perf: Add ability to attach user stack dump to sample

Introducing PERF_SAMPLE_STACK_USER sample type bit to trigger the dump
of the user level stack on sample. The size of the dump is specified by
sample_stack_user value.

Being able to dump parts of the user stack, starting from the stack
pointer, will be useful to make a post mortem dwarf CFI based stack
unwinding.

Added HAVE_PERF_USER_STACK_DUMP config option to determine if the
architecture provides user stack dump on perf event samples.  This needs
access to the user stack pointer which is not unified across
architectures. Enabling this for x86 architecture.

Signed-off-by: Jiri Olsa <jolsa@redhat.com>
Original-patch-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: "Frank Ch. Eigler" <fche@redhat.com>
Cc: Arun Sharma <asharma@fb.com>
Cc: Benjamin Redelings <benjamin.redelings@nescent.org>
Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com>
Cc: Cyrill Gorcunov <gorcunov@openvz.org>
Cc: Frank Ch. Eigler <fche@redhat.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Robert Richter <robert.richter@amd.com>
Cc: Stephane Eranian <eranian@google.com>
Cc: Tom Zanussi <tzanussi@gmail.com>
Cc: Ulrich Drepper <drepper@gmail.com>
Link: http://lkml.kernel.org/r/1344345647-11536-6-git-send-email-jolsa@redhat.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

5 files changed, 190 insertions, 2 deletions

diff --git a/arch/Kconfig b/arch/Kconfig
index 68d827b7ae82..2a83a3f6a615 100644
--- a/arch/Kconfig
+++ b/arch/Kconfig
@@ -228,6 +228,13 @@ config HAVE_PERF_REGS
           Support selective register dumps for perf events. This includes
           bit-mapping of each registers and a unique architecture id.
 
+config HAVE_PERF_USER_STACK_DUMP
+        bool
+        help
+          Support user stack dumps for perf event samples. This needs
+          access to the user stack pointer which is not unified across
+          architectures.
+
 config HAVE_ARCH_JUMP_LABEL
         bool
 
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index 3fab6ec9edc4..a2d19ee750ca 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -61,6 +61,7 @@ config X86
         select PERF_EVENTS
         select HAVE_PERF_EVENTS_NMI
         select HAVE_PERF_REGS
+        select HAVE_PERF_USER_STACK_DUMP
         select ANON_INODES
         select HAVE_ALIGNED_STRUCT_PAGE if SLUB && !M386
         select HAVE_CMPXCHG_LOCAL if !M386
diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
index 8a73f75beb16..d1d25f6a5e24 100644
--- a/include/linux/perf_event.h
+++ b/include/linux/perf_event.h
@@ -131,8 +131,9 @@ enum perf_event_sample_format {
         PERF_SAMPLE_RAW                         = 1U << 10,
         PERF_SAMPLE_BRANCH_STACK                = 1U << 11,
         PERF_SAMPLE_REGS_USER                   = 1U << 12,
+        PERF_SAMPLE_STACK_USER                  = 1U << 13,
 
-        PERF_SAMPLE_MAX = 1U << 13,             /* non-ABI */
+        PERF_SAMPLE_MAX = 1U << 14,             /* non-ABI */
 };
 
 /*
@@ -205,6 +206,7 @@ enum perf_event_read_format {
 #define PERF_ATTR_SIZE_VER1     72      /* add: config2 */
 #define PERF_ATTR_SIZE_VER2     80      /* add: branch_sample_type */
 #define PERF_ATTR_SIZE_VER3     88      /* add: sample_regs_user */
+#define PERF_ATTR_SIZE_VER4     96      /* add: sample_stack_user */
 
 /*
  * Hardware event_id to monitor via a performance monitoring event:
@@ -289,6 +291,14 @@ struct perf_event_attr {
          * See asm/perf_regs.h for details.
          */
         __u64   sample_regs_user;
+
+        /*
+         * Defines size of the user stack to dump on samples.
+         */
+        __u32   sample_stack_user;
+
+        /* Align to u64. */
+        __u32   __reserved_2;
 };
 
