x86/tsc: Force inlining of cyc2ns bits (4907c68a) · Commits · 戴 / test

Looking at the asm for native_sched_clock() I noticed we don't inline enough. Mostly caused by sharing code with cyc2ns_read_begin(), which we didn't used to do. So mark all that __force_inline to make it DTRT. Fixes: ("x86/tsc: Remodel cyc2ns to use seqcount_latch()") Reported-by:

Eric Dumazet <edumazet@google.com> Signed-off-by:

Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by:

Thomas Gleixner <tglx@linutronix.de> Cc: hpa@zytor.com Cc: eric.dumazet@gmail.com Cc: bp@alien8.de Cc: stable@vger.kernel.org Link: https://lkml.kernel.org/r/20181011104019.695196158@infradead.org

arch/x86/kernel/tsc.c

+3 −3

Original line number	Diff line number	Diff line
		@@ -58,7 +58,7 @@ struct cyc2ns {

		static DEFINE_PER_CPU_ALIGNED(struct cyc2ns, cyc2ns);

		void cyc2ns_read_begin(struct cyc2ns_data *data)
		void __always_inline cyc2ns_read_begin(struct cyc2ns_data *data)
		{
		int seq, idx;

		@@ -75,7 +75,7 @@ void cyc2ns_read_begin(struct cyc2ns_data *data)
		} while (unlikely(seq != this_cpu_read(cyc2ns.seq.sequence)));
		}

		void cyc2ns_read_end(void)
		void __always_inline cyc2ns_read_end(void)
		{
		preempt_enable_notrace();
		}
		@@ -104,7 +104,7 @@ void cyc2ns_read_end(void)
		* -johnstul@us.ibm.com "math is hard, lets go shopping!"
		*/

		static inline unsigned long long cycles_2_ns(unsigned long long cyc)
		static __always_inline unsigned long long cycles_2_ns(unsigned long long cyc)
		{
		struct cyc2ns_data data;
		unsigned long long ns;

Admin message