This patch add support for perf callchain sampling on riscv platform.
The return address of leaf function is retrieved from pt_regs as
it is not saved in the outmost frame.
Signed-off-by: Mao Han <[email protected]>
Cc: Paul Walmsley <[email protected]>
Cc: Greentime Hu <[email protected]>
Cc: Palmer Dabbelt <[email protected]>
Cc: linux-riscv <[email protected]>
Cc: Christoph Hellwig <[email protected]>
Cc: Guo Ren <[email protected]>
---
arch/riscv/Makefile | 3 ++
arch/riscv/kernel/Makefile | 3 +-
arch/riscv/kernel/perf_callchain.c | 95 ++++++++++++++++++++++++++++++++++++++
arch/riscv/kernel/stacktrace.c | 2 +-
4 files changed, 101 insertions(+), 2 deletions(-)
create mode 100644 arch/riscv/kernel/perf_callchain.c
diff --git a/arch/riscv/Makefile b/arch/riscv/Makefile
index 7a117be..946565b 100644
--- a/arch/riscv/Makefile
+++ b/arch/riscv/Makefile
@@ -54,6 +54,9 @@ endif
ifeq ($(CONFIG_MODULE_SECTIONS),y)
KBUILD_LDFLAGS_MODULE += -T $(srctree)/arch/riscv/kernel/module.lds
endif
+ifeq ($(CONFIG_PERF_EVENTS),y)
+ KBUILD_CFLAGS += -fno-omit-frame-pointer
+endif
KBUILD_CFLAGS_MODULE += $(call cc-option,-mno-relax)
diff --git a/arch/riscv/kernel/Makefile b/arch/riscv/kernel/Makefile
index 2420d37..b1bea89 100644
--- a/arch/riscv/kernel/Makefile
+++ b/arch/riscv/kernel/Makefile
@@ -38,6 +38,7 @@ obj-$(CONFIG_MODULE_SECTIONS) += module-sections.o
obj-$(CONFIG_FUNCTION_TRACER) += mcount.o ftrace.o
obj-$(CONFIG_DYNAMIC_FTRACE) += mcount-dyn.o
-obj-$(CONFIG_PERF_EVENTS) += perf_event.o
+obj-$(CONFIG_PERF_EVENTS) += perf_event.o
+obj-$(CONFIG_PERF_EVENTS) += perf_callchain.o
clean:
diff --git a/arch/riscv/kernel/perf_callchain.c b/arch/riscv/kernel/perf_callchain.c
new file mode 100644
index 0000000..72168e1
--- /dev/null
+++ b/arch/riscv/kernel/perf_callchain.c
@@ -0,0 +1,95 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (C) 2019 Hangzhou C-SKY Microsystems co.,ltd. */
+
+#include <linux/perf_event.h>
+#include <linux/uaccess.h>
+
+/* Kernel callchain */
+struct stackframe {
+ unsigned long fp;
+ unsigned long ra;
+};
+
+/*
+ * Get the return address for a single stackframe and return a pointer to the
+ * next frame tail.
+ */
+static unsigned long user_backtrace(struct perf_callchain_entry_ctx *entry,
+ unsigned long fp, unsigned long reg_ra)
+{
+ struct stackframe buftail;
+ unsigned long ra = 0;
+ unsigned long *user_frame_tail =
+ (unsigned long *)(fp - sizeof(struct stackframe));
+
+ /* Check accessibility of one struct frame_tail beyond */
+ if (!access_ok(user_frame_tail, sizeof(buftail)))
+ return 0;
+ if (__copy_from_user_inatomic(&buftail, user_frame_tail,
+ sizeof(buftail)))
+ return 0;
+
+ if (reg_ra != 0)
+ ra = reg_ra;
+ else
+ ra = buftail.ra;
+
+ fp = buftail.fp;
+ if (ra != 0)
+ perf_callchain_store(entry, ra);
+ else
+ return 0;
+
+ return fp;
+}
+
+/*
+ * This will be called when the target is in user mode
+ * This function will only be called when we use
+ * "PERF_SAMPLE_CALLCHAIN" in
+ * kernel/events/core.c:perf_prepare_sample()
+ *
+ * How to trigger perf_callchain_[user/kernel] :
+ * $ perf record -e cpu-clock --call-graph fp ./program
+ * $ perf report --call-graph
+ *
+ * On RISC-V platform, the program being sampled and the C library
+ * need to be compiled with -fno-omit-frame-pointer, otherwise
+ * the user stack will not contain function frame.
