From: James Gowans <jgowans@amazon.com>
To: <linux-kernel@vger.kernel.org>
Cc: Eric Biederman <ebiederm@xmission.com>,
<kexec@lists.infradead.org>, "Joerg Roedel" <joro@8bytes.org>,
Will Deacon <will@kernel.org>, <iommu@lists.linux.dev>,
Alexander Viro <viro@zeniv.linux.org.uk>,
"Christian Brauner" <brauner@kernel.org>,
<linux-fsdevel@vger.kernel.org>,
Paolo Bonzini <pbonzini@redhat.com>,
Sean Christopherson <seanjc@google.com>, <kvm@vger.kernel.org>,
Andrew Morton <akpm@linux-foundation.org>, <linux-mm@kvack.org>,
Alexander Graf <graf@amazon.com>,
David Woodhouse <dwmw@amazon.co.uk>,
"Jan H . Schoenherr" <jschoenh@amazon.de>,
Usama Arif <usama.arif@bytedance.com>,
Anthony Yznaga <anthony.yznaga@oracle.com>,
Stanislav Kinsburskii <skinsburskii@linux.microsoft.com>,
<madvenka@linux.microsoft.com>, <steven.sistare@oracle.com>,
<yuleixzhang@tencent.com>
Subject: [RFC 10/18] iommu/intel: zap context table entries on kexec
Date: Mon, 5 Feb 2024 12:01:55 +0000 [thread overview]
Message-ID: <20240205120203.60312-11-jgowans@amazon.com> (raw)
In-Reply-To: <20240205120203.60312-1-jgowans@amazon.com>
In the next commit the IOMMU shutdown function will be modified to not
actually shut down the IOMMU when doing a kexec. To prevent leaving DMA
mappings for non-persistent devices around during kexec we add a
function to the kexec flow which iterates though all IOMMU domains and
zaps the context entries for the devices belonging to those domain.
A list of domains for the IOMMU is added and maintained.
---
drivers/iommu/intel/dmar.c | 1 +
drivers/iommu/intel/iommu.c | 34 ++++++++++++++++++++++++++++++----
drivers/iommu/intel/iommu.h | 2 ++
3 files changed, 33 insertions(+), 4 deletions(-)
diff --git a/drivers/iommu/intel/dmar.c b/drivers/iommu/intel/dmar.c
index 23cb80d62a9a..00f69f40a4ac 100644
--- a/drivers/iommu/intel/dmar.c
+++ b/drivers/iommu/intel/dmar.c
@@ -1097,6 +1097,7 @@ static int alloc_iommu(struct dmar_drhd_unit *drhd)
iommu->segment = drhd->segment;
iommu->node = NUMA_NO_NODE;
+ INIT_LIST_HEAD(&iommu->domains);
ver = readl(iommu->reg + DMAR_VER_REG);
pr_info("%s: reg_base_addr %llx ver %d:%d cap %llx ecap %llx\n",
diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c
index 2dd3f055dbce..315c6b7f901c 100644
--- a/drivers/iommu/intel/iommu.c
+++ b/drivers/iommu/intel/iommu.c
@@ -1831,6 +1831,7 @@ static int domain_attach_iommu(struct dmar_domain *domain,
goto err_clear;
}
domain_update_iommu_cap(domain);
+ list_add(&domain->domains, &iommu->domains);
spin_unlock(&iommu->lock);
return 0;
@@ -3608,6 +3609,33 @@ static void intel_disable_iommus(void)
iommu_disable_translation(iommu);
}
+void zap_context_table_entries(struct intel_iommu *iommu)
+{
+ struct context_entry *context;
+ struct dmar_domain *domain;
+ struct device_domain_info *device;
+ int bus, devfn;
+ u16 did_old;
+
+ list_for_each_entry(domain, &iommu->domains, domains) {
+ list_for_each_entry(device, &domain->devices, link) {
+ context = iommu_context_addr(iommu, device->bus, device->devfn, 0);
+ if (!context || !context_present(context))
+ continue;
+ context_domain_id(context);
+ context_clear_entry(context);
+ __iommu_flush_cache(iommu, context, sizeof(*context));
+ iommu->flush.flush_context(iommu,
+ did_old,
+ (((u16)bus) << 8) | devfn,
+ DMA_CCMD_MASK_NOBIT,
+ DMA_CCMD_DEVICE_INVL);
+ iommu->flush.flush_iotlb(iommu, did_old, 0, 0,
+ DMA_TLB_DSI_FLUSH);
+ }
+ }
+}
+
void intel_iommu_shutdown(void)
{
struct dmar_drhd_unit *drhd;
@@ -3620,10 +3648,8 @@ void intel_iommu_shutdown(void)
/* Disable PMRs explicitly here. */
for_each_iommu(iommu, drhd)
- iommu_disable_protect_mem_regions(iommu);
-
- /* Make sure the IOMMUs are switched off */
- intel_disable_iommus();
+ zap_context_table_entries(iommu);
+ return
up_write(&dmar_global_lock);
}
