Merge tag 'powerpc-5.7-1' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux

Pull powerpc updates from Michael Ellerman:
 "Slightly late as I had to rebase mid-week to insert a bug fix:

   - A large series from Nick for 64-bit to further rework our exception
     vectors, and rewrite portions of the syscall entry/exit and
     interrupt return in C. The result is much easier to follow code
     that is also faster in general.

   - Cleanup of our ptrace code to split various parts out that had
     become badly intertwined with #ifdefs over the years.

   - Changes to our NUMA setup under the PowerVM hypervisor which should
     hopefully avoid non-sensical topologies which can lead to warnings
     from the workqueue code and other problems.

   - MAINTAINERS updates to remove some of our old orphan entries and
     update the status of others.

   - Quite a few other small changes and fixes all over the map.

  Thanks to: Abdul Haleem, afzal mohammed, Alexey Kardashevskiy, Andrew
  Donnellan, Aneesh Kumar K.V, Balamuruhan S, Cédric Le Goater, Chen
  Zhou, Christophe JAILLET, Christophe Leroy, Christoph Hellwig, Clement
  Courbet, Daniel Axtens, David Gibson, Douglas Miller, Fabiano Rosas,
  Fangrui Song, Ganesh Goudar, Gautham R. Shenoy, Greg Kroah-Hartman,
  Greg Kurz, Gustavo Luiz Duarte, Hari Bathini, Ilie Halip, Jan Kara,
  Joe Lawrence, Joe Perches, Kajol Jain, Larry Finger, Laurentiu Tudor,
  Leonardo Bras, Libor Pechacek, Madhavan Srinivasan, Mahesh Salgaonkar,
  Masahiro Yamada, Masami Hiramatsu, Mauricio Faria de Oliveira, Michael
  Neuling, Michal Suchanek, Mike Rapoport, Nageswara R Sastry, Nathan
  Chancellor, Nathan Lynch, Naveen N. Rao, Nicholas Piggin, Nick
  Desaulniers, Oliver O'Halloran, Po-Hsu Lin, Pratik Rajesh Sampat,
  Rasmus Villemoes, Ravi Bangoria, Roman Bolshakov, Sam Bobroff,
  Sandipan Das, Santosh S, Sedat Dilek, Segher Boessenkool, Shilpasri G
  Bhat, Sourabh Jain, Srikar Dronamraju, Stephen Rothwell, Tyrel
  Datwyler, Vaibhav Jain, YueHaibing"

* tag 'powerpc-5.7-1' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux: (158 commits)
  powerpc: Make setjmp/longjmp signature standard
  powerpc/cputable: Remove unnecessary copy of cpu_spec->oprofile_type
  powerpc: Suppress .eh_frame generation
  powerpc: Drop -fno-dwarf2-cfi-asm
  powerpc/32: drop unused ISA_DMA_THRESHOLD
  powerpc/powernv: Add documentation for the opal sensor_groups sysfs interfaces
  selftests/powerpc: Fix try-run when source tree is not writable
  powerpc/vmlinux.lds: Explicitly retain .gnu.hash
  powerpc/ptrace: move ptrace_triggered() into hw_breakpoint.c
  powerpc/ptrace: create ppc_gethwdinfo()
  powerpc/ptrace: create ptrace_get_debugreg()
  powerpc/ptrace: split out ADV_DEBUG_REGS related functions.
  powerpc/ptrace: move register viewing functions out of ptrace.c
  powerpc/ptrace: split out TRANSACTIONAL_MEM related functions.
  powerpc/ptrace: split out SPE related functions.
  powerpc/ptrace: split out ALTIVEC related functions.
  powerpc/ptrace: split out VSX related functions.
  powerpc/ptrace: drop PARAMETER_SAVE_AREA_OFFSET
  powerpc/ptrace: drop unnecessary #ifdefs CONFIG_PPC64
  powerpc/ptrace: remove unused header includes
  ...