+ */
+void perf_callchain_user(struct perf_callchain_entry_ctx *entry,
+ struct pt_regs *regs)
+{
+ unsigned long fp = 0;
+
+ /* RISC-V does not support perf in guest mode. */
+ if (perf_guest_cbs && perf_guest_cbs->is_in_guest())
+ return;
+
+ fp = regs->s0;
+ perf_callchain_store(entry, regs->sepc);
+
+ fp = user_backtrace(entry, fp, regs->ra);
+ while (fp && !(fp & 0x3) && entry->nr < entry->max_stack)
+ fp = user_backtrace(entry, fp, 0);
+}
+
+bool fill_callchain(unsigned long pc, void *entry)
+{
+ return perf_callchain_store(entry, pc);
+}
+
+void notrace walk_stackframe(struct task_struct *task,
+ struct pt_regs *regs, bool (*fn)(unsigned long, void *), void *arg);
+void perf_callchain_kernel(struct perf_callchain_entry_ctx *entry,
+ struct pt_regs *regs)
+{
+
+ /* RISC-V does not support perf in guest mode. */
+ if (perf_guest_cbs && perf_guest_cbs->is_in_guest()) {
+ pr_warn("RISC-V does not support perf in guest mode!");
+ return;
+ }
+
+ walk_stackframe(NULL, regs, fill_callchain, entry);
+}
diff --git a/arch/riscv/kernel/stacktrace.c b/arch/riscv/kernel/stacktrace.c
index f156427..9b376c1 100644
--- a/arch/riscv/kernel/stacktrace.c
+++ b/arch/riscv/kernel/stacktrace.c
@@ -19,7 +19,7 @@ struct stackframe {
unsigned long ra;
};
-static void notrace walk_stackframe(struct task_struct *task,
+void notrace walk_stackframe(struct task_struct *task,
struct pt_regs *regs, bool (*fn)(unsigned long, void *), void *arg)
{
unsigned long fp, sp, pc;
--
2.7.4
Hello Mao Han,
On Thu, 29 Aug 2019, Mao Han wrote:
> This patch add support for perf callchain sampling on riscv platform.
> The return address of leaf function is retrieved from pt_regs as
> it is not saved in the outmost frame.
>
> Signed-off-by: Mao Han <[email protected]>
> Cc: Paul Walmsley <[email protected]>
> Cc: Greentime Hu <[email protected]>
> Cc: Palmer Dabbelt <[email protected]>
> Cc: linux-riscv <[email protected]>
> Cc: Christoph Hellwig <[email protected]>
> Cc: Guo Ren <[email protected]>
There are some 'checkpatch.pl --strict' warnings with this patch (below).
These have been fixed here. The following patch has been queued for
v5.4-rc1 with Greentime's Tested-by:. Thanks for your hard work following
up on the feedback with these patches -
- Paul
CHECK: Alignment should match open parenthesis
#77: FILE: arch/riscv/kernel/perf_callchain.c:18:
+static unsigned long user_backtrace(struct perf_callchain_entry_ctx
*entry,
+ unsigned long fp, unsigned long reg_ra)
CHECK: Blank lines aren't necessary after an open brace '{'
#146: FILE: arch/riscv/kernel/perf_callchain.c:87:
+{
+
CHECK: Alignment should match open parenthesis
#165: FILE: arch/riscv/kernel/stacktrace.c:23:
+void notrace walk_stackframe(struct task_struct *task,
struct pt_regs *regs, bool (*fn)(unsigned long, void *), void *arg)
------
From: Mao Han <[email protected]>
Date: Thu, 29 Aug 2019 14:57:00 +0800
Subject: [PATCH] riscv: Add perf callchain support
This patch adds support for perf callchain sampling on riscv platforms.
