Ed Czeck [Thu, 18 Mar 2021 17:36:58 +0000 (13:36 -0400)]
net/ark: cleanup dynamic extension interface
- Rename extension functions with rte_pmd_ark prefix
- Update local function documentation
Signed-off-by: Ed Czeck <ed.czeck@atomicrules.com>
Ed Czeck [Thu, 18 Mar 2021 17:36:57 +0000 (13:36 -0400)]
net/ark: update to reflect FPGA updates
- New PCIe IDs using net/ark driver
- Update Version IDs and structures specified by hardware
- New internal descriptor status for TX
- Adjust data placement in RX operations, headroom in retained for
segmented mbufs
Signed-off-by: Ed Czeck <ed.czeck@atomicrules.com>
Ed Czeck [Thu, 18 Mar 2021 17:36:56 +0000 (13:36 -0400)]
net/ark: refactor Rx buffer recovery
Allocate mbufs for Rx path in bulk of at least 64 buffers
to improve performance. Allow recovery even without
a Rx operation to support lack of buffers in pool.
Fixes:
be410a861598 ("net/ark: add recovery for lack of mbufs")
Cc: stable@dpdk.org
Signed-off-by: Ed Czeck <ed.czeck@atomicrules.com>
Ed Czeck [Thu, 18 Mar 2021 17:36:55 +0000 (13:36 -0400)]
net/ark: update packet director initial state
Fixes:
b33ccdb17f55 ("net/ark: provide API for hardware modules MPU RQP and pktdir")
Cc: stable@dpdk.org
Signed-off-by: Ed Czeck <ed.czeck@atomicrules.com>
David Marchand [Fri, 26 Mar 2021 08:24:47 +0000 (09:24 +0100)]
telemetry: cleanup internal header
The experimental banner can be removed.
Every in-tree file is compiled with _GNU_SOURCE, so RTE_HAS_CPUSET is
unneeded for an internal header.
Fixes:
0e64ae618e10 ("telemetry: move init function to internal header")
Signed-off-by: David Marchand <david.marchand@redhat.com>
Acked-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Ciara Power <ciara.power@intel.com>
Dmitry Kozlyuk [Wed, 24 Mar 2021 19:32:26 +0000 (22:32 +0300)]
mem: fix cleanup when multi-process is disabled
rte_eal_memory_detach() did not account for cases where multi-process
mode is disabled: --in-memory and --no-shconf. This resulted
in unmapping memory that had not been mapped, which caused errors:
EAL: Could not unmap memory: No error (Windows)
EAL: Cannot munmap(0x1d47f40, 0x7000): Invalid argument (Linux)
Confusing "No error" was caused by using errno instead of rte_errno
set by rte_mem_unmap().
Skip detaching memory altogether when --in-memory is specified.
Skip unmapping configuration when it's not shared.
Fix and add error handling to produce proper log messages.
Fixes:
dfbc61a2f9a6 ("mem: detach memsegs on cleanup")
Reported-by: Jie Zhou <jizh@microsoft.com>
Suggested-by: David Marchand <david.marchand@redhat.com>
Signed-off-by: Dmitry Kozlyuk <dmitry.kozliuk@gmail.com>
Acked-by: Ranjit Menon <ranjit.menon@intel.com>
Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tal Shnaiderman [Tue, 16 Mar 2021 13:28:19 +0000 (15:28 +0200)]
eal: rename thread TLS API
Rename the key opaque pointer from rte_tls_key to
rte_thread_key to avoid confusion with transport layer security.
Also rename and remove the "_tls" term from the following
functions to avoid redundancy:
rte_thread_tls_key_create
rte_thread_tls_key_delete
rte_thread_tls_value_set
rte_thread_tls_value_get
Suggested-by: Vladimir Medvedkin <vladimir.medvedkin@intel.com>
Suggested-by: Morten Brørup <mb@smartsharesystems.com>
Signed-off-by: Tal Shnaiderman <talshn@nvidia.com>
Acked-by: Morten Brørup <mb@smartsharesystems.com>
Tal Shnaiderman [Tue, 16 Mar 2021 13:28:18 +0000 (15:28 +0200)]
eal: add error numbers in thread TLS API
Add error number reporting to rte_errno in all
functions in the rte_thread_tls_* API.
Suggested-by: Anatoly Burakov <anatoly.burakov@intel.com>
Signed-off-by: Tal Shnaiderman <talshn@nvidia.com>
Acked-by: Dmitry Kozlyuk <dmitry.kozliuk@gmail.com>
Thomas Monjalon [Mon, 8 Mar 2021 22:25:20 +0000 (23:25 +0100)]
usertools: show hugepages requested/set on failure
In case the number of requested hugepages cannot be set,
a more detailed error message is printed.
The new message does not mention "reserve" because setting
can be reserving or clearing.
The filename and numbers requested/set are printed to ease debugging.
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Thomas Monjalon [Sun, 21 Mar 2021 09:09:03 +0000 (10:09 +0100)]
usertools: check 0-division with hugepage size
The default page size can be None, and the page size from user request
can be 0 kB if lower than 1024. In these cases, a division will fail.
In order to avoid a Python exception, the page size is checked
and an error message "Invalid page size" is printed.
A similar error message is printed in set_hugepages()
if the size is not supported, except at this stage the message can be
completed with "Valid page sizes".
Unfortunately the first check is too early to print such information.
A third error message can be printed in a different place (get_memsize)
in case of a format issue, e.g. a negative size.
The function get_memsize() is also used for total requested size,
so the error message "not a valid page size" was potentially wrong.
This message is replaced with the more general "is not a valid size".
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>
Bruce Richardson [Tue, 16 Feb 2021 11:39:21 +0000 (11:39 +0000)]
usertools: print process name when telemetry connects
When the dpdk-telemetry client connects to a DPDK instance, we can use the
PID provided in the initial connection message to query from /proc the name
of the process we are connected to, and display that to the user. We use
the "cmdline" procfs entry for the query since that is available on both
Linux and FreeBSD (assuming procfs is mounted on the BSD instance).
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Kevin Laatz <kevin.laatz@intel.com>
Kevin Laatz [Tue, 16 Feb 2021 11:50:08 +0000 (11:50 +0000)]
usertools: add file-prefix option for telemetry
Currently the dpdk-telemetry.py script connects to all running DPDK apps
consecutively. With the addition of this file-prefix argument, we can limit
the amount of information returned providing improved consumability and
precision to the user.
Signed-off-by: Kevin Laatz <kevin.laatz@intel.com>
Reviewed-by: Bruce Richardson <bruce.richardson@intel.com>
Tested-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>
Ciara Power [Thu, 25 Mar 2021 11:53:11 +0000 (11:53 +0000)]
maintainers: update for telemetry
Replace Kevin Laatz with Ciara Power.