15 files changed, 465 insertions, 41 deletions

diff --git a/tools/testing/selftests/powerpc/benchmarks/Makefile b/tools/testing/selftests/powerpc/benchmarks/Makefile
index d40300a65b42..a32a6ab89914 100644
--- a/tools/testing/selftests/powerpc/benchmarks/Makefile
+++ b/tools/testing/selftests/powerpc/benchmarks/Makefile
@@ -2,6 +2,8 @@
 TEST_GEN_PROGS := gettimeofday context_switch fork mmap_bench futex_bench null_syscall
 TEST_GEN_FILES := exec_target
 
+TEST_FILES := settings
+
 CFLAGS += -O2
 
 top_srcdir = ../../../../..
diff --git a/tools/testing/selftests/powerpc/benchmarks/settings b/tools/testing/selftests/powerpc/benchmarks/settings
new file mode 100644
index 000000000000..e7b9417537fb
--- /dev/null
+++ b/tools/testing/selftests/powerpc/benchmarks/settings
@@ -0,0 +1 @@
+timeout=0
diff --git a/tools/testing/selftests/powerpc/dscr/Makefile b/tools/testing/selftests/powerpc/dscr/Makefile
index 5df476364b4d..cfa6eedcb66c 100644
--- a/tools/testing/selftests/powerpc/dscr/Makefile
+++ b/tools/testing/selftests/powerpc/dscr/Makefile
@@ -3,6 +3,8 @@ TEST_GEN_PROGS := dscr_default_test dscr_explicit_test dscr_user_test	\
 	      dscr_inherit_test dscr_inherit_exec_test dscr_sysfs_test	\
 	      dscr_sysfs_thread_test
 
+TEST_FILES := settings
+
 top_srcdir = ../../../../..
 include ../../lib.mk
 
diff --git a/tools/testing/selftests/powerpc/dscr/settings b/tools/testing/selftests/powerpc/dscr/settings
new file mode 100644
index 000000000000..e7b9417537fb
--- /dev/null
+++ b/tools/testing/selftests/powerpc/dscr/settings
@@ -0,0 +1 @@
+timeout=0
diff --git a/tools/testing/selftests/powerpc/mm/.gitignore b/tools/testing/selftests/powerpc/mm/.gitignore
index 7cf7ad261d02..2ca523255b1b 100644
--- a/tools/testing/selftests/powerpc/mm/.gitignore
+++ b/tools/testing/selftests/powerpc/mm/.gitignore
@@ -7,3 +7,4 @@ segv_errors
 wild_bctr
 large_vm_fork_separation
 bad_accesses
+tlbie_test
diff --git a/tools/testing/selftests/powerpc/pmu/ebb/Makefile b/tools/testing/selftests/powerpc/pmu/ebb/Makefile
index 417306353e07..ca35dd8848b0 100644
--- a/tools/testing/selftests/powerpc/pmu/ebb/Makefile
+++ b/tools/testing/selftests/powerpc/pmu/ebb/Makefile
@@ -7,6 +7,7 @@ noarg:
 # The EBB handler is 64-bit code and everything links against it
 CFLAGS += -m64
 
+TMPOUT = $(OUTPUT)/
 # Toolchains may build PIE by default which breaks the assembly
 no-pie-option := $(call try-run, echo 'int main() { return 0; }' | \
         $(CC) -Werror $(KBUILD_CPPFLAGS) $(CC_OPTION_CFLAGS) -no-pie -x c - -o "$$TMP", -no-pie)
diff --git a/tools/testing/selftests/powerpc/signal/.gitignore b/tools/testing/selftests/powerpc/signal/.gitignore
index f897b55a44dd..405b5364044c 100644
--- a/tools/testing/selftests/powerpc/signal/.gitignore
+++ b/tools/testing/selftests/powerpc/signal/.gitignore
@@ -2,3 +2,4 @@
 signal
 signal_tm
 sigfuz
+sigreturn_vdso
diff --git a/tools/testing/selftests/powerpc/signal/Makefile b/tools/testing/selftests/powerpc/signal/Makefile
index 113838fbbe7f..932a032bf036 100644
--- a/tools/testing/selftests/powerpc/signal/Makefile
+++ b/tools/testing/selftests/powerpc/signal/Makefile
@@ -1,10 +1,12 @@
 # SPDX-License-Identifier: GPL-2.0
-TEST_GEN_PROGS := signal signal_tm sigfuz
+TEST_GEN_PROGS := signal signal_tm sigfuz sigreturn_vdso
 
