mem: add function to check if memory is contiguous

For now, memory is always contiguous because legacy mem mode is
enabled unconditionally, but this function will be helpful down
the line when we implement support for allocating physically
non-contiguous memory. We can no longer guarantee physically
contiguous memory unless we're in legacy or IOVA_AS_VA mode, but
we can certainly try and see if we succeed.

In addition, this would be useful for e.g. PMD's who may allocate
chunks that are smaller than the pagesize, but they must not cross
the page boundary, in which case we will be able to accommodate
that request. This function will also support non-hugepage memory.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

eal: add single file segments option

Currently, DPDK stores all pages as separate files in hugetlbfs.
This option will allow storing all pages in one file (one file
per memseg list).

We do this by using fallocate() calls on FreeBSD, however this is
only supported on fairly recent (4.3+) kernels, so ftruncate()
fallback is provided to grow (but not shrink) hugepage files.
Naming scheme is deterministic, so both primary and secondary
processes will be able to easily map needed files and offsets.

For multi-file segments, we can close fd's right away. For
single-file segments, we can reuse the same fd and reduce the
amount of fd's needed to map/use hugepages. However, we need to
store the fd's somewhere, so we add a tailq.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

mem: support unmapping pages at runtime

This isn't used anywhere yet, but the support is now there. Also,
adding cleanup to allocation procedures, so that if we fail to
allocate everything we asked for, we can free all of it back.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

mem: support mapping hugepages at runtime

Nothing uses this code yet. The bulk of it is copied from old
memory allocation code (linuxapp eal_memory.c). We provide an
EAL-internal API to allocate either one page or multiple pages,
guaranteeing that we'll get contiguous VA for all of the pages
that we requested.

Not supported on FreeBSD.

Locking is done via fcntl() because that way, when it comes to
taking out write locks or unlocking on deallocation, we don't
have to keep original fd's around. Plus, using fcntl() gives us
ability to lock parts of a file, which is useful for single-file
segments, which are coming down the line.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

memzone: replace memzone array with fbarray

It's there, so we might as well use it. Some operations will be
sped up by that.

Since we have to allocate an fbarray for memzones, we have to do
it before we initialize memory subsystem, because that, in
secondary processes, will (later) allocate more fbarrays than the
primary process, which will result in inability to attach to
memzone fbarray if we do it after the fact.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

mem: replace memseg with memseg lists

Before, we were aggregating multiple pages into one memseg, so the
number of memsegs was small. Now, each page gets its own memseg,
so the list of memsegs is huge. To accommodate the new memseg list
size and to keep the under-the-hood workings sane, the memseg list
is now not just a single list, but multiple lists. To be precise,
each hugepage size available on the system gets one or more memseg
lists, per socket.

In order to support dynamic memory allocation, we reserve all
memory in advance (unless we're in 32-bit legacy mode, in which
case we do not preallocate memory). As in, we do an anonymous
mmap() of the entire maximum size of memory per hugepage size, per
socket (which is limited to either RTE_MAX_MEMSEG_PER_TYPE pages or
RTE_MAX_MEM_MB_PER_TYPE megabytes worth of memory, whichever is the
smaller one), split over multiple lists (which are limited to
either RTE_MAX_MEMSEG_PER_LIST memsegs or RTE_MAX_MEM_MB_PER_LIST
megabytes per list, whichever is the smaller one). There is also
a global limit of CONFIG_RTE_MAX_MEM_MB megabytes, which is mainly
used for 32-bit targets to limit amounts of preallocated memory,
but can be used to place an upper limit on total amount of VA
memory that can be allocated by DPDK application.

So, for each hugepage size, we get (by default) up to 128G worth
of memory, per socket, split into chunks of up to 32G in size.
The address space is claimed at the start, in eal_common_memory.c.
The actual page allocation code is in eal_memalloc.c (Linux-only),
and largely consists of copied EAL memory init code.

Pages in the list are also indexed by address. That is, in order
to figure out where the page belongs, one can simply look at base
address for a memseg list. Similarly, figuring out IOVA address
of a memzone is a matter of finding the right memseg list, getting
offset and dividing by page size to get the appropriate memseg.

This commit also removes rte_eal_dump_physmem_layout() call,
according to deprecation notice [1], and removes that deprecation
notice as well.

On 32-bit targets due to limited VA space, DPDK will no longer
spread memory to different sockets like before. Instead, it will
(by default) allocate all of the memory on socket where master
lcore is. To override this behavior, --socket-mem must be used.

The rest of the changes are really ripple effects from the memseg
change - heap changes, compile fixes, and rewrites to support
fbarray-backed memseg lists. Due to earlier switch to _walk()
functions, most of the changes are simple fixes, however some
of the _walk() calls were switched to memseg list walk, where
it made sense to do so.

Additionally, we are also switching locks from flock() to fcntl().
Down the line, we will be introducing single-file segments option,
and we cannot use flock() locks to lock parts of the file. Therefore,
we will use fcntl() locks for legacy mem as well, in case someone is
unfortunate enough to accidentally start legacy mem primary process
alongside an already working non-legacy mem-based primary process.

[1] http://dpdk.org/dev/patchwork/patch/34002/

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

eal: add shared indexed file-backed array

rte_fbarray is a simple indexed array stored in shared memory
via mapping files into memory. Rationale for its existence is the
following: since we are going to map memory page-by-page, there
could be quite a lot of memory segments to keep track of (for
smaller page sizes, page count can easily reach thousands). We
can't really make page lists truly dynamic and infinitely expandable,
because that involves reallocating memory (which is a big no-no in
multiprocess). What we can do instead is have a maximum capacity as
something really, really large, and decide at allocation time how
big the array is going to be. We map the entire file into memory,
which makes it possible to use fbarray as shared memory, provided
the structure itself is allocated in shared memory. Per-fbarray
locking is also used to avoid index data races (but not contents
data races - that is up to user application to synchronize).

In addition, in understanding that we will frequently need to scan
this array for free space and iterating over array linearly can
become slow, rte_fbarray provides facilities to index array's
usage. The following use cases are covered:
 - find next free/used slot (useful either for adding new elements
   to fbarray, or walking the list)
 - find starting index for next N free/used slots (useful for when
   we want to allocate chunk of VA-contiguous memory composed of
   several pages)
 - find how many contiguous free/used slots there are, starting
   from specified index (useful for when we want to figure out
   how many pages we have until next hole in allocated memory, to
   speed up some bulk operations where we would otherwise have to
   walk the array and add pages one by one)

This is accomplished by storing a usage mask in-memory, right
after the data section of the array, and using some bit-level
magic to figure out the info we need.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

eal: add legacy memory option

This adds a "--legacy-mem" command-line switch. It will be used to
go back to the old memory behavior, one where we can't dynamically
allocate/free memory (the downside), but one where the user can
get physically contiguous memory, like before (the upside).

For now, nothing but the legacy behavior exists, non-legacy
memory init sequence will be added later. For FreeBSD, non-legacy
memory init will never be enabled, while for Linux, it is
disabled in this patch to avoid breaking bisect, but will be
enabled once non-legacy mode will be fully operational.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

vfio: allow to map other memory regions

Currently it is not possible to use memory that is not owned by DPDK to
perform DMA. This scenarion might be used in vhost applications (like
SPDK) where guest send its own memory table. To fill this gap provide
API to allow registering arbitrary address in VFIO container.

Signed-off-by: Pawel Wodkowski <pawelx.wodkowski@intel.com>
Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Signed-off-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

memzone: use walk instead of iteration for dumping

Simplify memzone dump code to use memzone walk, to not maintain
the same memzone iteration code twice.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

net/mlx5: use virt2memseg instead of iteration

Reduce dependency on internal details of EAL memory subsystem, and
simplify code.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

net/mlx4: use virt2memseg instead of iteration

Reduce dependency on internal details of EAL memory subsystem, and
simplify code.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

crypto/dpaa_sec: use virt2memseg instead of iteration

Reduce dependency on internal details of EAL memory subsystem, and
simplify code.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

