From: Joseph Cihula <joseph.cihula@linux.intel.com>
To: linux-kernel@vger.kernel.org
Cc: shane.wang@intel.com, gang.wei@intel.com, arjan@linux.intel.com,
asit.k.mallick@intel.com, jun.nakajima@intel.com,
chrisw@sous-sol.org, jbeulich@novell.com, mingo@elte.hu,
tytso@mit.edu, linux-kernel@vger.kernel.org,
joseph.cihula@intel.com
Subject: [RFC][PATCH 2/3] TXT: Intel IOMMU PMR Disable
Date: Tue, 07 Oct 2008 16:22:59 -0700 [thread overview]
Message-ID: <48EBEF53.9060904@linux.intel.com> (raw)
(re-posted with apologies for formatting and email addresses)
Disable Intel VT-d (IOMMU) Protected Memory Regions (PMRs) when DMA
remapping is enabled.
Leaving the PMRs enabled may add more delay to the VT-d DMA request
processing (as another lookup may be performed) but whether they are
even used is not deterministic:
DMA requests that are subject to address remapping, and accessing the
protected memory regions may or may not be blocked by hardware. For
such requests, software must not depend on hardware protection of the
protected memory regions, and instead program the DMA-remapping
page-tables to block DMA to protected memory regions.
-- from the description field of the PMEN_REG register in the v1.2
(May '08) VT-d spec.
Signed-off-by: Shane Wang <shane.wang@intel.com>
Signed-off-by: Joseph Cihula <joseph.cihula@intel.com>
---
diff -r 29cfdc52a076 drivers/pci/intel-iommu.c
--- a/drivers/pci/intel-iommu.c Mon Oct 06 12:50:47 2008 -0700
+++ b/drivers/pci/intel-iommu.c Mon Oct 06 15:54:59 2008 -0700
@@ -742,6 +742,22 @@ static void iommu_disable_protect_mem_re
spin_unlock_irqrestore(&iommu->register_lock, flags);
}
+static void disable_pmr(struct intel_iommu *iommu)
+{
+ u32 data;
+
+ data = readl(iommu->reg + DMAR_PMEN_REG);
+ if (!(data & DMA_PMEN_PRS))
+ return;
+
+ data = data & ~DMA_PMEN_EPM;
+ writel(data, iommu->reg + DMAR_PMEN_REG);
+
+ IOMMU_WAIT_OP(iommu, DMAR_PMEN_REG,
+ readl, !(data & DMA_PMEN_PRS), data);
+ return;
+}
+
static int iommu_enable_translation(struct intel_iommu *iommu)
{
u32 sts;
@@ -755,6 +771,7 @@ static int iommu_enable_translation(stru
readl, (sts & DMA_GSTS_TES), sts);
iommu->gcmd |= DMA_GCMD_TE;
+ disable_pmr(iommu);
spin_unlock_irqrestore(&iommu->register_lock, flags);
return 0;
}
diff -r 29cfdc52a076 drivers/pci/intel-iommu.h
--- a/drivers/pci/intel-iommu.h Mon Oct 06 12:50:47 2008 -0700
+++ b/drivers/pci/intel-iommu.h Mon Oct 06 15:54:59 2008 -0700
@@ -158,6 +158,10 @@ static inline void dmar_writeq(void __io
#define DMA_GSTS_FLS (((u32)1) << 29)
#define DMA_GSTS_AFLS (((u32)1) << 28)
#define DMA_GSTS_WBFS (((u32)1) << 27)
+
+/* PMEN_REG */
+#define DMA_PMEN_EPM (((u32)1) << 31)
+#define DMA_PMEN_PRS (((u32)1) << 0)
/* CCMD_REG */
#define DMA_CCMD_ICC (((u64)1) << 63)
next reply other threads:[~2008-10-07 23:24 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-10-07 23:22 Joseph Cihula [this message]
-- strict thread matches above, loose matches on Subject: below --
2008-10-07 20:34 [RFC][PATCH 2/3] TXT: Intel IOMMU PMR Disable Cihula, Joseph
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=48EBEF53.9060904@linux.intel.com \
--to=joseph.cihula@linux.intel.com \
--cc=arjan@linux.intel.com \
--cc=asit.k.mallick@intel.com \
--cc=chrisw@sous-sol.org \
--cc=gang.wei@intel.com \
--cc=jbeulich@novell.com \
--cc=joseph.cihula@intel.com \
--cc=jun.nakajima@intel.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=shane.wang@intel.com \
--cc=tytso@mit.edu \
/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