diff --git a/drivers/iommu/intel/iommu.h b/drivers/iommu/intel/iommu.h
index a2338e398ba3..4a2f163a86f3 100644
--- a/drivers/iommu/intel/iommu.h
+++ b/drivers/iommu/intel/iommu.h
@@ -600,6 +600,7 @@ struct dmar_domain {
spinlock_t lock; /* Protect device tracking lists */
struct list_head devices; /* all devices' list */
struct list_head dev_pasids; /* all attached pasids */
+ struct list_head domains; /* all struct dmar_domains on this IOMMU */
struct dma_pte *pgd; /* virtual address */
int gaw; /* max guest address width */
@@ -700,6 +701,7 @@ struct intel_iommu {
void *perf_statistic;
struct iommu_pmu *pmu;
+ struct list_head domains; /* all struct dmar_domains on this IOMMU */
};
/* PCI domain-device relationship */
--
2.40.1
next prev parent reply other threads:[~2024-02-05 12:04 UTC|newest]
Thread overview: 29+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-02-05 12:01 [RFC 00/18] Pkernfs: Support persistence for live update James Gowans
2024-02-05 12:01 ` [RFC 01/18] pkernfs: Introduce filesystem skeleton James Gowans
2024-02-05 12:01 ` [RFC 02/18] pkernfs: Add persistent inodes hooked into directies James Gowans
2024-02-05 12:01 ` [RFC 03/18] pkernfs: Define an allocator for persistent pages James Gowans
2024-02-05 12:01 ` [RFC 04/18] pkernfs: support file truncation James Gowans
2024-02-05 12:01 ` [RFC 05/18] pkernfs: add file mmap callback James Gowans
2024-02-05 23:34 ` Dave Chinner
2024-02-05 12:01 ` [RFC 06/18] init: Add liveupdate cmdline param James Gowans
2024-02-05 12:01 ` [RFC 07/18] pkernfs: Add file type for IOMMU root pgtables James Gowans
2024-02-05 12:01 ` [RFC 08/18] iommu: Add allocator for pgtables from persistent region James Gowans
2024-02-05 12:01 ` [RFC 09/18] intel-iommu: Use pkernfs for root/context pgtable pages James Gowans
2024-02-05 12:01 ` James Gowans [this message]
2024-02-05 12:01 ` [RFC 11/18] dma-iommu: Always enable deferred attaches for liveupdate James Gowans
2024-02-05 17:45 ` Jason Gunthorpe
2024-02-05 12:01 ` [RFC 12/18] pkernfs: Add IOMMU domain pgtables file James Gowans
2024-02-05 12:01 ` [RFC 13/18] vfio: add ioctl to define persistent pgtables on container James Gowans
2024-02-05 17:08 ` Jason Gunthorpe
2024-02-05 12:01 ` [RFC 14/18] intel-iommu: Allocate domain pgtable pages from pkernfs James Gowans
2024-02-05 17:12 ` Jason Gunthorpe
2024-02-05 12:02 ` [RFC 15/18] pkernfs: register device memory for IOMMU domain pgtables James Gowans
2024-02-05 12:02 ` [RFC 16/18] vfio: support not mapping IOMMU pgtables on live-update James Gowans
2024-02-05 12:02 ` [RFC 17/18] pci: Don't clear bus master is persistence enabled James Gowans
2024-02-05 12:02 ` [RFC 18/18] vfio-pci: Assume device working after liveupdate James Gowans
2024-02-05 17:10 ` [RFC 00/18] Pkernfs: Support persistence for live update Alex Williamson
2024-02-07 14:56 ` Gowans, James
2024-02-07 15:28 ` Jason Gunthorpe
2024-02-05 17:42 ` Jason Gunthorpe
2024-02-07 14:45 ` Gowans, James
2024-02-07 15:22 ` Jason Gunthorpe
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=20240205120203.60312-11-jgowans@amazon.com \
--to=jgowans@amazon.com \
--cc=akpm@linux-foundation.org \
--cc=anthony.yznaga@oracle.com \
--cc=brauner@kernel.org \
--cc=dwmw@amazon.co.uk \
--cc=ebiederm@xmission.com \
--cc=graf@amazon.com \
--cc=iommu@lists.linux.dev \
--cc=joro@8bytes.org \
--cc=jschoenh@amazon.de \
--cc=kexec@lists.infradead.org \
--cc=kvm@vger.kernel.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=madvenka@linux.microsoft.com \
--cc=pbonzini@redhat.com \
--cc=seanjc@google.com \
--cc=skinsburskii@linux.microsoft.com \
--cc=steven.sistare@oracle.com \
--cc=usama.arif@bytedance.com \
--cc=viro@zeniv.linux.org.uk \
--cc=will@kernel.org \
--cc=yuleixzhang@tencent.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