bus/fslmc: use virt2memseg instead of iteration

Reduce dependency on internal details of EAL memory subsystem, and
simplify code.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

mem: add virt2memseg function

This can be used as a virt2iova function that only looks up
memory that is owned by DPDK (as opposed to doing pagemap walks).
Using this will result in less dependency on internals of mem API.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

crypto/dpaa_sec: use iova2virt instead of memseg iteration

Reduce dependency on internal details of EAL memory subsystem, and
simplify code.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

bus/fslmc: use iova2virt instead of memseg iteration

Reduce dependency on internal details of EAL memory subsystem, and
simplify code.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

bus/dpaa: use iova2virt instead of memseg iteration

Reduce dependency on internal details of EAL memory subsystem, and
simplify code.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

mem: add iova2virt function

This is reverse lookup of PA to VA. Using this will make
other code less dependent on internals of mem API.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

net/virtio: use memseg contig walk instead of iteration

Reduce dependency on internal details of EAL memory subsystem, and
simplify code.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

mem: add contig walk function

This function is meant to walk over first segment of each
VA-contiguous group of memsegs.

For future users of this function, this is done so that
there is less dependency on internals of mem API and less
noise later change sets.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

vfio/spapr: use memseg walk instead of iteration

Reduce dependency on internal details of EAL memory subsystem, and
simplify code.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

vfio/type1: use memseg walk instead of iteration

Reduce dependency on internal details of EAL memory subsystem, and
simplify code.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

test: use memseg walk instead of iteration

Reduce dependency on internal details of EAL memory subsystem, and
simplify code.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

mempool: use memseg walk instead of iteration

Reduce dependency on internal details of EAL memory subsystem, and
simplify code.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

eal: use memseg walk instead of iteration

Reduce dependency on internal details of EAL memory subsystem, and
simplify code.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

net/mlx5: use memseg walk instead of iteration

Reduce dependency on internal details of EAL memory subsystem, and
simplify code.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

bus/pci: use memseg walk instead of iteration

Reduce dependency on internal details of EAL memory subsystem, and
simplify code.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

bus/fslmc: use memseg walk instead of iteration

Reduce dependency on internal details of EAL memory subsystem, and
simplify code.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

mem: add function to walk all memsegs

For code that might need to iterate over list of allocated
segments, using this API will make it more resilient to
internal API changes and will prevent copying the same
iteration code over and over again.

Additionally, down the line there will be locking implemented,
so users of this API will not need to care about locking
either.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

mempool: support new allocation methods

If a user has specified that the zone should have contiguous memory,
add a memzone flag to request contiguous memory. Otherwise, account
for the fact that unless we're in IOVA_AS_VA mode, we cannot
guarantee that the pages would be physically contiguous, so we
calculate the memzone size and alignments as if we were getting
the smallest page size available.

However, for the non-IOVA contiguous case, existing mempool size
calculation function doesn't give us expected results, because it
will return memzone sizes aligned to page size (e.g. a 1MB mempool
may use an entire 1GB page), therefore in cases where we weren't
specifically asked to reserve non-contiguous memory, first try
reserving a memzone as IOVA-contiguous, and if that fails, then
try reserving with page-aligned size/alignment.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Olivier Matz <olivier.matz@6wind.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

net/vmxnet3: use contiguous allocation for DMA memory

All hardware drivers should allocate IOVA-contiguous
memzones for their hardware resources.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

net/virtio: use contiguous allocation for DMA memory

All hardware drivers should allocate IOVA-contiguous
memzones for their hardware resources.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Reviewed-by: Venkatesh Srinivas <venkateshs@google.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

net/qede: use contiguous allocation for DMA memory

All hardware drivers should allocate IOVA-contiguous
memzones for their hardware resources.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Harish Patil <harish.patil@cavium.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

net/i40e: use contiguous allocation for DMA memory

All hardware drivers should allocate IOVA-contiguous
memzones for their hardware resources.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

net/enic: use contiguous allocation for DMA memory

All hardware drivers should allocate IOVA-contiguous
memzones for their hardware resources.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: John Daley <johndale@cisco.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

net/ena: use contiguous allocation for DMA memory

All hardware drivers should allocate IOVA-contiguous
memzones for their hardware resources.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Michal Krawczyk <mk@semihalf.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

net/cxgbe: use contiguous allocation for DMA memory

All hardware drivers should allocate IOVA-contiguous
memzones for their hardware resources.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

net/bnxt: use contiguous allocation for DMA memory

All hardware drivers should allocate IOVA-contiguous
memzones for their hardware resources.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

net/bnx2x: use contiguous allocation for DMA memory

All hardware drivers should allocate IOVA-contiguous
memzones for their hardware resources.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

net/avf: use contiguous allocation for DMA memory

All hardware drivers should allocate IOVA-contiguous
memzones for their hardware resources.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

crypto/qat: use contiguous allocation for DMA memory

All hardware drivers should allocate IOVA-contiguous
memzones for their hardware resources.

Also, remove the weird page alignment code.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Fiona Trahe <fiona.trahe@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

ethdev: use contiguous allocation for DMA memory

All hardware drivers should allocate IOVA-contiguous
memzones for their hardware resources.

This fixes the following drivers in one go:

grep -Rl rte_eth_dma_zone_reserve drivers/

drivers/net/avf/avf_rxtx.c
drivers/net/thunderx/nicvf_ethdev.c
drivers/net/e1000/igb_rxtx.c
drivers/net/e1000/em_rxtx.c
drivers/net/fm10k/fm10k_ethdev.c
drivers/net/vmxnet3/vmxnet3_rxtx.c
drivers/net/liquidio/lio_rxtx.c
drivers/net/i40e/i40e_rxtx.c
drivers/net/sfc/sfc.c
drivers/net/ixgbe/ixgbe_rxtx.c
drivers/net/nfp/nfp_net.c

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

memzone: enable IOVA-contiguous reserving

This adds a new flag to request reserved memzone to be IOVA
contiguous. This is useful for allocating hardware resources like
NIC rings/queues etc.For now, hugepage memory is always contiguous,
but we need to prepare the drivers for the switch.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

malloc: support contiguous allocation

No major changes, just add some checks in a few key places, and
a new parameter to pass around.

Also, add a function to check malloc element for physical
contiguousness. For now, assume hugepage memory is always
contiguous, while non-hugepage memory will be checked.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

malloc: replace panics with error messages

We shouldn't ever panic in libraries, let alone in EAL, so
replace all panic messages with error messages.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

malloc: make free return resulting element

This will be needed because we need to know how big is the
new empty space, to check whether we can free some pages as
a result.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

malloc: make free list removal function public

We will need to be able to remove entries from free lists from
heaps during certain events, such as rollbacks, or when freeing
memory to the system (where a previously element disappears and
thus can no longer be in the free list).

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

malloc: make join elements function public

Down the line, we will need to join free segments to determine
whether the resulting contiguous free space is bigger than a
page size, allowing to free some memory back to the system.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

test: add command to dump malloc heap contents

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

malloc: add function to dump heap contents

Malloc heap is now a doubly linked list, so it's now possible to
iterate over each malloc element regardless of its state.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

malloc: make heap a doubly-linked list

As we are preparing for dynamic memory allocation, we need to be
able to handle holes in our malloc heap, hence we're switching to
doubly linked list, and prepare infrastructure to support it.

Since our heap is now aware where are our first and last elements,
there is no longer any need to have a dummy element at the end of
each heap, so get rid of that as well. Instead, let insert/remove/
join/split operations handle end-of-list conditions automatically.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

malloc: move all locking to heap

Down the line, we will need to do everything from the heap as any
alloc or free may trigger alloc/free OS memory, which would involve
growing/shrinking heap.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

mem: move virtual area function in common directory

Move get_virtual_area out of linuxapp EAL memory and make it
common to EAL, so that other code could reserve virtual areas
as well.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Tested-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Tested-by: Gowrishankar Muthukrishnan <gowrishankar.m@linux.vnet.ibm.com>

vfio: do not needlessly check for IOVA mode

We already set IOVA addresses of memsegs and memzones to VA
address during initialization, so we don't need to check
whether we're in RTE_IOVA_VA mode anywhere else.

