public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: linas <linas@austin.ibm.com>
To: paulus@samba.org
Cc: linuxppc64-dev@ozlabs.org, linux-kernel@vger.kernel.org,
	linux-pci@atrey.karlin.mff.cuni.cz
Subject: [PATCH 4/22] ppc64: EEH Recovery support routines
Date: Thu, 6 Oct 2005 18:28:31 -0500	[thread overview]
Message-ID: <20051006232831.GE29826@austin.ibm.com> (raw)
In-Reply-To: <20051006232032.GA29826@austin.ibm.com>


04-eeh-recovery-support-routines.patch

EEH Recovery support routines

This patch adds routines required to help drive the recovery of
EEH-frozen slots.  The main function is to drive the PCI #RST
signal line high for a qurter of a second, and then allow for 
a second & a half of settle time.

Signed-off-by: Linas Vepstas <linas@austin.ibm.com>


Index: linux-2.6.14-rc2-git6/arch/ppc64/kernel/pci.h
===================================================================
--- linux-2.6.14-rc2-git6.orig/arch/ppc64/kernel/pci.h	2005-10-06 17:50:31.847694674 -0500
+++ linux-2.6.14-rc2-git6/arch/ppc64/kernel/pci.h	2005-10-06 17:51:58.844488173 -0500
@@ -51,4 +51,18 @@
 extern unsigned long pci_assign_all_buses;
 extern int pci_read_irq_line(struct pci_dev *pci_dev);
 
+/* ---- EEH internal-use-only related routines ---- */
+#ifdef CONFIG_EEH
+/**
+ * rtas_set_slot_reset -- unfreeze a frozen slot
+ *
+ * Clear the EEH-frozen condition on a slot.  This routine
+ * does this by asserting the PCI #RST line for 1/8th of
+ * a second; this routine will sleep while the adapter is
+ * being reset.
+ */
+void rtas_set_slot_reset (struct pci_dn *);
+
+#endif
+
 #endif /* __PPC_KERNEL_PCI_H__ */
Index: linux-2.6.14-rc2-git6/arch/ppc64/kernel/eeh.c
===================================================================
--- linux-2.6.14-rc2-git6.orig/arch/ppc64/kernel/eeh.c	2005-10-06 17:50:31.847694674 -0500
+++ linux-2.6.14-rc2-git6/arch/ppc64/kernel/eeh.c	2005-10-06 17:52:27.908410223 -0500
@@ -17,6 +17,7 @@
  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
  */
 
+#include <linux/delay.h>
 #include <linux/init.h>
 #include <linux/list.h>
 #include <linux/pci.h>
@@ -678,6 +679,104 @@
 EXPORT_SYMBOL(eeh_check_failure);
 
 /* ------------------------------------------------------------- */
+/* The code below deals with error recovery */
+
+/** Return negative value if a permanent error, else return
+ * a number of milliseconds to wait until the PCI slot is
+ * ready to be used.
+ */
+static int
+eeh_slot_availability(struct pci_dn *pdn)
+{
+	int rc;
+	int rets[3];
+
+	rc = read_slot_reset_state(pdn, rets);
+
+	if (rc) return rc;
+
+	if (rets[1] == 0) return -1;  /* EEH is not supported */
+	if (rets[0] == 0)  return 0;  /* Oll Korrect */
+	if (rets[0] == 5) {
+		if (rets[2] == 0) return -1; /* permanently unavailable */
+		return rets[2]; /* number of millisecs to wait */
+	}
+	return -1;
+}
+
+/** rtas_pci_slot_reset raises/lowers the pci #RST line
+ *  state: 1/0 to raise/lower the #RST
+ *
+ * Clear the EEH-frozen condition on a slot.  This routine
+ * asserts the PCI #RST line if the 'state' argument is '1',
+ * and drops the #RST line if 'state is '0'.  This routine is
+ * safe to call in an interrupt context.
+ *
+ */
+
+static void
+rtas_pci_slot_reset(struct pci_dn *pdn, int state)
+{
+	int rc;
+
+	BUG_ON (pdn==NULL); 
+
+	if (!pdn->phb) {
+		printk (KERN_WARNING "EEH: in slot reset, device node %s has no phb\n",
+		        pdn->node->full_name);
+		return;
+	}
+
+	rc = rtas_call(ibm_set_slot_reset,4,1, NULL,
+	               pdn->eeh_config_addr,
+	               BUID_HI(pdn->phb->buid),
+	               BUID_LO(pdn->phb->buid),
+	               state);
+	if (rc) {
+		printk (KERN_WARNING "EEH: Unable to reset the failed slot, (%d) #RST=%d dn=%s\n", 
+		        rc, state, pdn->node->full_name);
+		return;
+	}
+
+	if (state == 0)
+		eeh_clear_slot (pdn->node->parent->child);
+}
+
+/** rtas_set_slot_reset -- assert the pci #RST line for 1/4 second
+ *  dn -- device node to be reset.
+ */
+
+void
+rtas_set_slot_reset(struct pci_dn *pdn)
+{
+	int i, rc;
+
+	rtas_pci_slot_reset (pdn, 1);
+
+	/* The PCI bus requires that the reset be held high for at least
+	 * a 100 milliseconds. We wait a bit longer 'just in case'.  */
+
+#define PCI_BUS_RST_HOLD_TIME_MSEC 250
+	msleep (PCI_BUS_RST_HOLD_TIME_MSEC);
+	rtas_pci_slot_reset (pdn, 0);
+
+	/* After a PCI slot has been reset, the PCI Express spec requires
+	 * a 1.5 second idle time for the bus to stabilize, before starting
+	 * up traffic. */
+#define PCI_BUS_SETTLE_TIME_MSEC 1800
+	msleep (PCI_BUS_SETTLE_TIME_MSEC);
+
+	/* Now double check with the firmware to make sure the device is
+	 * ready to be used; if not, wait for recovery. */
+	for (i=0; i<10; i++) {
+		rc = eeh_slot_availability (pdn);
+		if (rc <= 0) break;
+
+		msleep (rc+100);
+	}
+}
+
+/* ------------------------------------------------------------- */
 /* The code below deals with enabling EEH for devices during  the
  * early boot sequence.  EEH must be enabled before any PCI probing
  * can be done.

