diff options
| author | Linas Vepstas <linas@linas.org> | 2005-11-03 18:50:48 -0600 |
|---|---|---|
| committer | Paul Mackerras <paulus@samba.org> | 2005-11-10 16:00:32 +1100 |
| commit | d9564ad11454581f85b77026f290f4bb24eecf25 (patch) | |
| tree | 33aac0695124ad73a16a9b3d15458c4463804783 /arch/powerpc/platforms/pseries/eeh.c | |
| parent | 0c95fbb25602caa02ef697c3852cd61f9829e6ff (diff) | |
[PATCH] ppc64: mark failed devices
17-eeh-slot-marking-bug.patch
A device that experiences a PCI outage may be just one deivce out
of many that was affected. In order to avoid repeated reports of
a failure, the entire tree of affected devices should be marked
as failed. This patch marks up the entire tree.
Signed-off-by: Linas Vepstas <linas@linas.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
Diffstat (limited to 'arch/powerpc/platforms/pseries/eeh.c')
| -rw-r--r-- | arch/powerpc/platforms/pseries/eeh.c | 48 |
1 files changed, 32 insertions, 16 deletions
diff --git a/arch/powerpc/platforms/pseries/eeh.c b/arch/powerpc/platforms/pseries/eeh.c index 3a857b22aa1e..79de2310e70b 100644 --- a/arch/powerpc/platforms/pseries/eeh.c +++ b/arch/powerpc/platforms/pseries/eeh.c @@ -478,32 +478,47 @@ static struct device_node * find_device_pe(struct device_node *dn) * an interrupt context, which is bad. */ -static void __eeh_mark_slot (struct device_node *dn) +static void __eeh_mark_slot (struct device_node *dn, int mode_flag) { while (dn) { - PCI_DN(dn)->eeh_mode |= EEH_MODE_ISOLATED; + if (PCI_DN(dn)) { + PCI_DN(dn)->eeh_mode |= mode_flag; - if (dn->child) - __eeh_mark_slot (dn->child); + if (dn->child) + __eeh_mark_slot (dn->child, mode_flag); + } dn = dn->sibling; } } -static void __eeh_clear_slot (struct device_node *dn) +void eeh_mark_slot (struct device_node *dn, int mode_flag) +{ + dn = find_device_pe (dn); + PCI_DN(dn)->eeh_mode |= mode_flag; + __eeh_mark_slot (dn->child, mode_flag); +} + +static void __eeh_clear_slot (struct device_node *dn, int mode_flag) { while (dn) { - PCI_DN(dn)->eeh_mode &= ~EEH_MODE_ISOLATED; - if (dn->child) - __eeh_clear_slot (dn->child); + if (PCI_DN(dn)) { + PCI_DN(dn)->eeh_mode &= ~mode_flag; + PCI_DN(dn)->eeh_check_count = 0; + if (dn->child) + __eeh_clear_slot (dn->child, mode_flag); + } dn = dn->sibling; } } -static inline void eeh_clear_slot (struct device_node *dn) +void eeh_clear_slot (struct device_node *dn, int mode_flag) { unsigned long flags; spin_lock_irqsave(&confirm_error_lock, flags); - __eeh_clear_slot (dn); + dn = find_device_pe (dn); + PCI_DN(dn)->eeh_mode &= ~mode_flag; + PCI_DN(dn)->eeh_check_count = 0; + __eeh_clear_slot (dn->child, mode_flag); spin_unlock_irqrestore(&confirm_error_lock, flags); } @@ -528,7 +543,6 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) int rets[3]; unsigned long flags; struct pci_dn *pdn; - struct device_node *pe_dn; int rc = 0; __get_cpu_var(total_mmio_ffs)++; @@ -630,8 +644,7 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) /* Avoid repeated reports of this failure, including problems * with other functions on this device, and functions under * bridges. */ - pe_dn = find_device_pe (dn); - __eeh_mark_slot (pe_dn); + eeh_mark_slot (dn, EEH_MODE_ISOLATED); spin_unlock_irqrestore(&confirm_error_lock, flags); eeh_send_failure_event (dn, dev, rets[0], rets[2]); @@ -743,9 +756,6 @@ rtas_pci_slot_reset(struct pci_dn *pdn, int state) 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 @@ -764,6 +774,12 @@ rtas_set_slot_reset(struct pci_dn *pdn) #define PCI_BUS_RST_HOLD_TIME_MSEC 250 msleep (PCI_BUS_RST_HOLD_TIME_MSEC); + + /* We might get hit with another EEH freeze as soon as the + * pci slot reset line is dropped. Make sure we don't miss + * these, and clear the flag now. */ + eeh_clear_slot (pdn->node, EEH_MODE_ISOLATED); + rtas_pci_slot_reset (pdn, 0); /* After a PCI slot has been reset, the PCI Express spec requires |