All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH] Use maximum latency when determining L1 ASPM
@ 2020-10-07 13:28 Ian Kumlien
  2020-10-08  4:20 ` Kai-Heng Feng
  2020-10-08 16:13 ` Bjorn Helgaas
  0 siblings, 2 replies; 54+ messages in thread
From: Ian Kumlien @ 2020-10-07 13:28 UTC (permalink / raw)
  To: helgaas, linux-pci, alexander.duyck, refactormyself, puranjay12,
	kai.heng.feng
  Cc: Ian Kumlien

Make pcie_aspm_check_latency comply with the PCIe spec, specifically:
"5.4.1.2.2. Exit from the L1 State"

Which makes it clear that each switch is required to initiate a
transition within 1μs from receiving it, accumulating this latency and
then we have to wait for the slowest link along the path before
entering L0 state from L1.

The current code doesn't take the maximum latency into account.

From the example:
   +----------------+
   |                |
   |  Root complex  |
   |                |
   |    +-----+     |
   |    |32 μs|     |
   +----------------+
           |
           |  Link 1
           |
   +----------------+
   |     |8 μs|     |
   |     +----+     |
   |    Switch A    |
   |     +----+     |
   |     |8 μs|     |
   +----------------+
           |
           |  Link 2
           |
   +----------------+
   |    |32 μs|     |
   |    +-----+     |
   |    Switch B    |
   |    +-----+     |
   |    |32 μs|     |
   +----------------+
           |
           |  Link 3
           |
   +----------------+
   |     |8μs|      |
   |     +---+      |
   |   Endpoint C   |
   |                |
   |                |
   +----------------+

Links 1, 2 and 3 are all in L1 state - endpoint C initiates the
transition to L0 at time T. Since switch B takes 32 μs to exit L1 on
it's ports, Link 3 will transition to L0 at T+32 (longest time
considering T+8 for endpoint C and T+32 for switch B).

Switch B is required to initiate a transition from the L1 state on it's
upstream port after no more than 1 μs from the beginning of the
transition from L1 state on the downstream port. Therefore, transition from
L1 to L0 will begin on link 2 at T+1, this will cascade up the path.

The path will exit L1 at T+34.

On my specific system:
lspci -PP -s 04:00.0
00:01.2/01:00.0/02:04.0/04:00.0 Unassigned class [ff00]: Realtek Semiconductor Co., Ltd. Device 816e (rev 1a)

lspci -vvv -s 04:00.0
		DevCap:	MaxPayload 128 bytes, PhantFunc 0, Latency L0s <512ns, L1 <64us
...
		LnkCap:	Port #0, Speed 5GT/s, Width x1, ASPM L0s L1, Exit Latency L0s unlimited, L1 <64us
...

Which means that it can't be followed by any switch that is in L1 state.

This patch fixes it by disabling L1 on 02:04.0, 01:00.0 and 00:01.2.

                                                    LnkCtl    LnkCtl
           ------DevCap-------  ----LnkCap-------  -Before-  -After--
  00:01.2                                L1 <32us       L1+       L1-
  01:00.0                                L1 <32us       L1+       L1-
  02:04.0                                L1 <32us       L1+       L1-
  04:00.0  L0s <512 L1 <64us             L1 <64us       L1+       L1-

Signed-off-by: Ian Kumlien <ian.kumlien@gmail.com>
---
 drivers/pci/pcie/aspm.c | 23 +++++++++++++++--------
 1 file changed, 15 insertions(+), 8 deletions(-)

diff --git a/drivers/pci/pcie/aspm.c b/drivers/pci/pcie/aspm.c
index 253c30cc1967..893b37669087 100644
--- a/drivers/pci/pcie/aspm.c
+++ b/drivers/pci/pcie/aspm.c
@@ -434,7 +434,7 @@ static void pcie_get_aspm_reg(struct pci_dev *pdev,
 
 static void pcie_aspm_check_latency(struct pci_dev *endpoint)
 {
-	u32 latency, l1_switch_latency = 0;
+	u32 latency, l1_max_latency = 0, l1_switch_latency = 0;
 	struct aspm_latency *acceptable;
 	struct pcie_link_state *link;
 
@@ -456,10 +456,14 @@ static void pcie_aspm_check_latency(struct pci_dev *endpoint)
 		if ((link->aspm_capable & ASPM_STATE_L0S_DW) &&
 		    (link->latency_dw.l0s > acceptable->l0s))
 			link->aspm_capable &= ~ASPM_STATE_L0S_DW;
+
 		/*
 		 * Check L1 latency.
-		 * Every switch on the path to root complex need 1
-		 * more microsecond for L1. Spec doesn't mention L0s.
+		 *
+		 * PCIe r5.0, sec 5.4.1.2.2 states:
+		 * A Switch is required to initiate an L1 exit transition on its
+		 * Upstream Port Link after no more than 1 μs from the beginning of an
+		 * L1 exit transition on any of its Downstream Port Links.
 		 *
 		 * The exit latencies for L1 substates are not advertised
 		 * by a device.  Since the spec also doesn't mention a way
@@ -469,11 +473,14 @@ static void pcie_aspm_check_latency(struct pci_dev *endpoint)
 		 * L1 exit latencies advertised by a device include L1
 		 * substate latencies (and hence do not do any check).
 		 */
-		latency = max_t(u32, link->latency_up.l1, link->latency_dw.l1);
-		if ((link->aspm_capable & ASPM_STATE_L1) &&
-		    (latency + l1_switch_latency > acceptable->l1))
-			link->aspm_capable &= ~ASPM_STATE_L1;
-		l1_switch_latency += 1000;
+		if (link->aspm_capable & ASPM_STATE_L1) {
+			latency = max_t(u32, link->latency_up.l1, link->latency_dw.l1);
+			l1_max_latency = max_t(u32, latency, l1_max_latency);
+			if (l1_max_latency + l1_switch_latency > acceptable->l1)
+				link->aspm_capable &= ~ASPM_STATE_L1;
+
+			l1_switch_latency += 1000;
+		}
 
