Piotr Azarewicz [Mon, 23 May 2016 12:19:56 +0000 (14:19 +0200)]
examples/quota_watermark: fix memory overflow
qw app at its init stage reserve 2*sizeof(int) memory space for quota
and low_watermark shared variables, but both apps (qw and qwctl) assign
wrong address for low_watermark pointer (out of reserved memzone space)
due to wrong pointer arithmetic.
CID 30709 : Extra sizeof expression (SIZEOF_MISMATCH)
suspicious_pointer_arithmetic: Adding 4UL /* sizeof (int) */ to pointer
(unsigned int *)(*qw_memzone).addr of type unsigned int * is suspicious
because adding an integral value to this pointer automatically scales
that value by the size, 4 bytes, of the pointed-to type, unsigned int.
Most likely, sizeof (int) is extraneous and should be replaced with 1.
Tomasz Kulasek [Fri, 29 Apr 2016 16:30:24 +0000 (18:30 +0200)]
examples/performance-thread: fix size of destination port ids
After extending IPv4 next hop in lpm library, size of dst_port array was
changed from 16 to 32 bits in l3fwd-thread example, without modification
of the rest of path written for 16 bit value.
This patch uses similar approach for fix, like in commit 8353a36a9b4b
("examples/l3fwd: fix size of destination port ids"), restoring 16 bit size
for destination port ids and doing necessary conversion from 32 to 16 bit
after lpm_lookupx4.
Fixes: dc81ebbacaeb ("lpm: extend IPv4 next hop field") Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com>
Jerin Jacob [Tue, 19 Apr 2016 09:35:28 +0000 (15:05 +0530)]
examples/l2fwd: increase mempool cache size for performance
l3fwd sets the mempool cache size to 256, selected the same value for l2fwd
Signed-off-by: Jerin Jacob <jerin.jacob@caviumnetworks.com> Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com> Acked-by: Reshma Pattan <reshma.pattan@intel.com>
Jerin Jacob [Tue, 19 Apr 2016 09:35:27 +0000 (15:05 +0530)]
examples/l2fwd: remove hardcoded cycles per second
Signed-off-by: Jerin Jacob <jerin.jacob@caviumnetworks.com> Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com> Acked-by: Reshma Pattan <reshma.pattan@intel.com>
Jasvinder Singh [Wed, 1 Jun 2016 14:01:02 +0000 (15:01 +0100)]
examples/ip_pipeline: add sample configs for various layers
The sample configuration file demonstrates that network layer components such
as TCP, UDP, ICMP etc, can be easily integrated into ip pipeline infrastructure.
Similarily, various other functionalities such as IP Reassembly for input
traffic with local destination and IP Fragmentation to enforce the MTU for
the routed output traffic, can be added using SWQs enabled with
reassembly and fragmentation features.
Jasvinder Singh [Wed, 1 Jun 2016 14:01:01 +0000 (15:01 +0100)]
examples/ip_pipeline: update routes when ports state change
The routing pipeline registers a callback function with the nic ports and
this function is invoked for updating the routing entries (corresponding to
local host and directly attached network) tables whenever the nic ports
change their states (up/down).
Jasvinder Singh [Wed, 1 Jun 2016 14:01:00 +0000 (15:01 +0100)]
examples/ip_pipeline: assign MAC address to routing ports
As a result of tracking, output ports of routing pipelines are linked with
physical nic ports (potentially through other pipeline instances).
Thus, the mac addresses of the NIC ports are assigned to routing pipeline
out ports which are connected to them and are further used in routing table
entries instead of hardcoded default values.
Jasvinder Singh [Wed, 1 Jun 2016 14:00:59 +0000 (15:00 +0100)]
examples/ip_pipeline: link routing output ports to devices
This commit implements tracking mechanism for linking routing pipeline
output ports to physical NIC ports.
Once all the pipelines of the application are initialised, mechanism is
invoked during post initialisation phase for relating routing pipeline
output with NIC ports by navigating through the intermediate pipelines,
if present.
The tracking functions of the pipelines which help in navigating through
the intermediate pipelines are moved from pipeline_<pipeline_name>_be.c
to pipeline_<pipeline_name>.c. All pipelines except passthrough pipelines
use default tracking function (pipeline/pipeline_common_fe.c).