Signed-off-by: Ciara Power <ciara.power@intel.com>
Acked-by: Kevin Laatz <kevin.laatz@intel.com>
Bruce Richardson [Thu, 25 Mar 2021 13:57:07 +0000 (13:57 +0000)]
telemetry: move init function to internal header
The rte_telemetry_init() function is for EAL use only, so can be moved to
the internal header rather than being in the public one.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Ciara Power <ciara.power@intel.com>
Bruce Richardson [Thu, 25 Mar 2021 13:57:06 +0000 (13:57 +0000)]
telemetry: rename internal-only header file
The header file containing the legacy telemetry function prototypes was all
internal-only, so we rename the file to be an internal-only one to make it
clearer it's not for installation.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Ciara Power <ciara.power@intel.com>
Bruce Richardson [Thu, 25 Mar 2021 13:57:05 +0000 (13:57 +0000)]
telemetry: make the legacy registration function internal
The function for registration of callbacks for legacy telemetry was
documented as internal-only in the API documents, but marked as
experimental in the version.map file. Since this is an internal-only
function, for consistency we update the version mapping to have it as
internal.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Ciara Power <ciara.power@intel.com>
Bruce Richardson [Thu, 25 Mar 2021 13:57:04 +0000 (13:57 +0000)]
telemetry: use log function from pointer
Rather than passing back an error string to the caller, take as input the
rte_log function to use, and just use regular logging.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Ciara Power <ciara.power@intel.com>
Thinh Tran [Mon, 2 Mar 2020 20:41:32 +0000 (15:41 -0500)]
test: fix autotest handling of skipped tests
- When running the default configuration of autotest of the make test
it'd take 900 seconds (15 minutes) for the script TIMEOUT and marks
Failed for a specific device test that is not supported on the
system under test.
- Adding the checking for those tests, print out as "Skipped [Not Run]"
quickly return and continue for next test
Fixes:
da0af48a67a5 ("test: add skipped return result")
Cc: stable@dpdk.org
Signed-off-by: Thinh Tran <thinhtr@linux.vnet.ibm.com>
Reviewed-by: David Christensen <drc@linux.vnet.ibm.com>
David Hunt [Tue, 16 Mar 2021 09:38:11 +0000 (09:38 +0000)]
power: check sysfs base frequency
Some kernels may show in incorrect value for base frequency in
sysfs (e.g. 15 GHz). This throws off the SST-BF algorithm for
high and low priority cores. So if base_frequency is greater
than max turbo frequency, ignore, and handle it as a normal
core.
Known Kernel version with issue: Linux 5.8.7
Signed-off-by: David Hunt <david.hunt@intel.com>
Cristian Dumitrescu [Tue, 16 Mar 2021 11:40:39 +0000 (11:40 +0000)]
pipeline: add meter array to SWX
Meter arrays are stateful objects that are updated by the data plane
and configured & monitored by the control plane. The meters implement
the RFC 2698 Two Rate Three Color Marker (trTCM) algorithm.
Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Cristian Dumitrescu [Tue, 16 Mar 2021 11:40:38 +0000 (11:40 +0000)]
pipeline: add register array to SWX
Register arrays are stateful objects that can be read & modified by
both the data plane and the control plane, as opposed to tables, which
are read-only for data plane. One key use-case is the implementation
of stats counters.
Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Cristian Dumitrescu [Wed, 24 Mar 2021 11:36:03 +0000 (11:36 +0000)]
pipeline: fix instruction translation
The SWX pipeline instructions work with operands of different types:
header fields (h.header.field), packet meta-data (m.field), extern
object mailbox field (e.obj.field), extern function (f.field), action
data read from table entries (t.field), or immediate values; hence the
HMEFTI acronym.
For some pipeline instructions (add/sub, srl/shr, jmplt/jmpgt), only
the H, M and I cases were handled, while the E, F and T cases were
disregarded. This is what we fix here.
Fixes:
baf7999303d0 ("pipeline: introduce SWX add instruction")
Fixes:
c88c62943818 ("pipeline: introduce SWX subtract instruction")
Fixes:
b09ba6d0a3c2 ("pipeline: introduce SWX SHL instruction")
Fixes:
e0f51638b715 ("pipeline: introduce SWX SHR instruction")
Fixes:
b3947e25bed4 ("pipeline: introduce SWX jump and return instructions")
Cc: stable@dpdk.org
Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Churchill Khangar [Wed, 24 Mar 2021 12:22:33 +0000 (08:22 -0400)]
examples/pipeline: support hexadecimal argument in CLI
This patch allows specification of integer arguments for the
CLI commands in hexadecimal and octal as well as decimal.
Signed-off-by: Churchill Khangar <churchill.khangar@intel.com>
Acked-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Dapeng Yu [Fri, 19 Mar 2021 07:11:18 +0000 (15:11 +0800)]
examples/packet_ordering: fix port configuration
A global ethernet port configuration is assigned to local variable
because in this way the local variable may be updated as required. But
this local variable is not used as input of rte_eth_dev_configure() in
original implementation, and cause that fast mbuf free feature cannot
be enabled on port.
This patch use this local variable as input of rte_eth_dev_configure().
Fixes:
6833f919f56b ("examples/packet_ordering: convert to new ethdev offloads API")
Cc: stable@dpdk.org
Signed-off-by: Dapeng Yu <dapengx.yu@intel.com>
Acked-by: Reshma Pattan <reshma.pattan@intel.com>
Cristian Dumitrescu [Fri, 19 Feb 2021 14:26:19 +0000 (14:26 +0000)]
examples/pipeline: make quanta configurable
Make the pipeline instruction quanta configurable at build time.
Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Venkata Suresh Kumar P [Tue, 23 Mar 2021 18:05:03 +0000 (14:05 -0400)]
port: add file descriptor SWX port
Add the file descriptor input/output port type for the SWX pipeline.
File descriptor port type provides interface with the kernel network
stack. Example file descriptor port is TAP device.
Signed-off-by: Venkata Suresh Kumar P <venkata.suresh.kumar.p@intel.com>
Signed-off-by: Churchill Khangar <churchill.khangar@intel.com>
Acked-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Cristian Dumitrescu [Tue, 16 Feb 2021 20:46:46 +0000 (20:46 +0000)]
table: add wildcard match table type
Add the widlcard match/ACL table type for the SWX pipeline, which is
used under the hood by the table instruction.
Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Churchill Khangar <churchill.khangar@intel.com>
Cristian Dumitrescu [Tue, 16 Feb 2021 20:46:45 +0000 (20:46 +0000)]
table: add table entry priority
Add support for table entry priority, which is required for the
wildcard match/ACL table type.
Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Cristian Dumitrescu [Tue, 16 Feb 2021 20:46:44 +0000 (20:46 +0000)]
pipeline: support non-incremental table updates
Some table types (e.g. exact match/hash) allow for incremental table
updates, while others (e.g. wildcard match/ACL) do not. The former is
already supported, the latter is enabled by this patch.
Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Cristian Dumitrescu [Tue, 16 Feb 2021 20:46:43 +0000 (20:46 +0000)]
pipeline: improve table entry parsing
Improve the table entry parsing: better code structure, enable parsing
for the key field masks, allow comments and empty lines in the table
entry files.
Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Venkata Suresh Kumar P <venkata.suresh.kumar.p@intel.com>
Signed-off-by: Churchill Khangar <churchill.khangar@intel.com>
Cristian Dumitrescu [Tue, 16 Feb 2021 20:46:42 +0000 (20:46 +0000)]
pipeline: improve table entry helpers
Improve the internal table entry helper routines for key comparison,
entry duplication and checks.
Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Cristian Dumitrescu [Wed, 17 Mar 2021 11:42:07 +0000 (11:42 +0000)]
table: fix actions with different data size
The table layer provisions an action_id and action_data_size data
bytes for each table key. This action_data_size is a maximal amount,
as some actions (depending on action_id) can require zero or less data
bytes than the maximal action_data_size. This fix allows for actions
with different data sizes to co-exist within the same table.
Fixes:
d0a00966618b ("table: add exact match SWX table")
Cc: stable@dpdk.org
Signed-off-by: Churchill Khangar <churchill.khangar@intel.com>
Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Acked-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Cristian Dumitrescu [Thu, 28 Jan 2021 18:55:12 +0000 (18:55 +0000)]
port: add ring SWX port
Add the ring input/output port type for the SWX pipeline.
Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Ibtisam Tariq [Thu, 4 Feb 2021 07:34:20 +0000 (07:34 +0000)]
examples/vhost_crypto: enhance getopt_long usage
Instead of using getopt_long return value, strcmp was used to
compare the input parameters with the struct option array. This
patch get rid of all those strcmp by directly binding each longopt
with an int enum. This is to improve readability and consistency in
all examples.
Bugzilla ID: 238
Reported-by: David Marchand <david.marchand@redhat.com>
Signed-off-by: Ibtisam Tariq <ibtisam.tariq@emumba.com>
Acked-by: Fan Zhang <roy.fan.zhang@intel.com>
Reviewed-by: David Marchand <david.marchand@redhat.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Ibtisam Tariq [Thu, 4 Feb 2021 07:34:19 +0000 (07:34 +0000)]
examples/vhost: enhance getopt_long usage
Instead of using getopt_long return value, strcmp was used to
compare the input parameters with the struct option array. This
patch get rid of all those strcmp by directly binding each longopt
with an int enum. This is to improve readability and consistency in
all examples.
Bugzilla ID: 238
Reported-by: David Marchand <david.marchand@redhat.com>
Signed-off-by: Ibtisam Tariq <ibtisam.tariq@emumba.com>
Reviewed-by: David Marchand <david.marchand@redhat.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Ibtisam Tariq [Thu, 4 Feb 2021 07:34:18 +0000 (07:34 +0000)]
examples/qos_sched: enhance getopt_long usage
Instead of using getopt_long return value, strcmp was used to
compare the input parameters with the struct option array. This
patch get rid of all those strcmp by directly binding each longopt
with an int enum. This is to improve readability and consistency in
all examples.
Bugzilla ID: 238
Reported-by: David Marchand <david.marchand@redhat.com>
Signed-off-by: Ibtisam Tariq <ibtisam.tariq@emumba.com>
Reviewed-by: David Marchand <david.marchand@redhat.com>
Ibtisam Tariq [Thu, 4 Feb 2021 07:34:17 +0000 (07:34 +0000)]
examples/performance-thread: enhance getopt_long usage
Instead of using getopt_long return value, strcmp was used to
compare the input parameters with the struct option array. This
patch get rid of all those strcmp by directly binding each longopt
with an int enum. This is to improve readability and consistency in
all examples.
Bugzilla ID: 238
Reported-by: David Marchand <david.marchand@redhat.com>
Signed-off-by: Ibtisam Tariq <ibtisam.tariq@emumba.com>
Reviewed-by: David Marchand <david.marchand@redhat.com>
Ibtisam Tariq [Thu, 4 Feb 2021 07:34:16 +0000 (07:34 +0000)]
examples/packet_ordering: enhance getopt_long usage
Instead of using getopt_long return value, strcmp was used to
compare the input parameters with the struct option array. This
patch get rid of all those strcmp by directly binding each longopt
with an int enum. This is to improve readability and consistency in
all examples.
Bugzilla ID: 238
Reported-by: David Marchand <david.marchand@redhat.com>
Signed-off-by: Ibtisam Tariq <ibtisam.tariq@emumba.com>
Reviewed-by: David Marchand <david.marchand@redhat.com>
Ibtisam Tariq [Thu, 4 Feb 2021 07:34:15 +0000 (07:34 +0000)]
examples/l3fwd-acl: enhance getopt_long usage
Instead of using getopt_long return value, strcmp was used to
compare the input parameters with the struct option array. This
patch get rid of all those strcmp by directly binding each longopt
with an int enum. This is to improve readability and consistency in
all examples.
Bugzilla ID: 238
Reported-by: David Marchand <david.marchand@redhat.com>
Signed-off-by: Ibtisam Tariq <ibtisam.tariq@emumba.com>
Reviewed-by: David Marchand <david.marchand@redhat.com>
Ibtisam Tariq [Thu, 4 Feb 2021 07:34:14 +0000 (07:34 +0000)]
examples/fips_validation: enhance getopt_long usage
Instead of using getopt_long return value, strcmp was used to
compare the input parameters with the struct option array. This
patch get rid of all those strcmp by directly binding each longopt
with an int enum. This is to improve readability and consistency in
all examples.
Bugzilla ID: 238
Reported-by: David Marchand <david.marchand@redhat.com>
Signed-off-by: Ibtisam Tariq <ibtisam.tariq@emumba.com>
Reviewed-by: David Marchand <david.marchand@redhat.com>
Thomas Monjalon [Thu, 25 Feb 2021 08:43:35 +0000 (09:43 +0100)]
examples/bbdev: fix header include for musl
The header file unistd.h should not be included from sys/ directory,
it is an error with musl libc.
Fixes:
1ffee690eaa1 ("examples/bbdev: add sample app")
Cc: stable@dpdk.org
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Acked-by: David Marchand <david.marchand@redhat.com>
Natanael Copa [Thu, 5 Nov 2020 21:17:09 +0000 (22:17 +0100)]
app/testpmd: fix build with musl
1/ Improve portability by avoiding use of non-standard 'uint'.
Use uint8_t for hash_key_len as rss_key_len is a uint8_t type.
This solves following build error when building with musl libc:
app/test-pmd/testpmd.h:813:29: error: unknown type name 'uint'
2/ In musl libc, stdout is of type (FILE * const).
Because of the const qualifier, a dark magic cast
must be achieved through uintptr_t.
