Rewrite vlib_buffer_chain_linearize function so that it works as intended.
Linearize buffer chains coming out of reassembly to work around some
dpdk-tx issues. Note that this is not a complete workaround
as a sufficiently large packet will still cause the resulting chain to
be too long.
Drop features from reassembly code which relies on knowing which and how
many buffers were freed during linearization, buffer counts and tracing
capabilities for these cases.
Change-Id: Ic65de53ecb5c78cd96b178033f6a576ab4060ed1
Signed-off-by: Klement Sekera <ksekera@cisco.com>
Symptom
-------
With NDR traffic blasting at VPP, bringing up a new VM with vhost
connection to VPP causes packet drops. I am able to recreate this
problem easily using a simple setup like this.
TREX-------------- switch ---- VPP
|---------------| |-------|
Cause
-----
The reason for the packet drops is due to vhost holding onto the worker
barrier lock for too long in vhost_user_socket_read(). There are quite a
few of system calls inside the routine. At the end of the routine, it
unconditionally calls vhost_user_update_iface_state() for all message
types. vhost_user_update_iface_state() also unconditionally calls
vhost_user_rx_thread_placement() and vhost_user_tx_thread_placement().
vhost_user_rx_thread_placement scraps out all existing cpu/queue mappings
for the interface and creates brand new cpu/queue mappings for the
interface. This process is very disruptive and very expensive. In my
opinion, this area of code needs a makeover.
Fixes
-----
* vhost_user_socket_read() is rewritten that it should not hold
onto the worker barrier lock for system calls, or at least minimize the
need for doing it.
* Remove the call to vhost_user_update_iface_state as a default route at
the end of vhost_user_socket_read(). There is only a couple of message
types which really need to call vhost_user_update_iface_state(). We put
the call to those message types which need it.
* Remove vhost_user_rx_thread_placement() and
vhost_user_tx_thread_placement from vhost_user_update_iface_state().
There is no need to repetatively change the cpu/queue mappings.
* vhost_user_rx_thread_placement() is actually quite expensive. It should
be called only once per queue for the interface. There is no need to
scrap the existing cpu/queue mappings and create new cpu/queue mappings
when the additional queues becomes active/enable.
* Change to create the cpu/queue mappings for the first RX when the
interface is created. Dont remove the cpu/queue mapping when the
interface is disconnected. Remove the cpu/queue mapping only when the
interface is deleted.
The create vhost user interface CLI also has some very expensive system
calls if the command is entered with the optional keyword "server"
As a bonus, This patch makes the create vhost user interface binary-api and
CLI thread safe. Do the protection for the small amount of code which is
thread unsafe.
Change-Id: I4a19cbf7e9cc37ea01286169882e5603e6d7eb77
Signed-off-by: Steven Luong <sluong@cisco.com>
(cherry picked from commit 67f935ec6eb9ec37b7d73029c5afa89cbf4a9aa2)
Some x86 CPUs have IOMMU capable dealing only with 39-bit address space
This patch also adds option to specify physmem base address from
startup.conf
Change-Id: I9e8abd26efb60e9c4ad54c035fb1751a4a61f4dc
Signed-off-by: Damjan Marion <damarion@cisco.com>
allow routes that are local and connected to be added via the API.
this emulates the addition of a second address in the same subnet
added to an interface.
Change-Id: Ib18a08c26956be9a07b3360664210c8cf6734c84
Signed-off-by: Neale Ranns <nranns@cisco.com>
(cherry picked from commit 7425f9289bdc17d330044b0df0fc31cdeeca626b)
since it can realloc when new ctx are added. If
not we can get some nasty memory corruption.
Change-Id: I617709c3013acbcb8aee07dc147894f0de896555
Signed-off-by: Neale Ranns <nranns@cisco.com>
(cherry picked from commit 58085f2f6757e464196b283cc8335ab5cf34a012)
During CSIT testing we discovered that LACP tests were failing and
producing coredumps. Reverting this patch fix the problem with VPP
crashing.
This reverts commit f23890138e02d4218c828c427f687f8ecdb0e165.
Change-Id: Icf97053ce1473350add885cbebe591f7f3efcbea
Signed-off-by: Peter Mikus <pmikus@cisco.com>
We register callback for VNET_HW_INTERFACE_LINK_UP_DOWN_FUNCTION and
VNET_SW_INTERFACE_ADMIN_UP_DOWN_FUNCTION to add and remove the slave
interface from the bond interface accordingly. For static bonding without
lacp, one would think that it is good enough to put the slave interface into
the ective slave set as soon as it is configured. Wrong, sometimes the slave
interface is configured to be part of the bonding without ever bringing up the
hardware carrier or setting the admin state to up. In that case, we send
traffic to the "dead" slave interface.
The fix is to make sure both the carrier and admin state are up before we put
the slave into the active set for forwarding traffic.