Jasvinder Singh [Mon, 30 May 2016 17:14:49 +0000 (18:14 +0100)]
examples/ip_pipeline: support RSS
This patch enables rss (receive side scaling) per network interface
through the configuration file. The user can specify following
parameters in LINK section for enabling the rss feature - rss_qs,
rss_proto_ipv4, rss_proto_ipv6 and ip_proto_l2.
The "rss_qs" is mandatory parameter which indicates the queues to be
used for rss, while rest of the parameters are optional. When optional
parameters are not provided in the configuration file, default setting
(ETH_RSS_IPV4 | ETH_RSS_IPV6) is assumed for "rss_hf" field of the
rss_conf structure.
For example, following configuration can be applied for using the rss
on port 0 of the network interface;
Jasvinder Singh [Wed, 8 Jun 2016 16:30:56 +0000 (17:30 +0100)]
examples/ip_pipeline: clean up configuration parser
This commit adds following changes to configuration file parsing of
the ip pipeline application;
1. Parsing routines related to packet queues (pktq_in/out fields in the
PIPELINE section) and message queues (msgq_in/out fields of in the MSGQ
Section) are updated.
In the parsing routines, function "strtok_r()" is used for parsing the
string instead of manually checking the string termination, white
spaces, tabs etc., between the string tokens. Each call to strtok_r()
returns a pointer to a null-terminated string containing the next token.
If no more tokens are found, strtok_r() returns NULL. As a result of
using strtok_r(), the code size of the parsing routines is reduced
significantly.
2. Replace PARSER_PARAM_ADD_CHECK macro by more specific macros such as
PARSE_CHECK_DUPLICATE_SECTION, PARSE_CHECK_DUPLICATE_SECTION_EAL to detect
duplicate entries in the various sections of the configuration file
3. Add new macros PARSER_ERROR_NO_ELEMENTS and PARSE_ERROR_TOO_MANY_ELEMENTS
for detecting no element and more elements than allowed situations
respectively, in the section entry.
4. Add new macros APP_PARAM_ADD_LINK_FOR_RXQ, APP_PARAM_ADD_LINK_FOR_TXQ
and APP_PARAM_ADD_LINK_FOR_TM which add corresponding nic ports entry to
the application param structure while parsing rx/tx queues, TM (Traffic
Manager) port sections and pktq_in/out entries of pipeline sections
Piotr Azarewicz [Wed, 8 Jun 2016 10:35:24 +0000 (12:35 +0200)]
examples/ip_pipeline: rework routing CLI
Several routing commands are merged into two commands:
route and arp - these two commands are handled by cli library.
Rest of the commands are handled internaly by the pipeline code.
Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com> Acked-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
This patch modifies flow classifications pipeline command line
interface. All commands are merged into one cmd_fc_parsed.
Additionally a classification for ipv6, ipv4 and qinq can be added from
configuration file.
1. flow add qinq bulk
File line format:
qinq <svlan> <cvlan> port <port ID> id <flow ID>
File line example:
qinq 1 2 port 3 id 0
2. flow add ipv4 bulk
File line format:
ipv4 <sipaddr> <dipaddr> <sport> <dport> <proto> port <port ID> id
<flowID>
File line example:
ipv4 1.2.3.4 10.20.30.40 63 127 6 port 2 id 999
3. flow add ipv6 bulk
File line format:
ipv6 <sipaddr> <dipaddr> <sport> <dport> <proto> port <port ID> id
<flowID>
Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Signed-off-by: Marcin Kerlin <marcinx.kerlin@intel.com> Signed-off-by: Slawomir Mrozowicz <slawomirx.mrozowicz@intel.com> Signed-off-by: Michal Kobylinski <michalx.kobylinski@intel.com> Acked-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Daniel Mrzyglod [Wed, 8 Jun 2016 10:35:21 +0000 (12:35 +0200)]
examples/ip_pipeline: rework firewall CLI
Each command are merged into one: cmd_firewall_parsed.
