Age | Commit message (Collapse) | Author | Files | Lines |
|
Introduce async model into memif by utilizing new DMA API. Original
process is broken down to submission stage and completion stage. As
multiple submissions may in flight simultaneously, per thread data is
no longer safe, now replace thread data into each dma data structure.
As slave side already support zero copy mode, DMA option is only added
in master side.
Type: feature
Signed-off-by: Marvin Liu <yong.liu@intel.com>
Change-Id: I084f253866f5127cdc73b9a08c8ce73b091488f3
|
|
This patch adds an API memif_socket_filename_add_del_v2
that allows autogenerating memif socket_id when passing
~0 in the socket_id field.
It opportunistically walks the hash to find a free ID
to use, and returns it in the reply.
socket_filename also becomes a variable length string,
to accomodate for longer names (in case a netns gets
passed)
Type: feature
Change-Id: I33fc3e1cf553af27579d6bad8691b22b530531cc
Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
|
|
Type: improvement
Change-Id: Ida2d044bccf0bc8914b4fe7d383f827400fa6a52
Signed-off-by: Damjan Marion <dmarion@me.com>
|
|
Type: improvement
Change-Id: I12b120d988347cced3df82810e86dc2fd5cfca80
Signed-off-by: Damjan Marion <dmarion@me.com>
|
|
Use of _vec_len() to set vector length breaks address sanitizer.
Users should use vec_set_len(), vec_inc_len(), vec_dec_len () instead.
Type: improvement
Change-Id: I441ae948771eb21c23a61f3ff9163bdad74a2cb8
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
Type: fix
Calling vnet_hw_if_register_tx_queue should
be done with the worker barrier held, as
virtio-pre-input might be grabbing a queue
while a memif connect event is triggered.
Change-Id: Ie1272cdfd2477faf7a4e10f30778279872f04916
Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
|
|
This is required after distinguishing between max_frame_size and MTU
Type: fix
Signed-off-by: Artem Glazychev <artem.glazychev@xored.com>
Change-Id: Ie642bee4e30ca76903bb8be5eeb6914c2c09bf35
|
|
When deleting a memif interface, there is a memory leak in
mif->local_disc_string. The code that leads to memory leak is
memif_delete_if -> memif_disconect -> vec_dup
Type: fix
Signed-off-by: Steven Luong <sluong@cisco.com>
Change-Id: I579c743c3d47770131153e083e65f6cd899067b3
|
|
Prep for supporting multiple callbacks, optional args, etc.
Type: improvement
Change-Id: I96244c098712e8213374678623f12527b0e7f387
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
Make it shorter to type, easier to debug, make adding callbacks in
future simpler.
Type: improvement
Change-Id: I6cdd6375e36da23bd452a7c7273ff42789e94433
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
Type: improvement
Change-Id: I06130447b4e6f4726c4f5bffbe606385c45b8bd4
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
Type: feature
Change-Id: I231f782b3c56dc2b10321e4569ac7acdad1c11da
Signed-off-by: Mohsin Kazmi <sykazmi@cisco.com>
|
|
Type: fix
Fixes: 2bae16b238bd ("memif: fix the default txq placement")
Signed-off-by: Mohsin Kazmi <sykazmi@cisco.com>
Change-Id: I8fbb2aa92dc31d84a5cd0f7b9a3c7f39dfb064ee
|
|
Type: fix
Fixes: 3effb4e63068 ("memif: integrate with new tx infra")
"memif: integrate with new tx infra" patch integrated memif
with new tx infra. There might be scenarios when txqs were
less than vpp threads, in which case, txqs should be shared
among threads. This patch fixes it.
Signed-off-by: Mohsin Kazmi <sykazmi@cisco.com>
Change-Id: I1c64a1370f5024240ab56311f75665db31714b60
|
|
Admin down implies Link down but nothing came with admin up.
Ticket: VPP-1959
Type: fix
Change-Id: I43725329ae7918659c73d703280c25de5f0b1d14
Signed-off-by: Daniel Béreš <daniel.beres@pantheon.tech>
|
|
Type: fix
Fixes: 3effb4e63068 ("memif: integrate with new tx infra")
memif is recently integrated with new tx infra. But it
introduces a crash when slave disconnect from master but
interface is not deleted. Disconnect routine was missing
unregister of all tx queues. This patch fixes it.
