xsk: add functions to fill control buffer

Commit 94ecc5ca4dbf ("xsk: Add cb area to struct xdp_buff_xsk") has added
a buffer for custom data to xdp_buff_xsk. Particularly, this memory is used
for data, consumed by XDP hints kfuncs. It does not always change on
a per-packet basis and some parts can be set for example, at the same time
as RX queue info.

Add functions to fill all cbs in xsk_buff_pool with the same metadata.

Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>
Acked-by: Magnus Karlsson <magnus.karlsson@intel.com>
Link: https://lore.kernel.org/r/20231205210847.28460-8-larysa.zaremba@intel.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>

xsk: Add option to calculate TX checksum in SW

For XDP_COPY mode, add a UMEM option XDP_UMEM_TX_SW_CSUM
to call skb_checksum_help in transmit path. Might be useful
to debugging issues with real hardware. I also use this mode
in the selftests.

Signed-off-by: Stanislav Fomichev <sdf@google.com>
Link: https://lore.kernel.org/r/20231127190319.1190813-9-sdf@google.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>

xsk: Add TX timestamp and TX checksum offload support

This change actually defines the (initial) metadata layout
that should be used by AF_XDP userspace (xsk_tx_metadata).
The first field is flags which requests appropriate offloads,
followed by the offload-specific fields. The supported per-device
offloads are exported via netlink (new xsk-flags).

The offloads themselves are still implemented in a bit of a
framework-y fashion that's left from my initial kfunc attempt.
I'm introducing new xsk_tx_metadata_ops which drivers are
supposed to implement. The drivers are also supposed
to call xsk_tx_metadata_request/xsk_tx_metadata_complete in
the right places. Since xsk_tx_metadata_{request,_complete}
are static inline, we don't incur any extra overhead doing
indirect calls.

The benefit of this scheme is as follows:
- keeps all metadata layout parsing away from driver code
- makes it easy to grep and see which drivers implement what
- don't need any extra flags to maintain to keep track of what
offloads are implemented; if the callback is implemented - the offload
is supported (used by netlink reporting code)

Two offloads are defined right now:
1. XDP_TXMD_FLAGS_CHECKSUM: skb-style csum_start+csum_offset
2. XDP_TXMD_FLAGS_TIMESTAMP: writes TX timestamp back into metadata
area upon completion (tx_timestamp field)

XDP_TXMD_FLAGS_TIMESTAMP is also implemented for XDP_COPY mode: it writes
SW timestamp from the skb destructor (note I'm reusing hwtstamps to pass
metadata pointer).

The struct is forward-compatible and can be extended in the future
by appending more fields.

Reviewed-by: Song Yoong Siang <yoong.siang.song@intel.com>
Signed-off-by: Stanislav Fomichev <sdf@google.com>
Acked-by: Jakub Kicinski <kuba@kernel.org>
Link: https://lore.kernel.org/r/20231127190319.1190813-3-sdf@google.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>

xsk: Support tx_metadata_len

For zerocopy mode, tx_desc->addr can point to an arbitrary offset
and carry some TX metadata in the headroom. For copy mode, there
is no way currently to populate skb metadata.

Introduce new tx_metadata_len umem config option that indicates how many
bytes to treat as metadata. Metadata bytes come prior to tx_desc address
(same as in RX case).

The size of the metadata has mostly the same constraints as XDP:
- less than 256 bytes
- 8-byte aligned (compared to 4-byte alignment on xdp, due to 8-byte
timestamp in the completion)
- non-zero

This data is not interpreted in any way right now.

Reviewed-by: Song Yoong Siang <yoong.siang.song@intel.com>
Signed-off-by: Stanislav Fomichev <sdf@google.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Link: https://lore.kernel.org/r/20231127190319.1190813-2-sdf@google.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>

xsk: support mbuf on ZC RX

Given that skb_shared_info relies on skb_frag_t, in order to support
xskb chaining, introduce xdp_buff_xsk::xskb_list_node and
xsk_buff_pool::xskb_list.

This is needed so ZC drivers can add frags as xskb nodes which will make
it possible to handle it both when producing AF_XDP Rx descriptors as
well as freeing/recycling all the frags that a single frame carries.