Fixes:
8205e241b2b0 ("app/testpmd: add missing type to RSS hash commands")
Fixes:
e977e4199a8d ("app/testpmd: add commands to load/unload BPF filters")
Cc: stable@dpdk.org
Signed-off-by: Natanael Copa <ncopa@alpinelinux.org>
Reviewed-by: Morten Brørup <mb@smartsharesystems.com>
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Acked-by: David Marchand <david.marchand@redhat.com>
Thomas Monjalon [Fri, 19 Mar 2021 09:20:42 +0000 (10:20 +0100)]
event/dlb: fix header includes for musl
The header file fcntl.h should not be included from sys/ directory
as done in dlb drivers, it is an error with musl libc.
Fixes:
19980083fd57 ("event/dlb: add eventdev probe")
Fixes:
5433956d5185 ("event/dlb2: add eventdev probe")
Cc: stable@dpdk.org
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Acked-by: David Marchand <david.marchand@redhat.com>
Acked-by: Timothy McDaniel <timothy.mcdaniel@intel.com>
Thomas Monjalon [Wed, 24 Feb 2021 23:05:10 +0000 (00:05 +0100)]
vdpa/mlx5: replace pthread functions unavailable in musl
1/ The function pthread_yield() does not exist in musl libc,
and can be replaced with sched_yield() after including sched.h.
2/ The function pthread_attr_setaffinity_np() does not exist in musl libc,
and can be replaced with pthread_setaffinity_np() after pthread_create().
Fixes:
b7fa0bf4d5c6 ("vdpa/mlx5: fix polling threads scheduling")
Fixes:
5cf3fd3af4df ("vdpa/mlx5: add CPU core parameter to bind polling thread")
Cc: stable@dpdk.org
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Matan Azrad <matan@nvidia.com>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Acked-by: David Marchand <david.marchand@redhat.com>
Thomas Monjalon [Thu, 25 Feb 2021 09:45:09 +0000 (10:45 +0100)]
net/igc: remove use of uint type
Improve portability (especially with musl libc)
by replacing the non-standard type 'uint' with 'size_t'.
Fixes:
746664d546fb ("net/igc: support flow API")
Cc: stable@dpdk.org
Suggested-by: David Marchand <david.marchand@redhat.com>
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Haiyue Wang <haiyue.wang@intel.com>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Acked-by: David Marchand <david.marchand@redhat.com>
Natanael Copa [Thu, 5 Nov 2020 21:17:10 +0000 (22:17 +0100)]
net/cxgbe: remove use of uint type
Improve portability by replacing non-standard 'uint' with 'unsigned int'.
This solves the build error with musl libc:
In file included from ../drivers/net/cxgbe/cxgbe.h:9,
from ../drivers/net/cxgbe/cxgbe_ethdev.c:37:
../drivers/net/cxgbe/base/common.h:201:4: error: unknown type name 'uint'
201 | uint synmapen:1; /* SYN Map Enable */
| ^~~~
Fixes:
bfcb257d3014 ("net/cxgbe: enable RSS for VF")
Cc: stable@dpdk.org
Signed-off-by: Natanael Copa <ncopa@alpinelinux.org>
Reviewed-by: Morten Brørup <mb@smartsharesystems.com>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Acked-by: David Marchand <david.marchand@redhat.com>
Thomas Monjalon [Thu, 5 Nov 2020 21:17:11 +0000 (22:17 +0100)]
bus/pci: support I/O port operations with musl
Add a fallback for non-GNU libc systems like musl libc for the
non-standard functions outl_p, outw_p and outb_p.
It solves the following errors when building with musl libc:
pci_uio.c: undefined reference to 'outw_p'
pci_uio.c: undefined reference to 'outl_p'
pci_uio.c: undefined reference to 'outb_p'
Bugzilla ID: 35
Fixes:
756ce64b1ecd ("eal: introduce PCI ioport API")
Cc: stable@dpdk.org
Reported-by: Natanael Copa <ncopa@alpinelinux.org>
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Acked-by: David Marchand <david.marchand@redhat.com>
Natanael Copa [Thu, 5 Nov 2020 21:17:12 +0000 (22:17 +0100)]
bus/dpaa: fix build with musl
The header files argp.h and error.h do not exist in musl libc.
Fix build with musl libc by using err(3) instead of
the GNU-specific error(3).
We could have used the identical errx("...: %s", strerror(ret))` but
strerror(3) is not thread-safe and the strerror_r variant has two
incompatible versions, one GNU specific and one XSI-compliant.
Avoid the mess by letting "err" use the thread-local errno.
This also fixes error message for kzmalloc failures which previously
would always have given "Unknown error -1", since that is what
strerror(-1) returns. Let "err" use the proper error message from errno
which is set by kzalloc.
Fixes:
9d32ef0f5d61 ("bus/dpaa: support creating dynamic HW portal")
Fixes:
f09ede6c8fd1 ("bus/dpaa: add BMAN driver core")
Fixes:
5b22cf744689 ("bus/dpaa: introducing FMan configurations")
Fixes:
39f373cf015a ("bus/dpaa: add compatibility and helper macros")
Cc: stable@dpdk.org
Signed-off-by: Natanael Copa <ncopa@alpinelinux.org>
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Acked-by: David Marchand <david.marchand@redhat.com>
Natanael Copa [Thu, 5 Nov 2020 21:17:13 +0000 (22:17 +0100)]
bus/dpaa: fix 64-bit arch detection
There is no standard saying that __WORDSIZE should be defined or in
what include it should be defined. Use RTE_ARCH_64 instead.
This solves a warning when building with musl libc:
warning: "__WORDSIZE" is not defined, evaluates to 0 [-Wundef]
Fixes:
847ee3bd0d1f ("bus/dpaa: support FMAN frame queue lookup")
Cc: stable@dpdk.org
Signed-off-by: Natanael Copa <ncopa@alpinelinux.org>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Acked-by: David Marchand <david.marchand@redhat.com>
Natanael Copa [Thu, 5 Nov 2020 21:17:15 +0000 (22:17 +0100)]
common/dpaax/caamflib: simplify log macros
Simplify pr_debug, pr_err and pr_warn macros by adding an intermediate
pr_msg macro. This way we only need to test for SUPPRESS_PRINTS once.
Signed-off-by: Natanael Copa <ncopa@alpinelinux.org>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Acked-by: David Marchand <david.marchand@redhat.com>
Natanael Copa [Thu, 5 Nov 2020 21:17:14 +0000 (22:17 +0100)]
common/dpaax/caamflib: fix build with musl
The swab16/swab32/swab64 are Linux specific and not GNU libc specific.
Keep the check for __GLIBC__ just in case other GNU systems depends on
this (Hurd or GNU/kFreeBSD).
This fixes a build error with musl libc.
Fixes:
04711d41a872 ("crypto/dpaa2_sec: add run-time assembler for descriptor")
Cc: stable@dpdk.org
Signed-off-by: Natanael Copa <ncopa@alpinelinux.org>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Acked-by: David Marchand <david.marchand@redhat.com>
Thomas Monjalon [Thu, 25 Feb 2021 00:07:56 +0000 (01:07 +0100)]
drivers: replace page size definitions with function
The page size is often retrieved from the macro PAGE_SIZE.
If PAGE_SIZE is not defined, it is either using hard coded default,
or getting the system value from the UNIX-only function sysconf().