Cc: stable@dpdk.org
Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>

bus/fslmc: do not needlessly check for IOVA mode

We already set IOVA addresses of memsegs and memzones to VA
address during initialization, so we don't need to check
whether we're in RTE_IOVA_VA mode anywhere else.

Cc: stable@dpdk.org
Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>

mem: do not use physical addresses in IOVA as VA mode

We already use VA addresses for IOVA purposes everywhere if we're in
RTE_IOVA_VA mode:
 1) rte_malloc_virt2phy()/rte_malloc_virt2iova() always return VA addresses
 2) Because of 1), memzone's IOVA is set to VA address on reserve
 3) Because of 2), mempool's IOVA addresses are set to VA addresses

The only place where actual physical addresses are stored is in memsegs at
init time, but we're not using them anywhere, and there is no external API
to get those addresses (aside from manually iterating through memsegs), nor
should anyone care about them in RTE_IOVA_VA mode.

So, fix EAL initialization to allocate VA-contiguous segments at the start
without regard for physical addresses (as if they weren't available), and
use VA to set final IOVA addresses for all pages.

Fixes: 62196f4e0941 ("mem: rename address mapping function to IOVA")
Cc: stable@dpdk.org
Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>

app/crypto-perf: fix IOVA translation

IOVA addresses should be found by calling rte_virt2iova() as
opposed to rte_virt2phy(), as physical address may not be
equal to IOVA address.

Fixes: 2eb6a1a3e5fc ("app/crypto-perf: fix crypto op init")
Cc: stable@dpdk.org
Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>

usertools: change to SPDX license identifier

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>

usertools: add missing SPDX identifier

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>

align SPDX Mellanox copyrights

Aligning Mellanox SPDX copyrights to a single format.
In addition replace to SPDX licence files which were missed.

Signed-off-by: Shahaf Shuler <shahafs@mellanox.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>

eal/arm: use SPDX tag for Cavium and RehiveTech copyright file

Replace the BSD license header with the SPDX tag for files
with a RehiveTech and Cavium copyright on them.

Signed-off-by: Jan Viktorin <viktorin@rehivetech.com>
Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>

use SPDX tag for RehiveTech copyright files

Replace the BSD license header with the SPDX tag for files
with only an RehiveTech copyright on them.

Signed-off-by: Jan Viktorin <viktorin@rehivetech.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>

maintainers: call out subtree committers

The MAINTAINERS file contains information of the maintainers
of the different components on DPDK.
However, it does not give any information on who maintains the
different subtrees which accept new commits for these components.

This commit adds a list of the subtree committers.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Ferruh Yigit <ferruh.yigit@intel.com>

maintainers: maintain avf PMD in next-net-intel

Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>

maintainers: claim responsibility for vmxnet3

Signed-off-by: Yong Wang <yongwang@vmware.com>

hash: fix missing spinlock unlock in add key

Fix missing spinlock unlock during add key when key is already present.

Fixes: be856325cba3 ("hash: add scalable multi-writer insertion with Intel TSX")
Cc: stable@dpdk.org
Signed-off-by: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

examples/ip_pipeline: skip in FreeBSD build

IP_Pipeline app is not supported in FreeBSD environment. Therefore,
skip it while building the sample apps on FreeBSD.

Fixes: 4bbf8e30aa5e ("examples/ip_pipeline: add CLI interface")
Fixes: 2f74ae28e23f ("examples/ip_pipeline: add tap object")

Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Acked-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>

doc: add meter API change to release notes

Update the release notes with meter api change to support configuration
profiles.

Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Acked-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>

examples/ip_pipeline: add ECMP route example

Add example to build pipeline to demonstrate equal-cost multi-path
routing example.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>

examples/ip_pipeline: fix load balance table action

Fix bug in load balance action for pipeline table.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>

examples/ip_pipeline: add flow classification example

Add example to build pipeline with hash table to classify the
ingress traffic.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Signed-off-by: Fan Zhang <roy.fan.zhang@intel.com>

examples/ip_pipeline: add firewall example

Add example to built pipeline with ACL table to demonstrate
the firewall operation.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>

examples/ip_pipeline: add route example

Add example to built pipeline with LPM table to demonstrate layer 3
routing.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Signed-off-by: Reshma Pattan <reshma.pattan@intel.com>

examples/ip_pipeline: add TAP port example

Add example to illustrate the pipeline functioning with TAP
interface.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Signed-off-by: Kevin Laatz <kevin.laatz@intel.com>

examples/ip_pipeline: add KNI port example

Add example to illustrate the pipeline functioning with KNI
interface.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Signed-off-by: Kevin Laatz <kevin.laatz@intel.com>

examples/ip_pipeline: add l2fwd example

This patch add the configuration file for l2fwd example. It
includes commands to build the packet processing stage (pipeline),
defining action, add rules to its table, etc.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Signed-off-by: Kevin Laatz <kevin.laatz@intel.com>

examples/ip_pipeline: add load balance action command

Add command for load balance action.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>

examples/ip_pipeline: add TTL stats command

Add command to read the ttl stats.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>

examples/ip_pipeline: add DSCP table update command

Add command to update the dscp table for traffic meter and traffic
manager.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>

examples/ip_pipeline: add meter stats command

Add command to read traffic meter stats.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>

examples/ip_pipeline: add meter profile commands

Add commands to configure the traffic meter profile.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>

examples/ip_pipeline: add table entry stats command

Add command to read the pipeline table entry stats.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Signed-off-by: Fan Zhang <roy.fan.zhang@intel.com>

examples/ip_pipeline: add table bulk add command

Add cli for adding bulk entries to pipeline table.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Signed-off-by: Kevin Laatz <kevin.laatz@intel.com>

examples/ip_pipeline: add table entry delete command

Add command to delete the pipeline table entry.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>

examples/ip_pipeline: add table entry commands

Add commands to add pipeline table entries which contains match and
action part.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Signed-off-by: Kevin Laatz <kevin.laatz@intel.com>

examples/ip_pipeline: add stats read commands

Add commands to read the pipeline  port in, port out
and table stats.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Signed-off-by: Fan Zhang <roy.fan.zhang@intel.com>

examples/ip_pipeline: add port enable and disable commands

Add commands to enable and disable the pipeline ports.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Signed-off-by: Fan Zhang <roy.fan.zhang@intel.com>

examples/ip_pipeline: add enable and disable commands

Add commands to enable and disable the pipeline on the thread.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Signed-off-by: Fan Zhang <roy.fan.zhang@intel.com>

examples/ip_pipeline: add thread runtime

Add runtime thread functions for the pipeline.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>

examples/ip_pipeline: add threads

Add threads data structure and initialisation functions to run
the pipeline.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>

examples/ip_pipeline: add pipeline object

Add pipeline object implementation to the application.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>

examples/ip_pipeline: add action profile objects

Add action profile object implementation to the application.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Signed-off-by: Fan Zhang <roy.fan.zhang@intel.com>

examples/ip_pipeline: add KNI object

Add kni object implementation to the application.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Signed-off-by: Kevin Laatz <kevin.laatz@intel.com>

examples/ip_pipeline: add tap object

Add tap object implementation to the application

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Signed-off-by: Kevin Laatz <kevin.laatz@intel.com>

examples/ip_pipeline: add traffic manager object

Add traffic manager object implementation to the application.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>

examples/ip_pipeline: add sw queue object

Add swq object implementation to the application.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Signed-off-by: Kevin Laatz <kevin.laatz@intel.com>

examples/ip_pipeline: add link object

Add link object implementation to the application.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Signed-off-by: Fan Zhang <roy.fan.zhang@intel.com>

examples/ip_pipeline: add mempool object

Add mempool object implementation to the application.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Signed-off-by: Fan Zhang <roy.fan.zhang@intel.com>

examples/ip_pipeline: add CLI interface

CLI interface allowing connectivity with external agent (e.g. telnet,
netcat, Python script, etc) is added to the application.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>