Added unit test for adding queues to Rx adapter in bursts using
rte_event_eth_rx_adapter_queues_add().
Signed-off-by: Shijith Thotton
---
app/test/test_event_eth_rx_adapter.c | 86
1 file changed, 86 insertions(+)
diff --git a/app/test/test_event_eth_rx_adapter.c
Implemented PMD support for the eventdev PMD operation to burst add
queues to the Rx adapter.
Signed-off-by: Shijith Thotton
---
drivers/event/cnxk/cn10k_eventdev.c | 82 --
drivers/event/cnxk/cn20k_eventdev.c | 195 ---
drivers/event/cnxk/cn9k_eventdev.c
apply specific configurations to each of the added queues.
A new internal PMD operation, eventdev_eth_rx_adapter_queues_add_t, has
been introduced to enable this functionality. It takes an array of
receive queue IDs along with their corresponding queue configurations.
Signed-off-by: Shijith Thotton
/
v2:
- Updated documentation to explain the new API.
v3:
- Improved MULTI_EVENTQ check.
Shijith Thotton (3):
eventdev/eth_rx: add API to burst add queues to Rx adapter
event/cnxk: enable PMD op to burst add queues to Rx adapter
test/event: unit test to burst add Rx queues to adapter
app
y of receive
>> queue
>> IDs along with their corresponding queue configurations.
>>
>> Signed-off-by: Shijith Thotton
>> ---
>> .../eventdev/event_ethernet_rx_adapter.rst| 60 +--
>> lib/eventdev/eventdev_pmd.h | 34
>> lib/ev
Added unit test for adding queues to Rx adapter in bursts using
rte_event_eth_rx_adapter_queues_add().
Signed-off-by: Shijith Thotton
---
app/test/test_event_eth_rx_adapter.c | 86
1 file changed, 86 insertions(+)
diff --git a/app/test/test_event_eth_rx_adapter.c
Implemented PMD support for the eventdev PMD operation to burst add
queues to the Rx adapter.
Signed-off-by: Shijith Thotton
---
drivers/event/cnxk/cn10k_eventdev.c | 82 --
drivers/event/cnxk/cn20k_eventdev.c | 195 ---
drivers/event/cnxk/cn9k_eventdev.c
specific configurations to each of the added queues.
A new internal PMD operation, eventdev_eth_rx_adapter_queues_add_t, has
been introduced to enable this functionality. It takes an array of
receive queue IDs along with their corresponding queue configurations.
Signed-off-by: Shijith Thotton
/
v1:
- Updated documentation to explain the new API.
Shijith Thotton (3):
eventdev/eth_rx: add API to burst add queues to Rx adapter
event/cnxk: enable PMD op to burst add queues to Rx adapter
test/event: unit test to burst add Rx queues to adapter
app/test/test_event_eth_rx_adap
configurations.
>
>Signed-off-by: Shijith Thotton
>---
> lib/eventdev/eventdev_pmd.h | 34 ++
> lib/eventdev/eventdev_trace.h | 14 +++
> lib/eventdev/eventdev_trace_points.c| 3 +
> lib/eventdev/rte_event_eth_rx_adapter.c | 146
>+++
>
> Documentation update is missing.
>
Missed it, I will update in v2.
Thanks,
Shijith
>This patch series introduces the rte_event_eth_rx_adapter_queues_add()
>API, allowing multiple Rx queues to be added to an Rx adapter in a
>single burst. This enhancement benefits applications that require bu
Implemented PMD support for the eventdev PMD operation to burst add
queues to the Rx adapter.
Signed-off-by: Shijith Thotton
---
drivers/event/cnxk/cn10k_eventdev.c | 82 --
drivers/event/cnxk/cn20k_eventdev.c | 195 ---
drivers/event/cnxk/cn9k_eventdev.c
Added unit test for adding queues to Rx adapter in bursts using
rte_event_eth_rx_adapter_queues_add().
Signed-off-by: Shijith Thotton
---
app/test/test_event_eth_rx_adapter.c | 86
1 file changed, 86 insertions(+)
diff --git a/app/test/test_event_eth_rx_adapter.c
specific configurations to each of the added queues.
A new internal PMD operation, eventdev_eth_rx_adapter_queues_add_t, has
been introduced to enable this functionality. It takes an array of
receive queue IDs along with their corresponding queue configurations.