Speaking of latter, update xsk_buff_free() to take care of list nodes.
For the former (adding as frags), introduce xsk_buff_add_frag() for ZC
drivers usage that is going to be used to add a frag to xskb list from
pool.

xsk_buff_get_frag() will be utilized by XDP_TX and, on contrary, will
return xdp_buff.

One of the previous patches added a wrapper for ZC Rx so implement xskb
list walk and production of Rx descriptors there.

On bind() path, bail out if socket wants to use ZC multi-buffer but
underlying netdev does not support it.

Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Link: https://lore.kernel.org/r/20230719132421.584801-12-maciej.fijalkowski@intel.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>

xsk: allow core/drivers to test EOP bit

Drivers are used to check for EOP bit whereas AF_XDP operates on
inverted logic - user space indicates that current frag is not the last
one and packet continues. For AF_XDP core needs, add xp_mb_desc() that
will simply test XDP_PKT_CONTD from xdp_desc::options, but in order to
preserve drivers default behavior, introduce an interface for ZC drivers
that will negate xp_mb_desc() result and therefore make it easier to
test EOP bit from during production of HW Tx descriptors.

Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Link: https://lore.kernel.org/r/20230719132421.584801-8-maciej.fijalkowski@intel.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>

xsk: Use pool->dma_pages to check for DMA

Compare pool->dma_pages instead of pool->dma_pages_cnt to check for an
active DMA mapping. pool->dma_pages needs to be read anyway to access
the map so this compiles to more efficient code.

Signed-off-by: Kal Conley <kal.conley@dectris.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Xuan Zhuo <xuanzhuo@linux.alibaba.com>
Acked-by: Magnus Karlsson <magnus.karlsson@intel.com>
Link: https://lore.kernel.org/bpf/20230423180157.93559-1-kal.conley@dectris.com

xsk: Fix unaligned descriptor validation

Make sure unaligned descriptors that straddle the end of the UMEM are
considered invalid. Currently, descriptor validation is broken for
zero-copy mode which only checks descriptors at page granularity.
For example, descriptors in zero-copy mode that overrun the end of the
UMEM but not a page boundary are (incorrectly) considered valid. The
UMEM boundary check needs to happen before the page boundary and
contiguity checks in xp_desc_crosses_non_contig_pg(). Do this check in
xp_unaligned_validate_desc() instead like xp_check_unaligned() already
does.

Fixes: 2b43470add8c ("xsk: Introduce AF_XDP buffer allocation API")
Signed-off-by: Kal Conley <kal.conley@dectris.com>
Acked-by: Magnus Karlsson <magnus.karlsson@intel.com>
Link: https://lore.kernel.org/r/20230405235920.7305-2-kal.conley@dectris.com
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>

xsk: Add cb area to struct xdp_buff_xsk

Add an area after the xdp_buff in struct xdp_buff_xsk that drivers can use
to stash extra information to use in metadata kfuncs. The maximum size of
24 bytes means the full xdp_buff_xsk structure will take up exactly two
cache lines (with the cb field spanning both). Also add a macro drivers can
use to check their own wrapping structs against the available size.

Cc: John Fastabend <john.fastabend@gmail.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Martin KaFai Lau <martin.lau@linux.dev>
Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Willem de Bruijn <willemb@google.com>
Cc: Jesper Dangaard Brouer <brouer@redhat.com>
Cc: Anatoly Burakov <anatoly.burakov@intel.com>
Cc: Alexander Lobakin <alexandr.lobakin@intel.com>
Cc: Magnus Karlsson <magnus.karlsson@gmail.com>
Cc: Maryam Tahhan <mtahhan@redhat.com>
Cc: xdp-hints@xdp-project.net
Cc: netdev@vger.kernel.org
Suggested-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: Stanislav Fomichev <sdf@google.com>
Link: https://lore.kernel.org/r/20230119221536.3349901-15-sdf@google.com
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>

xsk: Inherit need_wakeup flag for shared sockets

The flag for need_wakeup is not set for xsks with `XDP_SHARED_UMEM`
flag and of different queue ids and/or devices. They should inherit
the flag from the first socket buffer pool since no flags can be
specified once `XDP_SHARED_UMEM` is specified.

Fixes: b5aea28dca134 ("xsk: Add shared umem support between queue ids")
Signed-off-by: Jalal Mostafa <jalal.a.mostapha@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Magnus Karlsson <magnus.karlsson@intel.com>
Link: https://lore.kernel.org/bpf/20220921135701.10199-1-jalal.a.mostapha@gmail.com

xsk: Fix possible crash when multiple sockets are created

Fix a crash that happens if an Rx only socket is created first, then a
second socket is created that is Tx only and bound to the same umem as
the first socket and also the same netdev and queue_id together with the
XDP_SHARED_UMEM flag. In this specific case, the tx_descs array page
pool was not created by the first socket as it was an Rx only socket.
When the second socket is bound it needs this tx_descs array of this
shared page pool as it has a Tx component, but unfortunately it was
never allocated, leading to a crash. Note that this array is only used
for zero-copy drivers using the batched Tx APIs, currently only ice and
i40e.

[ 5511.150360] BUG: kernel NULL pointer dereference, address: 0000000000000008
[ 5511.158419] #PF: supervisor write access in kernel mode
[ 5511.164472] #PF: error_code(0x0002) - not-present page
[ 5511.170416] PGD 0 P4D 0
[ 5511.173347] Oops: 0002 [#1] PREEMPT SMP PTI
[ 5511.178186] CPU: 0 PID: 0 Comm: swapper/0 Tainted: G E 5.18.0-rc1+ #97
[ 5511.187245] Hardware name: Intel Corp. GRANTLEY/GRANTLEY, BIOS GRRFCRB1.86B.0276.D07.1605190235 05/19/2016
[ 5511.198418] RIP: 0010:xsk_tx_peek_release_desc_batch+0x198/0x310
[ 5511.205375] Code: c0 83 c6 01 84 c2 74 6d 8d 46 ff 23 07 44 89 e1 48 83 c0 14 48 c1 e1 04 48 c1 e0 04 48 03 47 10 4c 01 c1 48 8b 50 08 48 8b 00 <48> 89 51 08 48 89 01 41 80 bd d7 00 00 00 00 75 82 48 8b 19 49 8b
[ 5511.227091] RSP: 0018:ffffc90000003dd0 EFLAGS: 00010246
[ 5511.233135] RAX: 0000000000000000 RBX: ffff88810c8da600 RCX: 0000000000000000
[ 5511.241384] RDX: 000000000000003c RSI: 0000000000000001 RDI: ffff888115f555c0
[ 5511.249634] RBP: ffffc90000003e08 R08: 0000000000000000 R09: ffff889092296b48
[ 5511.257886] R10: 0000ffffffffffff R11: ffff889092296800 R12: 0000000000000000
[ 5511.266138] R13: ffff88810c8db500 R14: 0000000000000040 R15: 0000000000000100
[ 5511.274387] FS: 0000000000000000(0000) GS:ffff88903f800000(0000) knlGS:0000000000000000
[ 5511.283746] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 5511.290389] CR2: 0000000000000008 CR3: 00000001046e2001 CR4: 00000000003706f0
[ 5511.298640] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 5511.306892] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[ 5511.315142] Call Trace:
[ 5511.317972] <IRQ>
[ 5511.320301] ice_xmit_zc+0x68/0x2f0 [ice]
[ 5511.324977] ? ktime_get+0x38/0xa0
[ 5511.328913] ice_napi_poll+0x7a/0x6a0 [ice]
[ 5511.333784] __napi_poll+0x2c/0x160
[ 5511.337821] net_rx_action+0xdd/0x200
[ 5511.342058] __do_softirq+0xe6/0x2dd
[ 5511.346198] irq_exit_rcu+0xb5/0x100
[ 5511.350339] common_interrupt+0xa4/0xc0
[ 5511.354777] </IRQ>
[ 5511.357201] <TASK>
[ 5511.359625] asm_common_interrupt+0x1e/0x40
[ 5511.364466] RIP: 0010:cpuidle_enter_state+0xd2/0x360
[ 5511.370211] Code: 49 89 c5 0f 1f 44 00 00 31 ff e8 e9 00 7b ff 45 84 ff 74 12 9c 58 f6 c4 02 0f 85 72 02 00 00 31 ff e8 02 0c 80 ff fb 45 85 f6 <0f> 88 11 01 00 00 49 63 c6 4c 2b 2c 24 48 8d 14 40 48 8d 14 90 49
[ 5511.391921] RSP: 0018:ffffffff82a03e60 EFLAGS: 00000202
[ 5511.397962] RAX: ffff88903f800000 RBX: 0000000000000001 RCX: 000000000000001f
[ 5511.406214] RDX: 0000000000000000 RSI: ffffffff823400b9 RDI: ffffffff8234c046
[ 5511.424646] RBP: ffff88810a384800 R08: 000005032a28c046 R09: 0000000000000008
[ 5511.443233] R10: 000000000000000b R11: 0000000000000006 R12: ffffffff82bcf700
[ 5511.461922] R13: 000005032a28c046 R14: 0000000000000001 R15: 0000000000000000
[ 5511.480300] cpuidle_enter+0x29/0x40
[ 5511.494329] do_idle+0x1c7/0x250
[ 5511.507610] cpu_startup_entry+0x19/0x20
[ 5511.521394] start_kernel+0x649/0x66e
[ 5511.534626] secondary_startup_64_no_verify+0xc3/0xcb
[ 5511.549230] </TASK>

