Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1164443

[PATCH 16/17] x86/tsc: Use rdtsc_ordered() in check_tsc_warp() and drop extra barriers

Show all headers | View raw

Using get_cycles was unnecessary: check_tsc_warp() is not called on
TSC-less systems.  Replace barrier_before_rdtsc(); get_cycles() with
rdtsc_ordered().

While we're at it, make the somewhat more dangerous change of
removing barrier_before_rdtsc after RDTSC in the TSC warp check
code.  This should be okay, though -- the vDSO TSC code doesn't have
that barrier, so, if removing the barrier from the warp check would
cause us to detect a warp that we otherwise wouldn't detect, then we
have a genuine bug.

Signed-off-by: Andy Lutomirski <luto@kernel.org>
---
 arch/x86/kernel/tsc_sync.c | 14 ++++++--------
 1 file changed, 6 insertions(+), 8 deletions(-)

diff --git a/arch/x86/kernel/tsc_sync.c b/arch/x86/kernel/tsc_sync.c
index cf36e9befb39..78083bf23ed1 100644
--- a/arch/x86/kernel/tsc_sync.c
+++ b/arch/x86/kernel/tsc_sync.c
@@ -39,16 +39,15 @@ static cycles_t max_warp;
 static int nr_warps;
 
 /*
- * TSC-warp measurement loop running on both CPUs:
+ * TSC-warp measurement loop running on both CPUs.  This is not called
+ * if there is no TSC.
  */
 static void check_tsc_warp(unsigned int timeout)
 {
 	cycles_t start, now, prev, end;
 	int i;
 
-	barrier_before_rdtsc();
-	start = get_cycles();
-	barrier_before_rdtsc();
+	start = rdtsc_ordered();
 	/*
 	 * The measurement runs for 'timeout' msecs:
 	 */
@@ -63,9 +62,7 @@ static void check_tsc_warp(unsigned int timeout)
 		 */
 		arch_spin_lock(&sync_lock);
 		prev = last_tsc;
-		barrier_before_rdtsc();
-		now = get_cycles();
-		barrier_before_rdtsc();
+		now = rdtsc_ordered();
 		last_tsc = now;
 		arch_spin_unlock(&sync_lock);
 
@@ -126,7 +123,7 @@ void check_tsc_sync_source(int cpu)
 
 	/*
 	 * No need to check if we already know that the TSC is not
-	 * synchronized:
+	 * synchronized or if we have no TSC.
 	 */
 	if (unsynchronized_tsc())
 		return;
@@ -190,6 +187,7 @@ void check_tsc_sync_target(void)
 {
 	int cpus = 2;
 
+	/* Also aborts if there is no TSC. */
 	if (unsynchronized_tsc() || tsc_clocksource_reliable)
 		return;
 
-- 
2.4.2

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

[PATCH 00/17] x86/tsc: Clean up rdtsc helpers Andy Lutomirski <luto@kernel.org> - 2015-06-13 01:50 +0200
  [PATCH 13/17] x86/tsc: Rename native_read_tsc() to rdtsc_unordered() Andy Lutomirski <luto@kernel.org> - 2015-06-13 01:50 +0200
  [PATCH 16/17] x86/tsc: Use rdtsc_ordered() in check_tsc_warp() and drop extra barriers Andy Lutomirski <luto@kernel.org> - 2015-06-13 01:50 +0200
  [PATCH 04/17] x86/tsc: Remove the rdtscp and rdtscpll macros Andy Lutomirski <luto@kernel.org> - 2015-06-13 01:50 +0200
  [PATCH 02/17] x86/msr/kvm: Remove vget_cycles() Andy Lutomirski <luto@kernel.org> - 2015-06-13 01:50 +0200
  [PATCH 14/17] x86/tsc: Move rdtsc_barrier() and rename it to barrier_before_rdtsc() Andy Lutomirski <luto@kernel.org> - 2015-06-13 01:50 +0200
  [PATCH 15/17] x86: Add rdtsc_ordered() and use it in trivial call sites Andy Lutomirski <luto@kernel.org> - 2015-06-13 01:50 +0200
  [PATCH 10/17] input/joystick/analog: Switch from rdtscl() to native_read_tsc() Andy Lutomirski <luto@kernel.org> - 2015-06-13 01:50 +0200
  [PATCH 07/17] x86/cpu/amd: Use the full 64-bit TSC to detect the 2.6.2 bug Andy Lutomirski <luto@kernel.org> - 2015-06-13 01:50 +0200
  [PATCH 11/17] drivers/input/gameport: Replace rdtscl() with native_read_tsc() Andy Lutomirski <luto@kernel.org> - 2015-06-13 01:50 +0200
  [PATCH 17/17] x86/tsc: In read_tsc, use rdtsc_ordered() instead of get_cycles() Andy Lutomirski <luto@kernel.org> - 2015-06-13 01:50 +0200
  [PATCH 09/17] staging/lirc_serial: Remove TSC-based timing Andy Lutomirski <luto@kernel.org> - 2015-06-13 01:50 +0200
  [PATCH 16/17] x86/tsc: Use rdtsc_unordered() in check_tsc_warp() Andy Lutomirski <luto@kernel.org> - 2015-06-13 01:50 +0200
  [PATCH 12/17] x86/tsc: Remove rdtscl() Andy Lutomirski <luto@kernel.org> - 2015-06-13 01:50 +0200
  [PATCH 03/17] x86/tsc: Replace rdtscll with native_read_tsc Andy Lutomirski <luto@kernel.org> - 2015-06-13 02:00 +0200
  [PATCH 01/17] x86/tsc: Inline native_read_tsc and remove __native_read_tsc Andy Lutomirski <luto@kernel.org> - 2015-06-13 02:00 +0200
  [PATCH 08/17] baycom_epp: Replace rdtscl() with native_read_tsc() Andy Lutomirski <luto@kernel.org> - 2015-06-13 02:00 +0200
  [PATCH 03/17] x86/tsc/paravirt: Remove the read_tsc and read_tscp paravirt hooks Andy Lutomirski <luto@kernel.org> - 2015-06-13 02:00 +0200
  [PATCH 05/17] x86/tsc: Remove the rdtscp and rdtscpll macros Andy Lutomirski <luto@kernel.org> - 2015-06-13 02:00 +0200
  [PATCH 06/17] x86/tsc: Use the full 64-bit tsc in tsc_delay Andy Lutomirski <luto@kernel.org> - 2015-06-13 02:00 +0200
  [PATCH 04/17] x86/tsc: Replace rdtscll with native_read_tsc Andy Lutomirski <luto@kernel.org> - 2015-06-13 02:00 +0200
  [PATCH 05/17] x86/tsc/paravirt: Remove the read_tsc and read_tscp paravirt hooks Andy Lutomirski <luto@kernel.org> - 2015-06-13 02:00 +0200

csiph-web