  parent reply	other threads:[~2005-10-06 23:28 UTC|newest]

Thread overview: 30+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2005-10-06 23:20 [PATCH 0/22] ppc64: Full sequence of PCI Error recovery patches linas
2005-10-06 23:23 ` [PATCH 1/22] ppc64: Dynamic LPAR bugfix linas
2005-10-06 23:25 ` [PATCH 2/22] ppc64: Enable detection bugfix linas
2005-10-06 23:26 ` [PATCH 3/22] ppc64: EEH Recovery dispatcher thread linas
2005-10-06 23:28 ` linas [this message]
2005-10-06 23:29 ` [PATCH 5/22] ppc64: Device BAR save and restore linas
2005-10-06 23:31 ` [PATCH 6/22] ppc64: PCI Error Recovery: documentation patch linas
2005-10-06 23:32 ` [PATCH 7/22] PCI Error Recovery: header file patch linas
2005-10-06 23:33 ` [PATCH 8/22] ppc64: Slot Marking Bugfix linas
2005-10-06 23:35 ` [PATCH 9/22] ppc64: DLPAR slot add and remove bugfixes linas
2005-10-06 23:36 ` [PATCH 10/22] ppc64: Crash on DLPAR PHB add linas
2005-10-06 23:39 ` [PATCH 11/22] ppc64: RPA PHP and EEH common code linas
2005-10-06 23:40 ` [PATCH 12/22] ppc64: RPA PHP cleanup linas
2005-10-06 23:44 ` [PATCH 13/22] ppc64: RPAPHP duplicated code removal linas
2005-10-06 23:46 ` [PATCH 14/22] ppc64: RPA PHP to EEH code movement linas
2006-01-07 21:28   ` Olaf Hering
2006-01-09 19:58     ` [PATCH]: ppowerpc: fix compile-time failure when EEH disabled linas
2005-10-06 23:47 ` [PATCH 15/22] ppc64: PCI Error Recovery: PPC64 core recovery routines linas
2005-10-12  9:49   ` Paul Mackerras
2005-10-13 16:03     ` linas
2005-10-06 23:53 ` [PATCH 16/22] PCI Address cache lookup code linas
2005-10-06 23:54 ` [PATCH 17/22] ppc64: New Partition Endpoin support linas
2005-10-06 23:55 ` [PATCH 18/22] PCI Error Recovery: IPR SCSI device driver linas
2005-10-06 23:56 ` [PATCH 19/22] PCI Error Recovery: Symbios " linas
2005-10-06 23:57 ` [PATCH 20/22] PCI Error Recovery: e100 network " linas
2005-10-11  0:10   ` Greg KH
2005-10-11 23:04     ` linas
2005-10-11 23:41       ` Paul Mackerras
2005-10-06 23:58 ` [PATCH 21/22] PCI Error Recovery: e1000 " linas
2005-10-06 23:59 ` [PATCH 22/22] PCI Error Recovery: ixgb " linas

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20051006232831.GE29826@austin.ibm.com \
    --to=linas@austin.ibm.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pci@atrey.karlin.mff.cuni.cz \
    --cc=linuxppc64-dev@ozlabs.org \
    --cc=paulus@samba.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox