2020-09-29 08:22:42

by quanyang wang

Subject: [PATCH V3] time/sched_clock: mark sched_clock_read_begin/retry as notrace

From: Quanyang Wang <[email protected]>

Since sched_clock_read_begin and sched_clock_read_retry are called
by notrace function sched_clock, they shouldn't be traceable either,
or else ftrace_graph_caller will run into a dead loop on the path
as below (arm for instance):

ftrace_graph_caller
prepare_ftrace_return
function_graph_enter
ftrace_push_return_trace
trace_clock_local
sched_clock
sched_clock_read_begin/retry

Fixes: 1b86abc1c645 ("sched_clock: Expose struct clock_read_data")
Signed-off-by: Quanyang Wang <[email protected]>
Acked-by: Peter Zijlstra (Intel) <[email protected]>
---
Changes:
V2: Add notrace to sched_clock_read_retry according to Peter's suggestion.
V3: Adjust the placement of notrace according to Peter's suggestion.

kernel/time/sched_clock.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/kernel/time/sched_clock.c b/kernel/time/sched_clock.c
index 1c03eec6ca9b..f629e3f5afbe 100644
--- a/kernel/time/sched_clock.c
+++ b/kernel/time/sched_clock.c
@@ -68,13 +68,13 @@ static inline u64 notrace cyc_to_ns(u64 cyc, u32 mult, u32 shift)
return (cyc * mult) >> shift;
}

-struct clock_read_data *sched_clock_read_begin(unsigned int *seq)
+notrace struct clock_read_data *sched_clock_read_begin(unsigned int *seq)
{
*seq = raw_read_seqcount_latch(&cd.seq);
return cd.read_data + (*seq & 1);
}

-int sched_clock_read_retry(unsigned int seq)
+notrace int sched_clock_read_retry(unsigned int seq)
{
return read_seqcount_retry(&cd.seq, seq);
}
--
2.17.1

2020-10-26 11:27:10

by tip-bot2 for Jacob Pan

Subject: [tip: timers/urgent] time/sched_clock: Mark sched_clock_read_begin/retry() as notrace

The following commit has been merged into the timers/urgent branch of tip:

Commit-ID: 4cd2bb12981165f865d2b8ed92b446b52310ef74
Gitweb: https://git.kernel.org/tip/4cd2bb12981165f865d2b8ed92b446b52310ef74
Author: Quanyang Wang <[email protected]>
AuthorDate: Tue, 29 Sep 2020 16:20:27 +08:00
Committer: Thomas Gleixner <[email protected]>
CommitterDate: Mon, 26 Oct 2020 11:34:31 +01:00

time/sched_clock: Mark sched_clock_read_begin/retry() as notrace

Since sched_clock_read_begin() and sched_clock_read_retry() are called
by notrace function sched_clock(), they shouldn't be traceable either,
or else ftrace_graph_caller will run into a dead loop on the path
as below (arm for instance):

ftrace_graph_caller()
prepare_ftrace_return()
function_graph_enter()
ftrace_push_return_trace()
trace_clock_local()
sched_clock()
sched_clock_read_begin/retry()

Fixes: 1b86abc1c645 ("sched_clock: Expose struct clock_read_data")
Signed-off-by: Quanyang Wang <[email protected]>
Signed-off-by: Thomas Gleixner <[email protected]>
Acked-by: Peter Zijlstra (Intel) <[email protected]>
Cc: [email protected]
Link: https://lore.kernel.org/r/[email protected]
---
kernel/time/sched_clock.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/kernel/time/sched_clock.c b/kernel/time/sched_clock.c
index 0642013..b1b9b12 100644
--- a/kernel/time/sched_clock.c
+++ b/kernel/time/sched_clock.c
@@ -68,13 +68,13 @@ static inline u64 notrace cyc_to_ns(u64 cyc, u32 mult, u32 shift)
return (cyc * mult) >> shift;
}

-struct clock_read_data *sched_clock_read_begin(unsigned int *seq)
+notrace struct clock_read_data *sched_clock_read_begin(unsigned int *seq)
{
*seq = raw_read_seqcount_latch(&cd.seq);
return cd.read_data + (*seq & 1);
}

-int sched_clock_read_retry(unsigned int seq)
+notrace int sched_clock_read_retry(unsigned int seq)
{
return read_seqcount_latch_retry(&cd.seq, seq);
}