 		link = link->parent;
 	}
-- 
2.28.0


^ permalink raw reply related	[flat|nested] 54+ messages in thread

end of thread, other threads:[~2021-05-15 11:52 UTC | newest]

Thread overview: 54+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-10-07 13:28 [PATCH] Use maximum latency when determining L1 ASPM Ian Kumlien
2020-10-08  4:20 ` Kai-Heng Feng
2020-10-08 16:13 ` Bjorn Helgaas
2020-10-12 10:20   ` Ian Kumlien
2020-10-14  8:34     ` Kai-Heng Feng
2020-10-14 13:33       ` Ian Kumlien
2020-10-14 14:36         ` Bjorn Helgaas
2020-10-14 15:39           ` Ian Kumlien
2020-10-16 14:53             ` Ian Kumlien
2020-10-16 21:28         ` Bjorn Helgaas
2020-10-16 22:41           ` Ian Kumlien
2020-10-18 11:35             ` Ian Kumlien
2020-10-22 15:37               ` Bjorn Helgaas
2020-10-22 15:41                 ` Ian Kumlien
2020-10-22 18:30                   ` Bjorn Helgaas
2020-10-24 20:55                     ` [PATCH 1/3] PCI/ASPM: Use the path max in L1 ASPM latency check Ian Kumlien
2020-10-24 20:55                       ` [PATCH 2/3] PCI/ASPM: Fix L0s max " Ian Kumlien
2020-11-15 21:49                         ` Ian Kumlien
2020-10-24 20:55                       ` [PATCH 3/3] [RFC] PCI/ASPM: Print L1/L0s latency messages per endpoint Ian Kumlien
2020-11-15 21:49                       ` [PATCH 1/3] PCI/ASPM: Use the path max in L1 ASPM latency check Ian Kumlien
2020-12-07 11:04                         ` Ian Kumlien
2020-12-12 23:47                       ` Bjorn Helgaas
2020-12-13 21:39                         ` Ian Kumlien
2020-12-14  5:44                           ` Bjorn Helgaas
2020-12-14  5:44                             ` [Intel-wired-lan] " Bjorn Helgaas
2020-12-14  9:14                             ` Ian Kumlien
2020-12-14  9:14                               ` [Intel-wired-lan] " Ian Kumlien
2020-12-14 14:02                               ` Bjorn Helgaas
2020-12-14 14:02                                 ` [Intel-wired-lan] " Bjorn Helgaas
2020-12-14 15:47                                 ` Ian Kumlien
2020-12-14 15:47                                   ` [Intel-wired-lan] " Ian Kumlien
2020-12-14 19:19                                   ` Bjorn Helgaas
2020-12-14 19:19                                     ` [Intel-wired-lan] " Bjorn Helgaas
2020-12-14 22:56                                     ` Ian Kumlien
2020-12-14 22:56                                       ` [Intel-wired-lan] " Ian Kumlien
2020-12-15  0:40                                       ` Bjorn Helgaas
2020-12-15  0:40                                         ` [Intel-wired-lan] " Bjorn Helgaas
2020-12-15 13:09                                         ` Ian Kumlien
2020-12-15 13:09                                           ` [Intel-wired-lan] " Ian Kumlien
2020-12-16  0:08                                           ` Bjorn Helgaas
2020-12-16  0:08                                             ` [Intel-wired-lan] " Bjorn Helgaas
2020-12-16 11:20                                             ` Ian Kumlien
2020-12-16 11:20                                               ` [Intel-wired-lan] " Ian Kumlien
2020-12-16 23:21                                               ` Bjorn Helgaas
2020-12-16 23:21                                                 ` [Intel-wired-lan] " Bjorn Helgaas
2020-12-17 23:37                                                 ` Ian Kumlien
2020-12-17 23:37                                                   ` [Intel-wired-lan] " Ian Kumlien
2021-01-12 20:42                       ` Bjorn Helgaas
2021-01-28 12:41                         ` Ian Kumlien
2021-02-24 22:19                           ` Ian Kumlien
2021-02-25 22:03                             ` Bjorn Helgaas
2021-04-26 14:36                               ` Ian Kumlien
2021-04-28 21:15                                 ` Bjorn Helgaas
2021-05-15 11:52                                   ` Ian Kumlien

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.