ADD command format is changed:
p <pipeline ID> firewall add priority <priority> ipv4 <sipaddr>
<sipdepth> <dipaddr> <dipdepth> <sport0> <sport1> <dport0> <dport1>
<proto> <protomask> port <port ID>
and bulk command was modified:
1. firewall add bulk
File line format:
priority <priority> ipv4 <sipaddr> <sipdepth> <dipaddr> <dipdepth>
<sport0> <sport1> <dport0> <dport1> <proto> <protomask> port <port ID>
(protomask is a hex value)
File line example:
priority 0 ipv4 1.2.3.0 24 10.20.30.40 32 0 63 64 127 6 0xF port 3
Tomasz Kulasek [Wed, 8 Jun 2016 10:35:20 +0000 (12:35 +0200)]
examples/ip_pipeline: rework common CLI
All link commands are merged into one command:
cmd_link_parsed.
Improve run command to allow run periodically.
Adding static keyword to a lot of token declarations.
Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com> Signed-off-by: Michal Kobylinski <michalx.kobylinski@intel.com> Acked-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Fan Zhang [Tue, 17 May 2016 16:02:10 +0000 (17:02 +0100)]
examples/ip_pipeline: fix null packet in source port
This patch fixes the NULL packet processing problem. Originally,
pipeline's write attempt to NULL packets generated by source
port may crash the application.
This patch fixes the problem by enforcing each source port
defined in cfg file containing a user specified or default pcap
file path.
When initializing crypto devices within the app,
the provided key sizes are checked against the supported
sizes from the crypto device capabilities.
When the supported sizes are not a range, but a single value,
the check may become an infinite loop (when size is not supported).
Fixes: a061e50a0d97 ("examples/l2fwd-crypto: fix ambiguous input key size") Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com> Tested-by: Min Cao <min.cao@intel.com>
Arek Kusztal [Tue, 7 Jun 2016 16:58:13 +0000 (17:58 +0100)]
qat: avoid strict-aliasing break
To avoid GCC warning about "dereferencing type-punned pointer will break
strict-aliasing rules" aad_len pointer is dereferenced instead of direct
dereferencing of uint32_t* cast of the middle of an array.
Fix null pointer dereferencing by reporting if null and
exiting the function.
Coverity issue: 126584 Fixes: c0f87eb5252b ("cryptodev: change burst API to be crypto op oriented") Signed-off-by: Deepak Kumar Jain <deepak.k.jain@intel.com> Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Coverity issue: 126585 Fixes: 924e84f87306 ("aesni_mb: add driver for multi buffer based crypto") Signed-off-by: Deepak Kumar Jain <deepak.k.jain@intel.com> Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Anatoly Burakov [Wed, 1 Jun 2016 14:51:57 +0000 (15:51 +0100)]
ivshmem: document a potential crash in rte_ring
Commit 4768c475 added a pointer to the memzone in rte_ring. However,
all memzones are residing in local mem_config, therefore accessing
the memzone pointer inside the guest in an IVSHMEM-shared rte_ring
will cause segmentation fault. This issue is unlikely to ever get
fixed, as this would require lots of changes for very little benefit,
therefore we're documenting this limitation instead.
Anatoly Burakov [Tue, 24 May 2016 12:50:05 +0000 (13:50 +0100)]
ivshmem: fix overlap detection
Partial revert of an earlier ill-conceived "fix".
Adjacent segments can never be considered overlapping because we
are not comparing ends to starts, but rather starts to starts.
Therefore the earlier fix was wrong (plus it also had a typo).
Olivier Matz [Mon, 30 May 2016 11:56:46 +0000 (13:56 +0200)]
kni: fix build clean on Ubuntu
The $(comma) variable is not defined in this Makefile, nor in
any included Makefile. Seen while doing a "make clean" on ubuntu:
$ make clean
== Clean lib
== Clean lib/librte_compat
== Clean lib/librte_eal
== Clean lib/librte_eal/common
== Clean lib/librte_eal/linuxapp
== Clean lib/librte_eal/linuxapp/eal
== Clean lib/librte_eal/linuxapp/igb_uio
== Clean lib/librte_eal/linuxapp/kni
tr: missing operand after ‘.-’
Two strings must be given when translating.
Try 'tr --help' for more information.
This commit replaces $(comma) by a ',' character, it's not a problem in
that case since we are inside antiquotes.
Jerin Jacob [Thu, 26 May 2016 11:20:14 +0000 (16:50 +0530)]
mk: fix cross-compilation
Removed comparison against $CC in Makefiles as
in cross-compiling mode CC can be a different string
instead of string "gcc"
Suggested-by: Thomas Monjalon <thomas.monjalon@6wind.com> Signed-off-by: Jerin Jacob <jerin.jacob@caviumnetworks.com> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Thomas Monjalon [Thu, 2 Jun 2016 09:13:15 +0000 (11:13 +0200)]
scripts: check commits with checkpatch
The new option -n allows to give a number of commits to check
from the git HEAD.
If neither -n nor patch files are given, the commits after
origin/master are checked.
Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Jan Viktorin [Tue, 3 May 2016 19:15:27 +0000 (21:15 +0200)]
app/test: fix buffer overflow
A bug has been detected by valgrind:
Invalid write of size 1
by 0x86ECC76: sprintf (in /usr/lib/libc-2.23.so)
by 0x430B0A: commands_init (in build/app/test)
by 0x42F215: main (in build/app/test)
Address 0x9d72ff2 is 0 bytes after a block of size 1,346 alloc'd
at 0x78C1BD0: malloc (in vgpreload_memcheck-amd64-linux.so)
by 0x430AE4: commands_init (in build/app/test)
by 0x42F215: main (in build/app/test)
The commands buffer is exactly 1346 B long so there is an access just
after the buffer. The sprintf always writes '\0' at the end of the string.
The '#' separator adds 1 B more to each string. This is correct until the
last string is sprinted there. The last one is 1 B longer then expected,
i.e.:
strlen(t->command) + strlen("#") + ONE_FOR_ZERO
Fixes: 727909c59231 ("app/test: introduce dynamic commands list") Signed-off-by: Jan Viktorin <viktorin@rehivetech.com> Acked-by: David Marchand <david.marchand@6wind.com>
Thomas Monjalon [Wed, 4 May 2016 13:18:29 +0000 (15:18 +0200)]
app/test: fix log check when default level is high
The log unit test was checking display of low priority messages.
It was not working if RTE_LOG_LEVEL is not RTE_LOG_DEBUG.
It is even easier to see since the default level is INFO (9b9d7ca).
Now the test use ERR and CRIT levels which should be always enabled
while not trigerring syslog output on the console.
Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
Yuanhan Liu [Wed, 11 May 2016 04:09:01 +0000 (21:09 -0700)]
scripts: highlight bad patterns in commit check
I got a report like following:
Wrong headline lowercase:
xxx: move vhost device ctx to cuse
It takes a bit while (by checking the code) that it's "ctx" triggers
the warning. It could be spotted very quickly if "tx" is highlighted.
This patch adds such support.
Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Thomas Monjalon [Thu, 19 May 2016 18:00:04 +0000 (20:00 +0200)]
scripts: remove unused map files merger
This script was forgotten when dropping the combined library.
Fixes: 948fd64befc3 ("mk: replace the combined library with a linker script") Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com> Acked-by: Panu Matilainen <pmatilai@redhat.com>
Panu Matilainen [Tue, 24 May 2016 10:28:31 +0000 (13:28 +0300)]
qede: add missing external dependency and disable by default
The qede driver depends on libz but the LDLIBS entry in makefile
was missing. Also because of the external dependency, make it
disabled in default config as per common DPDK policy on external deps.
Fixes: ec94dbc57362 ("qede: add base driver") Signed-off-by: Panu Matilainen <pmatilai@redhat.com> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
The RTE_ETH_VALID_PORTID_OR_ERR_RET macro is used in some places
to check if a port id is valid or not. This commit makes use of it in
some new parts of the code.
Signed-off-by: Mauricio Vasquez B <mauricio.vasquezbernal@studenti.polito.it> Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
Olivier Matz [Wed, 18 May 2016 16:02:08 +0000 (18:02 +0200)]
mbuf: add prefetch helpers
Some architectures (ex: Power8) have a cache line size of 128 bytes,
so the drivers should not expect that prefetching the second part of
the mbuf with rte_prefetch0(&m->cacheline1) is valid.
This commit add helpers that can be used by drivers to prefetch the
rx or tx part of the mbuf, whatever the cache line size.
Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Reviewed-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
Olivier Matz [Wed, 18 May 2016 11:04:55 +0000 (13:04 +0200)]
app/test: rework mempool test
Rework the mempool test to better indicate where it failed,
and, now that this feature is available, add the freeing of the
mempool after the test is done.
Olivier Matz [Wed, 18 May 2016 11:04:50 +0000 (13:04 +0200)]
mempool: populate with anonymous memory
Now that we can populate a mempool with any virtual memory,
it is easier to introduce a function to populate a mempool
with memory coming from an anonymous mapping, as it's done
in test-pmd.
The next commit will replace test-pmd anonymous mapping by
this function.
Olivier Matz [Wed, 18 May 2016 11:04:49 +0000 (13:04 +0200)]
mempool: create the internal ring when populating
Instead of creating the internal ring at mempool creation, do
it when populating the mempool with the first memory chunk. The
objective here is to simplify the change of external handler
when it will be introduced.
Olivier Matz [Wed, 18 May 2016 11:04:48 +0000 (13:04 +0200)]
mempool: rework support of Xen dom0
Avoid to have a specific file for that, and remove #ifdefs.
Now that we have introduced a function to populate a mempool
with a virtual area, the support of xen dom0 is much easier.
The only thing we need to do is to convert the guest physical
address into the machine physical address using rte_mem_phy2mch().
This function does nothing when not running xen.
Olivier Matz [Wed, 18 May 2016 11:04:46 +0000 (13:04 +0200)]
mempool: introduce a function to create an empty pool
Introduce a new function rte_mempool_create_empty()
that allocates a mempool that is not populated.
The functions rte_mempool_create() and rte_mempool_xmem_create()
now make use of it, making their code much easier to read.
Currently, they are the only users of rte_mempool_create_empty()
but the function will be made public in next commits.
Olivier Matz [Wed, 18 May 2016 11:04:45 +0000 (13:04 +0200)]
mempool: introduce a function to free a pool
Introduce rte_mempool_free() that:
- unlink the mempool from the global list if it is found
- free all the memory chunks using their free callbacks
- free the internal ring
- free the memzone containing the mempool
Currently this function is only used in error cases when
creating a new mempool, but it will be made public later
in the patch series.
Olivier Matz [Wed, 18 May 2016 11:04:44 +0000 (13:04 +0200)]
mempool: replace physical address by a memzone pointer
Storing the pointer to the memzone instead of the physical address
provides more information than just the physical address: for instance,
the memzone flags.
Moreover, keeping the memzone pointer will allow us to free the mempool
(this is done later in the series).
Olivier Matz [Wed, 18 May 2016 11:04:43 +0000 (13:04 +0200)]
mempool: support no hugepage mode
Introduce a new function rte_mempool_populate_virt() that is now called
by default when hugepages are not supported. This function populate the
mempool with several physically contiguous chunks whose minimum size is
the page size of the system.
Thanks to this, rte_mempool_create() will work properly in without
hugepages (if the object size is smaller than a page size), and 2
specific workarouds can be removed:
- trailer_size was artificially extended to a page size
- rte_mempool_virt2phy() did not rely on object physical address
Olivier Matz [Wed, 18 May 2016 11:04:42 +0000 (13:04 +0200)]
mem: lock pages when not using hugepages
Although the physical address won't be correct in memory segment,
this allows at least to retrieve the physical address using
rte_mem_virt2phy(). Indeed, if the page is not locked, the page
may not be present in physical memory.
With next commit, it allows a mempool to have properly filled physical
addresses when using --no-huge option.
Olivier Matz [Wed, 18 May 2016 11:04:41 +0000 (13:04 +0200)]
mempool: allocate in several memory chunks by default
Introduce rte_mempool_populate_default() which allocates
mempool objects in several memzones.
The mempool header is now always allocated in a specific memzone
(not with its objects). Thanks to this modification, we can remove
many specific behavior that was required when hugepages are not
enabled in case we are using rte_mempool_xmem_create().
