Re: ahci: sata hotplug doesn't work after suspend

On Mon, 2012-02-20 at 13:52 +0100, Jiri Slaby wrote:

On 02/20/2012 09:30 AM, Lin Ming wrote:

quoted

On Sun, 2012-02-19 at 22:17 +0100, Holger Macht wrote:

quoted

On So 19. Feb - 21:11:33, Jiri Slaby wrote:

quoted

Hi,

ahci hotplug stopped working here. It always worked (-next-20111206 is
the last known working for sure). I'm using -next and after plugging a
sata disk in, nothing happens. I don't *think* (in the traffic, it's
hard to tell for sure) an interrupt is generated either. I also *think*
that it works before suspend/resume, but not after. (Or maybe uptime
matters?)

This happens before suspend on plug-in:
ata3: exception Emask 0x50 SAct 0x0 SErr 0x40d0800 action 0xe frozen
ata3: irq_stat 0x00000040, connection status changed
ata3: SError: { HostInt PHYRdyChg CommWake 10B8B DevExch }
ata3: hard resetting link
ata3: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
ata3.00: ATA-8: ST31000528AS, CC38, max UDMA/133
ata3.00: 1953525168 sectors, multi 0: LBA48 NCQ (depth 31/32)
ata3.00: configured for UDMA/133
ata3: EH complete
scsi 2:0:0:0: Direct-Access     ATA      ST31000528AS     CC38 PQ: 0 ANSI: 5
sd 2:0:0:0: [sdc] 1953525168 512-byte logical blocks: (1.00 TB/931 GiB)
sd 2:0:0:0: [sdc] Write Protect is off
sd 2:0:0:0: [sdc] Mode Sense: 00 3a 00 00
sd 2:0:0:0: [sdc] Write cache: enabled, read cache: enabled, doesn't
support DPO or FUA
 sdc: sdc1 sdc2
sd 2:0:0:0: [sdc] Attached SCSI disk

After suspend/resume and plug-in, nothing appears in dmesg. Storing "- -
-" to /sys/class/scsi_host/host*/scan doesn't help. It looks like the
line is completely dead, because only ata1 and ata2 say "hard resetting
link" after this (and another two disks are connected there).

I tried 'echo adapter >host_reset', but I get 'write error: Invalid
argument'.

00:1f.2 SATA controller [0106]: Intel Corporation 82801IR/IO/IH
(ICH9R/DO/DH) 6 port SATA Controller [AHCI mode] [8086:2922] (rev 02)
(prog-if 01 [AHCI 1.0])
        Subsystem: Intel Corporation 82801IR/IO/IH (ICH9R/DO/DH) 6 port
SATA Controller [AHCI mode] [8086:2922]
        Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop-
ParErr- Stepping- SERR- FastB2B- DisINTx+
        Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort-
<TAbort- <MAbort- >SERR- <PERR- INTx-
        Latency: 0
        Interrupt: pin B routed to IRQ 43
        Region 0: I/O ports at cc00 [size=8]
        Region 1: I/O ports at c880 [size=4]
        Region 2: I/O ports at c800 [size=8]
        Region 3: I/O ports at c480 [size=4]
        Region 4: I/O ports at c400 [size=32]
        Region 5: Memory at feb7c800 (32-bit, non-prefetchable) [size=2K]
        Capabilities: [80] MSI: Enable+ Count=1/16 Maskable- 64bit-
                Address: fee0100c  Data: 4171
        Capabilities: [70] Power Management version 3
                Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA
PME(D0-,D1-,D2-,D3hot+,D3cold-)
                Status: D0 NoSoftRst+ PME-Enable- DSel=0 DScale=0 PME-
        Capabilities: [a8] SATA HBA v1.0 BAR4 Offset=00000004
        Capabilities: [b0] Vendor Specific Information: Len=06 <?>
        Kernel driver in use: ahci

Any ideas what that could be?

I suspect this is caused by commit
9ee4f3933930abf5cc34f8e9d69fe0e08c18f602. Maybe you could verify? I
bisected this a couple of days ago but didn't have time to look into it
further. Maybe Lin can help here if this is the problematic commit.

I'll try to re-produce this issue first.

Ok, just a few more information. Suspend/resume cycle is needed to see
the issue. Reverting of 9ee4f3933930 helps when this is applied too (I
haven't tried one or the other alone):

For now, the simple fix is to disable runtime pm for hotpluggable port.

Could you please try below patch?