Such definitions are replaced with the generic function
rte_mem_page_size() defined for each supported OS.
Removing PAGE_SIZE definitions will fix dlb drivers for musl libc,
because #ifdef checks were missing, causing redefinition errors.
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Andrew Boyer <aboyer@pensando.io>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Acked-by: David Marchand <david.marchand@redhat.com>
Acked-by: Timothy McDaniel <timothy.mcdaniel@intel.com>
Thomas Monjalon [Wed, 24 Feb 2021 23:21:58 +0000 (00:21 +0100)]
eal: fix build with musl
In musl libc, cpu_set_t is defined only if _GNU_SOURCE is defined.
In case _GNU_SOURCE is undefined, as in eal_common_errno.c,
it was not possible to include rte_os.h which uses cpu_set_t.
This limitation is removed: if CPU_SETSIZE is not defined,
cpu_set_t related definitions and functions are skipped.
Note: such definitions are unneeded in eal_common_errno.c.
Applications which do not define _GNU_SOURCE may miss cpu_set_t related
features on musl. Such case is detected by RTE_HAS_CPUSET being undefined,
so functions which depend on rte_cpuset_t will be unavailable.
A missing include of fcntl.h is also added.
Bugzilla ID: 35
Fixes:
11b57c698005 ("eal: fix error string function")
Fixes:
176bb37ca6f3 ("eal: introduce internal wrappers for file operations")
Cc: stable@dpdk.org
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Signed-off-by: Natanael Copa <ncopa@alpinelinux.org>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Acked-by: David Marchand <david.marchand@redhat.com>
Thomas Monjalon [Thu, 25 Feb 2021 08:45:06 +0000 (09:45 +0100)]
build: remove redundant _GNU_SOURCE definitions
The feature macro _GNU_SOURCE is defined globally,
but there was some remaining useless settings.
The internal definition in config/meson.build is kept,
all other internal definitions of _GNU_SOURCE are removed,
except in examples, which can be built as external applications.
Note: external applications do not inherit of _GNU_SOURCE.
Fixes:
5d7b673d5fd6 ("mk: build with _GNU_SOURCE defined by default")
Fixes:
28188cee2aa0 ("build: enable BSD features visibility for FreeBSD")
Fixes:
e6cdc54cc0ef ("net/mlx5: add socket server for external tools")
Cc: stable@dpdk.org
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Acked-by: David Marchand <david.marchand@redhat.com>
Thomas Monjalon [Thu, 25 Feb 2021 01:49:19 +0000 (02:49 +0100)]
build: detect execinfo library on Linux
The library execinfo and its header file can be installed on Alpine Linux
where the backtrace feature is not part of musl libc:
apk add libexecinfo-dev
As a consequence, this library should not be restricted to BSD only.
At the same time, the library and header are detected once and added
globally to be linked with any application, internal or external.
Fixes:
9065b1fac65f ("build: fix dependency on execinfo for BSD meson builds")
Cc: stable@dpdk.org
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Acked-by: David Marchand <david.marchand@redhat.com>
Thomas Monjalon [Wed, 24 Feb 2021 22:47:56 +0000 (23:47 +0100)]
buildtools: fix build with busybox
If using busybox for mktemp and awk (as in Alpine),
some bugs prevent the script from running:
1/ It seems busybox mktemp requires the pattern to have at least
6 X and no other suffix.
The same has been fixed for other scripts in the past:
commit
3771edc35438 ("buildtools: fix build for some mktemp")
2/ It seems busybox awk does not accept the regex ^.*{
except if the opening curly brace is escaped.
Fixes:
4c82473412e8 ("build: add internal tag check")
Fixes:
68b1f1cda5b4 ("build: check AVX512 rather than binutils version")
Fixes:
3290ac14eb94 ("buildtools: detect discrepancies for experimental symbols")
Cc: stable@dpdk.org
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Acked-by: David Marchand <david.marchand@redhat.com>
Thomas Monjalon [Wed, 24 Feb 2021 23:16:22 +0000 (00:16 +0100)]
eal: fix comment of OS-specific header files
The same comment is on top of each rte_os.h file.
It is reworded to remove the mention of "future releases".
Fixes:
428eb983f5f7 ("eal: add OS specific header file")
Cc: stable@dpdk.org
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Acked-by: David Marchand <david.marchand@redhat.com>
Viacheslav Ovsiienko [Sun, 7 Mar 2021 11:45:47 +0000 (11:45 +0000)]
net/mlx5: fix Rx metadata leftovers
The Rx metadata might use the metadata register C0 to keep the
values. The same register C0 might be used by kernel for source
vport value handling, kernel uses upper half of the register,
leaving the lower half for application usage.
In the extended metadata mode 1 (dv_xmeta_en devarg is
assigned with value 1) the metadata width is 16 bits only,
the Rx datapath code fetched the entire 32-bit value of the
metadata register and presented one to application. The patch
provides data masking depending on the chosen metadata mode.
Fixes:
6c55b622a956 ("net/mlx5: set dynamic flow metadata in Rx queues")
Cc: stable@dpdk.org
Signed-off-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
Viacheslav Ovsiienko [Sun, 14 Mar 2021 12:13:02 +0000 (12:13 +0000)]
compress/mlx5: support timestamp format
This patch adds support for the timestamp format settings for
the receive and send queues. If the firmware version x.30.1000
or above is installed and the NIC timestamps are configured
with the real-time format, the default zero values for newly
added fields cause the queue creation to fail.
The patch queries the timestamp formats supported by the hardware
and sets the configuration values in queue context accordingly.
Fixes:
8619fcd5161b ("compress/mlx5: support queue pair operations")
Cc: stable@dpdk.org
Signed-off-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
Acked-by: Ori Kam <orika@nvidia.com>
Viacheslav Ovsiienko [Sun, 14 Mar 2021 12:13:01 +0000 (12:13 +0000)]
regex/mlx5: support timestamp format
This patch adds support for the timestamp format settings for
the receive and send queues. If the firmware version x.30.1000
or above is installed and the NIC timestamps are configured
with the real-time format, the default zero values for newly
added fields cause the queue creation to fail.
The patch queries the timestamp formats supported by the hardware
and sets the configuration values in queue context accordingly.
Fixes:
92f2c6a30fe0 ("regex/mlx5: add send queue")
Cc: stable@dpdk.org
Signed-off-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
Acked-by: Ori Kam <orika@nvidia.com>
Viacheslav Ovsiienko [Sun, 14 Mar 2021 12:13:00 +0000 (12:13 +0000)]
vdpa/mlx5: support timestamp format
This patch adds support for the timestamp format settings for
the receive and send queues. If the firmware version x.30.1000
or above is installed and the NIC timestamps are configured
with the real-time format, the default zero values for newly
added fields cause the queue creation to fail.
The patch queries the timestamp formats supported by the hardware
and sets the configuration values in queue context accordingly.