 CFLAGS += -maltivec
 $(OUTPUT)/signal_tm: CFLAGS += -mhtm
 $(OUTPUT)/sigfuz: CFLAGS += -pthread -m64
 
+TEST_FILES := settings
+
 top_srcdir = ../../../../..
 include ../../lib.mk
 
diff --git a/tools/testing/selftests/powerpc/signal/settings b/tools/testing/selftests/powerpc/signal/settings
new file mode 100644
index 000000000000..e7b9417537fb
--- /dev/null
+++ b/tools/testing/selftests/powerpc/signal/settings
@@ -0,0 +1 @@
+timeout=0
diff --git a/tools/testing/selftests/powerpc/signal/sigreturn_vdso.c b/tools/testing/selftests/powerpc/signal/sigreturn_vdso.c
new file mode 100644
index 000000000000..e282fff0fe25
--- /dev/null
+++ b/tools/testing/selftests/powerpc/signal/sigreturn_vdso.c
@@ -0,0 +1,127 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Test that we can take signals with and without the VDSO mapped, which trigger
+ * different paths in the signal handling code.
+ *
+ * See handle_rt_signal64() and setup_trampoline() in signal_64.c
+ */
+
+#define _GNU_SOURCE
+
+#include <errno.h>
+#include <stdio.h>
+#include <signal.h>
+#include <stdlib.h>
+#include <string.h>
+#include <sys/mman.h>
+#include <sys/types.h>
+#include <unistd.h>
+
+// Ensure assert() is not compiled out
+#undef NDEBUG
+#include <assert.h>
+
+#include "utils.h"
+
+static int search_proc_maps(char *needle, unsigned long *low, unsigned long *high)
+{
+	unsigned long start, end;
+	static char buf[4096];
+	char name[128];
+	FILE *f;
+	int rc = -1;
+
+	f = fopen("/proc/self/maps", "r");
+	if (!f) {
+		perror("fopen");
+		return -1;
+	}
+
+	while (fgets(buf, sizeof(buf), f)) {
+		rc = sscanf(buf, "%lx-%lx %*c%*c%*c%*c %*x %*d:%*d %*d %127s\n",
+			    &start, &end, name);
+		if (rc == 2)
+			continue;
+
+		if (rc != 3) {
+			printf("sscanf errored\n");
+			rc = -1;
+			break;
+		}
+
+		if (strstr(name, needle)) {
+			*low = start;
+			*high = end - 1;
+			rc = 0;
+			break;
+		}
+	}
+
+	fclose(f);
+
+	return rc;
+}
+
+static volatile sig_atomic_t took_signal = 0;
+
+static void sigusr1_handler(int sig)
+{
+	took_signal++;
+}
+
+int test_sigreturn_vdso(void)
+{
+	unsigned long low, high, size;
+	struct sigaction act;
+	char *p;
+
+	act.sa_handler = sigusr1_handler;
+	act.sa_flags = 0;
+	sigemptyset(&act.sa_mask);
+
+	assert(sigaction(SIGUSR1, &act, NULL) == 0);
+
+	// Confirm the VDSO is mapped, and work out where it is
+	assert(search_proc_maps("[vdso]", &low, &high) == 0);
+	size = high - low + 1;
+	printf("VDSO is at 0x%lx-0x%lx (%lu bytes)\n", low, high, size);
+
+	kill(getpid(), SIGUSR1);
+	assert(took_signal == 1);
+	printf("Signal delivered OK with VDSO mapped\n");
