (addr > ADDR_MARK_128TB) {
+		printf("Bad address %lx\n", addr);
+		return 1;
+	}
+	return 0;
+}
+
+static int validate_lower_address_hint(void)
+{
+	char *ptr;
+
+	ptr = mmap((void *) (1UL << 45), MAP_CHUNK_SIZE, PROT_READ |
+			PROT_WRITE, MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
+
+	if (ptr == MAP_FAILED)
+		return 0;
+
+	return 1;
+}
+
+int main(int argc, char *argv[])
+{
+	char *ptr[NR_CHUNKS_128TB];
+	char *hptr[NR_CHUNKS_384TB];
+	char *hint;
+	unsigned long i, lchunks, hchunks;
+
+	for (i = 0; i < NR_CHUNKS_128TB; i++) {
+		ptr[i] = mmap(NULL, MAP_CHUNK_SIZE, PROT_READ | PROT_WRITE,
+					MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
+
+		if (ptr[i] == MAP_FAILED) {
+			if (validate_lower_address_hint())
+				return 1;
+			break;
+		}
+
+		if (validate_addr(ptr[i], 0))
+			return 1;
+	}
+	lchunks = i;
+
+	for (i = 0; i < NR_CHUNKS_384TB; i++) {
+		hint = hind_addr();
+		hptr[i] = mmap(hint, MAP_CHUNK_SIZE, PROT_READ | PROT_WRITE,
+					MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
+
+		if (hptr[i] == MAP_FAILED)
+			break;
+
+		if (validate_addr(hptr[i], 1))
+			return 1;
+	}
+	hchunks = i;
+
+	for (i = 0; i < lchunks; i++)
+		munmap(ptr[i], MAP_CHUNK_SIZE);
+
+	for (i = 0; i < hchunks; i++)
+		munmap(hptr[i], MAP_CHUNK_SIZE);
+
+	return 0;
+}