And please also show the debug info.
dmesg |grep "DEBUG: port"

From 28bf2f0592f231c2caba816428beaffc28873d8e Mon Sep 17 00:00:00 2001

From: Lin Ming <redacted>
Date: Thu, 23 Feb 2012 10:21:18 +0800
Subject: [PATCH] libata: disable runtime pm for hotpluggable port

---
 drivers/ata/ahci.c             |    3 +++
 drivers/ata/ahci.h             |    3 +++
 drivers/ata/libahci.c          |   21 +++++++++++++++++++++
 drivers/ata/libata-transport.c |    6 ++++--
 include/linux/libata.h         |    4 +++-
 5 files changed, 34 insertions(+), 3 deletions(-)

diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index d07bf03..02e93ff 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c

@@ -1145,6 +1145,9 @@ static int ahci_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (hpriv->cap & HOST_CAP_PMP)
 		pi.flags |= ATA_FLAG_PMP;
 
+	if (hpriv->cap & HOST_CAP_SXS)
+		pi.flags |= ATA_FLAG_EXTERNAL;
+
 	ahci_set_em_messages(hpriv, &pi);
 
 	if (ahci_broken_system_poweroff(pdev)) {

diff --git a/drivers/ata/ahci.h b/drivers/ata/ahci.h
index b175000..92f7172 100644
--- a/drivers/ata/ahci.h
+++ b/drivers/ata/ahci.h

@@ -172,6 +172,9 @@ enum {
 	PORT_CMD_ALPE		= (1 << 26), /* Aggressive Link PM enable */
 	PORT_CMD_ATAPI		= (1 << 24), /* Device is ATAPI */
 	PORT_CMD_FBSCP		= (1 << 22), /* FBS Capable Port */
+	PORT_CMD_ESP		= (1 << 21), /* External SATA Port */
+	PORT_CMD_MPSP		= (1 << 19), /* Mechanical Presence Switch Attached to Port */
+	PORT_CMD_HPCP		= (1 << 18), /* Hot Plug Capable Port */
 	PORT_CMD_PMP		= (1 << 17), /* PMP attached */
 	PORT_CMD_LIST_ON	= (1 << 15), /* cmd list DMA engine running */
 	PORT_CMD_FIS_ON		= (1 << 14), /* FIS DMA engine running */

diff --git a/drivers/ata/libahci.c b/drivers/ata/libahci.c
index a72bfd0..a0222e9 100644
--- a/drivers/ata/libahci.c
+++ b/drivers/ata/libahci.c

@@ -1097,6 +1097,24 @@ static void ahci_port_init(struct device *dev, struct ata_port *ap,
 	writel(1 << port_no, mmio + HOST_IRQ_STAT);
 }
 
+static bool ahci_port_pluggable(struct ata_port *ap)
+{
+	void __iomem *port_mmio = ahci_port_base(ap);
+	u32 cmd;
+
+	cmd = readl(port_mmio + PORT_CMD);
+	printk(KERN_ERR "DEBUG: port %d, PORT_CMD=0x%x\n", ap->port_no, cmd);
+
+	if ((ap->flags & ATA_FLAG_EXTERNAL) &&
+	    (cmd & PORT_CMD_ESP))
+		return true;
+
+	if (cmd & (PORT_CMD_MPSP | PORT_CMD_HPCP))
+		return true;
+
+	return false;
+}
+
 void ahci_init_controller(struct ata_host *host)
 {
 	struct ahci_host_priv *hpriv = host->private_data;

@@ -1112,6 +1130,9 @@ void ahci_init_controller(struct ata_host *host)
 		if (ata_port_is_dummy(ap))
 			continue;
 
+		if (ahci_port_pluggable(ap))
+			ap->flags |= ATA_FLAG_PLUGGABLE;
+
 		ahci_port_init(host->dev, ap, i, mmio, port_mmio);
 	}
 

diff --git a/drivers/ata/libata-transport.c b/drivers/ata/libata-transport.c
index 74aaee3..a7166b9 100644
--- a/drivers/ata/libata-transport.c
+++ b/drivers/ata/libata-transport.c

@@ -292,8 +292,10 @@ int ata_tport_add(struct device *parent,
 	}
 
 	device_enable_async_suspend(dev);
-	pm_runtime_set_active(dev);
-	pm_runtime_enable(dev);
+	if (!(ap->flags & ATA_FLAG_PLUGGABLE)) {
+		pm_runtime_set_active(dev);
+		pm_runtime_enable(dev);
+	}
 
 	transport_add_device(dev);
 	transport_configure_device(dev);

diff --git a/include/linux/libata.h b/include/linux/libata.h
index cafc09a..e8221cf 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h

@@ -207,8 +207,10 @@ enum {
 	ATA_FLAG_SW_ACTIVITY	= (1 << 22), /* driver supports sw activity
 					      * led */
 	ATA_FLAG_NO_DIPM	= (1 << 23), /* host not happy with DIPM */
+	ATA_FLAG_EXTERNAL	= (1 << 24), /* controller supports external SATA */
+	ATA_FLAG_PLUGGABLE	= (1 << 25), /* port is hotpluggable */
 
-	/* bits 24:31 of ap->flags are reserved for LLD specific flags */
+	/* bits 26:31 of ap->flags are reserved for LLD specific flags */
 
 
 	/* struct ata_port pflags */

-- 
1.7.9