+
+	// Remap the VDSO somewhere else
+	p = mmap(NULL, size, PROT_READ|PROT_WRITE, MAP_ANONYMOUS|MAP_PRIVATE, -1, 0);
+	assert(p != MAP_FAILED);
+	assert(mremap((void *)low, size, size, MREMAP_MAYMOVE|MREMAP_FIXED, p) != MAP_FAILED);
+	assert(search_proc_maps("[vdso]", &low, &high) == 0);
+	size = high - low + 1;
+	printf("VDSO moved to 0x%lx-0x%lx (%lu bytes)\n", low, high, size);
+
+	kill(getpid(), SIGUSR1);
+	assert(took_signal == 2);
+	printf("Signal delivered OK with VDSO moved\n");
+
+	assert(munmap((void *)low, size) == 0);
+	printf("Unmapped VDSO\n");
+
+	// Confirm the VDSO is not mapped anymore
+	assert(search_proc_maps("[vdso]", &low, &high) != 0);
+
+	// Make the stack executable
+	assert(search_proc_maps("[stack]", &low, &high) == 0);
+	size = high - low + 1;
+	mprotect((void *)low, size, PROT_READ|PROT_WRITE|PROT_EXEC);
+	printf("Remapped the stack executable\n");
+
+	kill(getpid(), SIGUSR1);
+	assert(took_signal == 3);
+	printf("Signal delivered OK with VDSO unmapped\n");
+
+	return 0;
+}
+
+int main(void)
+{
+	return test_harness(test_sigreturn_vdso, "sigreturn_vdso");
+}
diff --git a/tools/testing/selftests/powerpc/tm/.gitignore b/tools/testing/selftests/powerpc/tm/.gitignore
index 7baf2a46002f..d8900a0c47a1 100644
--- a/tools/testing/selftests/powerpc/tm/.gitignore
+++ b/tools/testing/selftests/powerpc/tm/.gitignore
@@ -14,6 +14,7 @@ tm-signal-context-chk-vmx
 tm-signal-context-chk-vsx
 tm-signal-context-force-tm
 tm-signal-sigreturn-nt
+tm-signal-pagefault
 tm-vmx-unavail
 tm-unavailable
 tm-trap
diff --git a/tools/testing/selftests/powerpc/tm/Makefile b/tools/testing/selftests/powerpc/tm/Makefile
index b15a1a325bd0..0b0db8d3857c 100644
--- a/tools/testing/selftests/powerpc/tm/Makefile
+++ b/tools/testing/selftests/powerpc/tm/Makefile
@@ -5,7 +5,9 @@ SIGNAL_CONTEXT_CHK_TESTS := tm-signal-context-chk-gpr tm-signal-context-chk-fpu
 TEST_GEN_PROGS := tm-resched-dscr tm-syscall tm-signal-msr-resv tm-signal-stack \
 	tm-vmxcopy tm-fork tm-tar tm-tmspr tm-vmx-unavail tm-unavailable tm-trap \
 	$(SIGNAL_CONTEXT_CHK_TESTS) tm-sigreturn tm-signal-sigreturn-nt \
-	tm-signal-context-force-tm tm-poison
+	tm-signal-context-force-tm tm-poison tm-signal-pagefault
+
+TEST_FILES := settings
 
 top_srcdir = ../../../../..
 include ../../lib.mk
@@ -22,6 +24,7 @@ $(OUTPUT)/tm-resched-dscr: ../pmu/lib.c
 $(OUTPUT)/tm-unavailable: CFLAGS += -O0 -pthread -m64 -Wno-error=uninitialized -mvsx
 $(OUTPUT)/tm-trap: CFLAGS += -O0 -pthread -m64
 $(OUTPUT)/tm-signal-context-force-tm: CFLAGS += -pthread -m64
+$(OUTPUT)/tm-signal-pagefault: CFLAGS += -pthread -m64
 
