The patch updates the env_abstraction_layer.rst part in prog_guide.
It adds the RX interrupt event declaration and revises the others in
interrupt event section.
Signed-off-by: Cunming Liang <cunming.liang@intel.com>
Acked-by: Danny Zhou <danny.zhou@intel.com>
When using VM power manager app on systems with more than 64 cores,
app could not run even though user does not use cores 64 or higher.
The problem happens only in that case, in which case it will result
in an undefined behaviour.
Thefere, this patch allows the user to run the app on a system with more
than 64 cores, warning the user not to use cores higher than 64 in the VM(s).
Add new known issue where VM power manager app may not work
in a system with more than 64 cores, in release notes.
Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Tested-by: Marvin Liu <yong.liu@intel.com>
Acked-by: Sergio Gonzalez Monroy <sergio.gonzalez.monroy@intel.com>
L3fwd was trying to use an inexistent function "simple_ipv6_fwd_4pkts",
instead it should be "simple_ipv6_fwd_8pkts", and "simple_ipv8_fwd_4pkts"
instead of "simple_ipv4_fwd_8pkts".
clang reports some unused functions, used only for LPM lookup:
examples/l3fwd/main.c:545:1: error: unused function 'send_packetsx4'
examples/l3fwd/main.c:1165:1: error: unused function 'rfc1812_process'
Fixes: 80fcb4d4 ("examples/l3fwd: increase lookup burst size to 8")
Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Sergio Gonzalez Monroy <sergio.gonzalez.monroy@intel.com>
[Thomas: more #if to fix clang warnings]
Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
Test failed on verification if number of bytes
transmitted on each slave is not less than 90%
and greater than 110% of mean value of bytes transmitted
thru one slave. This was verified on a real system
but is difficult to achieve using virtualpmd.
That's why for unit tests only, it is sufficient to verify that with
high load (2 seconds transmission) all slaves are transmitting
so the traffic is balanced.
Fixes: 0c8396e6d7 ("bond: unit tests for mode 5")
Signed-off-by: Michal Jastrzebski <michalx.k.jastrzebski@intel.com>
If the name parameter to rte_eth_bond_create() was NULL,
there was a segmentation fault because eth_dev was also NULL.
Add error handling of mac_addrs memory allocation.
Add call to rte_eth_dev_release_port() in error handling.
Signed-off-by: Bernard Iremonger <bernard.iremonger@intel.com>
Tested-by: Marvin Liu <yong.liu@intel.com>
Acked-by: Michal Jastrzebski <michalx.k.jastrzebski@intel.com>
On slave activation in LACP (8023AD) SOCKET_ANY_ID (which is -1)
is being casted to unsigned char and then to signed int.
The result is that socket_id has value of 255, not -1.
This results to memory allocation failure.
Fixes: 46fb436836 ("bond: add mode 4")
Signed-off-by: Sergey Balabanov <balabanovsv@ecotelecom.ru>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
The i40e_xmit_pkts() is called, which often means HW offload is used here,
so we had better remove 'unlikely' check for checksum offload.
Signed-off-by: Jijiang Liu <jijiang.liu@intel.com>
Acked-by: Marvin Liu <yong.liu@intel.com>
To obtain detachable flag, pci_drv is accessed in rte_eth_dev_is_detachable().
But pci_drv is only valid if port is enabled. Not to cause illegal access,
add rte_eth_dev_is_valid_port() before accessing.
Signed-off-by: Tetsuya Mukawa <mukawa@igel.co.jp>
Acked-by: Bernard Iremonger <bernard.iremonger@intel.com>
There is no sync between host and guest to allow removal of memzones,
and freeing them result in undefined behavior.
In the guest, we identify IVSHMEM memsegs/memzones by having
ioremap_addr != 0. In the host, nothing is done to the memzone, meaning
ioremap_addr == 0.
As a solution, mark memzones being added to IVSHMEM in the host, by
setting ioremap_addr, then return an error whenever we try to free an
IVSHMEM memzone.
Fixes: ff909fe21f ("mem: introduce memzone freeing")
Signed-off-by: Sergio Gonzalez Monroy <sergio.gonzalez.monroy@intel.com>
After the changes introduced by Dynamic Memzones, all the memsegs were
added to the malloc heap during init.
Those changes did not account for IVSHMEM memsegs which should not be
added to the malloc heap as part of available memory.
Fixes: fafcc11985 ("mem: rework memzone to be allocated by malloc")
Signed-off-by: Sergio Gonzalez Monroy <sergio.gonzalez.monroy@intel.com>
Byte ordering macros were used without including the needed header.
Fixes: ce10b21bf6 ("eal/ppc: fix cpu cycle count for little endian")
Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
Acked-by: Chao Zhu <chaozhu@linux.vnet.ibm.com>
GCC 4.8 raises this error:
lib/librte_eal/bsdapp/eal/eal_pci.c:453:15: error: cast discards
'__attribute__((const))' qualifier from pointer target type
.pi_data = *(u_int32_t *)buf,
^
Note: this assignment seems useless because pi_data is filled
with memset later.
Fixes: 632b2d1dee ("eal: provide functions to access PCI config")
Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
'--no-huge' option now can workable with -m option.
Unit test for eal flag should change pass criterion.
Fixes: a7de7e6beb ("eal: allow combining -m and --no-huge")
Signed-off-by: Marvin Liu <yong.liu@intel.com>
In previous setting, mempool size and cache_size were both 32.
It does not satisfy with cache_size checking rule by now.
Cache size should be less than CONFIG_RTE_MEMPOOL_CACHE_MAX_SIZE and
mempool size / 1.5.
Fixes: 462321b44a ("mempool: limit cache size")
Signed-off-by: Marvin Liu <yong.liu@intel.com>
Acked-by: Jingjing Wu <jingjing.wu@intel.com>
[Thomas: remove unused PKT_BURST_SZ]
Niantic HW expects Header Buffer Address in the RXD to be word aligned.
So, if mbuf's buf_physaddr is not word aligned then
RX path will not work properly.
Right now, in ixgbe PMD we always setup Packet Buffer Address(PBA) and
Header Buffer Address (HBA) to the same value:
buf_physaddr + RTE_PKTMBUF_HEADROOM.
As ixgbe PMD doesn't support split header feature anyway,
the issue can be fixed just by always setting HBA in the RXD to zero.
Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
This patch adds management of PKT_RX_FDIR and PKT_RX_RSS_HASH ol_flags in
vPMD for unified packet type as well as for 16 bit field packet_type when
RTE_NEXT_ABI is not defined.
Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
The function documentation was obviously copied and not updated.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
This option permit to build librte_kni.so without building rte_kni.ko
so you can build a sdk without building kernel drivers.
Signed-off-by: Nikita Kozlov <nikita@elyzion.net>
The patch sets zero as the default value of pci device numa_node
if the socket could not be determined.
It provides the same default value as FreeBSD which has no NUMA support,
and makes the return value of rte_eth_dev_socket_id() be consistent
with the API description.
Signed-off-by: Cunming Liang <cunming.liang@intel.com>
Do some cleanup of pci scan loop.
* check errors first
* don't initialize variables where not necessary
* cuddle else (follow existing style)
* chop off conditional after return
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Build log:
lib/librte_eal/bsdapp/eal/eal_pci.c:462:9: error:
incompatible integer to pointer conversion passing 'u_int32_t'
(aka 'unsigned int') to parameter of type 'void *'
It is fixed by passing the pointer of pi.pi_data to memcpy.
By the way, it seems strange that pi_data is initialized twice:
.pi_data = *(u_int32_t *)buf
memcpy(&pi.pi_data, buf, len);
Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
Example of errors:
error: cast to pointer from integer of different size [-Werror=int-to-pointer-cast]
error: format ‘%lx’ expects argument of type ‘long unsigned int’, but argument 5 has type ‘uint64_t
error: format ‘%lu’ expects argument of type ‘long unsigned int’, but argument 6 has type ‘unsigned int’
Only 2 files are fixed. The others errors are left as exercise to the authors.
Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
Build log:
error: unused variable ‘cid’
error: ‘RTE_LOG_WARN’ undeclared
error: expected ‘)’ before ‘sc’
There were unused variables defined for debug but not used in debug log because
it was ifdef'ed a the wrong condition (RTE_LIBRTE_BNX2X_DEBUG_DRIVER).
The warning were using WARN instead of WARNING.
Some debug messages had some extra parameters.
Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
When a Tx queue fails to start in fm10k_dev_start, all Rx queues
and Tx queues that are started should be cleaned before the
function returns an error.
Signed-off-by: Wang Xiao W <xiao.w.wang@intel.com>
Acked-by: Jing Chen <jing.d.chen@intel.com>
In Rx and Tx queue_disable functions, the index of queue should
be qnum other than i which is the iteration of time expiration.
