[PATCH v2 29/33] iommu/omap: Convert to probe/release_device() call-backs

From: Joerg Roedel Convert the OMAP IOMMU driver to use the probe_device() and release_device() call-backs of iommu_ops, so that the iommu core code does the group and sysfs setup. Signed-off-by: Joerg Roedel --- drivers/iommu/omap-iommu.c | 49 ++ 1 file

[PATCH v2 03/33] iommu/amd: Implement iommu_ops->def_domain_type call-back

From: Joerg Roedel Implement the new def_domain_type call-back for the AMD IOMMU driver. Signed-off-by: Joerg Roedel --- drivers/iommu/amd_iommu.c | 15 +++ 1 file changed, 15 insertions(+) diff --git a/drivers/iommu/amd_iommu.c b/drivers/iommu/amd_iommu.c index

[PATCH v2 22/33] iommu/mediatek: Convert to probe/release_device() call-backs

From: Joerg Roedel Convert the Mediatek IOMMU driver to use the probe_device() and release_device() call-backs of iommu_ops, so that the iommu core code does the group and sysfs setup. Signed-off-by: Joerg Roedel --- drivers/iommu/mtk_iommu.c | 24 ++-- 1 file changed, 6

[PATCH v2 10/33] iommu: Move new probe_device path to separate function

From: Joerg Roedel This makes it easier to remove to old code-path when all drivers are converted. As a side effect that it also fixes the error cleanup path. Signed-off-by: Joerg Roedel --- drivers/iommu/iommu.c | 69 --- 1 file changed, 46

[PATCH v2 11/33] iommu: Split off default domain allocation from group assignment

From: Joerg Roedel When a bus is initialized with iommu-ops, all devices on the bus are scanned and iommu-groups are allocated for them, and each groups will also get a default domain allocated. Until now this happened as soon as the group was created and the first device added to it. When

[PATCH 13/29] mm: remove vmap_page_range_noflush and vunmap_page_range

These have non-static aliases called map_kernel_range_noflush and unmap_kernel_range_noflush that just differ slightly in the calling conventions that pass addr + size instead of an end. Signed-off-by: Christoph Hellwig Acked-by: Peter Zijlstra (Intel) --- mm/vmalloc.c | 98

[PATCH 28/29] powerpc: use __vmalloc_node in alloc_vm_stack

alloc_vm_stack can use a slightly higher level vmalloc function. Signed-off-by: Christoph Hellwig Acked-by: Peter Zijlstra (Intel) --- arch/powerpc/kernel/irq.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/arch/powerpc/kernel/irq.c b/arch/powerpc/kernel/irq.c index

[PATCH v2 07/33] iommu: Add probe_device() and remove_device() call-backs

From: Joerg Roedel Add call-backs to 'struct iommu_ops' as an alternative to the add_device() and remove_device() call-backs, which will be removed when all drivers are converted. The new call-backs will not setupt IOMMU groups and domains anymore, so also add a probe_finalize() call-back where

[PATCH v2 17/33] iommu/arm-smmu: Convert to probe/release_device() call-backs

From: Joerg Roedel Convert the arm-smmu and arm-smmu-v3 drivers to use the probe_device() and release_device() call-backs of iommu_ops, so that the iommu core code does the group and sysfs setup. Signed-off-by: Joerg Roedel --- drivers/iommu/arm-smmu-v3.c | 38

[PATCH v2 06/33] iommu/amd: Return -ENODEV in add_device when device is not handled by IOMMU

From: Joerg Roedel When check_device() fails on the device, it is not handled by the IOMMU and amd_iommu_add_device() needs to return -ENODEV. Signed-off-by: Joerg Roedel --- drivers/iommu/amd_iommu.c | 5 - 1 file changed, 4 insertions(+), 1 deletion(-) diff --git

[PATCH 29/29] s390: use __vmalloc_node in stack_alloc

stack_alloc can use a slightly higher level vmalloc function. Signed-off-by: Christoph Hellwig Acked-by: Christian Borntraeger Acked-by: Peter Zijlstra (Intel) --- arch/s390/kernel/setup.c | 9 +++-- 1 file changed, 3 insertions(+), 6 deletions(-) diff --git a/arch/s390/kernel/setup.c

[PATCH v2 32/33] iommu: Remove add_device()/remove_device() code-paths

From: Joerg Roedel All drivers are converted to use the probe/release_device() call-backs, so the add_device/remove_device() pointers are unused and the code using them can be removed. Signed-off-by: Joerg Roedel --- drivers/iommu/iommu.c | 149 --

[PATCH 24/29] mm: remove __vmalloc_node_flags_caller

Just use __vmalloc_node instead which gets and extra argument. To be able to to use __vmalloc_node in all caller make it available outside of vmalloc and implement it in nommu.c. Signed-off-by: Christoph Hellwig Acked-by: Peter Zijlstra (Intel) --- include/linux/vmalloc.h | 4 ++--

[PATCH v2 01/33] iommu: Move default domain allocation to separate function

From: Joerg Roedel Move the code out of iommu_group_get_for_dev() into a separate function. Signed-off-by: Joerg Roedel --- drivers/iommu/iommu.c | 74 ++- 1 file changed, 45 insertions(+), 29 deletions(-) diff --git a/drivers/iommu/iommu.c

[PATCH v2 16/33] iommu/vt-d: Convert to probe/release_device() call-backs

From: Joerg Roedel Convert the Intel IOMMU driver to use the probe_device() and release_device() call-backs of iommu_ops, so that the iommu core code does the group and sysfs setup. Signed-off-by: Joerg Roedel --- drivers/iommu/intel-iommu.c | 67 - 1 file

[PATCH v2 28/33] iommu/omap: Remove orphan_dev tracking

From: Joerg Roedel Remove the tracking of device which could not be probed because their IOMMU is not probed yet. Replace it with a call to bus_iommu_probe() when a new IOMMU is probed. Signed-off-by: Joerg Roedel --- drivers/iommu/omap-iommu.c | 54 +++--- 1

[PATCH v2 13/33] iommu: Export bus_iommu_probe() and make is safe for re-probing

From: Joerg Roedel Add a check to the bus_iommu_probe() call-path to make sure it ignores devices which have already been successfully probed. Then export the bus_iommu_probe() function so it can be used by IOMMU drivers. Signed-off-by: Joerg Roedel --- drivers/iommu/iommu.c | 6 +-

[PATCH v2 23/33] iommu/mediatek-v1 Convert to probe/release_device() call-backs

From: Joerg Roedel Convert the Mediatek-v1 IOMMU driver to use the probe_device() and release_device() call-backs of iommu_ops, so that the iommu core code does the group and sysfs setup. Signed-off-by: Joerg Roedel --- drivers/iommu/mtk_iommu_v1.c | 50 +++- 1

[PATCH 09/29] mm: unexport unmap_kernel_range_noflush

There are no modular users of this function. Signed-off-by: Christoph Hellwig Acked-by: Peter Zijlstra (Intel) --- mm/vmalloc.c | 1 - 1 file changed, 1 deletion(-) diff --git a/mm/vmalloc.c b/mm/vmalloc.c index d1534d610b48..3375f9508ef6 100644 --- a/mm/vmalloc.c +++ b/mm/vmalloc.c @@

[PATCH v2 19/33] iommu/s390: Convert to probe/release_device() call-backs

From: Joerg Roedel Convert the S390 IOMMU driver to use the probe_device() and release_device() call-backs of iommu_ops, so that the iommu core code does the group and sysfs setup. Signed-off-by: Joerg Roedel --- drivers/iommu/s390-iommu.c | 22 ++ 1 file changed, 6

[PATCH v2 05/33] iommu/amd: Remove dma_mask check from check_device()

From: Joerg Roedel The check was only needed for the DMA-API implementation in the AMD IOMMU driver, which no longer exists. Signed-off-by: Joerg Roedel --- drivers/iommu/amd_iommu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/iommu/amd_iommu.c

[PATCH 10/29] mm: rename CONFIG_PGTABLE_MAPPING to CONFIG_ZSMALLOC_PGTABLE_MAPPING

Rename the Kconfig variable to clarify the scope. Signed-off-by: Christoph Hellwig Acked-by: Minchan Kim Acked-by: Peter Zijlstra (Intel) --- arch/arm/configs/omap2plus_defconfig | 2 +- include/linux/zsmalloc.h | 2 +- mm/Kconfig | 2 +- mm/zsmalloc.c

[PATCH 07/29] powerpc: remove __ioremap_at and __iounmap_at

These helpers are only used for remapping the ISA I/O base. Replace the mapping side with a remap_isa_range helper in isa-bridge.c that hard codes all the known arguments, and just remove __iounmap_at in favour of open coding it in the only caller. Signed-off-by: Christoph Hellwig Acked-by:

[PATCH v2 15/33] iommu/amd: Convert to probe/release_device() call-backs

From: Joerg Roedel Convert the AMD IOMMU Driver to use the probe_device() and release_device() call-backs of iommu_ops, so that the iommu core code does the group and sysfs setup. Signed-off-by: Joerg Roedel --- drivers/iommu/amd_iommu.c | 71 --- 1 file

[PATCH v2 20/33] iommu/virtio: Convert to probe/release_device() call-backs

From: Joerg Roedel Convert the VirtIO IOMMU driver to use the probe_device() and release_device() call-backs of iommu_ops, so that the iommu core code does the group and sysfs setup. Signed-off-by: Joerg Roedel --- drivers/iommu/virtio-iommu.c | 41 +--- 1 file

[PATCH 04/29] staging: media: ipu3: use vmap instead of reimplementing it

Just use vmap instead of messing with vmalloc internals. Signed-off-by: Christoph Hellwig Acked-by: Peter Zijlstra (Intel) --- drivers/staging/media/ipu3/ipu3-css-pool.h | 4 +-- drivers/staging/media/ipu3/ipu3-dmamap.c | 30 ++ 2 files changed, 9 insertions(+), 25

[PATCH 21/29] mm: remove the pgprot argument to __vmalloc

The pgprot argument to __vmalloc is always PROT_KERNEL now, so remove it. Signed-off-by: Christoph Hellwig Reviewed-by: Michael Kelley [hyperv] Acked-by: Gao Xiang [erofs] Acked-by: Peter Zijlstra (Intel) --- arch/x86/hyperv/hv_init.c | 3 +-- arch/x86/include/asm/kvm_host.h

[PATCH v2 08/33] iommu: Move default domain allocation to iommu_probe_device()

From: Joerg Roedel Well, not really. The call to iommu_alloc_default_domain() in iommu_group_get_for_dev() has to stay around as long as there are IOMMU drivers using the add/remove_device() call-backs instead of probe/release_device(). Those drivers expect that iommu_group_get_for_dev()

[PATCH 08/29] mm: remove __get_vm_area

Switch the two remaining callers to use __get_vm_area_caller instead. Signed-off-by: Christoph Hellwig Acked-by: Peter Zijlstra (Intel) --- arch/powerpc/kernel/pci_64.c | 3 ++- arch/sh/kernel/cpu/sh4/sq.c | 3 ++- include/linux/vmalloc.h | 2 -- mm/vmalloc.c | 8

[PATCH 19/29] mm: enforce that vmap can't map pages executable

To help enforcing the W^X protection don't allow remapping existing pages as executable. x86 bits from Peter Zijlstra , arm64 bits from Mark Rutland . Signed-off-by: Christoph Hellwig Acked-by: Peter Zijlstra (Intel) --- arch/arm64/include/asm/pgtable.h | 3 +++

[PATCH v2 18/33] iommu/pamu: Convert to probe/release_device() call-backs

From: Joerg Roedel Convert the PAMU IOMMU driver to use the probe_device() and release_device() call-backs of iommu_ops, so that the iommu core code does the group and sysfs setup. Signed-off-by: Joerg Roedel --- drivers/iommu/fsl_pamu_domain.c | 22 +- 1 file changed, 5

[PATCH v2 09/33] iommu: Keep a list of allocated groups in __iommu_probe_device()

From: Joerg Roedel This is needed to defer default_domain allocation for new IOMMU groups until all devices have been added to the group. Signed-off-by: Joerg Roedel --- drivers/iommu/iommu.c | 9 +++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/drivers/iommu/iommu.c

[PATCH 22/29] mm: remove the prot argument to __vmalloc_node

This is always PAGE_KERNEL now. Signed-off-by: Christoph Hellwig Acked-by: Peter Zijlstra (Intel) --- mm/vmalloc.c | 35 ++- 1 file changed, 14 insertions(+), 21 deletions(-) diff --git a/mm/vmalloc.c b/mm/vmalloc.c index 466a449b3a15..de7952959e82 100644 ---

[PATCH 16/29] mm: remove map_vm_range

Switch all callers to map_kernel_range, which symmetric to the unmap side (as well as the _noflush versions). Signed-off-by: Christoph Hellwig Acked-by: Peter Zijlstra (Intel) --- Documentation/core-api/cachetlb.rst | 2 +- include/linux/vmalloc.h | 10 -- mm/vmalloc.c

[PATCH 14/29] mm: rename vmap_page_range to map_kernel_range

This matches the map_kernel_range_noflush API. Also change to pass a size instead of the end, similar to the noflush version. Signed-off-by: Christoph Hellwig Acked-by: Peter Zijlstra (Intel) --- mm/vmalloc.c | 11 +-- 1 file changed, 5 insertions(+), 6 deletions(-) diff --git

[PATCH 15/29] mm: don't return the number of pages from map_kernel_range{, _noflush}

None of the callers needs the number of pages, and a 0 / -errno return value is a lot more intuitive. Signed-off-by: Christoph Hellwig Acked-by: Peter Zijlstra (Intel) --- mm/vmalloc.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/mm/vmalloc.c b/mm/vmalloc.c index

[PATCH 11/29] mm: only allow page table mappings for built-in zsmalloc

This allows to unexport map_vm_area and unmap_kernel_range, which are rather deep internal and should not be available to modules, as they for example allow fine grained control of mapping permissions, and also allow splitting the setup of a vmalloc area and the actual mapping and thus expose

[PATCH 27/29] arm64: use __vmalloc_node in arch_alloc_vmap_stack

arch_alloc_vmap_stack can use a slightly higher level vmalloc function. Signed-off-by: Christoph Hellwig Acked-by: Peter Zijlstra (Intel) --- arch/arm64/include/asm/vmap_stack.h | 6 ++ 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/arch/arm64/include/asm/vmap_stack.h

[PATCH 23/29] mm: remove both instances of __vmalloc_node_flags

The real version just had a few callers that can open code it and remove one layer of indirection. The nommu stub was public but only had a single caller, so remove it and avoid a CONFIG_MMU ifdef in vmalloc.h. Signed-off-by: Christoph Hellwig Acked-by: Peter Zijlstra (Intel) ---

[PATCH 12/29] mm: pass addr as unsigned long to vb_free

Ever use of addr in vb_free casts to unsigned long first, and the caller has an unsigned long version of the address available anyway. Just pass that and avoid all the casts. Signed-off-by: Christoph Hellwig Acked-by: Peter Zijlstra (Intel) --- mm/vmalloc.c | 16 +++- 1 file

[PATCH 06/29] powerpc: add an ioremap_phb helper

Factor code shared between pci_64 and electra_cf into a ioremap_pbh helper that follows the normal ioremap semantics, and returns a useful __iomem pointer. Note that it opencodes __ioremap_at as we know from the callers the slab is available. Switch pci_64 to also store the result as __iomem

[PATCH 01/29] x86/hyperv: use vmalloc_exec for the hypercall page

Use the designated helper for allocating executable kernel memory, and remove the now unused PAGE_KERNEL_RX define. Signed-off-by: Christoph Hellwig Reviewed-by: Michael Kelley Acked-by: Wei Liu Acked-by: Peter Zijlstra (Intel) --- arch/x86/hyperv/hv_init.c| 2 +-

[PATCH 03/29] staging: android: ion: use vmap instead of vm_map_ram

vm_map_ram can keep mappings around after the vm_unmap_ram. Using that with non-PAGE_KERNEL mappings can lead to all kinds of aliasing issues. Signed-off-by: Christoph Hellwig Acked-by: Greg Kroah-Hartman Acked-by: Peter Zijlstra (Intel) --- drivers/staging/android/ion/ion_heap.c | 4 ++-- 1

[PATCH 02/29] x86: fix vmap arguments in map_irq_stack

vmap does not take a gfp_t, the flags argument is for VM_* flags. Signed-off-by: Christoph Hellwig --- arch/x86/kernel/irq_64.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/x86/kernel/irq_64.c b/arch/x86/kernel/irq_64.c index 12df3a4abfdd..6b32ab009c19 100644 ---

[PATCH 05/29] dma-mapping: use vmap insted of reimplementing it

Replace the open coded instance of vmap with the actual function. In the non-contiguous (IOMMU) case this requires an extra find_vm_area, but given that this isn't a fast path function that is a small price to pay. Signed-off-by: Christoph Hellwig Acked-by: Peter Zijlstra (Intel) ---

decruft the vmalloc API v2

Hi all, Peter noticed that with some dumb luck you can toast the kernel address space with exported vmalloc symbols. I used this as an opportunity to decruft the vmalloc.c API and make it much more systematic. This also removes any chance to create vmalloc mappings outside the designated areas

Re: [PATCH 3/4] dma-mapping: add a dma_ops_bypass flag to struct device

On Tue, Apr 14, 2020 at 02:25:05PM +0200, Christoph Hellwig wrote: > Several IOMMU drivers have a bypass mode where they can use a direct > mapping if the devices DMA mask is large enough. Add generic support > to the core dma-mapping code to do that to switch those drivers to > a common