 /*
@@ -568,6 +578,10 @@ enum perf_event_type {
          *
          *      { u64                   abi; # enum perf_sample_regs_abi
          *        u64                   regs[weight(mask)]; } && PERF_SAMPLE_REGS_USER
+         *
+         *      { u64                   size;
+         *        char                  data[size];
+         *        u64                   dyn_size; } && PERF_SAMPLE_STACK_USER
          * };
          */
         PERF_RECORD_SAMPLE                      = 9,
@@ -1160,6 +1174,7 @@ struct perf_sample_data {
         struct perf_raw_record          *raw;
         struct perf_branch_stack        *br_stack;
         struct perf_regs_user           regs_user;
+        u64                             stack_user_size;
 };
 
 static inline void perf_sample_data_init(struct perf_sample_data *data,
@@ -1172,6 +1187,7 @@ static inline void perf_sample_data_init(struct perf_sample_data *data,
         data->period = period;
         data->regs_user.abi = PERF_SAMPLE_REGS_ABI_NONE;
         data->regs_user.regs = NULL;
+        data->stack_user_size = 0;
 }
 
 extern void perf_output_sample(struct perf_output_handle *handle,
diff --git a/kernel/events/core.c b/kernel/events/core.c
index d3ce97525b9f..2ba890450d15 100644
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -36,6 +36,7 @@
 #include <linux/perf_event.h>
 #include <linux/ftrace_event.h>
 #include <linux/hw_breakpoint.h>
+#include <linux/mm_types.h>
 
 #include "internal.h"
 
@@ -3787,6 +3788,101 @@ static void perf_sample_regs_user(struct perf_regs_user *regs_user,
         }
 }
 
+/*
+ * Get remaining task size from user stack pointer.
+ *
+ * It'd be better to take stack vma map and limit this more
+ * precisly, but there's no way to get it safely under interrupt,
+ * so using TASK_SIZE as limit.
+ */
+static u64 perf_ustack_task_size(struct pt_regs *regs)
+{
+        unsigned long addr = perf_user_stack_pointer(regs);
+
+        if (!addr || addr >= TASK_SIZE)
+                return 0;
+
+        return TASK_SIZE - addr;
+}
+
+static u16
+perf_sample_ustack_size(u16 stack_size, u16 header_size,
+                        struct pt_regs *regs)
+{
+        u64 task_size;
+
+        /* No regs, no stack pointer, no dump. */
+        if (!regs)
+                return 0;
+
+        /*
+         * Check if we fit in with the requested stack size into the:
+         * - TASK_SIZE
+         *   If we don't, we limit the size to the TASK_SIZE.
+         *
+         * - remaining sample size
+         *   If we don't, we customize the stack size to
+         *   fit in to the remaining sample size.
+         */
+
+        task_size  = min((u64) USHRT_MAX, perf_ustack_task_size(regs));
+        stack_size = min(stack_size, (u16) task_size);
+
+        /* Current header size plus static size and dynamic size. */
+        header_size += 2 * sizeof(u64);
+
+        /* Do we fit in with the current stack dump size? */
+        if ((u16) (header_size + stack_size) < header_size) {
+                /*
+                 * If we overflow the maximum size for the sample,
+                 * we customize the stack dump size to fit in.
+                 */
+                stack_size = USHRT_MAX - header_size - sizeof(u64);
+                stack_size = round_up(stack_size, sizeof(u64));
+        }
+
+        return stack_size;
+}
+
+static void
+perf_output_sample_ustack(struct perf_output_handle *handle, u64 dump_size,
+                          struct pt_regs *regs)
+{
+        /* Case of a kernel thread, nothing to dump */
+        if (!regs) {
+                u64 size = 0;
+                perf_output_put(handle, size);
+        } else {
+                unsigned long sp;
+                unsigned int rem;
+                u64 dyn_size;
+
+                /*
+                 * We dump:
+                 * static size
+                 *   - the size requested by user or the best one we can fit
+                 *     in to the sample max size
+                 * data
+                 *   - user stack dump data
+                 * dynamic size
+                 *   - the actual dumped size
+                 */
+
+                /* Static size. */
+                perf_output_put(handle, dump_size);
+
+                /* Data. */
+                sp = perf_user_stack_pointer(regs);
+                rem = __output_copy_user(handle, (void *) sp, dump_size);
+                dyn_size = dump_size - rem;
+
+                perf_output_skip(handle, rem);
+
+                /* Dynamic size. */
+                perf_output_put(handle, dyn_size);
+        }
+}
+
 static void __perf_event_header__init_id(struct perf_event_header *header,
                                          struct perf_sample_data *data,
                                          struct perf_event *event)
@@ -4064,6 +4160,11 @@ void perf_output_sample(struct perf_output_handle *handle,
                                                 mask);
                 }
         }
+
+        if (sample_type & PERF_SAMPLE_STACK_USER)
+                perf_output_sample_ustack(handle,
+                                          data->stack_user_size,
+                                          data->regs_user.regs);
 }
 