 SIGNAL_CONTEXT_CHK_TESTS := $(patsubst %,$(OUTPUT)/%,$(SIGNAL_CONTEXT_CHK_TESTS))
 $(SIGNAL_CONTEXT_CHK_TESTS): tm-signal.S
diff --git a/tools/testing/selftests/powerpc/tm/settings b/tools/testing/selftests/powerpc/tm/settings
new file mode 100644
index 000000000000..e7b9417537fb
--- /dev/null
+++ b/tools/testing/selftests/powerpc/tm/settings
@@ -0,0 +1 @@
+timeout=0
diff --git a/tools/testing/selftests/powerpc/tm/tm-signal-context-force-tm.c b/tools/testing/selftests/powerpc/tm/tm-signal-context-force-tm.c
index 31717625f318..421cb082f6be 100644
--- a/tools/testing/selftests/powerpc/tm/tm-signal-context-force-tm.c
+++ b/tools/testing/selftests/powerpc/tm/tm-signal-context-force-tm.c
@@ -42,9 +42,10 @@
 #endif
 
 /* Setting contexts because the test will crash and we want to recover */
-ucontext_t init_context, main_context;
+ucontext_t init_context;
 
-static int count, first_time;
+/* count is changed in the signal handler, so it must be volatile */
+static volatile int count;
 
 void usr_signal_handler(int signo, siginfo_t *si, void *uc)
 {
@@ -98,11 +99,6 @@ void usr_signal_handler(int signo, siginfo_t *si, void *uc)
 
 void seg_signal_handler(int signo, siginfo_t *si, void *uc)
 {
-	if (count == COUNT_MAX) {
-		/* Return to tm_signal_force_msr() and exit */
-		setcontext(&main_context);
-	}
-
 	count++;
 
 	/* Reexecute the test */
@@ -126,37 +122,41 @@ void tm_trap_test(void)
 	 */
 	getcontext(&init_context);
 
-	/* Allocated an alternative signal stack area */
-	ss.ss_sp = mmap(NULL, SIGSTKSZ, PROT_READ | PROT_WRITE,
-			MAP_PRIVATE | MAP_ANONYMOUS, 0, 0);
-	ss.ss_size = SIGSTKSZ;
-	ss.ss_flags = 0;
+	while (count < COUNT_MAX) {
+		/* Allocated an alternative signal stack area */
+		ss.ss_sp = mmap(NULL, SIGSTKSZ, PROT_READ | PROT_WRITE,
+				MAP_PRIVATE | MAP_ANONYMOUS, 0, 0);
+		ss.ss_size = SIGSTKSZ;
+		ss.ss_flags = 0;
 
-	if (ss.ss_sp == (void *)-1) {
-		perror("mmap error\n");
-		exit(-1);
-	}
+		if (ss.ss_sp == (void *)-1) {
+			perror("mmap error\n");
+			exit(-1);
+		}
 
-	/* Force the allocation through a page fault */
-	if (madvise(ss.ss_sp, SIGSTKSZ, MADV_DONTNEED)) {
-		perror("madvise\n");
-		exit(-1);
-	}
+		/* Force the allocation through a page fault */
+		if (madvise(ss.ss_sp, SIGSTKSZ, MADV_DONTNEED)) {
+			perror("madvise\n");
+			exit(-1);
+		}
 
-	/* Setting an alternative stack to generate a page fault when
-	 * the signal is raised.
-	 */
-	if (sigaltstack(&ss, NULL)) {
-		perror("sigaltstack\n");
-		exit(-1);
+		/*
+		 * Setting an alternative stack to generate a page fault when
+		 * the signal is raised.
+		 */
+		if (sigaltstack(&ss, NULL)) {
+			perror("sigaltstack\n");
+			exit(-1);
+		}
+
+		/* The signal handler will enable MSR_TS */
+		sigaction(SIGUSR1, &usr_sa, NULL);
+		/* If it does not crash, it might segfault, avoid it to retest */
+		sigaction(SIGSEGV, &seg_sa, NULL);
+
+		raise(SIGUSR1);
+		count++;
 	}
-
-	/* The signal handler will enable MSR_TS */
-	sigaction(SIGUSR1, &usr_sa, NULL);
-	/* If it does not crash, it will segfault, avoid it to retest */
-	sigaction(SIGSEGV, &seg_sa, NULL);
-
-	raise(SIGUSR1);
 }
 
 int tm_signal_context_force_tm(void)
@@ -169,11 +169,7 @@ int tm_signal_context_force_tm(void)
 	 */
 	SKIP_IF(!is_ppc64le());
 
-	/* Will get back here after COUNT_MAX interactions */
-	getcontext(&main_context);
-
-	if (!first_time++)
-		tm_trap_test();
+	tm_trap_test();
 
 	return EXIT_SUCCESS;
 }
diff --git a/tools/testing/selftests/powerpc/tm/tm-signal-pagefault.c b/tools/testing/selftests/powerpc/tm/tm-signal-pagefault.c
new file mode 100644
index 000000000000..5908bc6abe60
--- /dev/null
+++ b/tools/testing/selftests/powerpc/tm/tm-signal-pagefault.c
@@ -0,0 +1,284 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright 2020, Gustavo Luiz Duarte, IBM Corp.
+ *
+ * This test starts a transaction and triggers a signal, forcing a pagefault to
+ * happen when the kernel signal handling code touches the user signal stack.
+ *
+ * In order to avoid pre-faulting the signal stack memory and to force the
+ * pagefault to happen precisely in the kernel signal handling code, the
+ * pagefault handling is done in userspace using the userfaultfd facility.
+ *
+ * Further pagefaults are triggered by crafting the signal handler's ucontext
+ * to point to additional memory regions managed by the userfaultfd, so using
+ * the same mechanism used to avoid pre-faulting the signal stack memory.
+ *
+ * On failure (bug is present) kernel crashes or never returns control back to
+ * userspace. If bug is not present, tests completes almost immediately.
+ */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <linux/userfaultfd.h>
+#include <poll.h>
+#include <unistd.h>
+#include <sys/ioctl.h>
+#include <sys/syscall.h>
+#include <fcntl.h>
+#include <sys/mman.h>
+#include <pthread.h>
+#include <signal.h>
+#include <errno.h>
+
+#include "tm.h"
+
+
+#define UF_MEM_SIZE 655360	/* 10 x 64k pages */
+
+/* Memory handled by userfaultfd */
+static char *uf_mem;
+static size_t uf_mem_offset = 0;
+
+/*
+ * Data that will be copied into the faulting pages (instead of zero-filled
+ * pages). This is used to make the test more reliable and avoid segfaulting
+ * when we return from the signal handler. Since we are making the signal
+ * handler's ucontext point to newly allocated memory, when that memory is
+ * paged-in it will contain the expected content.
+ */
+static char backing_mem[UF_MEM_SIZE];
+
+static size_t pagesize;
+
+/*
+ * Return a chunk of at least 'size' bytes of memory that will be handled by
+ * userfaultfd. If 'backing_data' is not NULL, its content will be save to
+ * 'backing_mem' and then copied into the faulting pages when the page fault
+ * is handled.
+ */
+void *get_uf_mem(size_t size, void *backing_data)
+{
+	void *ret;
+
+	if (uf_mem_offset + size > UF_MEM_SIZE) {
+		fprintf(stderr, "Requesting more uf_mem than expected!\n");
+		exit(EXIT_FAILURE);
+	}
+
+	ret = &uf_mem[uf_mem_offset];
+
+	/* Save the data that will be copied into the faulting page */
+	if (backing_data != NULL)
+		memcpy(&backing_mem[uf_mem_offset], backing_data, size);
+
+	/* Reserve the requested amount of uf_mem */
+	uf_mem_offset += size;
+	/* Keep uf_mem_offset aligned to the page size (round up) */
+	uf_mem_offset = (uf_mem_offset + pagesize - 1) & ~(pagesize - 1);
+
+	return ret;
+}
+
+void *fault_handler_thread(void *arg)
+{
+	struct uffd_msg msg;	/* Data read from userfaultfd */
+	long uffd;		/* userfaultfd file descriptor */
+	struct uffdio_copy uffdio_copy;
+	struct pollfd pollfd;
+	ssize_t nread, offset;
+
+	uffd = (long) arg;
+
+	for (;;) {
+		pollfd.fd = uffd;
+		pollfd.events = POLLIN;
+		if (poll(&pollfd, 1, -1) == -1) {
+			perror("poll() failed");
+			exit(EXIT_FAILURE);
+		}
+
+		nread = read(uffd, &msg, sizeof(msg));
+		if (nread == 0) {
+			fprintf(stderr, "read(): EOF on userfaultfd\n");
+			exit(EXIT_FAILURE);
+		}
+
+		if (nread == -1) {
+			perror("read() failed");
+			exit(EXIT_FAILURE);
+		}
+
+		/* We expect only one kind of event */
+		if (msg.event != UFFD_EVENT_PAGEFAULT) {
+			fprintf(stderr, "Unexpected event on userfaultfd\n");
+			exit(EXIT_FAILURE);
+		}
+
+		/*
+		 * We need to handle page faults in units of pages(!).
+		 * So, round faulting address down to page boundary.
+		 */
+		uffdio_copy.dst = msg.arg.pagefault.address & ~(pagesize-1);
+
+		offset = (char *) uffdio_copy.dst - uf_mem;
+		uffdio_copy.src = (unsigned long) &backing_mem[offset];
+
+		uffdio_copy.len = pagesize;
+		uffdio_copy.mode = 0;