Signed-off-by: Shijith Thotton
adding Rx
queues has been implemented. The second patch in the series enables this
operation for the CNXK PMD.
Additionally, a unit test has been added to validate the API's
functionality.
RFC:
https://patches.dpdk.org/project/dpdk/patch/20241219073405.1724200-1-sthot...@marvell.com/
Sh
;)
Cc: sta...@dpdk.org
Signed-off-by: Shijith Thotton
---
app/test/test_event_eth_rx_adapter.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/app/test/test_event_eth_rx_adapter.c
b/app/test/test_event_eth_rx_adapter.c
index 0233c87779..9a60e34352 100644
---
d, int32_t
>> >> >> rx_queue_id[],
>> >> >> const struct
>> >> >> rte_event_eth_rx_adapter_queue_conf conf[],
>> >> >> uint16_t nb_rx_queues); rx_queues_id is an
>&
ll use conf[i].
>> >>
>> >
>> >Ok
>> >
>> >> > Since this is still not possible with the proposed API, the
>> >> >existing queue_add API needs to be used with specific queue_ids and
>> >> >their configurations.
>> &
plore deprecation as an alternative.
>> >> >
>> >>
>> >> Please let me know if you have any suggestions/feedback on what I
>> >> said above.
>> >
>> >Still feel the new proposed API can be av
ding some new features.
>
>> If not, I can go ahead and send v1.
>>
>> >>> The proposed API, rte_event_eth_rx_adapter_queues_add, addresses
>> >>> this limitation by:
>> >>>
>> >>> - Enabling users to specify an array o
o
>>> configure. When set to 0, the API applies a common configuration to
>>> all queues, similar to the existing rx_queue_id = -1 behavior.
>>>
>>> This enhancement allows for more granular control when configuring
>multiple
>>> Rx queues. Additio
e. When set to 0, the API applies a common configuration to
>> all queues, similar to the existing rx_queue_id = -1 behavior.
>>
>> This enhancement allows for more granular control when configuring multiple
>> Rx queues. Additionally, the API can act as a replacement for the
ontrol when configuring
>multiple Rx queues. Additionally, the API can act as a replacement for
>the older API, offering both flexibility and improved functionality.
>
>Signed-off-by: Shijith Thotton
>---
> lib/eventdev/eventdev_pmd.h | 34 +
&g
a replacement for
the older API, offering both flexibility and improved functionality.
Signed-off-by: Shijith Thotton
---
lib/eventdev/eventdev_pmd.h | 34 +
lib/eventdev/rte_event_eth_rx_adapter.h | 34 +
2 files changed, 68 insertions
a replacement for
the older API, offering both flexibility and improved functionality.
Signed-off-by: Shijith Thotton
---
lib/eventdev/eventdev_pmd.h | 34 +
lib/eventdev/rte_event_eth_rx_adapter.h | 34 +
2 files changed, 68 insertions
: Shijith Thotton
---
v4:
- Keep trailing comma.
v3:
- Updated documentation.
v2:
- Don't override the default value of RTE_IOVA_IN_MBUF in CNXK platform.
- Print a warning message if IOVA in mbuf is enabled on CNXK platform.
config/arm/meson.build | 4
doc/guides/platform/cnx
: Shijith Thotton
---
v3:
- Updated documentation.
v2:
- Don't override the default value of RTE_IOVA_IN_MBUF in CNXK platform.
- Print a warning message if IOVA in mbuf is enabled on CNXK platform.
config/arm/meson.build | 8 ++--
doc/guides/platform/cnxk.rst
>On Fri, Oct 25, 2024 at 03: 20: 40PM +0530, Shijith Thotton wrote: > The value
>of RTE_IOVA_IN_MBUF has always been disabled on CNXK platforms, > as
>IOVA in the mbuf is unnecessary. This update changes that behavior to >
>respect the value
>
>On Fri, Oct 25, 2024 a
: Shijith Thotton
---
v2:
- Don't override the default value of RTE_IOVA_IN_MBUF in CNXK platform.
- Print a warning message if IOVA in mbuf is enabled on CNXK platform.
config/arm/meson.build | 8 ++--
drivers/common/cnxk/meson.build | 6 ++
2 files changed, 8 insertions(
build option
>> >> > -Denable_iova_as_pa=true is explicitly specified.
>> >> >
>> >> > Signed-off-by: Shijith Thotton
>> >> > ---
>> >> >
>> >> > diff --git a/config/arm/meson.build b/config/arm/meson.bu
>On Thu, Oct 24, 2024 at 04:40:40PM +0530, Jerin Jacob wrote:
>> On Mon, Oct 14, 2024 at 4:37 PM Shijith Thotton
>wrote:
>> >
>> > Value of RTE_IOVA_IN_MBUF was always disabled on cnxk platforms, as
>IOVA
>> > in the mbuf is not required. This ch
Value of RTE_IOVA_IN_MBUF was always disabled on cnxk platforms, as IOVA
in the mbuf is not required. This change modifies that behavior,
allowing RTE_IOVA_IN_MBUF to be enabled if the build option
-Denable_iova_as_pa=true is explicitly specified.
Signed-off-by: Shijith Thotton
---
config/arm
mbufs. This patch fixes the same.
see https://bugs.dpdk.org/show_bug.cgi?id=1472
Bugzilla ID: 1472
Fixes: 03b57eb7ab9a ("mbuf: add second dynamic field member")
Cc: sta...@dpdk.org
Signed-off-by: Shijith Thotton
Reviewed-by: Morten Brørup
Acked-by: Stephen Hemminger
---
v2:
- Mov
mbufs. This patch fixes the same.
see https://bugs.dpdk.org/show_bug.cgi?id=1472
Bugzilla ID: 1472
Fixes: 03b57eb7ab9a ("mbuf: add second dynamic field member")
Cc: sta...@dpdk.org
Signed-off-by: Shijith Thotton
---
lib/mbuf/rte_mbuf.h | 3 +++
1 file changed, 3 insertions(+)
diff -
>From: Pavan Nikhilesh
>
>Prevent mempool exhaustion due to elements being stuck in lcore
>local caches.
>
>Signed-off-by: Pavan Nikhilesh
Acked-by: Shijith Thotton
>---
> app/test-eventdev/test_perf_common.c | 11 ++-
> 1 file changed, 6 insertions(+), 5 d
>From: Pavan Nikhilesh
>
>Use the `rte_event_enqueue_new_burst` routine to enqueue events
>with rte_event::op as RTE_EVENT_OP_NEW. This allows PMDs to use
>optimized enqueue routines.
>
>Signed-off-by: Pavan Nikhilesh []
Acked-by: Shijith Thotton
>---
> app/
>From: Pavan Nikhilesh
>
>Use LMTST when all events in the burst are enqueue with
>rte_event:op as RTE_EVENT_OP_NEW i.e. events are enqueued
>with the `rte_event_enqueue_new_burst` API.
>
>Signed-off-by: Pavan Nikhilesh
Acked-by: Shijith Thotton
>---
>v2
Even when no events are dequeued, it is important to update event
attributes as they provide valuable information for checking the work
slot state and other attributes.
Fixes: c0a9774f030d ("event/cnxk: move post-processing to separate function")
Signed-off-by: Shijith Thotton
--
rypto: support runtime set/get parameters")
Signed-off-by: Shijith Thotton
---
app/test/test_event_crypto_adapter.c | 10 --
1 file changed, 8 insertions(+), 2 deletions(-)
diff --git a/app/test/test_event_crypto_adapter.c
b/app/test/test_event_crypto_adapter.c
index 264d6f731e..
behavior based on the
type of timer being used (software or hardware).
Fixes: d1f3385d0076 ("test: add event timer adapter auto-test")
Signed-off-by: Shijith Thotton
---
v2:
- Updated commit message and added fixed tag.
app/test/test_event_timer_adapter.c | 169 +++---
Replaced sleep with a function to dequeue timer expiry events until
either the expected number of events have been dequeued or the specified
time has elapsed. This change closely esembles the behavior of the
tested systems.
Signed-off-by: Shijith Thotton
---
app/test/test_event_timer_adapter.c
uot;)
Signed-off-by: Shijith Thotton
---
drivers/common/cnxk/roc_tim.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/common/cnxk/roc_tim.c b/drivers/common/cnxk/roc_tim.c
index 6f256c60fa..f8607b2852 100644
--- a/drivers/common/cnxk/roc_tim.c
+++ b/drivers/common/cnxk/
need to be skipped earlier
>to avoid defining the sub-libraries.