Signed-off-by: Wang Xiao W <xiao.w.wang@intel.com>
Acked-by: Jing Chen <jing.d.chen@intel.com>
Fixes issue where ieee15888 timestamping doesn't work for the i40e
pmd when RTE_ABI_NEXT is enabled.
Also refactors repeated ieee15888 flag checking and setting
code into a function.
Test report: http://dpdk.org/ml/archives/dev/2015-August/022496.html
Reported-by: Huilong Xu <huilongx.xu@intel.com>
Signed-off-by: John McNamara <john.mcnamara@intel.com>
Acked-by: Sergio Gonzalez Monroy <sergio.gonzalez.monroy@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
Tested-by: Huilong Xu <huilongx.xu@intel.com>
Ieee1588 reads system time to set its timestamp. On 1G NICs, for example,
i350, system time is disabled by default. It means the ieee1588 timestamp
will always be 0.
This patch enables system time when ieee1588 is enabled.
Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
Acked-by: John McNamara <john.mcnamara@intel.com>
When adding a "depth small" entry, if its extended flag is not set and
its depth is smaller than the one in the tbl24, nothing should be done
otherwise will operate on the wrong memory area.
Signed-off-by: Zhe Tao <zhe.tao@intel.com>
Acked-by: Cunming Liang <cunming.liang@intel.com>
Eliminate problematic race condition in rte_timer_manage() that can
lead to corruption of per-lcore pending-lists (implemented as
skip-lists). The race condition occurs when rte_timer_manage() expires
multiple timers on lcore A, while lcore B simultaneously invokes
rte_timer_reset() for one of the expiring timers (other than the first
one).
Lcore A splits its pending-list, creating a local list of expired timers
linked through their sl_next[0] pointers, and sets the first expired
timer to the RUNNING state, all during one list-lock round trip.
Lcore A then unlocks the list-lock to run the first callback, and that
is when A and B can have different interpretations of the subsequent
expired timers' true state. Lcore B sees an expired timer still in the
PENDING state, atomically changes the timer to the CONFIG state, locks
lcore A's list-lock, and reinserts the timer into A's pending-list.
The two lcores try to use the same next-pointers to maintain both lists!
Our solution is to remove expired timers from the pending-list and try
to set them all to the RUNNING state in one atomic step, i.e.,
rte_timer_manage() should perform these two actions within one
ownership of the list-lock.
After splitting the pending-list at the current point in time and trying
to set all expired timers to the RUNNING state, we must put back into
the pending-list any timers that we failed to set to the RUNNING state,
all while still holding the list-lock. It is then safe to release the
lock and run the callback functions for all expired timers that remain
on our local run-list.
Signed-off-by: Robert Sanford <rsanford@akamai.com>
Add new timer-manage race-condition test: We wrote a test to confirm
our suspicion that we could crash rte_timer_manage() under the right
circumstances. We repeatedly set several timers to expire at roughly
the same time on the master core. The master lcore just delays and runs
rte_timer_manage() about ten times per second. The slave lcores all
watch the first timer (timer-0) to see when rte_timer_manage() is
running on the master, i.e., timer-0's state is not PENDING.
At this point, each slave attempts to reset a subset of the timers to
a later expiration time. The goal here is to have the slaves moving
most of the timers to a different place in the master's pending-list,
while the master is traversing the same next-pointers (the slaves'
sl_next[0] pointers) and running callback functions. This eventually
results in the master traversing a corrupted linked-list.
In our observations, it results in an infinite loop.
Signed-off-by: Robert Sanford <rsanford@akamai.com>
Fix app/test timer stress test 2: Sometimes this test fails and
seg-faults because the slave lcores get out of phase with the master.
The master uses a single int, 'ready', to synchronize multiple slave
lcores through multiple phases of the test.
To resolve, we construct simple synchronization primitives that use one
atomic-int state variable per slave. The master tells the slaves when to
start, and then waits for all of them to finish. Each slave waits for
the master to tell it to start, and then tells the master when it has
finished.
Signed-off-by: Robert Sanford <rsanford@akamai.com>
On IBM POWER8 PPC64 little endian architecture, the definition of tsc
union will be different. This patch fix this to enable the right output
from rte_rdtsc().
Signed-off-by: Chao Zhu <chaozhu@linux.vnet.ibm.com>
Fix for ABI breakage introduced in LRO addition. Moves
lro bitfield to the end of the struct/member.