Detect such case during bind() and allocate this memory region via newly
introduced xp_alloc_tx_descs(). Also, use kvcalloc instead of kcalloc as
for other buffer pool allocations, so that it matches the kvfree() from
xp_destroy().

Fixes: d1bc532e99be ("i40e: xsk: Move tmp desc array from driver to pool")
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Magnus Karlsson <magnus.karlsson@intel.com>
Link: https://lore.kernel.org/bpf/20220425153745.481322-1-maciej.fijalkowski@intel.com

i40e: xsk: Move tmp desc array from driver to pool

Move desc_array from the driver to the pool. The reason behind this is
that we can then reuse this array as a temporary storage for descriptors
in all zero-copy drivers that use the batched interface. This will make
it easier to add batching to more drivers.

i40e is the only driver that has a batched Tx zero-copy
implementation, so no need to touch any other driver.

Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Alexander Lobakin <alexandr.lobakin@intel.com>
Link: https://lore.kernel.org/bpf/20220125160446.78976-6-maciej.fijalkowski@intel.com

xsk: Optimize for aligned case

Optimize for the aligned case by precomputing the parameter values of
the xdp_buff_xsk and xdp_buff structures in the heads array. We can do
this as the heads array size is equal to the number of chunks in the
umem for the aligned case. Then every entry in this array will reflect
a certain chunk/frame and can therefore be prepopulated with the
correct values and we can drop the use of the free_heads stack. Note
that it is not possible to allocate more buffers than what has been
allocated in the aligned case since each chunk can only contain a
single buffer.

We can unfortunately not do this in the unaligned case as one chunk
might contain multiple buffers. In this case, we keep the old scheme
of populating a heads entry every time it is used and using
the free_heads stack.

Also move xp_release() and xp_get_handle() to xsk_buff_pool.h. They
were for some reason in xsk.c even though they are buffer pool
operations.

Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20210922075613.12186-7-magnus.karlsson@gmail.com

xsk: Batched buffer allocation for the pool

Add a new driver interface xsk_buff_alloc_batch() offering batched
buffer allocations to improve performance. The new interface takes
three arguments: the buffer pool to allocated from, a pointer to an
array of struct xdp_buff pointers which will contain pointers to the
allocated xdp_buffs, and an unsigned integer specifying the max number
of buffers to allocate. The return value is the actual number of
buffers that the allocator managed to allocate and it will be in the
range 0 <= N <= max, where max is the third parameter to the function.

u32 xsk_buff_alloc_batch(struct xsk_buff_pool *pool, struct xdp_buff **xdp,
u32 max);

A second driver interface is also introduced that need to be used in
conjunction with xsk_buff_alloc_batch(). It is a helper that sets the
size of struct xdp_buff and is used by the NIC Rx irq routine when
receiving a packet. This helper sets the three struct members data,
data_meta, and data_end. The two first ones is in the xsk_buff_alloc()
case set in the allocation routine and data_end is set when a packet
is received in the receive irq function. This unfortunately leads to
worse performance since the xdp_buff is touched twice with a long time
period in between leading to an extra cache miss. Instead, we fill out
the xdp_buff with all 3 fields at one single point in time in the
driver, when the size of the packet is known. Hence this helper. Note
that the driver has to use this helper (or set all three fields
itself) when using xsk_buff_alloc_batch(). xsk_buff_alloc() works as
before and does not require this.

void xsk_buff_set_size(struct xdp_buff *xdp, u32 size);

Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20210922075613.12186-3-magnus.karlsson@gmail.com

xsk: Get rid of unused entry in struct xdp_buff_xsk

Get rid of the unused entry "unaligned" in struct xdp_buff_xsk.

Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20210922075613.12186-2-magnus.karlsson@gmail.com

xsk: Fix missing validation for skb and unaligned mode

Fix a missing validation of a Tx descriptor when executing in skb mode
and the umem is in unaligned mode. A descriptor could point to a
buffer straddling the end of the umem, thus effectively tricking the
kernel to read outside the allowed umem region. This could lead to a
kernel crash if that part of memory is not mapped.

In zero-copy mode, the descriptor validation code rejects such
descriptors by checking a bit in the DMA address that tells us if the
next page is physically contiguous or not. For the last page in the
umem, this bit is not set, therefore any descriptor pointing to a
packet straddling this last page boundary will be rejected. However,
the skb path does not use this bit since it copies out data and can do
so to two different pages. (It also does not have the array of DMA
address, so it cannot even store this bit.) The code just returned
that the packet is always physically contiguous. But this is
unfortunately also returned for the last page in the umem, which means
that packets that cross the end of the umem are being allowed, which
they should not be.

Fix this by introducing a check for this in the SKB path only, not
penalizing the zero-copy path.

Fixes: 2b43470add8c ("xsk: Introduce AF_XDP buffer allocation API")
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Björn Töpel <bjorn@kernel.org>
Link: https://lore.kernel.org/bpf/20210617092255.3487-1-magnus.karlsson@gmail.com

xsk: Fix race in SKB mode transmit with shared cq

Fix a race when multiple sockets are simultaneously calling sendto()
when the completion ring is shared in the SKB case. This is the case
when you share the same netdev and queue id through the
XDP_SHARED_UMEM bind flag. The problem is that multiple processes can
be in xsk_generic_xmit() and call the backpressure mechanism in
xskq_prod_reserve(xs->pool->cq). As this is a shared resource in this
specific scenario, a race might occur since the rings are
single-producer single-consumer.

Fix this by moving the tx_completion_lock from the socket to the pool
as the pool is shared between the sockets that share the completion
ring. (The pool is not shared when this is not the case.) And then
protect the accesses to xskq_prod_reserve() with this lock. The
tx_completion_lock is renamed cq_lock to better reflect that it
protects accesses to the potentially shared completion ring.

Fixes: 35fcde7f8deb ("xsk: support for Tx")
Reported-by: Xuan Zhuo <xuanzhuo@linux.alibaba.com>
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Björn Töpel <bjorn.topel@intel.com>
Link: https://lore.kernel.org/bpf/20201218134525.13119-2-magnus.karlsson@gmail.com

xsk: Fix possible memory leak at socket close

Fix a possible memory leak at xsk socket close that is caused by the
refcounting of the umem object being wrong. The reference count of the
umem was decremented only after the pool had been freed. Note that if
the buffer pool is destroyed, it is important that the umem is
destroyed after the pool, otherwise the umem would disappear while the
driver is still running. And as the buffer pool needs to be destroyed
in a work queue, the umem is also (if its refcount reaches zero)
destroyed after the buffer pool in that same work queue.

What was missing is that the refcount also needs to be decremented
when the pool is not freed and when the pool has not even been
created. The first case happens when the refcount of the pool is
higher than 1, i.e. it is still being used by some other socket using
the same device and queue id. In this case, it is safe to decrement
the refcount of the umem outside of the work queue as the umem will
never be freed because the refcount of the umem is always greater than
or equal to the refcount of the buffer pool. The second case is if the
buffer pool has not been created yet, i.e. the socket was closed
before it was bound but after the umem was created. In this case, it
is safe to destroy the umem outside of the work queue, since there is
no pool that can use it by definition.

Fixes: 1c1efc2af158 ("xsk: Create and free buffer pool independently from umem")
Reported-by: syzbot+eb71df123dc2be2c1456@syzkaller.appspotmail.com
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Björn Töpel <bjorn.topel@intel.com>
Link: https://lore.kernel.org/bpf/1603801921-2712-1-git-send-email-magnus.karlsson@gmail.com

xsk: Add shared umem support between queue ids

Add support to share a umem between queue ids on the same
device. This mode can be invoked with the XDP_SHARED_UMEM bind
flag. Previously, sharing was only supported within the same
queue id and device, and you shared one set of fill and
completion rings. However, note that when sharing a umem between
queue ids, you need to create a fill ring and a completion ring
and tie them to the socket before you do the bind with the
XDP_SHARED_UMEM flag. This so that the single-producer
single-consumer semantics can be upheld.

Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Björn Töpel <bjorn.topel@intel.com>
Link: https://lore.kernel.org/bpf/1598603189-32145-12-git-send-email-magnus.karlsson@intel.com

xsk: i40e: ice: ixgbe: mlx5: Test for dma_need_sync earlier for better performance

Test for dma_need_sync earlier to increase
performance. xsk_buff_dma_sync_for_cpu() takes an xdp_buff as
parameter and from that the xsk_buff_pool reference is dug out. Perf
shows that this dereference causes a lot of cache misses. But as the
buffer pool is now sent down to the driver at zero-copy initialization
time, we might as well use this pointer directly, instead of going via
the xsk_buff and we can do so already in xsk_buff_dma_sync_for_cpu()
instead of in xp_dma_sync_for_cpu. This gets rid of these cache
misses.

Throughput increases with 3% for the xdpsock l2fwd sample application
on my machine.

Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Björn Töpel <bjorn.topel@intel.com>
Link: https://lore.kernel.org/bpf/1598603189-32145-11-git-send-email-magnus.karlsson@intel.com

xsk: Rearrange internal structs for better performance

Rearrange the xdp_sock, xdp_umem and xsk_buff_pool structures so
that they get smaller and align better to the cache lines. In the
previous commits of this patch set, these structs have been
reordered with the focus on functionality and simplicity, not
performance. This patch improves throughput performance by around
3%.

Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Björn Töpel <bjorn.topel@intel.com>
Link: https://lore.kernel.org/bpf/1598603189-32145-10-git-send-email-magnus.karlsson@intel.com

xsk: Enable sharing of dma mappings

Enable the sharing of dma mappings by moving them out from the buffer
pool. Instead we put each dma mapped umem region in a list in the umem
structure. If dma has already been mapped for this umem and device, it
is not mapped again and the existing dma mappings are reused.

Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Björn Töpel <bjorn.topel@intel.com>
Link: https://lore.kernel.org/bpf/1598603189-32145-9-git-send-email-magnus.karlsson@intel.com

xsk: Move xsk_tx_list and its lock to buffer pool

Move the xsk_tx_list and the xsk_tx_list_lock from the umem to
the buffer pool. This so that we in a later commit can share the
umem between multiple HW queues. There is one xsk_tx_list per
device and queue id, so it should be located in the buffer pool.

Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Björn Töpel <bjorn.topel@intel.com>
Link: https://lore.kernel.org/bpf/1598603189-32145-7-git-send-email-magnus.karlsson@intel.com

xsk: Move queue_id, dev and need_wakeup to buffer pool

Move queue_id, dev, and need_wakeup from the umem to the
buffer pool. This so that we in a later commit can share the umem
between multiple HW queues. There is one buffer pool per dev and
queue id, so these variables should belong to the buffer pool, not
the umem. Need_wakeup is also something that is set on a per napi
level, so there is usually one per device and queue id. So move
this to the buffer pool too.

Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Björn Töpel <bjorn.topel@intel.com>
Link: https://lore.kernel.org/bpf/1598603189-32145-6-git-send-email-magnus.karlsson@intel.com

xsk: Move fill and completion rings to buffer pool

Move the fill and completion rings from the umem to the buffer
pool. This so that we in a later commit can share the umem
between multiple HW queue ids. In this case, we need one fill and
completion ring per queue id. As the buffer pool is per queue id
and napi id this is a natural place for it and one umem
struture can be shared between these buffer pools.

Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Björn Töpel <bjorn.topel@intel.com>
Link: https://lore.kernel.org/bpf/1598603189-32145-5-git-send-email-magnus.karlsson@intel.com

xsk: Create and free buffer pool independently from umem

Create and free the buffer pool independently from the umem. Move
these operations that are performed on the buffer pool from the
umem create and destroy functions to new create and destroy
functions just for the buffer pool. This so that in later commits
we can instantiate multiple buffer pools per umem when sharing a
umem between HW queues and/or devices. We also erradicate the
back pointer from the umem to the buffer pool as this will not
work when we introduce the possibility to have multiple buffer
pools per umem.

Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Björn Töpel <bjorn.topel@intel.com>
Link: https://lore.kernel.org/bpf/1598603189-32145-4-git-send-email-magnus.karlsson@intel.com

xsk: i40e: ice: ixgbe: mlx5: Pass buffer pool to driver instead of umem

Replace the explicit umem reference passed to the driver in AF_XDP
zero-copy mode with the buffer pool instead. This in preparation for
extending the functionality of the zero-copy mode so that umems can be
shared between queues on the same netdev and also between netdevs. In
this commit, only an umem reference has been added to the buffer pool
struct. But later commits will add other entities to it. These are
going to be entities that are different between different queue ids
and netdevs even though the umem is shared between them.

Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Björn Töpel <bjorn.topel@intel.com>
Link: https://lore.kernel.org/bpf/1598603189-32145-2-git-send-email-magnus.karlsson@intel.com

xsk: Replace the cheap_dma flag with a dma_need_sync flag

Invert the polarity and better name the flag so that the use case is
properly documented.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20200629130359.2690853-3-hch@lst.de

xsk: Explicitly inline functions and move definitions

In order to reduce the number of function calls, the struct
xsk_buff_pool definition is moved to xsk_buff_pool.h. The functions
xp_get_dma(), xp_dma_sync_for_cpu(), xp_dma_sync_for_device(),
xp_validate_desc() and various helper functions are explicitly
inlined.

Further, move xp_get_handle() and xp_release() to xsk.c, to allow for
the compiler to perform inlining.

rfc->v1: Make sure xp_validate_desc() is inlined for Tx perf. (Maxim)

Signed-off-by: Björn Töpel <bjorn.topel@intel.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200520192103.355233-15-bjorn.topel@gmail.com

xsk: Introduce AF_XDP buffer allocation API

In order to simplify AF_XDP zero-copy enablement for NIC driver
developers, a new AF_XDP buffer allocation API is added. The
implementation is based on a single core (single producer/consumer)
buffer pool for the AF_XDP UMEM.

A buffer is allocated using the xsk_buff_alloc() function, and
returned using xsk_buff_free(). If a buffer is disassociated with the
pool, e.g. when a buffer is passed to an AF_XDP socket, a buffer is
said to be released. Currently, the release function is only used by
the AF_XDP internals and not visible to the driver.

Drivers using this API should register the XDP memory model with the
new MEM_TYPE_XSK_BUFF_POOL type.

The API is defined in net/xdp_sock_drv.h.

The buffer type is struct xdp_buff, and follows the lifetime of
regular xdp_buffs, i.e. the lifetime of an xdp_buff is restricted to
a NAPI context. In other words, the API is not replacing xdp_frames.

In addition to introducing the API and implementations, the AF_XDP
core is migrated to use the new APIs.

rfc->v1: Fixed build errors/warnings for m68k and riscv. (kbuild test
robot)
Added headroom/chunk size getter. (Maxim/Björn)

v1->v2: Swapped SoBs. (Maxim)

v2->v3: Initialize struct xdp_buff member frame_sz. (Björn)
Add API to query the DMA address of a frame. (Maxim)
Do DMA sync for CPU till the end of the frame to handle
possible growth (frame_sz). (Maxim)

Signed-off-by: Björn Töpel <bjorn.topel@intel.com>
Signed-off-by: Maxim Mikityanskiy <maximmi@mellanox.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200520192103.355233-6-bjorn.topel@gmail.com