Signed-off-by: Mohsin Kazmi <sykazmi@cisco.com>
Change-Id: I81c59cc1a03561248ec8595d5e3caa54f421833e
|
|
Type: improvement
Signed-off-by: Mohsin Kazmi <sykazmi@cisco.com>
Change-Id: I7c2b9891e269f23c3aa2a0abfee3cf0a0f1e2135
|
|
Type: fix
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I3a33230de13fef613dc9523cf24a9968d200c2e5
|
|
Type: fix
Change-Id: I3cc09630cae999cc194a0eedb4d2cfee7a826b76
Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
|
|
For portabiliy reasons it is better to have all wrapped in clib code.
I.e. instead of using getcpu() we have clib_get_current_numa_node () and
clib_get_current_cpu_id().
Type: refactor
Change-Id: I29b52d7f29bc7f93873402c4070561f564b71c63
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
Type: improvement
Change-Id: If3da7d4338470912f37ff1794620418d928fb77f
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
Type: improvement
This patch adds flags to represent the modern NICs capabilities.
Change-Id: I96d38d9ab7eac55974d72795cd100d8337168e1e
Signed-off-by: Mohsin Kazmi <sykazmi@cisco.com>
|
|
Change-Id: Ifa8bccd8a34ec1b14e772ee53757e9083373e3de
Signed-off-by: Mohammed Hawari <mohammed@hawari.fr>
Type: feature
|
|
Type: improvement
Previously, memif would check every 3 seconds to see if an interface
had gone admin up. Now it does as soon as admin up occurs.
Signed-off-by: Ed Warnicke <hagbard@gmail.com>
Change-Id: I808699cd5980fa162c34ac977f43d00ed6b67115
Signed-off-by: Ed Warnicke <hagbard@gmail.com>
|
|
Type: refactor
Change-Id: Ie67dc579e88132ddb1ee4a34cb69f96920101772
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
Type: style
Change-Id: I0d25d922312dfd60a665916cf47c3d2faa29ea49
Signed-off-by: Paul Vinciguerra <pvinci@vinciconsulting.com>
|
|
Getting an unexpected EPOLLOUT for RX queue and crashing due to missing
callback function on write_function to handle EPOLLOUT.
(gdb) f 5
f 5
node=<optimized out>, vm=0x7f948bbe4540 <vlib_global_main>) at /vpp/src/vlib/unix/input.c:325
325 errors[n_errors] = f->write_function (f);
(gdb) p *e
p *e
$21 = {events = 4, data = {ptr = 0x23, fd = 35, u32 = 35, u64 = 35}}
(gdb) p file_main.file_pool[e->data.fd]
p file_main.file_pool[e->data.fd]
$22 = {file_descriptor = 37, flags = 1, polling_thread_index = 0, private_data = 65536,
read_function = 0x7f944a87b140 <memif_int_fd_read_ready>, write_function = 0x0,
error_function = 0x0, description = 0x7f944be22400 "memif0/1 rx 0 int", read_events = 0,
write_events = 1, error_events = 0}
(gdb)
File descriptors are recycled. It is likely that the EPOLLOUT is meant
for the previous owner of the same file descriptor number which memif
recycled and memif RX did not expect to receive an EPOLLOUT. For
defensive play, we create a dummy callback function to ignore the
EPOLLOUT instead of crashing.
Type: fix
Signed-off-by: Steven Luong <sluong@cisco.com>
Change-Id: I535cadbb9b71d3db2995b118c59a12c71f10af09
|
|
This is part of bigger refactor.
Type: refactor
Change-Id: I6fc2c0a1e2d217a70952901bcf775b8485bd3c20
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
Type: improvement
Change-Id: I91031d56aae884a0a42fca343a2fc24c442b1da8
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
code added to free the zero copy interface rx/tx queue buffers during disconnecting.
As ddc9eb4 find the last official solution introduced core in ut. This does not.
Type: fix
Signed-off-by: Changqian Wang <changqwa@cisco.com>
Change-Id: I971ee164e6d4331a85feb9e65d6702d771c86985
|
|
This reverts commit d570c4d18df404cb59ef72d20d56f8e31418c8e9.
That commit introduced crash into memif make test-debug.
To reproduce: "TEST=memif make test-debug".
With this commit reverted the make test-debug for memif passes.
