From: Gavin Shan <gwshan@linux.vnet.ibm.com>
To: linuxppc-dev@lists.ozlabs.org
Cc: Gavin Shan <gwshan@linux.vnet.ibm.com>
Subject: [PATCH 05/21] powerpc/eeh: Clear frozen state on passing device
Date: Tue, 30 Sep 2014 12:38:54 +1000 [thread overview]
Message-ID: <1412044750-24460-5-git-send-email-gwshan@linux.vnet.ibm.com> (raw)
In-Reply-To: <1412044750-24460-1-git-send-email-gwshan@linux.vnet.ibm.com>
When passing through device, its PE might have been put into frozen
state. One obvious example would be: the passed PE is forced to be
offline because of hitting maximal allowed EEH errors in userland.
In that case, the frozen state won't be cleared and then the PE is
returned back to host, which might not have chance detecting and
recovering from it.
The patch adds more check when passing through device and clear the
PE frozen state if necessary.
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
---
arch/powerpc/kernel/eeh.c | 36 +++++++++++++++++++++++++++++++++++-
1 file changed, 35 insertions(+), 1 deletion(-)
diff --git a/arch/powerpc/kernel/eeh.c b/arch/powerpc/kernel/eeh.c
index db2841c..e436d5e 100644
--- a/arch/powerpc/kernel/eeh.c
+++ b/arch/powerpc/kernel/eeh.c
@@ -1150,6 +1150,8 @@ void eeh_remove_device(struct pci_dev *dev)
int eeh_dev_open(struct pci_dev *pdev)
{
struct eeh_dev *edev;
+ int flag = (EEH_STATE_MMIO_ACTIVE | EEH_STATE_DMA_ACTIVE);
+ int ret = -ENODEV;
mutex_lock(&eeh_dev_mutex);
@@ -1162,6 +1164,38 @@ int eeh_dev_open(struct pci_dev *pdev)
if (!edev || !edev->pe)
goto out;
+ /*
+ * The PE might have been put into frozen state, but we
+ * didn't detect that yet. The passed through PCI devices
+ * in frozen PE won't work properly. Clear the frozen state
+ * in advance.
+ */
+ ret = eeh_ops->get_state(edev->pe, NULL);
+ if (ret > 0 && ret != EEH_STATE_NOT_SUPPORT &&
+ (ret & flag) != flag) {
+ ret = eeh_ops->set_option(edev->pe, EEH_OPT_THAW_MMIO);
+ if (ret) {
+ pr_warn("%s: Failure %d enabling MMIO "
+ "for PHB#%x-PE#%x\n",
+ __func__, ret, edev->phb->global_number,
+ edev->pe->addr);
+ goto out;
+ }
+
+ ret = eeh_ops->set_option(edev->pe, EEH_OPT_THAW_DMA);
+ if (ret) {
+ pr_warn("%s: Failure %d enabling DMA "
+ "for PHB#%x-PE#%x\n",
+ __func__, ret, edev->phb->global_number,
+ edev->pe->addr);
+ goto out;
+ }
+ }
+
+ /* Clear software isolated state */
+ if (edev->pe->state & EEH_PE_ISOLATED)
+ eeh_pe_state_clear(edev->pe, EEH_PE_ISOLATED);
+
/* Increase PE's pass through count */
atomic_inc(&edev->pe->pass_dev_cnt);
mutex_unlock(&eeh_dev_mutex);
@@ -1169,7 +1203,7 @@ int eeh_dev_open(struct pci_dev *pdev)
return 0;
out:
mutex_unlock(&eeh_dev_mutex);
- return -ENODEV;
+ return ret;
}
EXPORT_SYMBOL_GPL(eeh_dev_open);
--
1.8.3.2
next prev parent reply other threads:[~2014-09-30 2:39 UTC|newest]
Thread overview: 23+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-09-30 2:38 [PATCH 01/21] powerpc/eeh: Drop unused argument in eeh_check_failure() Gavin Shan
2014-09-30 2:38 ` [PATCH 02/21] powerpc/eeh: Add eeh_pe_state sysfs entry Gavin Shan
2014-10-01 3:43 ` [02/21] " Michael Ellerman
2014-10-01 4:20 ` Gavin Shan
2014-09-30 2:38 ` [PATCH 03/21] powerpc/eeh: Freeze PE before PE reset Gavin Shan
2014-09-30 2:38 ` [PATCH 04/21] powerpc/eeh: Reenable PCI devices after reset Gavin Shan
2014-09-30 2:38 ` Gavin Shan [this message]
2014-09-30 2:38 ` [PATCH 06/21] powerpc/powernv: Sync header with firmware Gavin Shan
2014-09-30 2:38 ` [PATCH 07/21] powerpc/eeh: Introduce eeh_ops::err_inject Gavin Shan
2014-09-30 2:38 ` [PATCH 08/21] powerpc/powernv: Add PCI error injection debugfs entry Gavin Shan
2014-09-30 2:38 ` [PATCH 09/21] powerpc/powernv: Clear PAPR error injection registers Gavin Shan
2014-09-30 2:38 ` [PATCH 10/21] powerpc/eeh: Clear frozen device state in time Gavin Shan
2014-09-30 2:39 ` [PATCH 11/21] powerpc/eeh: Fix improper condition in eeh_pci_enable() Gavin Shan
2014-09-30 2:39 ` [PATCH 12/21] powerpc/eeh: Unfreeze PE on enabling EEH functionality Gavin Shan
2014-09-30 2:39 ` [PATCH 13/21] powerpc/eeh: Use eeh_unfreeze_pe() Gavin Shan
2014-09-30 2:39 ` [PATCH 14/21] powerpc/eeh: Block PCI config access during reset Gavin Shan
2014-09-30 2:39 ` [PATCH 15/21] powerpc/pseries: Decrease message level on EEH initialization Gavin Shan
2014-09-30 2:39 ` [PATCH 16/21] powerpc/powernv: Sync OpalPciResetScope with firmware Gavin Shan
2014-09-30 2:39 ` [PATCH 17/21] powerpc/eeh: Tag reset state for user owned PE Gavin Shan
2014-09-30 2:39 ` [PATCH 18/21] powerpc/eeh: Emulate EEH recovery for VFIO devices Gavin Shan
2014-09-30 2:39 ` [PATCH 19/21] powerpc/eeh: Dump PCI config space for all child devices Gavin Shan
2014-09-30 2:39 ` [PATCH 20/21] powerpc/powernv: Fetch frozen PE on top level Gavin Shan
2014-09-30 2:39 ` [PATCH 21/21] powerpc/powernv: Override dma_get_required_mask() Gavin Shan
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=1412044750-24460-5-git-send-email-gwshan@linux.vnet.ibm.com \
--to=gwshan@linux.vnet.ibm.com \
--cc=linuxppc-dev@lists.ozlabs.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;
as well as URLs for NNTP newsgroup(s).