All of lore.kernel.org
 help / color / mirror / Atom feed
From: Tiejun Chen <tiejun.chen@intel.com>
To: yang.z.zhang@intel.com, kevin.tian@intel.com, jbeulich@suse.com,
	jinsong.liu@alibaba-inc.com, keir@xen.org,
	andrew.cooper3@citrix.com
Cc: xen-devel@lists.xen.org
Subject: [PATCH 3/3] xen/iommu: disable IOMMU engine completely before enter S5
Date: Mon,  4 May 2015 10:16:11 +0800	[thread overview]
Message-ID: <1430705771-6744-3-git-send-email-tiejun.chen@intel.com> (raw)
In-Reply-To: <1430705771-6744-1-git-send-email-tiejun.chen@intel.com>

S5 menas we also shut down IOMMU engine as well, so we should
disable IOMMU engine completely before enter S5, otherwise we
still probably receive some interrupts in this period. But
currently VT-D path depends on force_iommu and iommu_intremap
to determine if we should disable translation and queued
invalidation. So this unexpected behaviour can be fixed by
clearing force_iommu and iommu_intremap. Additionally, some
cleanups to code comments.

Signed-off-by: Tiejun Chen <tiejun.chen@intel.com>
---
 xen/arch/x86/acpi/power.c           |  2 ++
 xen/arch/x86/crash.c                | 12 ++++++++----
 xen/drivers/passthrough/vtd/iommu.c |  5 +++--
 3 files changed, 13 insertions(+), 6 deletions(-)

diff --git a/xen/arch/x86/acpi/power.c b/xen/arch/x86/acpi/power.c
index f41f0de..345470d 100644
--- a/xen/arch/x86/acpi/power.c
+++ b/xen/arch/x86/acpi/power.c
@@ -183,6 +183,8 @@ static int enter_state(u32 state)
         break;
     case ACPI_STATE_S5:
         acpi_enter_sleep_state(ACPI_STATE_S5);
+        /* Make sure we really disable VT-d engine completely. */
+        force_iommu = iommu_intremap = 0;
         break;
     default:
         error = -EINVAL;
diff --git a/xen/arch/x86/crash.c b/xen/arch/x86/crash.c
index eb7be9c..123fb12 100644
--- a/xen/arch/x86/crash.c
+++ b/xen/arch/x86/crash.c
@@ -171,15 +171,19 @@ static void nmi_shootdown_cpus(void)
         printk("Failed to shoot down CPUs {%s}\n", keyhandler_scratch);
     }
 
-    /* Crash shutdown any IOMMU functionality as the crashdump kernel is not
-     * happy when booting if interrupt/dma remapping is still enabled */
+    /*
+     * Crash shutdown any IOMMU functionality as the crashdump kernel is not
+     * happy when booting if interrupt/dma remapping is still enabled.
+     */
     iommu_crash_shutdown();
 
     __stop_this_cpu();
 
-    /* This is a bit of a hack due to the problems with the x2apic_enabled
+    /*
+     * This is a bit of a hack due to the problems with the x2apic_enabled
      * variable, but we can't do any better without a significant refactoring
-     * of the APIC code */
+     * of the APIC code.
+     */
     x2apic_enabled = (current_local_apic_mode() == APIC_MODE_X2APIC);
 
     disable_IO_APIC();
diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c
index 72cd854..28244ec 100644
--- a/xen/drivers/passthrough/vtd/iommu.c
+++ b/xen/drivers/passthrough/vtd/iommu.c
@@ -2367,9 +2367,10 @@ static void vtd_suspend(void)
 
         iommu_disable_translation(iommu);
 
-        /* If interrupt remapping is enabled, queued invalidation
+        /*
+         * If interrupt remapping is enabled, queued invalidation
          * will be disabled following interupt remapping disabling
-         * in local apic suspend
+         * in local apic suspend.
          */
         if ( !iommu_intremap && iommu_qinval )
             disable_qinval(iommu);
-- 
1.9.1

  parent reply	other threads:[~2015-05-04  2:16 UTC|newest]

Thread overview: 33+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-05-04  2:16 [PATCH 1/3] xen/vt-d: need barriers to workaround CLFLUSH Tiejun Chen
2015-05-04  2:16 ` [PATCH 2/3] xen/vt-d: mask interrupt message generation Tiejun Chen
2015-05-04  4:07   ` Zhang, Yang Z
2015-05-04  5:08     ` Chen, Tiejun
2015-05-04  6:41       ` Zhang, Yang Z
2015-05-04  8:57       ` Jan Beulich
2015-05-04 11:21         ` Chen, Tiejun
2015-05-04  2:16 ` Tiejun Chen [this message]
2015-05-04  9:00   ` [PATCH 3/3] xen/iommu: disable IOMMU engine completely before enter S5 Jan Beulich
2015-05-04  4:07 ` [PATCH 1/3] xen/vt-d: need barriers to workaround CLFLUSH Zhang, Yang Z
2015-05-04  8:52 ` Jan Beulich
2015-05-04  9:14   ` Andrew Cooper
2015-05-04  9:24     ` Jan Beulich
2015-05-05  1:13       ` Zhang, Yang Z
2015-05-05 15:46     ` Boris Ostrovsky
2015-05-05 15:58       ` Jan Beulich
2015-05-05 16:11         ` Boris Ostrovsky
2015-05-06  7:12           ` Jan Beulich
2015-05-06  7:26             ` Chen, Tiejun
2015-05-06  7:33               ` Jan Beulich
2015-05-06 15:11             ` Boris Ostrovsky
2015-05-06  7:09       ` Chen, Tiejun
2015-05-04 10:39   ` Chen, Tiejun
2015-05-04 10:43     ` Jan Beulich
2015-05-04 10:52       ` Tian, Kevin
2015-05-04 11:26         ` Chen, Tiejun
2015-05-05  1:13           ` Tian, Kevin
2015-05-05  2:45       ` Chen, Tiejun
2015-05-05  9:24         ` Jan Beulich
2015-05-06  6:47           ` Chen, Tiejun
2015-05-06  7:14             ` Jan Beulich
2015-05-04 15:22     ` Konrad Rzeszutek Wilk
2015-05-04 15:33       ` Andrew Cooper

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=1430705771-6744-3-git-send-email-tiejun.chen@intel.com \
    --to=tiejun.chen@intel.com \
    --cc=andrew.cooper3@citrix.com \
    --cc=jbeulich@suse.com \
    --cc=jinsong.liu@alibaba-inc.com \
    --cc=keir@xen.org \
    --cc=kevin.tian@intel.com \
    --cc=xen-devel@lists.xen.org \
    --cc=yang.z.zhang@intel.com \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.