The return address of leaf function is retrieved from pt_regs as
it is not saved in the outmost frame.
Signed-off-by: Mao Han <[email protected]>
Cc: Paul Walmsley <[email protected]>
Cc: Greentime Hu <[email protected]>
Cc: Palmer Dabbelt <[email protected]>
Cc: linux-riscv <[email protected]>
Cc: Christoph Hellwig <[email protected]>
Cc: Guo Ren <[email protected]>
Tested-by: Greentime Hu <[email protected]>
[[email protected]: fixed some 'checkpatch.pl --strict' issues;
fixed patch description spelling]
Signed-off-by: Paul Walmsley <[email protected]>
---
arch/riscv/Makefile | 3 +
arch/riscv/kernel/Makefile | 3 +-
arch/riscv/kernel/perf_callchain.c | 94 ++++++++++++++++++++++++++++++
arch/riscv/kernel/stacktrace.c | 4 +-
4 files changed, 101 insertions(+), 3 deletions(-)
create mode 100644 arch/riscv/kernel/perf_callchain.c
diff --git a/arch/riscv/Makefile b/arch/riscv/Makefile
index a73659e30f8d..4f0a3d2018d2 100644
--- a/arch/riscv/Makefile
+++ b/arch/riscv/Makefile
@@ -54,6 +54,9 @@ endif
ifeq ($(CONFIG_MODULE_SECTIONS),y)
KBUILD_LDFLAGS_MODULE += -T $(srctree)/arch/riscv/kernel/module.lds
endif
+ifeq ($(CONFIG_PERF_EVENTS),y)
+ KBUILD_CFLAGS += -fno-omit-frame-pointer
+endif
KBUILD_CFLAGS_MODULE += $(call cc-option,-mno-relax)
diff --git a/arch/riscv/kernel/Makefile b/arch/riscv/kernel/Makefile
index 2420d37d96de..b1bea89fc814 100644
--- a/arch/riscv/kernel/Makefile
+++ b/arch/riscv/kernel/Makefile
@@ -38,6 +38,7 @@ obj-$(CONFIG_MODULE_SECTIONS) += module-sections.o
obj-$(CONFIG_FUNCTION_TRACER) += mcount.o ftrace.o
obj-$(CONFIG_DYNAMIC_FTRACE) += mcount-dyn.o
-obj-$(CONFIG_PERF_EVENTS) += perf_event.o
+obj-$(CONFIG_PERF_EVENTS) += perf_event.o
+obj-$(CONFIG_PERF_EVENTS) += perf_callchain.o
clean:
diff --git a/arch/riscv/kernel/perf_callchain.c b/arch/riscv/kernel/perf_callchain.c
new file mode 100644
index 000000000000..8d2804f05cf9
--- /dev/null
+++ b/arch/riscv/kernel/perf_callchain.c
@@ -0,0 +1,94 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (C) 2019 Hangzhou C-SKY Microsystems co.,ltd. */
+
+#include <linux/perf_event.h>
+#include <linux/uaccess.h>
+
+/* Kernel callchain */
+struct stackframe {
+ unsigned long fp;
+ unsigned long ra;
+};
+
+/*
+ * Get the return address for a single stackframe and return a pointer to the
+ * next frame tail.