+		uffdio_copy.copy = 0;
+		if (ioctl(uffd, UFFDIO_COPY, &uffdio_copy) == -1) {
+			perror("ioctl-UFFDIO_COPY failed");
+			exit(EXIT_FAILURE);
+		}
+	}
+}
+
+void setup_uf_mem(void)
+{
+	long uffd;		/* userfaultfd file descriptor */
+	pthread_t thr;
+	struct uffdio_api uffdio_api;
+	struct uffdio_register uffdio_register;
+	int ret;
+
+	pagesize = sysconf(_SC_PAGE_SIZE);
+
+	/* Create and enable userfaultfd object */
+	uffd = syscall(__NR_userfaultfd, O_CLOEXEC | O_NONBLOCK);
+	if (uffd == -1) {
+		perror("userfaultfd() failed");
+		exit(EXIT_FAILURE);
+	}
+	uffdio_api.api = UFFD_API;
+	uffdio_api.features = 0;
+	if (ioctl(uffd, UFFDIO_API, &uffdio_api) == -1) {
+		perror("ioctl-UFFDIO_API failed");
+		exit(EXIT_FAILURE);
+	}
+
+	/*
+	 * Create a private anonymous mapping. The memory will be demand-zero
+	 * paged, that is, not yet allocated. When we actually touch the memory
+	 * the related page will be allocated via the userfaultfd mechanism.
+	 */
+	uf_mem = mmap(NULL, UF_MEM_SIZE, PROT_READ | PROT_WRITE,
+		      MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
+	if (uf_mem == MAP_FAILED) {
+		perror("mmap() failed");
+		exit(EXIT_FAILURE);
+	}
+
+	/*
+	 * Register the memory range of the mapping we've just mapped to be
+	 * handled by the userfaultfd object. In 'mode' we request to track
+	 * missing pages (i.e. pages that have not yet been faulted-in).
+	 */
+	uffdio_register.range.start = (unsigned long) uf_mem;
+	uffdio_register.range.len = UF_MEM_SIZE;
+	uffdio_register.mode = UFFDIO_REGISTER_MODE_MISSING;
+	if (ioctl(uffd, UFFDIO_REGISTER, &uffdio_register) == -1) {
+		perror("ioctl-UFFDIO_REGISTER");
+		exit(EXIT_FAILURE);
+	}
+
+	/* Create a thread that will process the userfaultfd events */
+	ret = pthread_create(&thr, NULL, fault_handler_thread, (void *) uffd);
+	if (ret != 0) {
+		fprintf(stderr, "pthread_create(): Error. Returned %d\n", ret);
+		exit(EXIT_FAILURE);
+	}
+}
+
+/*
+ * Assumption: the signal was delivered while userspace was in transactional or
+ * suspended state, i.e. uc->uc_link != NULL.
+ */
+void signal_handler(int signo, siginfo_t *si, void *uc)
+{
+	ucontext_t *ucp = uc;
+
+	/* Skip 'trap' after returning, otherwise we get a SIGTRAP again */
+	ucp->uc_link->uc_mcontext.regs->nip += 4;
+
+	ucp->uc_mcontext.v_regs =
+		get_uf_mem(sizeof(elf_vrreg_t), ucp->uc_mcontext.v_regs);
+
+	ucp->uc_link->uc_mcontext.v_regs =
+		get_uf_mem(sizeof(elf_vrreg_t), ucp->uc_link->uc_mcontext.v_regs);
+
+	ucp->uc_link = get_uf_mem(sizeof(ucontext_t), ucp->uc_link);
+}
+
+bool have_userfaultfd(void)
+{
+	long rc;
+
+	errno = 0;
+	rc = syscall(__NR_userfaultfd, -1);
+
+	return rc == 0 || errno != ENOSYS;
+}
+
+int tm_signal_pagefault(void)
+{
+	struct sigaction sa;
+	stack_t ss;
+
+	SKIP_IF(!have_htm());
+	SKIP_IF(!have_userfaultfd());
+
+	setup_uf_mem();
+
+	/*
+	 * Set an alternative stack that will generate a page fault when the
+	 * signal is raised. The page fault will be treated via userfaultfd,
+	 * i.e. via fault_handler_thread.
+	 */
+	ss.ss_sp = get_uf_mem(SIGSTKSZ, NULL);
+	ss.ss_size = SIGSTKSZ;
+	ss.ss_flags = 0;
+	if (sigaltstack(&ss, NULL) == -1) {
+		perror("sigaltstack() failed");
+		exit(EXIT_FAILURE);
+	}
+
+	sa.sa_flags = SA_SIGINFO | SA_ONSTACK;
+	sa.sa_sigaction = signal_handler;
+	if (sigaction(SIGTRAP, &sa, NULL) == -1) {
+		perror("sigaction() failed");
+		exit(EXIT_FAILURE);
+	}
+
+	/* Trigger a SIGTRAP in transactional state */
+	asm __volatile__(
+			"tbegin.;"
+			"beq    1f;"
+			"trap;"
+			"1: ;"
+			: : : "memory");
+
+	/* Trigger a SIGTRAP in suspended state */
+	asm __volatile__(
+			"tbegin.;"
+			"beq    1f;"
+			"tsuspend.;"
+			"trap;"
+			"tresume.;"
+			"1: ;"
+			: : : "memory");
+
+	return EXIT_SUCCESS;
+}
+
+int main(int argc, char **argv)
+{
+	/*
+	 * Depending on kernel config, the TM Bad Thing might not result in a
+	 * crash, instead the kernel never returns control back to userspace, so
+	 * set a tight timeout. If the test passes it completes almost
+	 * immediately.
+	 */
+	test_harness_set_timeout(2);
+	return test_harness(tm_signal_pagefault, "tm_signal_pagefault");
+}