Fixes: 8eecb3295a (ixgbe: add LRO support)
Signed-off-by: John McNamara <john.mcnamara@intel.com>
Acked-by: Vlad Zolotarov <vladz@cloudius-systems.com>
Enable vector ixgbe and i40e bulk alloc for bsd as it is
already done for linux.
Fixes: 304caba126 ("config: fix bsd options")
Fixes: 0ff3324da2 ("ixgbe: rework vector pmd following mbuf changes")
Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
The RX_OLFLAGS option was renamed from DISABLE to ENABLE in driver code
and linux config.
It is now renamed also in bsd config and documentation.
Fixes: 359f106a69 ("ixgbe: prefer enabling olflags rather than not disabling")
Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
Header buffer address for header split will be filled with the
physical address for DMA, which is actually not needed at all,
as header split hasn't been supported. Hardware requires the
least bit of header address which is 'Descriptor Done' bit when
write back should be set to 0 by driver.
The issue is that if the user wants to reserve an odd number of
bytes between the mbuf header and data buffer, the physical address
to be filled in the descriptor would happen to be odd. That means
the DD bit would be set to non-zero by driver. That will result in
reporting descriptor done wrongly.
Signed-off-by: Helin Zhang <helin.zhang@intel.com>
Header buffer address for header split will be filled with the physical
address for DMA, which is actually not needed at all, as header split
hasn't been supported. Hardware requires the least bit of header address
which is 'Descriptor Done' bit when write back should be set to 0 by driver.
The issue is that if the user wants to reserve an odd number of bytes between
the mbuf header and data buffer, the physical address to be filled in the
descriptor would happen to be odd. That means the DD bit would be set to
non-zero by driver. That will result in reporting descriptor done wrongly.
Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
It looks better to have a data buffer address that is aligned to
8 bytes. This is the case when there is no mbuf private area, but
if there is one, the alignment depends on the size of this area
that is located between the mbuf structure and the data buffer.
Indeed, some drivers expects to have the buffer address aligned
to an even address, and moreover an unaligned buffer may impact
the performance when accessing to network headers.
Add a check in rte_pktmbuf_pool_create() to verify the alignment
constraint before creating the mempool. For applications that use
the alternative way (direct call to rte_mempool_create), also
add an assertion in rte_pktmbuf_init().
By the way, also add the MBUF log type.
Signed-off-by: Olivier Matz <olivier.matz@6wind.com>
Acked-by: Helin Zhang <helin.zhang@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
librte_pmd_mlx4.so needs to be linked with libibverbs otherwise, the PMD is
not able to open Mellanox devices and the following message is printed by
testpmd at startup
"librte_pmd_mlx4: cannot access device, is mlx4_ib loaded?".
Applications dependency on libibverbs are moved to be only valid in static
mode, in shared mode, applications do not depend on it anymore,
librte_pmd_mlx4.so keeps this dependency and thus is linked with libibverbs.
MLX4 cannot be supported in combined shared library because there is no clean
way of adding -libverbs to the combined library.
Signed-off-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
Acked-by: Olivier Matz <olivier.matz@6wind.com>
Some .so libraries needs to be linked with external libraries. For that the
LDLIBS and EXTRA_LDFLAGS variables should be present on the link line when
those .so files are created. PMD Makefile is responsible for filling the
LDLIBS variable with the link to the external library it needs.
Signed-off-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
Acked-by: Olivier Matz <olivier.matz@6wind.com>
Compiling for dpdk x86_x32 gives the following error:
In file included from /usr/include/sys/sysctl.h:63:0,
from lib/librte_eal/common/eal_common_timer.c:39:
/usr/include/bits/sysctl.h:19:3: error: #error "sysctl system call is unsupported in x32 kernel"
# error "sysctl system call is unsupported in x32 kernel"
^
Including sysctl.h was added by mistake when merging bsd and linux EAL
timer code. It can be safely removed in this file, fixing the
compilation.
Fixes: 040cf8a411 ("eal: deduplicate timer functions")
Signed-off-by: Olivier Matz <olivier.matz@6wind.com>
Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
For consistency, RTE_LOG macro should be used instead of rte_log function.
The macro can be pruned at build time, though these logs have a high level
and should not pruned.
Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
Some drivers was not following DPDK convention and
was leaving logging always in even if LOG_LEVEL was configured
to disable debug logs.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
[Thomas: apply same fix to i40e, fm10k and bnx2x]
Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
Customers often screen off info level messages, so raise log
level of significant events.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
All the debug chatter messages in the system log causes
complaints from users. Change the INFO messages to DEBUG
for normal startup kind of stuff.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