+ */
+static unsigned long user_backtrace(struct perf_callchain_entry_ctx *entry,
+ unsigned long fp, unsigned long reg_ra)
+{
+ struct stackframe buftail;
+ unsigned long ra = 0;
+ unsigned long *user_frame_tail =
+ (unsigned long *)(fp - sizeof(struct stackframe));
+
+ /* Check accessibility of one struct frame_tail beyond */
+ if (!access_ok(user_frame_tail, sizeof(buftail)))
+ return 0;
+ if (__copy_from_user_inatomic(&buftail, user_frame_tail,
+ sizeof(buftail)))
+ return 0;
+
+ if (reg_ra != 0)
+ ra = reg_ra;
+ else
+ ra = buftail.ra;
+
+ fp = buftail.fp;
+ if (ra != 0)
+ perf_callchain_store(entry, ra);
+ else
+ return 0;
+
+ return fp;
+}
+
+/*
+ * This will be called when the target is in user mode
+ * This function will only be called when we use
+ * "PERF_SAMPLE_CALLCHAIN" in
+ * kernel/events/core.c:perf_prepare_sample()
+ *
+ * How to trigger perf_callchain_[user/kernel] :
+ * $ perf record -e cpu-clock --call-graph fp ./program
+ * $ perf report --call-graph
+ *
+ * On RISC-V platform, the program being sampled and the C library
+ * need to be compiled with -fno-omit-frame-pointer, otherwise
+ * the user stack will not contain function frame.
+ */
+void perf_callchain_user(struct perf_callchain_entry_ctx *entry,
+ struct pt_regs *regs)
+{
+ unsigned long fp = 0;
+
+ /* RISC-V does not support perf in guest mode. */
+ if (perf_guest_cbs && perf_guest_cbs->is_in_guest())
+ return;
+
+ fp = regs->s0;
+ perf_callchain_store(entry, regs->sepc);
+
+ fp = user_backtrace(entry, fp, regs->ra);
+ while (fp && !(fp & 0x3) && entry->nr < entry->max_stack)
+ fp = user_backtrace(entry, fp, 0);
+}
+
+bool fill_callchain(unsigned long pc, void *entry)
+{
+ return perf_callchain_store(entry, pc);
+}
+
+void notrace walk_stackframe(struct task_struct *task,
+ struct pt_regs *regs, bool (*fn)(unsigned long, void *), void *arg);
+void perf_callchain_kernel(struct perf_callchain_entry_ctx *entry,
+ struct pt_regs *regs)
+{
+ /* RISC-V does not support perf in guest mode. */
+ if (perf_guest_cbs && perf_guest_cbs->is_in_guest()) {
+ pr_warn("RISC-V does not support perf in guest mode!");
+ return;
+ }
+
+ walk_stackframe(NULL, regs, fill_callchain, entry);
+}
diff --git a/arch/riscv/kernel/stacktrace.c b/arch/riscv/kernel/stacktrace.c
index f15642715d1a..0940681d2f68 100644
--- a/arch/riscv/kernel/stacktrace.c
+++ b/arch/riscv/kernel/stacktrace.c
@@ -19,8 +19,8 @@ struct stackframe {
unsigned long ra;
};
-static void notrace walk_stackframe(struct task_struct *task,
- struct pt_regs *regs, bool (*fn)(unsigned long, void *), void *arg)
+void notrace walk_stackframe(struct task_struct *task, struct pt_regs *regs,
+ bool (*fn)(unsigned long, void *), void *arg)
{
unsigned long fp, sp, pc;
--
2.23.0
On Wed, Sep 04, 2019 at 12:54:41PM -0700, Paul Walmsley wrote:
> Hello Mao Han,
>
> On Thu, 29 Aug 2019, Mao Han wrote:
>
> > This patch add support for perf callchain sampling on riscv platform.
> > The return address of leaf function is retrieved from pt_regs as
> > it is not saved in the outmost frame.
> >
> > Signed-off-by: Mao Han <[email protected]>
> > Cc: Paul Walmsley <[email protected]>
> > Cc: Greentime Hu <[email protected]>
> > Cc: Palmer Dabbelt <[email protected]>
> > Cc: linux-riscv <[email protected]>
> > Cc: Christoph Hellwig <[email protected]>
> > Cc: Guo Ren <[email protected]>
>
> There are some 'checkpatch.pl --strict' warnings with this patch (below).
> These have been fixed here. The following patch has been queued for
> v5.4-rc1 with Greentime's Tested-by:. Thanks for your hard work following
> up on the feedback with these patches -
Thanks for the fixes. I'll check patches with --strict next time.
Thanks,
Mao Han