>
>Fixes: a986c2b7973d ("build: add option to configure IOVA mode as PA")
>Cc: sta...@dpdk.org
>
>Signed-off-by: Thomas Monjalon
>---
Acked-by: Shijith Thotton
--prod_type_cryptodev --crypto_adptr_mode 1 --test=perf_atq \
>--stlist=a --wlcores 1 --plcores 2 --prod_enq_burst_sz 32
>
>Signed-off-by: Volodymyr Fialko
Acked-by: Shijith Thotton
>---
> app/test-eventdev/test_perf_common.c | 235
>++-
> doc/guides
wlcores 1 --plcores 2 --prod_enq_burst_sz 32 \
>--enable_vector --vector_tmo_ns 0 --nb_flows 2
>
>Signed-off-by: Volodymyr Fialko
Acked-by: Shijith Thotton
>---
> app/test-eventdev/test_perf_atq.c| 62 ++--
> app/test-eventdev/test_perf_common.c | 68 +++---
. Modified code to match this logic.
Fixes: a1793ee8aba0 ("eventdev/eth_rx: add adapter instance get API")
Signed-off-by: Shijith Thotton
---
lib/eventdev/rte_event_eth_rx_adapter.c | 12
1 file changed, 4 insertions(+), 8 deletions(-)
diff --git a/li
Arm a bucket early to account for software delays in timer arm routine.
Signed-off-by: Shijith Thotton
---
drivers/event/cnxk/cnxk_tim_worker.h | 8 +---
1 file changed, 5 insertions(+), 3 deletions(-)
diff --git a/drivers/event/cnxk/cnxk_tim_worker.h
b/drivers/event/cnxk
Fixed function call arguments of l2fwd worker cleanup. Number of events
enqueued and dequeued are set the opposite.
Fixes: 622ebb6b4a7d ("examples/l2fwd-event: clean up worker state before exit")
Signed-off-by: Shijith Thotton
---
examples/l2fwd-event/l2fwd_event.c | 4 ++--
1 file
Hi Jingjing Wu/Beilei Xing.
>I guess driver may not handle the attribute enable_iova_as_pa well right now.
>Maybe you can have a try by disabling idpf driver by adding "-
>Ddisable_drivers=net/idpf".
>
Please send a fix. A check can be added similar to hns3 PMD.
+if dpdk_conf.get('RTE_IOVA_AS_P
: Shijith Thotton
---
drivers/common/cnxk/roc_dev.c | 12 ++---
drivers/common/cnxk/roc_dev_priv.h | 5 ++
drivers/common/cnxk/roc_irq.c | 11
drivers/common/cnxk/roc_sso.c | 78 -
drivers/common/cnxk/roc_sso.h | 4 +-
drivers/common/cnxk
>
>Fix stack overflow when event vector size is greater than
>MAX_BURST_SIZE.
>Add missing mac swap and rfc1812 stage.
>
>Fixes: e8adca1951d4 ("examples/l3fwd: support event vector")
>
>Signed-off-by: Pavan Nikhilesh
Acked-by: Shijith Thotton
>-
>
>Use lpm vector path to process event vector.
>
>Signed-off-by: Pavan Nikhilesh
Acked-by: Shijith Thotton
>---
> examples/l3fwd/l3fwd_altivec.h | 29 ++
> examples/l3fwd/l3fwd_event.h | 71
>++
> examples
>
>Split packet processing from packet send stage, as send stage
>is not common for poll and event mode.
>
>Signed-off-by: Pavan Nikhilesh
Acked-by: Shijith Thotton
>---
> examples/l3fwd/l3fwd_em_hlm.h | 39 +++---
> examples/l3fwd/
>
>Fix port group mask generation in altivec, vec_any_eq returns
>0 or 1 while port_groupx4 expects comparison mask result.
