Skip to contentSkip to conversation

AMD IOMMU fixes for 2.6.31-rc1

GIT PULL New Joerg Roedel 2 participants 4 messages Jul 3, 2009 x86

lore
Share & tools
lore
#0 lore raw

Hi Ingo,

The following changes since commit 7c5371c403abb29f01bc6cff6c5096abdf2dc524:
Yinghai Lu (1):
x86: add boundary check for 32bit res before expand e820 resource to alignment

are available in the git repository at:

git://git.kernel.org/pub/scm/linux/kernel/git/joro/linux-2.6-iommu.git amd-iommu/fixes

Joerg Roedel (2):
amd-iommu: handle alias entries correctly in init code
amd-iommu: set evt_buf_size correctly

 arch/x86/kernel/amd_iommu_init.c |   13 +++++++++----
 1 files changed, 9 insertions(+), 4 deletions(-)

These two patches fix an uncorrect device table initialization issue and a divide-by-zero error on an event log interrupt. Please pull.

Thanks,

Joerg

#1 lore raw

PATCH 1/2 amd-iommu: handle alias entries correctly in init code

An alias entry in the ACPI table means that the device can send requests to the IOMMU with both device ids, its own and the alias. This is not handled properly in the ACPI init code. This patch fixes the issue.

---
 arch/x86/kernel/amd_iommu_init.c |   11 +++++++----
 1 files changed, 7 insertions(+), 4 deletions(-)
arch/x86/kernel/amd_iommu_init.c +7 -4
index 10b2acc..ec72c77 100644
@@ -691,6 +691,7 @@ static void __init init_iommu_from_acpi(struct amd_iommu *iommu,
 
 			devid = e->devid;
 			devid_to = e->ext >> 8;
+			set_dev_entry_from_acpi(iommu, devid   , e->flags, 0);
 			set_dev_entry_from_acpi(iommu, devid_to, e->flags, 0);
 			amd_iommu_alias_table[devid] = devid_to;
 			break;
@@ -749,11 +750,13 @@ static void __init init_iommu_from_acpi(struct amd_iommu *iommu,
 
 			devid = e->devid;
 			for (dev_i = devid_start; dev_i <= devid; ++dev_i) {
-				if (alias)
+				if (alias) {
 					amd_iommu_alias_table[dev_i] = devid_to;
-				set_dev_entry_from_acpi(iommu,
-						amd_iommu_alias_table[dev_i],
-						flags, ext_flags);
+					set_dev_entry_from_acpi(iommu,
+						devid_to, flags, ext_flags);
+				}
+				set_dev_entry_from_acpi(iommu, dev_i,
+							flags, ext_flags);
 			}
 			break;
 		default:
Signature · 1 line
1.6.3.3
#2 lore raw

PATCH 2/2 amd-iommu: set evt_buf_size correctly

In reply to Joerg Roedel

The setting of this variable got lost during the suspend/resume implementation. But keeping this variable zero causes a divide-by-zero error in the interrupt handler. This patch fixes this.

---
 arch/x86/kernel/amd_iommu_init.c |    2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)
arch/x86/kernel/amd_iommu_init.c +2 -0
index ec72c77..c1b17e9 100644
@@ -472,6 +472,8 @@ static u8 * __init alloc_event_buffer(struct amd_iommu *iommu)
 	if (iommu->evt_buf == NULL)
 		return NULL;
 
+	iommu->evt_buf_size = EVT_BUFFER_SIZE;
+
 	return iommu->evt_buf;
 }
 
Signature · 1 line
1.6.3.3
#3 lore raw
In reply to Joerg Roedel
Quoted 14 lines · Joerg Roedel

Hi Ingo,

The following changes since commit 7c5371c403abb29f01bc6cff6c5096abdf2dc524:
Yinghai Lu (1):
x86: add boundary check for 32bit res before expand e820 resource to alignment

are available in the git repository at:

git://git.kernel.org/pub/scm/linux/kernel/git/joro/linux-2.6-iommu.git amd-iommu/fixes

Joerg Roedel (2):
amd-iommu: handle alias entries correctly in init code
amd-iommu: set evt_buf_size correctly

Quoted 6 lines
 arch/x86/kernel/amd_iommu_init.c |   13 +++++++++----
 1 files changed, 9 insertions(+), 4 deletions(-)

These two patches fix an uncorrect device table initialization issue and a divide-by-zero error on an event log interrupt. Please pull.

Pulled, thanks a lot Joerg!

Ingo