public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Ashok Raj <ashok.raj@intel.com>
To: linux-kernel@vger.kernel.org
Cc: akpm@osdl.org, ak@suse.de, gregkh@suse.de, muli@il.ibm.com,
	asit.k.mallick@intel.com, suresh.b.siddha@intel.com,
	anil.s.keshavamurthy@intel.com, arjan@linux.intel.com,
	ashok.raj@intel.com, shaohua.li@intel.com
Subject: [patch 4/8] [Intel IOMMU] Supporting Zero Length Reads in Intel IOMMU.
Date: Mon, 09 Apr 2007 14:55:56 -0700	[thread overview]
Message-ID: <20070409215723.795646000@intel.com> (raw)
In-Reply-To: 20070409215552.221374000@intel.com

[-- Attachment #1: rw-workaround.patch --]
[-- Type: text/plain, Size: 2481 bytes --]

PCI specs permit zero length reads (ZLR) even if the mapping for that region 
is write only. Support for this feature is indicated by the presence of a bit 
in the DMAR capability. If a particular DMAR does not support this capability
we map write-only regions as read-write.

This option can also provides a workaround for some drivers that request
a write-only mapping when they really should request a read-write.
(We ran into one such case in eepro100.c in handling rx_ring_dma)

Signed-off-by: Ashok Raj <ashok.raj@intel.com>
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
----------------------------------
 drivers/pci/intel-iommu.c |    7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

Index: linux-2.6.21-rc5/drivers/pci/intel-iommu.c
===================================================================
--- linux-2.6.21-rc5.orig/drivers/pci/intel-iommu.c	2007-04-09 03:05:25.000000000 -0700
+++ linux-2.6.21-rc5/drivers/pci/intel-iommu.c	2007-04-09 03:05:32.000000000 -0700
@@ -84,7 +84,7 @@
 	struct sys_device sysdev;
 };
 
-static int dmar_disabled;
+static int dmar_disabled, dmar_force_rw;
 
 static char *get_fault_reason(u8 fault_reason)
 {
@@ -102,6 +102,9 @@
 		if (!strncmp(str, "off", 3)) {
 			dmar_disabled = 1;
 			printk(KERN_INFO"Intel-IOMMU: disabled\n");
+		} else if (!strncmp(str, "forcerw", 7)) {
+			dmar_force_rw = 1;
+			printk(KERN_INFO"Intel-IOMMU: force R/W for W/O mapping\n");
 		}
 		str += strcspn(str, ",");
 		while (*str == ',')
@@ -1668,7 +1671,12 @@
 			goto error;
 	}
 
-	if (dir == DMA_TO_DEVICE || dir == DMA_BIDIRECTIONAL)
+	/*
+	 * Check if DMAR supports zero-length reads on write only
+	 * mappings..
+	 */
+	if (dir == DMA_TO_DEVICE || dir == DMA_BIDIRECTIONAL || \
+			!cap_zlr(domain->iommu->cap) || dmar_force_rw)
 		prot |= DMA_PTE_READ;
 	if (dir == DMA_FROM_DEVICE || dir == DMA_BIDIRECTIONAL)
 		prot |= DMA_PTE_WRITE;
Index: linux-2.6.21-rc5/include/linux/intel-iommu.h
===================================================================
--- linux-2.6.21-rc5.orig/include/linux/intel-iommu.h	2007-04-09 03:05:25.000000000 -0700
+++ linux-2.6.21-rc5/include/linux/intel-iommu.h	2007-04-09 03:05:32.000000000 -0700
@@ -79,6 +79,7 @@
 #define cap_max_fault_reg_offset(c) \
 	(cap_fault_reg_offset(c) + cap_num_fault_regs(c) * 16)
 
+#define cap_zlr(c)		(((c) >> 22) & 1)
 #define cap_isoch(c)		(((c) >> 23) & 1)
 #define cap_mgaw(c)		((((c) >> 16) & 0x3f) + 1)
 #define cap_sagaw(c)		(((c) >> 8) & 0x1f)

-- 

  parent reply	other threads:[~2007-04-09 22:49 UTC|newest]

Thread overview: 31+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-04-09 21:55 [patch 0/8] [Intel IOMMU] Support for Intel Virtualization Technology for Directed I/O Ashok Raj
2007-04-09 21:55 ` [patch 1/8] [Intel IOMMU] ACPI support " Ashok Raj
2007-04-10  3:39   ` Len Brown
2007-04-10 16:26     ` Ashok Raj
2007-04-09 21:55 ` [patch 2/8] [Intel IOMMU] Some generic search functions required to lookup device relationships Ashok Raj
2007-04-10  3:46   ` Greg KH
2007-04-10  8:11     ` Shaohua Li
2007-04-10 13:03       ` Greg KH
2007-04-11  1:40         ` Shaohua Li
2007-04-11  4:36           ` Greg KH
2007-04-11  6:10         ` Shaohua Li
2007-04-09 21:55 ` [patch 3/8] [Intel IOMMU] Generic hardware support for Intel IOMMU Ashok Raj
2007-04-09 21:55 ` Ashok Raj [this message]
2007-04-09 21:55 ` [patch 5/8] [Intel IOMMU] Graphics driver workarounds to provide unity map Ashok Raj
2007-04-10  8:33   ` Christoph Hellwig
2007-04-10  9:07     ` David Miller
2007-04-10  9:12       ` Andi Kleen
2007-04-11  2:40         ` Wang Zhenyu
2007-04-10 16:29       ` Arjan van de Ven
2007-04-09 21:55 ` [patch 6/8] [Intel IOMMU] Doc updates for Intel Virtualization Technology for Directed I/O Ashok Raj
2007-04-09 21:55 ` [patch 7/8] [Intel IOMMU] Support for legacy ISA devices Ashok Raj
2007-04-09 21:56 ` [patch 8/8] [Intel IOMMU] Preserve some Virtual Address when devices cannot address entire range Ashok Raj
2007-04-10  7:49 ` [patch 0/8] [Intel IOMMU] Support for Intel Virtualization Technology for Directed I/O Andi Kleen
2007-04-10  7:57   ` Shaohua Li
2007-04-10  8:09     ` Muli Ben-Yehuda
2007-04-10  8:20       ` Shaohua Li
2007-04-10 16:31   ` Ashok Raj
2007-04-10  8:21 ` Jeff Garzik
2007-04-10  8:27   ` Shaohua Li
2007-04-10  8:34     ` Jeff Garzik
2007-04-10 16:43       ` Ashok Raj

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=20070409215723.795646000@intel.com \
    --to=ashok.raj@intel.com \
    --cc=ak@suse.de \
    --cc=akpm@osdl.org \
    --cc=anil.s.keshavamurthy@intel.com \
    --cc=arjan@linux.intel.com \
    --cc=asit.k.mallick@intel.com \
    --cc=gregkh@suse.de \
    --cc=linux-kernel@vger.kernel.org \
    --cc=muli@il.ibm.com \
    --cc=shaohua.li@intel.com \
    --cc=suresh.b.siddha@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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox