From: Joerg Roedel <joerg.roedel@amd.com>
To: linux-kernel@vger.kernel.org
Cc: iommu@lists.linux-foundation.org, Joerg Roedel <joerg.roedel@amd.com>
Subject: [PATCH 14/23] AMD IOMMU: don't assing preallocated protection domains to devices
Date: Wed, 17 Sep 2008 18:52:48 +0200 [thread overview]
Message-ID: <1221670377-19295-15-git-send-email-joerg.roedel@amd.com> (raw)
In-Reply-To: <1221670377-19295-1-git-send-email-joerg.roedel@amd.com>
In isolation mode the protection domains for the devices are
preallocated and preassigned. This is bad if a device should be passed
to a virtualization guest because the IOMMU code does not know if it is
in use by a driver. This patch changes the code to assign the device to
the preallocated domain only if there were dma mapping requests for it.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
---
arch/x86/kernel/amd_iommu.c | 43 ++++++++++++++++++++++++++++++++----
include/asm-x86/amd_iommu_types.h | 6 +++++
2 files changed, 44 insertions(+), 5 deletions(-)
diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c
index 0410be1..09d5e92 100644
--- a/arch/x86/kernel/amd_iommu.c
+++ b/arch/x86/kernel/amd_iommu.c
@@ -33,6 +33,10 @@
static DEFINE_RWLOCK(amd_iommu_devtable_lock);
+/* A list of preallocated protection domains */
+static LIST_HEAD(iommu_pd_list);
+static DEFINE_SPINLOCK(iommu_pd_list_lock);
+
/*
* general struct to manage commands send to an IOMMU
*/
@@ -663,6 +667,7 @@ static struct dma_ops_domain *dma_ops_domain_alloc(struct amd_iommu *iommu,
dma_dom->next_bit = 0;
dma_dom->need_flush = false;
+ dma_dom->target_dev = 0xffff;
/* Intialize the exclusion range if necessary */
if (iommu->exclusion_start &&
@@ -769,6 +774,33 @@ static bool check_device(struct device *dev)
}
/*
+ * In this function the list of preallocated protection domains is traversed to
+ * find the domain for a specific device
+ */
+static struct dma_ops_domain *find_protection_domain(u16 devid)
+{
+ struct dma_ops_domain *entry, *ret = NULL;
+ unsigned long flags;
+
+ if (list_empty(&iommu_pd_list))
+ return NULL;
+
+ spin_lock_irqsave(&iommu_pd_list_lock, flags);
+
+ list_for_each_entry(entry, &iommu_pd_list, list) {
+ if (entry->target_dev == devid) {
+ ret = entry;
+ list_del(&ret->list);
+ break;
+ }
+ }
+
+ spin_unlock_irqrestore(&iommu_pd_list_lock, flags);
+
+ return ret;
+}
+
+/*
* In the dma_ops path we only have the struct device. This function
* finds the corresponding IOMMU, the protection domain and the
* requestor id for a given device.
@@ -803,9 +835,11 @@ static int get_device_resources(struct device *dev,
*iommu = amd_iommu_rlookup_table[*bdf];
if (*iommu == NULL)
return 0;
- dma_dom = (*iommu)->default_dom;
*domain = domain_for_device(*bdf);
if (*domain == NULL) {
+ dma_dom = find_protection_domain(*bdf);
+ if (!dma_dom)
+ dma_dom = (*iommu)->default_dom;
*domain = &dma_dom->domain;
set_device_domain(*iommu, *domain, *bdf);
printk(KERN_INFO "AMD IOMMU: Using protection domain %d for "
@@ -1257,10 +1291,9 @@ void prealloc_protection_domains(void)
if (!dma_dom)
continue;
init_unity_mappings_for_device(dma_dom, devid);
- set_device_domain(iommu, &dma_dom->domain, devid);
- printk(KERN_INFO "AMD IOMMU: Allocated domain %d for device ",
- dma_dom->domain.id);
- print_devid(devid, 1);
+ dma_dom->target_dev = devid;
+
+ list_add_tail(&dma_dom->list, &iommu_pd_list);
}
}
diff --git a/include/asm-x86/amd_iommu_types.h b/include/asm-x86/amd_iommu_types.h
index 1c9d8d4..56dc6bd 100644
--- a/include/asm-x86/amd_iommu_types.h
+++ b/include/asm-x86/amd_iommu_types.h
@@ -227,6 +227,12 @@ struct dma_ops_domain {
/* This will be set to true when TLB needs to be flushed */
bool need_flush;
+
+ /*
+ * if this is a preallocated domain, keep the device for which it was
+ * preallocated in this variable
+ */
+ u16 target_dev;
};
/*
--
1.5.6.4
next prev parent reply other threads:[~2008-09-17 16:56 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-09-17 16:52 [PATCH 0/23] AMD IOMMU 2.6.28 updates for review Joerg Roedel
2008-09-17 16:52 ` [PATCH 01/23] AMD IOMMU: check for invalid device pointers Joerg Roedel
2008-09-17 16:52 ` [PATCH 02/23] AMD IOMMU: move TLB flushing to the map/unmap helper functions Joerg Roedel
2008-09-17 16:52 ` [PATCH 03/23] AMD IOMMU: implement lazy IO/TLB flushing Joerg Roedel
2008-09-17 19:20 ` FUJITA Tomonori
2008-09-17 19:28 ` Joerg Roedel
2008-09-18 1:29 ` FUJITA Tomonori
2008-09-18 10:13 ` Joerg Roedel
2008-09-18 14:03 ` Joerg Roedel
2008-09-18 23:10 ` FUJITA Tomonori
2008-09-19 6:29 ` Joerg Roedel
2008-09-19 10:21 ` FUJITA Tomonori
2008-09-19 17:43 ` Joerg Roedel
2008-09-19 18:40 ` FUJITA Tomonori
2008-09-19 19:27 ` Joerg Roedel
2008-09-19 18:47 ` Keshavamurthy, Anil S
2008-09-19 18:47 ` Keshavamurthy, Anil S
2008-09-19 18:48 ` Keshavamurthy, Anil S
2008-09-21 9:05 ` Joerg Roedel
2008-09-22 16:26 ` David Woodhouse
2008-09-21 5:27 ` Muli Ben-Yehuda
2008-09-17 16:52 ` [PATCH 04/23] AMD IOMMU: add branch hints to completion wait checks Joerg Roedel
2008-09-17 16:52 ` [PATCH 05/23] AMD IOMMU: align alloc_coherent addresses properly Joerg Roedel
2008-09-17 16:52 ` [PATCH 06/23] AMD IOMMU: add event buffer allocation Joerg Roedel
2008-09-17 16:52 ` [PATCH 07/23] AMD IOMMU: save pci segment from ACPI tables Joerg Roedel
2008-09-17 16:52 ` [PATCH 08/23] AMD IOMMU: save pci_dev instead of devid Joerg Roedel
2008-09-17 16:52 ` [PATCH 09/23] AMD IOMMU: add MSI interrupt support Joerg Roedel
2008-09-17 16:52 ` [PATCH 10/23] AMD IOMMU: add event handling code Joerg Roedel
2008-09-17 16:52 ` [PATCH 11/23] AMD IOMMU: enable event logging Joerg Roedel
2008-09-17 16:52 ` [PATCH 12/23] AMD IOMMU: allow IO page faults from devices Joerg Roedel
2008-09-17 16:52 ` [PATCH 13/23] AMD IOMMU: add dma_supported callback Joerg Roedel
2008-09-17 16:52 ` Joerg Roedel [this message]
2008-09-17 16:52 ` [PATCH 15/23] AMD IOMMU: some set_device_domain cleanups Joerg Roedel
2008-09-17 16:52 ` [PATCH 16/23] AMD IOMMU: avoid unnecessary low zone allocation in alloc_coherent Joerg Roedel
2008-09-17 16:52 ` [PATCH 17/23] AMD IOMMU: replace memset with __GFP_ZERO " Joerg Roedel
2008-09-17 16:52 ` [PATCH 18/23] AMD IOMMU: simplify dma_mask_to_pages Joerg Roedel
2008-09-17 19:20 ` FUJITA Tomonori
2008-09-18 7:32 ` Joerg Roedel
2008-09-18 15:57 ` FUJITA Tomonori
2008-09-18 16:39 ` Joerg Roedel
2008-09-17 16:52 ` [PATCH 19/23] AMD IOMMU: free domain bitmap with its allocation order Joerg Roedel
2008-09-17 16:52 ` [PATCH 20/23] AMD IOMMU: remove unnecessary cast to u64 in the init code Joerg Roedel
2008-09-17 16:52 ` [PATCH 21/23] AMD IOMMU: calculate IVHD size with a function Joerg Roedel
2008-09-17 16:52 ` [PATCH 22/23] AMD IOMMU: use cmd_buf_size when freeing the command buffer Joerg Roedel
2008-09-17 16:52 ` [PATCH 23/23] add AMD IOMMU tree to MAINTAINERS file Joerg Roedel
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=1221670377-19295-15-git-send-email-joerg.roedel@amd.com \
--to=joerg.roedel@amd.com \
--cc=iommu@lists.linux-foundation.org \
--cc=linux-kernel@vger.kernel.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
all inboxes | Powered by JetHome®