Fixes:
95276abaaf0a ("vdpa/mlx5: introduce Mellanox vDPA driver")
Cc: stable@dpdk.org
Signed-off-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
Acked-by: Ori Kam <orika@nvidia.com>
Viacheslav Ovsiienko [Sun, 14 Mar 2021 12:12:59 +0000 (12:12 +0000)]
net/mlx5: support timestamp format
This patch adds support for the timestamp format settings for
the receive and send queues. If the firmware version x.30.1000
or above is installed and the NIC timestamps are configured
with the real-time format, the default zero values for newly
added fields cause the queue creation to fail.
The patch queries the timestamp formats supported by the hardware
and sets the configuration values in queue context accordingly.
Fixes:
86fc67fc9315 ("net/mlx5: create advanced RxQ object via DevX")
Fixes:
ae18a1ae9692 ("net/mlx5: support Tx hairpin queues")
Fixes:
15c3807e86ab ("common/mlx5: support DevX QP operations")
Cc: stable@dpdk.org
Signed-off-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
Acked-by: Ori Kam <orika@nvidia.com>
Viacheslav Ovsiienko [Sun, 14 Mar 2021 12:12:58 +0000 (12:12 +0000)]
common/mlx5: add timestamp format support to DevX
This patch handles the NIC-supported timestamp formats via DevX.
Two different timestamp formats can be provided potentially.
The free-running format provides opaque values captured from
the internal clock counter fed by some independent oscillator.
The free-running frequency is not pre-defined and should be
queried from the NIC. The real-time timestamps are expressed
in nanoseconds, captured from the dedicated UTC counter, that
can be adjusted on the fly and synchronized with some external
reference clock.
Depending on the version and configuration the hardware might
support either FR (free-running) or RT (real-time) timestamps,
per queue basis.
The commit provides the querying information about the supported
timestamp formats and provides the means to configure ones
at queue creation time.
Fixes:
e2b4925ef7c1 ("net/mlx5: support Direct Rules E-Switch")
Cc: stable@dpdk.org
Signed-off-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
Acked-by: Ori Kam <orika@nvidia.com>
Thomas Monjalon [Tue, 9 Mar 2021 09:48:36 +0000 (10:48 +0100)]
net/mlx5: reduce log level of alignment message
Having to force an alignment does not impact the user,
so it should not be a warning.
The log level is reduced from warning to debug.
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Matan Azrad <matan@nvidia.com>
Thomas Monjalon [Tue, 9 Mar 2021 09:48:35 +0000 (10:48 +0100)]
common/mlx5: remove extra line feed in log messages
The macro DRV_LOG already includes a terminating line feed character
defined in PMD_DRV_LOG_.
The extra line feeds added in some messages are removed.
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Matan Azrad <matan@nvidia.com>
Thomas Monjalon [Tue, 9 Mar 2021 09:48:34 +0000 (10:48 +0100)]
net/mlx5: enable debug logs dynamically
Most debug logs are using DRV_LOG(DEBUG,)
but some were using DEBUG().
The macro DEBUG is doing nothing if not compiled with
RTE_LIBRTE_MLX5_DEBUG.
As it is not used in the data path, the macro DEBUG
can be replaced with DRV_LOG.
Then all debug logs can be enabled at runtime with:
--log-level pmd.net.mlx5:debug
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Matan Azrad <matan@nvidia.com>
Thomas Monjalon [Tue, 9 Mar 2021 09:48:33 +0000 (10:48 +0100)]
net/mlx4: enable debug logs dynamically
The macro DEBUG was doing nothing if not compiled with
RTE_LIBRTE_MLX4_DEBUG.
As it is not used in the data path, it can be always enabled at
compilation time. Then it can be enabled at runtime with:
--log-level pmd.net.mlx4:debug
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Matan Azrad <matan@nvidia.com>
Wenjun Wu [Fri, 26 Feb 2021 07:22:00 +0000 (15:22 +0800)]
net/ice: check some functions return
Fix unchecked return values reported by coverity.
Coverity issue: 349907
Fixes:
03a05924dad0 ("net/ice: support device-specific DDP package loading")
Cc: stable@dpdk.org
Signed-off-by: Wenjun Wu <wenjun1.wu@intel.com>
Xueming Li [Thu, 11 Mar 2021 13:13:32 +0000 (13:13 +0000)]
ethdev: add helper function to get representor ID
The NIC can have multiple PCIe links and can be attached to multiple
hosts, for example the same single NIC can be shared for multiple server
units in the rack. On each PCIe link NIC can provide multiple PFs and
VFs/SFs based on these ones. The full representor identifier consists of
three indices - controller index, PF index, and VF or SF index (if any).
SR-IOV and SubFunction are created on top of PF. PF index is introduced
because there might be multiple PFs in the bonding configuration and
only bonding device is probed.
In eth representor comparator callback, ethdev representor ID was
compared with devarg. Since controller index and PF index not compared,
callback returned representor from other PF or controller.
This patch adds new API to get representor ID from controller, pf and
vf/sf index. Representor comparer callback get representor ID then
compare with device representor ID.
Signed-off-by: Xueming Li <xuemingl@nvidia.com>
Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Xueming Li [Thu, 11 Mar 2021 13:13:31 +0000 (13:13 +0000)]
ethdev: add API to get representor info
The NIC can have multiple PCIe links and can be attached to multiple
hosts, for example the same single NIC can be shared for multiple server
units in the rack. On each PCIe link NIC can provide multiple PFs and
VFs/SFs based on these ones. The full representor identifier consists of
three indices - controller index, PF index, and VF or SF index (if any).
This patch introduces a new API rte_eth_representor_info_get() to
retrieve representor corresponding info mapping:
- caller controller index and pf index.
- supported representor ID ranges.
- type, controller, pf and start vf/sf ID of each range.
The API is useful to calculate representor from devargs to representor
ID.
New ethdev callback representor_info_get() is added to retrieve info
from PMD driver, optional for PMD that doesn't support new devargs
representor syntax.
Signed-off-by: Xueming Li <xuemingl@nvidia.com>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Xueming Li [Thu, 11 Mar 2021 13:13:30 +0000 (13:13 +0000)]
ethdev: support multi-host in representor
The NIC can have multiple PCIe links and can be attached to the multiple
hosts, for example the same single NIC can be shared for multiple server
units in the rack. On each PCIe link NIC can provide multiple PFs and
VFs/SFs based on these ones. To provide the unambiguous identification
of the PCIe function the controller index is added. The full representor
identifier consists of three indices - controller index, PF index, and
VF or SF index (if any).
This patch introduces controller index to ethdev representor syntax,
examples:
[[c#]pf#]vf#: VF port representor/s, example: pf0vf1
[[c#]pf#]sf#: SF port representor/s, example: c1pf1sf[0-3]
c# is controller(host) ID/range in case of multi-host, optional.
For user application (e.g. OVS), PMD is responsible to interpret and
locate representor device based on controller ID, PF ID and VF/SF ID in
representor syntax.
Signed-off-by: Xueming Li <xuemingl@nvidia.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Acked-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Xueming Li [Thu, 11 Mar 2021 13:13:29 +0000 (13:13 +0000)]
ethdev: support PF index in representor
With Kernel bonding, multiple underlying PFs are bonded, VFs come
from different PF, need to identify representor of VFs unambiguously by
adding PF index.
This patch introduces optional 'pf' section to representor devargs
syntax, examples:
representor=pf0vf0 - single VF representor
representor=pf[0-1]sf[0-1023] - SF representors from 2 PFs
PF type representor is supported by using standalone 'pf' section:
representor=pf1 - PF representor
Signed-off-by: Xueming Li <xuemingl@nvidia.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Acked-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Xueming Li [Thu, 11 Mar 2021 13:13:28 +0000 (13:13 +0000)]
kvargs: support multiple lists
This patch updates kvargs parser to support value of multiple lists or
ranges:
k1=v[1,2]v[3-5]
Signed-off-by: Xueming Li <xuemingl@nvidia.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Acked-by: Thomas Monjalon <thomas@monjalon.net>
Xueming Li [Thu, 11 Mar 2021 13:13:27 +0000 (13:13 +0000)]
ethdev: support sub-function representor
SubFunction is a portion of the PCI device, created on demand, a SF
netdev has its own dedicated queues(txq, rxq). A SF netdev supports
eswitch representation offload similar to existing PF and VF
representors.
To support SF representor, this patch introduces new devargs syntax,
examples:
representor=sf0 - single SubFunction representor
representor=sf[1,3,5] - single list
representor=sf[0-3], - single range
representor=sf[0,2-6,8,10-12] - list with singles and ranges
Signed-off-by: Xueming Li <xuemingl@nvidia.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Acked-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Xueming Li [Thu, 11 Mar 2021 13:13:26 +0000 (13:13 +0000)]
ethdev: support new VF representor syntax
Current VF representor syntax:
representor=2 - single representor
representor=[0-3] - single range
To prepare for more representor types, this patch adds compatible VF
representor devargs syntax:
vf#:
representor=vf2 - single representor
representor=vf[1,3,5] - single list
representor=vf[0-3] - single range
representor=vf[0,1,4-7] - list with singles and range
For backwards compatibility, representor "#" is interpreted as "vf#".
Signed-off-by: Xueming Li <xuemingl@nvidia.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Acked-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Xueming Li [Thu, 11 Mar 2021 13:13:25 +0000 (13:13 +0000)]
ethdev: refactor representor port list parsing
To the extended representor syntax which need to reuse the value parsing
function for controller and PF section, this patch refactors the port
list parsing.
Signed-off-by: Xueming Li <xuemingl@nvidia.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Acked-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Xueming Li [Thu, 11 Mar 2021 13:13:24 +0000 (13:13 +0000)]
ethdev: introduce representor type
To support more representor type, this patch introduces representor type
enum. The enum is subject to be extended to support new representor in
patches upcoming.
For each devarg structure, only one type supported.
Signed-off-by: Xueming Li <xuemingl@nvidia.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Acked-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
Vijay Kumar Srivastava [Tue, 16 Mar 2021 08:58:32 +0000 (11:58 +0300)]
common/sfc_efx: move function to get family
Move function to get efx family from net driver into common driver.
Signed-off-by: Vijay Kumar Srivastava <vsrivast@xilinx.com>
Signed-off-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Vijay Kumar Srivastava [Tue, 16 Mar 2021 08:58:31 +0000 (11:58 +0300)]
net/sfc: skip driver probe for incompatible device class
Driver would be probed only for the net device class.
Signed-off-by: Vijay Kumar Srivastava <vsrivast@xilinx.com>
Signed-off-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Vijay Kumar Srivastava [Tue, 16 Mar 2021 08:58:30 +0000 (11:58 +0300)]
common/sfc_efx: support getting device class
Device class argument would be used to select compatible driver.
Driver probe would be skipped for incompatible device class.
Signed-off-by: Vijay Kumar Srivastava <vsrivast@xilinx.com>
Signed-off-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Vijay Kumar Srivastava [Tue, 16 Mar 2021 08:58:29 +0000 (11:58 +0300)]
common/sfc_efx/base: support verifying virtio features
Add an API to verify virtio features supported by device.
Signed-off-by: Vijay Kumar Srivastava <vsrivast@xilinx.com>
Signed-off-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Vijay Kumar Srivastava [Tue, 16 Mar 2021 08:58:28 +0000 (11:58 +0300)]
common/sfc_efx/base: support getting virtio features
Add an API to get virtio features supported by device.
Signed-off-by: Vijay Kumar Srivastava <vsrivast@xilinx.com>
Signed-off-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Vijay Kumar Srivastava [Tue, 16 Mar 2021 08:58:27 +0000 (11:58 +0300)]
common/sfc_efx/base: add virtio build dependency
Add EFSYS_HAS_UINT64 build dependency on EFSYS_OPT_VIRTIO.
virtio features are represented as bitmask in 64-bit unsigned
integer.
Signed-off-by: Vijay Kumar Srivastava <vsrivast@xilinx.com>
Signed-off-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Vijay Srivastava [Tue, 16 Mar 2021 08:58:26 +0000 (11:58 +0300)]
common/sfc_efx/base: support getting virtq doorbell offset
Add an API to query the virtqueue doorbell offset in the BAR for a VI.
For vDPA, the virtio net driver notifies the device directly by writing
doorbell. This API would be invoked from vDPA client driver.
Signed-off-by: Vijay Srivastava <vijays@solarflare.com>
Signed-off-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Vijay Srivastava [Tue, 16 Mar 2021 08:58:25 +0000 (11:58 +0300)]
common/sfc_efx/base: add base virtio support for vDPA
In the vDPA mode, only data path is offloaded in the hardware and
control path still goes through the hypervisor and it configures
virtqueues via vDPA driver so new virtqueue APIs are required.
Implement virtio init/fini and virtqueue create/destroy APIs.
Signed-off-by: Vijay Srivastava <vijays@solarflare.com>
Signed-off-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Kalesh AP [Tue, 16 Mar 2021 06:51:36 +0000 (12:21 +0530)]
app/testpmd: check MAC address query
This patch checks return value for rte_eth_dev_info_get() in show_macs().
Coverity issue: 353629
Fixes:
e1d44d0ad623 ("app/testpmd: show MAC addresses added to a port")
Cc: stable@dpdk.org
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Acked-by: Xiaoyun Li <xiaoyun.li@intel.com>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Somnath Kotur [Fri, 12 Mar 2021 05:21:09 +0000 (10:51 +0530)]
net/bnxt: fix Rx and Tx timestamps
timesync adjust and write_time APIs needed to account for the Rx and Tx
timestamp counters as well. Fix it since it was not done earlier.
Fixes:
b11cceb83a34 ("net/bnxt: support timesync")
Cc: stable@dpdk.org
Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Somnath Kotur [Fri, 12 Mar 2021 05:21:08 +0000 (10:51 +0530)]
net/bnxt: fix xstats get
Fix to return count in xstats get op in all cases.