[PATCH 2/4] dma-mapping: inline the fast path dma-direct calls

Inline the single page map/unmap/sync dma-direct calls into the now out of line generic wrappers. This restores the behavior of a single function call that we had before moving the generic calls out of line. Besides the dma-mapping callers there are just a few callers in IOMMU drivers that have a

generic DMA bypass flag v3

Hi all, I've recently beeing chatting with Lu about using dma-iommu and per-device DMA ops in the intel IOMMU driver, and one missing feature in dma-iommu is a bypass mode where the direct mapping is used even when an iommu is attached to improve performance. The powerpc code already has a

[PATCH 1/4] dma-mapping: move the remaining DMA API calls out of line

For a long time the DMA API has been implemented inline in dma-mapping.h, but the function bodies can be quite large. Move them all out of line. Signed-off-by: Christoph Hellwig --- include/linux/dma-direct.h | 58 + include/linux/dma-mapping.h | 247

[PATCH 4/4] powerpc: use the generic dma_ops_bypass mode

Use the DMA API bypass mechanism for direct window mappings. This uses common code and speed up the direct mapping case by avoiding indirect calls just when not using dma ops at all. It also fixes a problem where the sync_* methods were using the bypass check for DMA allocations, but those are

[PATCH 3/4] dma-mapping: add a dma_ops_bypass flag to struct device

Several IOMMU drivers have a bypass mode where they can use a direct mapping if the devices DMA mask is large enough. Add generic support to the core dma-mapping code to do that to switch those drivers to a common solution. Signed-off-by: Christoph Hellwig --- include/linux/device.h | 6

Re: [PATCH v2 1/3] iommu/uapi: Define uapi version and capabilities

On Mon, Apr 13, 2020 at 04:21:29PM -0600, Alex Williamson wrote: > Is the objection to a global version or to any version fields? I don't > really understand the global version, I'd think a mechanism to check > extensions plus a per structure flags/version would be preferred. The > former should

Re: [PATCH v2 1/3] iommu/uapi: Define uapi version and capabilities

On Mon, Apr 13, 2020 at 01:41:57PM -0700, Jacob Pan wrote: > Hi All, > > Just a gentle reminder, any feedback on the options I listed below? New > ideas will be even better. > > Christoph, does the explanation make sense to you? We do have the > capability/flag based scheme for IOMMU API

Re: [rfc v2 3/6] dma-pool: dynamically expanding atomic pools

On Fri, Apr 10, 2020 at 12:37:20PM -0700, David Rientjes wrote: > I'll rely on Christoph to determine whether it makes sense to add some > periodic scavening of the atomic pools, whether that's needed for this to > be merged, or wheter we should enforce some maximum pool size. I don't really

Re: [rfc v2 4/6] dma-direct: atomic allocations must come from atomic coherent pools

> + /* > + * Unencrypted memory must come directly from DMA atomic pools if > + * blocking is not allowed. > + */ > + if (IS_ENABLED(CONFIG_DMA_COHERENT_POOL) && > + force_dma_unencrypted(dev) && !gfpflags_allow_blocking(gfp)) { > + ret =

Re: [rfc v2 1/6] dma-remap: separate DMA atomic pools from direct remap code

On Wed, Apr 08, 2020 at 02:21:03PM -0700, David Rientjes wrote: > +++ b/kernel/dma/pool.c > @@ -0,0 +1,125 @@ > +// SPDX-License-Identifier: GPL-2.0 > +/* > + * Copyright (C) 2012 ARM Ltd. > + * Copyright (c) 2014 The Linux Foundation > + * Copyright (C) 2020 Google LLC Of the old copyrights the

Re: [PATCH 1/2] dma-mapping: add a dma_ops_bypass flag to struct device

On Tue, Apr 14, 2020 at 04:21:27PM +1000, Alexey Kardashevskiy wrote: > > Good points, I'll post revised version when you post your v3 of this. > > > > Any plans on posting v3 of this? Thanks, Just back from a long weekend. I'll take a stab at it soon.

Re: [PATCH 1/2] dma-mapping: add a dma_ops_bypass flag to struct device

On 07/04/2020 20:12, Alexey Kardashevskiy wrote: > > > On 07/04/2020 03:17, Christoph Hellwig wrote: >> On Mon, Apr 06, 2020 at 11:25:09PM +1000, Alexey Kardashevskiy wrote: > Do you see any serious problem with this approach? Thanks! Do you have a link to the whole branch? The

<    1   2