From: Nishanth Aravamudan <nacc@linux.vnet.ibm.com>
To: benh@kernel.crashing.org
Cc: nfont@linux.vnet.ibm.com, paulus@samba.org,
linuxppc-dev@lists.ozlabs.org, miltonm@bga.com, anton@samba.org
Subject: [PATCH 2/2] pseries/iommu: remove DDW on kexec
Date: Mon, 28 Jan 2013 18:03:58 -0800 [thread overview]
Message-ID: <20130129020357.GB12156@linux.vnet.ibm.com> (raw)
In-Reply-To: <20130129020245.GA12156@linux.vnet.ibm.com>
pseries/iommu: remove DDW on kexec
We currently insert a property in the device-tree when we successfully
configure DDW for a given slot. This was meant to be an optimization to
speed up kexec/kdump, so that we don't need to make the RTAS calls again
to re-configured DDW in the new kernel.
However, we end up tripping a plpar_tce_stuff failure on kexec/kdump
because we unconditionally parse the ibm,dma-window property for the
node at bus/dev setup time. This property contains the 32-bit DMA window
LIOBN, which is distinct from the DDW window's. We pass that LIOBN (via
iommu_table_init -> iommu_table_clear -> tce_free ->
tce_freemulti_pSeriesLP) to plpar_tce_stuff, which fails because that
32-bit window is no longer present after
25ebc45b93452d0bc60271f178237123c4b26808 ("powerpc/pseries/iommu: remove
default window before attempting DDW manipulation").
I believe the simplest, easiest-to-maintain fix is to just change our
initcall to, rather than detecting and updating the new kernel's DDW
knowledge, just remove all DDW configurations. When the drivers
re-initialize, we will set everything back up as it was before.
Signed-off-by: Nishanth Aravamudan <nacc@us.ibm.com>
diff --git a/arch/powerpc/platforms/pseries/iommu.c b/arch/powerpc/platforms/pseries/iommu.c
index a8e99f9..1b2a174 100644
--- a/arch/powerpc/platforms/pseries/iommu.c
+++ b/arch/powerpc/platforms/pseries/iommu.c
@@ -787,33 +787,68 @@ static u64 find_existing_ddw(struct device_node *pdn)
return dma_addr;
}
+static void __restore_default_window(struct eeh_dev *edev,
+ u32 ddw_restore_token)
+{
+ u32 cfg_addr;
+ u64 buid;
+ int ret;
+
+ /*
+ * Get the config address and phb buid of the PE window.
+ * Rely on eeh to retrieve this for us.
+ * Retrieve them from the pci device, not the node with the
+ * dma-window property
+ */
+ cfg_addr = edev->config_addr;
+ if (edev->pe_config_addr)
+ cfg_addr = edev->pe_config_addr;
+ buid = edev->phb->buid;
+
+ do {
+ ret = rtas_call(ddw_restore_token, 3, 1, NULL, cfg_addr,
+ BUID_HI(buid), BUID_LO(buid));
+ } while (rtas_busy_delay(ret));
+ pr_info("ibm,reset-pe-dma-windows(%x) %x %x %x returned %d\n",
+ ddw_restore_token, cfg_addr, BUID_HI(buid), BUID_LO(buid), ret);
+}
+
static int find_existing_ddw_windows(void)
{
- int len;
struct device_node *pdn;
- struct direct_window *window;
const struct dynamic_dma_window_prop *direct64;
+ const u32 *ddw_extensions;
if (!firmware_has_feature(FW_FEATURE_LPAR))
return 0;
for_each_node_with_property(pdn, DIRECT64_PROPNAME) {
- direct64 = of_get_property(pdn, DIRECT64_PROPNAME, &len);
+ direct64 = of_get_property(pdn, DIRECT64_PROPNAME, NULL);
if (!direct64)
continue;
- window = kzalloc(sizeof(*window), GFP_KERNEL);
- if (!window || len < sizeof(struct dynamic_dma_window_prop)) {
- kfree(window);
- remove_ddw(pdn);
- continue;
- }
+ /*
+ * We need to ensure the IOMMU table is active when we
+ * return from the IOMMU setup so that the common code
+ * can clear the table or find the holes. To that end,
+ * first, remove any existing DDW configuration.
+ */
+ remove_ddw(pdn);
- window->device = pdn;
- window->prop = direct64;
- spin_lock(&direct_window_list_lock);
- list_add(&window->list, &direct_window_list);
- spin_unlock(&direct_window_list_lock);
+ /*
+ * Second, if we are running on a new enough level of
+ * firmware where the restore API is present, use it to
+ * restore the 32-bit window, which was removed in
+ * create_ddw.
+ * If the API is not present, then create_ddw couldn't
+ * have removed the 32-bit window in the first place, so
+ * removing the DDW configuration should be sufficient.
+ */
+ ddw_extensions = of_get_property(pdn, "ibm,ddw-extensions",
+ NULL);
+ if (ddw_extensions && ddw_extensions[0] > 0)
+ __restore_default_window(of_node_to_eeh_dev(pdn),
+ ddw_extensions[1]);
}
return 0;
@@ -886,30 +921,7 @@ static int create_ddw(struct pci_dev *dev, const u32 *ddw_avail,
static void restore_default_window(struct pci_dev *dev,
u32 ddw_restore_token)
{
- struct eeh_dev *edev;
- u32 cfg_addr;
- u64 buid;
- int ret;
-
- /*
- * Get the config address and phb buid of the PE window.
- * Rely on eeh to retrieve this for us.
- * Retrieve them from the pci device, not the node with the
- * dma-window property
- */
- edev = pci_dev_to_eeh_dev(dev);
- cfg_addr = edev->config_addr;
- if (edev->pe_config_addr)
- cfg_addr = edev->pe_config_addr;
- buid = edev->phb->buid;
-
- do {
- ret = rtas_call(ddw_restore_token, 3, 1, NULL, cfg_addr,
- BUID_HI(buid), BUID_LO(buid));
- } while (rtas_busy_delay(ret));
- dev_info(&dev->dev,
- "ibm,reset-pe-dma-windows(%x) %x %x %x returned %d\n",
- ddw_restore_token, cfg_addr, BUID_HI(buid), BUID_LO(buid), ret);
+ __restore_default_window(pci_dev_to_eeh_dev(dev), ddw_restore_token);
}
/*
next prev parent reply other threads:[~2013-01-29 2:04 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-01-29 2:02 [PATCH 1/2] pseries/iommu: restore_default_window does not use liobn parameter Nishanth Aravamudan
2013-01-29 2:03 ` Nishanth Aravamudan [this message]
2013-01-29 10:58 ` [PATCH 2/2] pseries/iommu: remove DDW on kexec Michael Ellerman
2013-01-29 20:33 ` Nishanth Aravamudan
2013-02-04 3:38 ` Michael Ellerman
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=20130129020357.GB12156@linux.vnet.ibm.com \
--to=nacc@linux.vnet.ibm.com \
--cc=anton@samba.org \
--cc=benh@kernel.crashing.org \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=miltonm@bga.com \
--cc=nfont@linux.vnet.ibm.com \
--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;
as well as URLs for NNTP newsgroup(s).