Driver was returning 0 if the 'xstats' parameter being passed to
xstats_get_op was NULL. This won't work on some applications that
rely on a valid count being passed even in this case so that it can
allocate memory accordingly followed by a reissue of the xstats_get_op
to get the actual stats populated by the driver.
Fixes:
063e59ddd28e ("net/bnxt: fix crash in xstats get")
Cc: stable@dpdk.org
Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Lance Richardson [Sat, 6 Mar 2021 15:19:11 +0000 (10:19 -0500)]
net/bnxt: optimize Tx completion handling
Avoid copying mbuf pointers to separate array for bulk
mbuf free when handling transmit completions for vector
mode transmit.
Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Ajit Khaparde [Thu, 11 Mar 2021 23:30:33 +0000 (15:30 -0800)]
net/bnxt: rename a member to avoid conflict
Address build issues with Clang and without glibc on ppc64le.
Vector can be a keyword and should not be used in code.
Renaming it to avoid conflict.
Reported-by: Piotr Kubaj <pkubaj@freebsd.org>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Kalesh AP [Wed, 10 Mar 2021 07:50:54 +0000 (13:20 +0530)]
net/bnxt: mute some failure logs
In the init path, driver ignores few of the HWRM command failures.
There is no need to log the error message in those cases.
Fixes:
3fb93bc7c349 ("net/bnxt: initialize parent PF information")
Fixes:
4e3f887bec4b ("net/bnxt: support HWRM port PHY qcaps")
Cc: stable@dpdk.org
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Kalesh AP [Wed, 10 Mar 2021 07:50:53 +0000 (13:20 +0530)]
net/bnxt: fix HWRM and FW incompatibility handling
Fix to return an error when the HWRM version that the driver
is compiled against is incompatible with the FW that is actually
running on the card. This is determined based on the req length
indicated by FW against the value supported in the HWRM.
Fixes:
804e746c7b73 ("net/bnxt: add hardware resource manager init code")
Cc: stable@dpdk.org
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Kalesh AP [Thu, 4 Mar 2021 09:07:28 +0000 (14:37 +0530)]
net/bnxt: fix VF info allocation
1. Renamed bnxt_hwrm_alloc_vf_info()/bnxt_hwrm_free_vf_info to
bnxt_alloc_vf_info()/bnxt_free_vf_info as it does not
issue any HWRM command to fw.
2. Fix missing unlock when memory allocation fails.
Fixes:
b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF")
Cc: stable@dpdk.org
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Kalesh AP [Thu, 4 Mar 2021 09:07:27 +0000 (14:37 +0530)]
net/bnxt: fix device readiness check
Fix HWRM_VER_GET command to handle DEV_NOT_RDY state.
Driver should fail probe if the device is not ready.
Conversely, the HWRM_VER_GET poll after reset can safely
retry until the existing timeout is exceeded.
Fixes:
804e746c7b73 ("net/bnxt: add hardware resource manager init code")
Cc: stable@dpdk.org
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Ajit Khaparde [Thu, 4 Mar 2021 09:07:26 +0000 (14:37 +0530)]
net/bnxt: check flush status during ring free
When host SW issues a HWRM_RING_FREE for Tx/Rx/AGG ring in HW,
the FW flushes the BDs associated with the ring and performs other
cleanup in the HW. The host software should ideally check for an
indication from the FW indicating this step has been completed
successfully to avoid unexpected errors during cleanup.
The FW issues a HWRM_DONE response to the RING_FREE request on
the corresponding CQ ring. Poll the CQs during cleanup and
ensure the HWRM_FREE command is completed not just based on the
value of valid bit but also the HWRM_DONE response for the ring.
If the HWRM_DONE response is not seen, force the cleanup to
complete just based on the valid bit.
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Lance Richardson [Tue, 2 Mar 2021 15:16:08 +0000 (10:16 -0500)]
net/bnxt: refactor mbuf pointer reset
Remove code for setting consumed mbuf pointers to NULL from the
vector receive functions as a minor performance optimization.
Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Lance Richardson [Tue, 2 Mar 2021 14:29:27 +0000 (09:29 -0500)]
net/bnxt: fix Rx descriptor status
Fix a number of issues in the bnxt receive descriptor status
function, including:
- Provide status of receive descriptor instead of completion
descriptor.
- Remove invalid comparison of raw ring index with masked ring
index.
- Correct misinterpretation of offset parameter as ring index.
- Correct misuse of completion ring index for mbuf ring (the
two rings have different sizes).
Fixes:
0fe613bb87b2 ("net/bnxt: support Rx descriptor status")
Cc: stable@dpdk.org
Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
Reviewed-by: Andy Gospodarek <gospo@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Kalesh AP [Wed, 24 Feb 2021 15:55:53 +0000 (21:25 +0530)]
net/bnxt: fix PTP support for Thor
On Thor, Rx timestamp is present in the Rx completion record.
Only 32 bits of the timestamp is present in the completion.
The driver needs to periodically poll the current 48 bit
free running timer using the HWRM_PORT_TS_QUERY command.
It can combine the upper 16 bits from the HWRM response
with the lower 32 bits in the Rx completion to produce
the 48 bit timestamp for the Rx packet.
This patch adds an alarm thread to periodically poll the current 48 bit
free running timer using the HWRM_PORT_TS_QUERY command.
This avoids issuing the hwrm command from the Rx handler.
This patch also handles the timer roll over condition.
Fixes:
6cbd89f9f3d8 ("net/bnxt: support PTP for Thor")
Cc: stable@dpdk.org
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Kalesh AP [Wed, 24 Feb 2021 15:55:52 +0000 (21:25 +0530)]
net/bnxt: fix FW readiness check during recovery
Moved fw readiness check to a new routine bnxt_check_fw_ready().
During error recovery, driver needs to wait for fw readiness.
For that, it uses bnxt_hwrm_ver_get() function now and that
function does parsing of the VER_GET response as well.
Added a new lightweight function bnxt_hwrm_poll_ver_get() for polling
the firmware readiness which issues VER_GET and checks for success
without processing the command response.
Fixes:
df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW")
Cc: stable@dpdk.org
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Kalesh AP [Wed, 24 Feb 2021 15:55:51 +0000 (21:25 +0530)]
net/bnxt: fix firmware fatal error handling
During some fatal firmware error conditions, the PCI config space
register 0x2e which normally contains the subsystem ID will become
0xffff. This register will revert back to the normal value after
the chip has completed core reset. If we detect this condition,
we can poll this config register immediately for the value to revert.
Because we use config read cycles to poll this register, there is no
possibility of Master Abort if we happen to read it during core reset.
This speeds up recovery significantly as we don't have to wait for the
conservative min_time before polling to see if the firmware has come
out of reset. As soon as this register changes value we can proceed
to re-initialize the device.
Fixes:
df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW")
Cc: stable@dpdk.org
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>