Type: fix
Change-Id: Ic335359e2050bb8a5d530b01c595b081476fa197
Signed-off-by: Andrew Yourtchenko <ayourtch@gmail.com>
|
|
- code added to release the mbuf's to main pool during memif disconnect.
Type: fix
Change-Id: I504dacc9a40bc1702831cf3208fd532d1ae68a6f
Signed-off-by: Padma Ram <padmaram86.raja@gmail.com>
|
|
- Make plugin descriptions more consistent
so the output of "show plugin" can be
used in the wiki.
Change-Id: I4c6feb11e7dcc5a4cf0848eed37f1d3b035c7dda
Signed-off-by: Dave Wallace <dwallacelf@gmail.com>
|
|
This patch introduces following changes:
- deprecated free lists which are not used and not compatible
with external buffer managers (i.e. DPDK)
- introduces native support for per-numa buffer pools
- significantly improves performance of buffer alloc and free
Change-Id: I4a8e723ae47056717afd6cac0efe87cb731b5be7
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
Change-Id: I3bb1d9f83dd08f4b93acd4a281bfec0674e39c2e
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
Storing buffer in local template seems to be better option....
Change-Id: I1a2fdd68cb956f99a5b36d2cd810fc623e089bcf
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
Change-Id: Ic4c46bc733afae8bf0d8146623ed15633928de30
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
Change-Id: Ied34720ca5a6e6e717eea4e86003e854031b6eab
Signed-off-by: Dave Barach <dave@barachs.net>
|
|
Change-Id: I2b14e2a035a37b04816aac1e610146be58ad1e92
Signed-off-by: Jakub Grajciar <jgrajcia@cisco.com>
|
|
Check access rights using effective user/group IDs
Change-Id: I3683258c24bcd7817024bffbd56b54b2f596fdd7
Signed-off-by: Jakub Grajciar <jgrajcia@cisco.com>
|
|
Change-Id: I67f58aa93c721dbd518cc2ddcea330006693662f
Signed-off-by: Juraj Sloboda <jsloboda@cisco.com>
|
|
Change-Id: Ie315d4e83c0575d82aee0369ef50e8dd2dad6a2a
Signed-off-by: Dave Barach <dave@barachs.net>
|
|
Change-Id: I085615fde1f966490f30ed5d32017b8b088cfd59
Signed-off-by: Paul Vinciguerra <pvinci@vinciconsulting.com>
|
|
create runtime dir when adding default socket filename.
Change-Id: Iad7e751fcab7ce773e0907302414a06ea44dd6cd
Signed-off-by: Jakub Grajciar <jgrajcia@cisco.com>
|
|
in case of relative path, create subdirs inside runtime dir.
in case of absolute path, if folder does not exist return error.
Change-Id: I01fe018102c69deb105160cfa18b741541d76c82
Signed-off-by: Jakub Grajciar <jgrajcia@cisco.com>
|
|
Change-Id: I0fe60a639c7589dc842d85db092c81c1a7441cb7
Signed-off-by: Jakub Grajciar <jgrajcia@cisco.com>
|
|
Change-Id: I65306fb1f8e39221dd1d8c00737a7fb1c0129ba8
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
In version 1 of the protocol sender was always ring producer and
receiver was consumer. In version 2 slave is always producer,
and in case of master-to-slave rings, slave is responsible for
populating ring with empty buffers.
As this is major change, we need to bump version number.
In addition, descriptor size is reduced to 16 bytes.
This change allows zero-copy-slave operation (to be privided in the separate
patch).
Change-Id: I02115d232f455ffc05c0bd247f7d03f47252cfaf
Signed-off-by: Damjan Marion <damarion@cisco.com>
Signed-off-by: Jakub Grajciar <jgrajcia@cisco.com>
|
|
interface reconnect.
Change-Id: Ifc7eb2494a22c334d8899422545fca1a4bba4d05
Signed-off-by: Chun Li <chunl2@cisco.com>
|
|
This patch teaches worer threads to sleep and to be waken up by
kernel if there is activity on file desctiptors assigned to that thread.
It also adds counters to epoll file descriptors and new
debug cli 'show unix file'.
Change-Id: Iaf67869f4aa88ff5b0a08982e1c08474013107c4
Signed-off-by: Damjan Marion <damarion@cisco.com>
|