Change-Id: I93b1c36d5481ca76cc8b87e8ca1b375ca3bd453b
Signed-off-by: Steven <sluong@cisco.com>
(cherry picked from commit e43278f75fe3188551580c7d7991958805756e2f)
The 'example' keyword triggers the creation of very out-of-context
example in the "Examples" top level of documentation. Rename the
keyword in the comment so the comment is still readable, but
does not trigger the doxygen.
Change-Id: Iecbdc236918f9178a034817aa6cea7ab6b2c1654
Signed-off-by: Andrew Yourtchenko <ayourtch@gmail.com>
Allocate physmem on the same numa
set link speed
cap vmxnet3 support version to 3 if the device supports higher version
Change-Id: Id47dd83b0da53e623fd6cf3ad010e58158b1ab15
Signed-off-by: Steven <sluong@cisco.com>
Some dpdk driver does not support rte_eth_dev_rss_hash_conf_get. When that
happens, we display whatever that was in the stack variable for rss_hf which
is confusing.
The fix is to clear rss_hf prior to the call.
Before the fix
--------------
DBGvpp# sh hard
Name Idx Link Hardware
format_dpdk_device:599: rte_eth_dev_rss_hash_conf_get returned -95
GigabitEthernet13/0/0 2 up GigabitEthernet13/0/0
Link speed: 10 Gbps
Ethernet address 00:0c:29:c4:8c:eb
VMware VMXNET3
...
rss avail: ipv4 ipv4-tcp ipv6 ipv6-tcp
rss active: ipv4 ipv4-frag ipv4-tcp ipv4-udp ipv4-other
After the fix
-------------
DBGvpp# sh hard
Name Idx Link Hardware
format_dpdk_device:600: rte_eth_dev_rss_hash_conf_get returned -95
GigabitEthernet13/0/0 2 up GigabitEthernet13/0/0
Link speed: 10 Gbps
Ethernet address 00:0c:29:c4:8c:eb
VMware VMXNET3
..
rss avail: ipv4 ipv4-tcp ipv6 ipv6-tcp
rss active: none
Change-Id: If8f8327f0012eecc8d23cd7f3f9cc581ca025654
Signed-off-by: Steven <sluong@cisco.com>
Allow to display longer interface names, e.g. VirtualEthernet0/0/0.102
The field length (32) is now the same as for 'show interface'.
Change-Id: I1cb1efd459acb800bfaeeec40b672c8b17cd8c3d
Signed-off-by: Igor Mikhailov (imichail) <imichail@cisco.com>
Given n equals to the maximum number of bytes to copy from src in the API,
or the rough estimate strlen of src, strncpy_s_inline should not copy more
than the number of bytes, computed by strlen(src), to dst if n is greater than
strlen(src). The number of bytes to copy is computed by strnlen(src,n), not n.
Change-Id: I088b46125d9776962750e121f1fbf441952efc2b
Signed-off-by: Steven <sluong@cisco.com>
Remove the needless tests and checks which coverity complains about in
string_test.c
Change-Id: I971650cada77136f06528a65625ef99bd3d7e915
Signed-off-by: Steven <sluong@cisco.com>
Implement API/CLI to clear IP-MAC tables used for ARP-termination
in the specified bridge domain.
The CLI to flush MAC IP tables for a BD is:
set bridge-domain arp entry <bd-id> del-all
The API added is bd_ip_mac_flush.
Change-Id: I34ceb87c0f480c7102f6559312c24081ed485af8
Signed-off-by: John Lo <loj@cisco.com>
PR in openssl community is almost done, and need to change some code in VPP
to align with the openssl interface.
Change-Id: Ic7da53e507b67b53958760d07738dd774b1c526d
Signed-off-by: Ping Yu <ping.yu@intel.com>
It's not OK to crash due to a transient buffer allocation failure.
Return 1 if the requested operation failed, otherwise 0.
Buffer index parameter change to a value-result, so the caller can
differentiate between partial and complete allocation failure: callers
which request an initial allocation (inbound bi = ~0) need to check
the (out) value to decide whether or not to call vlib_buffer_free(...).
Change-Id: I03029d7f2714c17dca4630dfd95a1eb578b68384
Signed-off-by: Dave Barach <dave@barachs.net>
Set TCP checksum to zero, before calculate the TCP checksum.
Change-Id: Id96743334481804b8ebb74afef2f1dbfed29e0cb
Signed-off-by: Andrej Kozemcak <andrej.kozemcak@pantheon.tech>
Moves LDP logic that allows sharing of sessions between multi-process
app workers into a separate VCL shim layer. Also refactors LDP to use
the new layer.
Change-Id: I8198b51eae7d099a8c486e36b29e3a0cb8cee8e9
Signed-off-by: Florin Coras <fcoras@cisco.com>
