OF/DT core library provides architecture specific hook to match the
logical cpu index with the corresponding physical identifier.
On ARM64, the MPIDR_EL1 contains specific bitfields(MPIDR_EL1.Aff{3..0})
which uniquely identify a CPU, in addition to some non-identifying
information and reserved bits. The ARM cpu binding defines the 'reg'
property to only contain the affinity bits, and any cpu nodes with other
bits set in their 'reg' entry are skipped.
This patch overrides the weak definition of arch_match_cpu_phys_id
with ARM64 specific version using MPIDR_EL1.Aff{3..0} as cpu physical
identifiers.
Signed-off-by: Sudeep KarkadaNagesha <sudeep.karkadanagesha@arm.com>
Acked-by: Mark Rutland <mark.rutland@arm.com>
Cc: Will Deacon <will.deacon@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
(cherry picked from commit 6e15d0e04b)
Enable additional use of additional scheduler features with the topology
information.
Reported-by: Hanjun Guo <hanjun.guo@linaro.org>
Signed-off-by: Mark Brown <broonie@linaro.org>
Mostly just to quiet checkpatch warnings, be more verbose
in describing the ION config option.
Change-Id: I194235f1a68623dca15ae6e658fc99d00943a827
Signed-off-by: John Stultz <john.stultz@linaro.org>
Just some simple cleanups to address whitespace issues and
other issues found w/ checkpatch.
Change-Id: I181444505627894b8f3bbf59192703b0f65736ee
Signed-off-by: John Stultz <john.stultz@linaro.org>
Implement ion_cma_unmap_kernel, ion will call it unconditionally.
Use correct gfp flags when calling dma_alloc_coherent so it doesn't
try to use atomic DMA memory.
Check for invalid alignment when allocating.
Reject cached allocations - the cpu address returned by
dma_alloc_coherent is always going to be an uncached mapping, so
map_kernel will not see data written by a cached userspace mapping.
Change-Id: I2ea03f28fae3749f6de0b89700b69da3845926ea
Signed-off-by: Colin Cross <ccross@android.com>
Add ion_heap_pages_zero for ion heaps to use to zero pages
during initialization or allocation, when a struct ion_buffer
may not be available. Use it from the chunk heap and carveout
heaps.
Change-Id: Ic6c921943a8820cf9896da5164f2d9794d0fe91f
Signed-off-by: Colin Cross <ccross@android.com>
The carveout heap wasn't zeroing its buffers after use.
Create the sg_table during allocate instead of map_dma, to allow
using the sg_table during free, and call ion_heap_buffer_zero
during free. Also fixes a missing kfree when destroying the
table.
Change-Id: I318a8493cce32580250884cae336dd2e2c28e73b
Signed-off-by: Colin Cross <ccross@android.com>
There is no reason to use kzalloc, just call alloc_pages directly.
Change the GFP from GFP_KERNEL to include __GFP_HIGH, to allow it
to return contiguous pages from highmem. virt_to_* functions
aren't valid on highmem pages, so store the struct page * in an
sg_table in buffer->priv_virt like most other heaps, and replace
virt_to_* with page_to_*.
Change-Id: Ida78888b101f080883716e1fa5038dfc4dbabd16
Signed-off-by: Colin Cross <ccross@android.com>
Now that ion_vm_fault uses vm_insert_pfn instead of vm_insert_page
cached buffers can be supported in any heap. Remove the checks
in the chunk and system heaps.
Change-Id: I371a44c400ed8a342c3b0eed90d0fb7060537697
Signed-off-by: Colin Cross <ccross@android.com>
Now that ion_vm_fault doesn't need a struct page with a nonzero
refcount, there is no need allocate heap memory for cached pages using
split_page. Remove the ion_heap_alloc_pages and ion_heap_free_pages
helpers in favor of direct calls to alloc_pages and __free_pages,
and remove the special handling in the system heap.
Change-Id: I5966a798f48df2d56642e662a69c1495944f6509
Signed-off-by: Colin Cross <ccross@android.com>
Most ion userspace mappings are created with remap_pfn_range. Use
vm_insert_pfn instead of vm_insert_page to make faulted cached
mappings look more like uncached mappings.
Change-Id: I9ec5cad3fef54f3b80be8b306d7ff2f1fe3f0e66
Signed-off-by: Colin Cross <ccross@android.com>
Check the return value of remap_pfn_range and return an error if
it fails.
Change-Id: I206cf95a24607ebe1c80274e3ed15cc7c076d007
Signed-off-by: Colin Cross <ccross@android.com>
When the shrinkers are called with GFP_HIGH free low memory first,
it is more important to have free than high memory.
Change-Id: I7ad8a9c133830f04d429c3d87b781b3e862ccedb
Signed-off-by: Colin Cross <ccross@android.com>
ion_heap_buffer_zero can spend a long time in unmap_kernel_range
if it has to broadcast a tlb flush to every cpu for every page.
Modify it to batch pages into a larger region to clear using a
single mapping. This may cause the mapping size to change if
the buffer size is not a multiple of the mapping size, so
switch to allocating the address space for each chunk. This
allows us to use vm_map_ram to handle the allocation and mapping
together.
The number of pages to zero using a single mapping is set to 32
to hit the fastpath in vm_map_ram.
Change-Id: I1accfe67b285cbc9e95e387bea4246864197827d
Signed-off-by: Colin Cross <ccross@android.com>
Ion will compile and run on other platforms now, remove the
dependency on ARM.
Change-Id: I9da0ab686708bdab575a021031392b4402cce090
Signed-off-by: Colin Cross <ccross@android.com>
In testing ion system heap allocations, I ran across two issues:
1) Not k*z*allocing the sg table. This can cause trouble if
we end up trying call sg_alloc_table() with too many entries,
then sg_alloc_table() internally fails and tries to free what it
thinks is internal table structure, which causes bad pointer
traversals.
2) The second list_for_each_entry probably should be _safe,
since I was seeing strange lock warnings and oopses on occasion.
This seems to resolve it, but could use some extra checking.
Change-Id: I59d4c90104a8cf23dc4ae814d0b17348f1b68ac0
Signed-off-by: John Stultz <john.stultz@linaro.org>
The `buffer' variable is being used after being freed. Fix this.
Change-Id: Iea3471fa7dc7535bbf0620c1639fea2008d7cf19
Signed-off-by: Mitchel Humpherys <mitchelh@codeaurora.org>
Convert the ion ioctls to use _IOW instead of _IOWR where
appropriate, and factor out the copy_from_user and copy_to_user
based on the _IOC_DIR bits. For the existing incorrect ioctls,
add a function to wrap _IOC_DIR to return the corrected value.
Change-Id: I3cc34c84b9c52305bdbec27a9224447b102fadcd
Signed-off-by: Colin Cross <ccross@android.com>
ion_system_contig_heap buffers have an sglist, just call
ion_heap_map_user to map it.
Change-Id: I6dea383955834613fa8833659b31533c957c2b0b
Signed-off-by: Colin Cross <ccross@android.com>
Use %z for size_t and %pa for dma_addr_t to avoid warnings in printks.
Change-Id: I2c72874acd0b69cb35fca691928783817deb9394
Signed-off-by: Colin Cross <ccross@android.com>
phys_to_page and __phys_to_pfn don't exist on all platforms.
Use a combination of pfn_to_page, PFN_DOWN, page_to_pfn, and
virt_to_page to get the same results.
Change-Id: I53cef26059800bc8b7fb85ae458741574c97c257
Signed-off-by: Colin Cross <ccross@android.com>
ion_heap_map_kernel already implements mapping a scatterlist of
pages into the kernel, and all heaps are required to have struct
pages associated with them, so delete the functions that use
__arm_ioremap and use ion_heap_map_kernel instead.
Change-Id: Ia2dfd8d8c6e719d7d2f68dd4c458826fdb938260
Signed-off-by: Colin Cross <ccross@android.com>
Use idr_alloc instead if idr_pre_get/idr_get_new_above, and
remove idr_remove_all.
Change-Id: I675b789879549bd3767ed3ef2016cf108eb622d2
Signed-off-by: Colin Cross <ccross@android.com>
Add a /dev/ion-test device that will be created if CONFIG_ION_TEST
is set. The device accepts a dma_buf fd and allows reading and
writing to the backing memory using DMA-like apis or kernel mapping
apis. Can be used to test the dma_buf mapping ops, including
the ion implementations, from userspace.
Change-Id: I30703ba69cd75bdfe7767ac642e5f0cacd8d0478
Signed-off-by: Colin Cross <ccross@android.com>
The CLCD driver is used on ARM reference models for ARMv8 so add ARM64
to the list of dependencies. The driver also has no build time dependencies
on ARM (stubs are provided for ARM-specific DMA functions in the code) so
make it available with COMPILE_TEST in order to maximise build coverage.
Signed-off-by: Mark Brown <broonie@linaro.org>
commit 389a539058 upstream.
Now that scatterwalk_sg_chain sets the chain pointer bit the sg_page
call in scatterwalk_sg_next hits a BUG_ON when CONFIG_DEBUG_SG is
enabled. Use sg_chain_ptr instead of sg_page on a chain entry.
Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 12b69a5997 upstream.
Various Marvell datasheets advertise second PCIe unit of mv78230
flavour of Armada XP as x4/quad x1 capable. This second unit is in
fact only x1 capable. This patch fixes current mv78230 .dtsi to
reflect that, i.e. makes 1.0 the second interface (instead of 2.0
at the moment). This was successfully tested on a mv78230-based
ReadyNAS 2120 platform with a x1 device (FL1009 XHCI controller)
connected to this second interface.
Signed-off-by: Arnaud Ebalard <arno@natisbad.org>
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Jason Cooper <jason@lakedaemon.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 2163e61c92 upstream.
mv78260 flavour of Marvell Armada XP SoC has 3 PCIe units. The
two first units are both x4 and quad x1 capable. The third unit
is only x4 capable. This patch fixes mv78260 .dtsi to reflect
those capabilities.
Signed-off-by: Arnaud Ebalard <arno@natisbad.org>
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Jason Cooper <jason@lakedaemon.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