This change requires to update how kni and mellanox drivers lookup for
mbuf memory. For now, this will only work if there is only one memory
chunk (like today), but we could make use of rte_mempool_mem_iter() to
support more memory chunks.
We can also remove RTE_MEMPOOL_OBJ_NAME that is not required anymore for
the lookup, as memory chunks are referenced by the mempool.
Note that rte_mempool_create() is still broken (it was the case before)
when there is no hugepages support (rte_mempool_create_xmem() has to be
used). This is fixed in next commit.
Olivier Matz [Wed, 18 May 2016 11:04:40 +0000 (13:04 +0200)]
mempool: get memory size with unspecified page size
Update rte_mempool_xmem_size() so that when the page_shift argument is
set to 0, assume that memory is physically contiguous, allowing to
ignore page boundaries. This will be used in the next commits.
By the way, rename the variable 'n' as 'obj_per_page' and avoid the
affectation inside the if().
Olivier Matz [Wed, 18 May 2016 11:04:39 +0000 (13:04 +0200)]
mempool: introduce a free callback for memory chunks
Introduce a free callback that is passed to the populate* functions,
which is used when freeing a mempool. This is unused now, but as next
commits will populate the mempool with several chunks of memory, we
need a way to free them properly on error.
Later in the series, we will also introduce a public rte_mempool_free()
and the ability for the user to populate a mempool with its own memory.
For that, we also need a free callback.
Olivier Matz [Wed, 18 May 2016 11:04:38 +0000 (13:04 +0200)]
mempool: simplify memory usage calculation
This commit simplifies rte_mempool_xmem_usage().
Since previous commit, the function rte_mempool_xmem_usage() is
now the last user of rte_mempool_obj_mem_iter(). This complex
code can now be moved inside the function. We can get rid of the
callback and do some simplification to make the code more readable.
Olivier Matz [Wed, 18 May 2016 11:04:35 +0000 (13:04 +0200)]
mempool: remove macro to check if contiguous
This commit removes MEMPOOL_IS_CONTIG().
The next commits will change the behavior of the mempool library so that
the objects will never be allocated in the same memzone than the mempool
header. Therefore, there is no reason to keep this macro that would
always return 0.
Olivier Matz [Wed, 18 May 2016 11:04:34 +0000 (13:04 +0200)]
mempool: store physical address in objects
Store the physical address of the object in its header. It simplifies
rte_mempool_virt2phy() and prepares the removing of the paddr[] table
in the mempool header.
Olivier Matz [Wed, 18 May 2016 11:04:33 +0000 (13:04 +0200)]
mempool: create internal ring in a specific function
This makes the code of rte_mempool_create() clearer, and it will make
the introduction of external mempool handler easier (in another patch
series). Indeed, this function contains the specific part when a ring is
used, but it could be replaced by something else in the future.
This commit also adds a socket_id field in the mempool structure that
is used by this new function.
Olivier Matz [Wed, 18 May 2016 11:04:32 +0000 (13:04 +0200)]
mempool: use the list to initialize objects
Before this patch, the mempool elements were initialized at the time
they were added to the mempool. This patch changes this to do the
initialization of all objects once the mempool is populated, using
rte_mempool_obj_iter() introduced in previous commits.
Thanks to this modification, we are getting closer to a new API
that would allow us to do:
mempool_init()
mempool_populate(mem1)
mempool_populate(mem2)
mempool_populate(mem3)
mempool_init_obj()
Olivier Matz [Wed, 18 May 2016 11:04:30 +0000 (13:04 +0200)]
mempool: use the list to iterate the elements
Now that the mempool objects are chained into a list, we can use it to
browse them. This implies a rework of rte_mempool_obj_iter() API, that
does not need to take as many arguments as before. The previous function
is kept as a private function, and renamed in this commit. It will be
removed in a next commit of the patch series.
The only internal users of this function are the mellanox drivers. The
code is updated accordingly.
Introducing an API compatibility for this function has been considered,
but it is not easy to do without keeping the old code, as the previous
function could also be used to browse elements that were not added in a
mempool. Moreover, the API is already be broken by other patches in this
version.
The library version was already updated in
commit 213af31e0960 ("mempool: reduce structure size if no cache needed")
git.droids-corp.org / dpdk.git / log