 void perf_prepare_sample(struct perf_event_header *header,
@@ -4129,6 +4230,35 @@ void perf_prepare_sample(struct perf_event_header *header,
 
                 header->size += size;
         }
+
+        if (sample_type & PERF_SAMPLE_STACK_USER) {
+                /*
+                 * Either we need PERF_SAMPLE_STACK_USER bit to be allways
+                 * processed as the last one or have additional check added
+                 * in case new sample type is added, because we could eat
+                 * up the rest of the sample size.
+                 */
+                struct perf_regs_user *uregs = &data->regs_user;
+                u16 stack_size = event->attr.sample_stack_user;
+                u16 size = sizeof(u64);
+
+                if (!uregs->abi)
+                        perf_sample_regs_user(uregs, regs);
+
+                stack_size = perf_sample_ustack_size(stack_size, header->size,
+                                                     uregs->regs);
+
+                /*
+                 * If there is something to dump, add space for the dump
+                 * itself and for the field that tells the dynamic size,
+                 * which is how many have been actually dumped.
+                 */
+                if (stack_size)
+                        size += sizeof(u64) + stack_size;
+
+                data->stack_user_size = stack_size;
+                header->size += size;
+        }
 }
 
 static void perf_event_output(struct perf_event *event,
@@ -6205,8 +6335,26 @@ static int perf_copy_attr(struct perf_event_attr __user *uattr,
                 }
         }
 
-        if (attr->sample_type & PERF_SAMPLE_REGS_USER)
+        if (attr->sample_type & PERF_SAMPLE_REGS_USER) {
                 ret = perf_reg_validate(attr->sample_regs_user);
+                if (ret)
+                        return ret;
+        }
+
+        if (attr->sample_type & PERF_SAMPLE_STACK_USER) {
+                if (!arch_perf_have_user_stack_dump())
+                        return -ENOSYS;
+
+                /*
+                 * We have __u32 type for the size, but so far
+                 * we can only use __u16 as maximum due to the
+                 * __u16 sample size limit.
+                 */
+                if (attr->sample_stack_user >= USHRT_MAX)
+                        ret = -EINVAL;
+                else if (!IS_ALIGNED(attr->sample_stack_user, sizeof(u64)))
+                        ret = -EINVAL;
+        }
 
 out:
         return ret;
diff --git a/kernel/events/internal.h b/kernel/events/internal.h
index ce7bdfc1d045..d56a64c99a8b 100644
--- a/kernel/events/internal.h
+++ b/kernel/events/internal.h
@@ -158,4 +158,20 @@ static inline void put_recursion_context(int *recursion, int rctx)
         recursion[rctx]--;
 }
 
+#ifdef CONFIG_HAVE_PERF_USER_STACK_DUMP
+static inline bool arch_perf_have_user_stack_dump(void)
+{
+        return true;
+}
+
+#define perf_user_stack_pointer(regs) user_stack_pointer(regs)
+#else
+static inline bool arch_perf_have_user_stack_dump(void)
+{
+        return false;
+}
+
+#define perf_user_stack_pointer(regs) 0
+#endif /* CONFIG_HAVE_PERF_USER_STACK_DUMP */
+
 #endif /* _KERNEL_EVENTS_INTERNAL_H */