When unbinding a device so that I could pass it through to a KVM VM, I
got the lockdep report below. It looks like a legitimate lock
ordering problem:
- domain_context_mapping_one() takes iommu->lock and calls
iommu_support_dev_iotlb(), which takes device_domain_lock (inside
iommu->lock).
- domain_remove_one_dev_info() starts by taking device_domain_lock
then takes iommu->lock inside it (near the end of the function).
So this is the classic AB-BA deadlock. It looks like a safe fix is to
simply release device_domain_lock a bit earlier, since as far as I can
tell, it doesn't protect any of the stuff accessed at the end of
domain_remove_one_dev_info() anyway.
BTW, the use of device_domain_lock looks a bit unsafe to me... it's
at least not obvious to me why we aren't vulnerable to the race below:
iommu_support_dev_iotlb()
domain_remove_dev_info()
lock device_domain_lock
find info
unlock device_domain_lock
lock device_domain_lock
find same info
unlock device_domain_lock
free_devinfo_mem(info)
do stuff with info after it's free
However I don't understand the locking here well enough to know if
this is a real problem, let alone what the best fix is.
Anyway here's the full lockdep output that prompted all of this:
=======================================================
[ INFO: possible circular locking dependency detected ]
2.6.39.1+ #1
-------------------------------------------------------
bash/13954 is trying to acquire lock:
(&(&iommu->lock)->rlock){......}, at: [<ffffffff812f6421>] domain_remove_one_dev_info+0x121/0x230
but task is already holding lock:
(device_domain_lock){-.-...}, at: [<ffffffff812f6508>] domain_remove_one_dev_info+0x208/0x230
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #1 (device_domain_lock){-.-...}:
[<ffffffff8109ca9d>] lock_acquire+0x9d/0x130
[<ffffffff81571475>] _raw_spin_lock_irqsave+0x55/0xa0
[<ffffffff812f8350>] domain_context_mapping_one+0x600/0x750
[<ffffffff812f84df>] domain_context_mapping+0x3f/0x120
[<ffffffff812f9175>] iommu_prepare_identity_map+0x1c5/0x1e0
[<ffffffff81ccf1ca>] intel_iommu_init+0x88e/0xb5e
[<ffffffff81cab204>] pci_iommu_init+0x16/0x41
[<ffffffff81002165>] do_one_initcall+0x45/0x190
[<ffffffff81ca3d3f>] kernel_init+0xe3/0x168
[<ffffffff8157ac24>] kernel_thread_helper+0x4/0x10
-> #0 (&(&iommu->lock)->rlock){......}:
[<ffffffff8109bf3e>] __lock_acquire+0x195e/0x1e10
[<ffffffff8109ca9d>] lock_acquire+0x9d/0x130
[<ffffffff81571475>] _raw_spin_lock_irqsave+0x55/0xa0
[<ffffffff812f6421>] domain_remove_one_dev_info+0x121/0x230
[<ffffffff812f8b42>] device_notifier+0x72/0x90
[<ffffffff8157555c>] notifier_call_chain+0x8c/0xc0
[<ffffffff81089768>] __blocking_notifier_call_chain+0x78/0xb0
[<ffffffff810897b6>] blocking_notifier_call_chain+0x16/0x20
[<ffffffff81373a5c>] __device_release_driver+0xbc/0xe0
[<ffffffff81373ccf>] device_release_driver+0x2f/0x50
[<ffffffff81372ee3>] driver_unbind+0xa3/0xc0
[<ffffffff813724ac>] drv_attr_store+0x2c/0x30
[<ffffffff811e4506>] sysfs_write_file+0xe6/0x170
[<ffffffff8117569e>] vfs_write+0xce/0x190
[<ffffffff811759e4>] sys_write+0x54/0xa0
[<ffffffff81579a82>] system_call_fastpath+0x16/0x1b
other info that might help us debug this:
6 locks held by bash/13954:
#0: (&buffer->mutex){+.+.+.}, at: [<ffffffff811e4464>] sysfs_write_file+0x44/0x170
#1: (s_active#3){++++.+}, at: [<ffffffff811e44ed>] sysfs_write_file+0xcd/0x170
#2: (&__lockdep_no_validate__){+.+.+.}, at: [<ffffffff81372edb>] driver_unbind+0x9b/0xc0
#3: (&__lockdep_no_validate__){+.+.+.}, at: [<ffffffff81373cc7>] device_release_driver+0x27/0x50
#4: (&(&priv->bus_notifier)->rwsem){.+.+.+}, at: [<ffffffff8108974f>] __blocking_notifier_call_chain+0x5f/0xb0
#5: (device_domain_lock){-.-...}, at: [<ffffffff812f6508>] domain_remove_one_dev_info+0x208/0x230
stack backtrace:
Pid: 13954, comm: bash Not tainted 2.6.39.1+ #1
Call Trace:
[<ffffffff810993a7>] print_circular_bug+0xf7/0x100
[<ffffffff8109bf3e>] __lock_acquire+0x195e/0x1e10
[<ffffffff810972bd>] ? trace_hardirqs_off+0xd/0x10
[<ffffffff8109d57d>] ? trace_hardirqs_on_caller+0x13d/0x180
[<ffffffff8109ca9d>] lock_acquire+0x9d/0x130
[<ffffffff812f6421>] ? domain_remove_one_dev_info+0x121/0x230
[<ffffffff81571475>] _raw_spin_lock_irqsave+0x55/0xa0
[<ffffffff812f6421>] ? domain_remove_one_dev_info+0x121/0x230
[<ffffffff810972bd>] ? trace_hardirqs_off+0xd/0x10
[<ffffffff812f6421>] domain_remove_one_dev_info+0x121/0x230
[<ffffffff812f8b42>] device_notifier+0x72/0x90
[<ffffffff8157555c>] notifier_call_chain+0x8c/0xc0
[<ffffffff81089768>] __blocking_notifier_call_chain+0x78/0xb0
[<ffffffff810897b6>] blocking_notifier_call_chain+0x16/0x20
[<ffffffff81373a5c>] __device_release_driver+0xbc/0xe0
[<ffffffff81373ccf>] device_release_driver+0x2f/0x50
[<ffffffff81372ee3>] driver_unbind+0xa3/0xc0
[<ffffffff813724ac>] drv_attr_store+0x2c/0x30
[<ffffffff811e4506>] sysfs_write_file+0xe6/0x170
[<ffffffff8117569e>] vfs_write+0xce/0x190
[<ffffffff811759e4>] sys_write+0x54/0xa0
[<ffffffff81579a82>] system_call_fastpath+0x16/0x1b
Signed-off-by: Roland Dreier <roland@purestorage.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
The iommu module on omap contains a few functions that are
only used by the debug module. These are however only there
when the debug code is built as a module. Since it is possible
to build the debug code into the kernel, the functions should
also be provided for the built-in case.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Make report_iommu_fault() return -ENOSYS whenever an iommu fault
handler isn't installed, so IOMMU drivers can then do their own
platform-specific default behavior if they wanted.
Fault handlers can still return -ENOSYS in case they want to elicit the
default behavior of the IOMMU drivers.
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
commit 4f3f8d9 "iommu/core: Add fault reporting mechanism" added
the public iommu_set_fault_handler() symbol but forgot to export it.
Fix that.
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Before the restructruing of the x86 IOMMU code,
intel_iommu_init() was getting called directly from
pci_iommu_init() and hence needed to explicitly set
dmar_disabled to 1 for the failure conditions of
check_zero_address().
Recent changes don't call intel_iommu_init() if the intel iommu
detection fails as a result of failure in check_zero_address().
So no need for this ifdef and the code inside it.
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: yinghai@kernel.org
Cc: youquan.song@intel.com
Cc: joerg.roedel@amd.com
Cc: tony.luck@intel.com
Cc: dwmw2@infradead.org
Link: http://lkml.kernel.org/r/20110824001456.334878686@sbsiddha-desk.sc.intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Both DMA-remapping aswell as Interrupt-remapping depend on the
dmar dev scope to be initialized. When both DMA and
IRQ-remapping are enabled, we depend on DMA-remapping init code
to call dmar_dev_scope_init(). This resulted in not doing this
init when DMA-remapping was turned off but interrupt-remapping
turned on in the kernel config.
This caused interrupt routing to break with CONFIG_INTR_REMAP=y
and CONFIG_DMAR=n.
This issue was introduced by this commit:
| commit 9d5ce73a64
| Author: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
| Date: Tue Nov 10 19:46:16 2009 +0900
|
| x86: intel-iommu: Convert detect_intel_iommu to use iommu_init hook
Fix this by calling dmar_dev_scope_init() explicitly from the
interrupt remapping code too.
Reported-by: Andrew Vasquez <andrew.vasquez@qlogic.com>
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: yinghai@kernel.org
Cc: youquan.song@intel.com
Cc: joerg.roedel@amd.com
Cc: tony.luck@intel.com
Cc: dwmw2@infradead.org
Link: http://lkml.kernel.org/r/20110824001456.229207526@sbsiddha-desk.sc.intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
On the platforms which are x2apic and interrupt-remapping
capable, Linux kernel is enabling x2apic even if the BIOS
doesn't. This is to take advantage of the features that x2apic
brings in.
Some of the OEM platforms are running into issues because of
this, as their bios is not x2apic aware. For example, this was
resulting in interrupt migration issues on one of the platforms.
Also if the BIOS SMI handling uses APIC interface to send SMI's,
then the BIOS need to be aware of x2apic mode that OS has
enabled.
On some of these platforms, BIOS doesn't have a HW mechanism to
turnoff the x2apic feature to prevent OS from enabling it.
To resolve this mess, recent changes to the VT-d2 specification:
http://download.intel.com/technology/computing/vptech/Intel(r)_VT_for_Direct_IO.pdf
includes a mechanism that provides BIOS a way to request system
software to opt out of enabling x2apic mode.
Look at the x2apic optout flag in the DMAR tables before
enabling the x2apic mode in the platform. Also print a warning
that we have disabled x2apic based on the BIOS request.
Kernel boot parameter "intremap=no_x2apic_optout" can be used to
override the BIOS x2apic optout request.
Signed-off-by: Youquan Song <youquan.song@intel.com>
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: yinghai@kernel.org
Cc: joerg.roedel@amd.com
Cc: tony.luck@intel.com
Cc: dwmw2@infradead.org
Link: http://lkml.kernel.org/r/20110824001456.171766616@sbsiddha-desk.sc.intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Without this patch it is possible to select the VIDEO_OMAP3
driver which then selects OMAP_IOVMM. But the omap iommu
driver is not compiled without IOMMU_SUPPORT enabled. Fix
that by forcing OMAP_IOMMU and OMAP_IOVMM are enabled before
VIDEO_OMAP3 can be selected.
Cc: Ohad Ben-Cohen <ohad@wizery.com>
Cc: iommu@lists.linux-foundation.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Start using the generic fault report mechanism, as provided by
the IOMMU core, and remove its now-redundant omap_iommu_set_isr API.
Currently we're only interested in letting upper layers know about the
fault, so in case the faulting device is a remote processor, they could
restart it.
Dynamic PTE/TLB loading is not supported.
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Add iommu fault report mechanism to the IOMMU API, so implementations
could report about mmu faults (translation errors, hardware errors,
etc..).
Fault reports can be used in several ways:
- mere logging
- reset the device that accessed the faulting address (may be necessary
in case the device is a remote processor for example)
- implement dynamic PTE/TLB loading
A dedicated iommu_set_fault_handler() API has been added to allow
users, who are interested to receive such reports, to provide
their handler.
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Mark this lowlevel IRQ handler as non-threaded. This prevents a boot
crash when "threadirqs" is on the kernel commandline. Also the
interrupt handler is handling hardware critical events which should
not be delayed into a thread.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The qi->q_lock lock can be taken in atomic context and therefore
cannot be preempted on -rt - annotate it.
In mainline this change documents the low level nature of
the lock - otherwise there's no functional difference. Lockdep
and Sparse checking will work as usual.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The irq_2_ir_lock can be taken in atomic context and therefore
cannot be preempted on -rt - annotate it.
In mainline this change documents the low level nature of
the lock - otherwise there's no functional difference. Lockdep
and Sparse checking will work as usual.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The iommu->register_lock can be taken in atomic context and therefore
must not be preempted on -rt - annotate it.
In mainline this change documents the low level nature of
the lock - otherwise there's no functional difference. Lockdep
and Sparse checking will work as usual.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Replace the hard-coded 4kb by PAGE_SIZE to make iommu-api
implementations possible on architectures where
PAGE_SIZE != 4kb.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Replace iommu's alignment checks with the existing IS_ALIGNED macro,
to drop a few lines of code and utilize IS_ALIGNED's type safety.
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Users of the IOMMU API (kvm specifically) assume that iommu_unmap()
returns the order of the unmapped page (on success).
Fix msm_iommu_unmap() accordingly.
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
Cc: Stepan Moskovchenko <stepanm@codeaurora.org>
Cc: David Brown <davidb@codeaurora.org>
Acked-by: David Brown <davidb@codeaurora.org>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Users of the IOMMU API (kvm specifically) assume that iommu_unmap()
returns the order of the unmapped page.
Fix omap_iommu_unmap() to do so and adopt omap-iovmm accordingly.
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
omap_iovmm requires page-aligned buffers, and that sometimes causes
omap3isp failures (i.e. whenever the buffer passed from userspace is not
page-aligned).
Remove this limitation by rounding the address of the first page entry
down, and adding the offset back to the device address.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Acked-by: Hiroshi DOYU <Hiroshi.DOYU@nokia.com>
[ohad@wizery.com: rebased, but tested only with aligned buffers]
[ohad@wizery.com: slightly edited the commit log]
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
The domain_flush_devices() function takes the domain->lock.
But this function is only called from update_domain() which
itself is already called unter the domain->lock. This causes
a deadlock situation when the dma-address-space of a domain
grows larger than 1GB.
Cc: stable@kernel.org
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
The value is only set to true but never set back to false,
which causes to many completion-wait commands to be sent to
hardware. Fix it with this patch.
Cc: stable@kernel.org
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Make CONFIG_OMAP_IOMMU depend on CONFIG_ARCH_OMAP so other
allmodconfig builds won't fail.
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
The omap_iommu_set_isr() was still using the mutex functions
but the iommu_lock was converted to a spin_lock. Fix that
up.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Prepend 'omap_' to OMAP's 'struct iommu' and exposed API, to prevent
namespace pollution and generally to improve readability of the code
that still uses the driver directly.
Update the users as needed as well.
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Acked-by: Hiroshi DOYU <Hiroshi.DOYU@nokia.com>
Acked-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Remove unused public APIs from OMAP's iommu driver.
IOMMU functionality should be exposed only via the generic IOMMU API;
this way drivers stay generic, and different IOMMU drivers
don't need to duplicate similar functionalities.
The rest of the API still exposed by OMAP's iommu will be evaluated
and eventually either added to the generic IOMMU API (if relevant),
or completely removed.
The intention is that OMAP's iommu driver will eventually not expose
any public API.
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
Acked-by: Hiroshi DOYU <Hiroshi.DOYU@nokia.com>
Acked-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Remove unused functionality from OMAP's iovmm module.
The intention is to eventually completely replace iovmm with the
generic DMA-API, so new code that'd need this iovmm functionality
will have to extend the DMA-API instead.
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
Acked-by: Hiroshi DOYU <Hiroshi.DOYU@nokia.com>
Acked-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Use PREFETCH_IOTLB to control the content of the called function,
instead of inlining it in the code.
This improves readability of the code, and also prevents an "unused
function" warning to show up when PREFETCH_IOTLB isn't set.
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
Acked-by: Hiroshi DOYU <Hiroshi.DOYU@nokia.com>
Acked-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Stop exporting functions that are used only within the iommu
driver itself.
Eventually OMAP's iommu driver should only expose API via the generic
IOMMU framework.
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
Acked-by: Hiroshi DOYU <Hiroshi.DOYU@nokia.com>
Acked-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Move OMAP's iommu drivers to the dedicated iommu drivers folder.
While OMAP's iovmm (virtual memory manager) driver does not strictly
belong to the iommu drivers folder, move it there as well, because
it's by no means OMAP-specific (in concept. technically it is still
coupled with OMAP's iommu).
Eventually, iovmm will be completely replaced with the generic,
iommu-based, dma-mapping API.
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Acked-by: Hiroshi DOYU <Hiroshi.DOYU@nokia.com>
Acked-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Reserve the MSI address range in the address allocator so
that MSI addresses are not handed out as dma handles.
Cc: stable@kernel.org
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
A few parts of the driver were missing in drivers/iommu.
Move them there to have the complete driver in that
directory.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
This should ease finding similarities with different platforms,
with the intention of solving problems once in a generic framework
which everyone can use.
Note: to move intel-iommu.c, the declaration of pci_find_upstream_pcie_bridge()
has to move from drivers/pci/pci.h to include/linux/pci.h. This is handled
in this patch, too.
As suggested, also drop DMAR's EXPERIMENTAL tag while we're at it.
Compile-tested on x86_64.
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
This should ease finding similarities with different platforms,
with the intention of solving problems once in a generic framework
which everyone can use.
Compile-tested on x86_64.
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
This should ease finding similarities with different platforms,
with the intention of solving problems once in a generic framework
which everyone can use.
Compile-tested for MSM8X60.
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
Acked-by: David Brown <davidb@codeaurora.org>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Create a dedicated folder for iommu drivers, and move the base
iommu implementation over there.
Grouping the various iommu drivers in a single location will help
finding similar problems shared by different platforms, so they
could be solved once, in the iommu framework, instead of solved
differently (or duplicated) in each driver.
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