>
>Fixes: 2193b7467f7a ("examples/l3fwd: optimize packet processing on powerpc")
>Cc: sta...@dpdk.org
>
>Signed-off-by: Pavan Nik
0; i != j; i += FWDSTEP)
>+ processx4_step3(&vec->mbufs[i], &dst_ports[i]);
>+ for (; i < vec->nb_elem; i++)
>+ process_packet(vec->mbufs[i], &dst_ports[i]);
>+
>+ process_event_vector(vec, dst_ports);
> }
>
> #endif /* __L3FWD_EM_SEQUENTIAL_H__ */
>diff --git a/examples/l3fwd/l3fwd_event.h b/examples/l3fwd/l3fwd_event.h
>index 3fe38aada0..e21817c36b 100644
>--- a/examples/l3fwd/l3fwd_event.h
>+++ b/examples/l3fwd/l3fwd_event.h
>@@ -103,27 +103,6 @@ process_dst_port(uint16_t *dst_ports, uint16_t
>nb_elem)
> }
> #endif
>
>-static inline void
>-event_vector_attr_validate(struct rte_event_vector *vec, struct rte_mbuf
>*mbuf)
>-{
>- /* l3fwd application only changes mbuf port while processing */
>- if (vec->attr_valid && (vec->port != mbuf->port))
>- vec->attr_valid = 0;
>-}
>-
>-static inline void
>-event_vector_txq_set(struct rte_event_vector *vec, uint16_t txq)
>-{
>- if (vec->attr_valid) {
>- vec->queue = txq;
>- } else {
>- int i;
>-
>- for (i = 0; i < vec->nb_elem; i++)
>- rte_event_eth_tx_adapter_txq_set(vec->mbufs[i], txq);
>- }
>-}
>-
> static inline uint16_t
> filter_bad_packets(struct rte_mbuf **mbufs, uint16_t *dst_port,
> uint16_t nb_pkts)
>--
>2.25.1
For the series:
Acked-by: Shijith Thotton
>> Swapped position of mbuf next pointer and second dynamic field (dynfield2)
>> if the build is configured to disable IOVA as PA. This is to move the
>> mbuf next pointer to first cache line.
>>
>> Signed-off-by: Shijith Thotton
>> Acked-by: Olivier Matz
Enabled software PMDs in IOVA as PA disabled build as they work with
IOVA as VA.
Signed-off-by: Shijith Thotton
Acked-by: Olivier Matz
---
drivers/crypto/armv8/meson.build| 1 +
drivers/crypto/ipsec_mb/meson.build | 1 +
drivers/crypto/null/meson.build | 1 +
drivers/crypto/openssl
Enabled the flag pmd_supports_disable_iova_as_pa in cnxk driver build
files as they work with IOVA as VA. Updated cn9k and cn10k soc build
configurations to disable the IOVA as PA build by default.
Signed-off-by: Shijith Thotton
Acked-by: Olivier Matz
---
config/arm/meson.build
Swapped position of mbuf next pointer and second dynamic field (dynfield2)
if the build is configured to disable IOVA as PA. This is to move the
mbuf next pointer to first cache line.
Signed-off-by: Shijith Thotton
Acked-by: Olivier Matz
---
doc/guides/rel_notes/release_22_11.rst | 3 +++
lib
If IOVA as PA is disabled during build, mbuf physical address field is
undefined. This space is used to add the second dynamic field.
Signed-off-by: Shijith Thotton
Acked-by: Olivier Matz
---
lib/mbuf/rte_mbuf_core.h | 6 +-
lib/mbuf/rte_mbuf_dyn.c | 3 +++
2 files changed, 8 insertions
pmd_supports_disable_iova_as_pa in its build file.
mbuf structure holds the physical (PA) and virtual address (VA) of a
buffer. if IOVA as PA is disabled at compile time, PA field (buf_iova)
of mbuf is redundant as it is the same as VA and is replaced by a dummy
field.
Signed-off-by: Shijith Thotton
Acked-by: Olivier
Used rte_mbuf_data_iova API to get the physical address of mbuf data.
Signed-off-by: Shijith Thotton
Acked-by: Olivier Matz
---
app/test/test_dmadev.c | 33 ++---
1 file changed, 14 insertions(+), 19 deletions(-)
diff --git a/app/test/test_dmadev.c b/app/test
Added APIs rte_mbuf_iova_set and rte_mbuf_iova_get to set and get the
physical address of an mbuf respectively. Updated applications and
library to use the same.
Signed-off-by: Shijith Thotton
Acked-by: Olivier Matz
---
app/test-bbdev/test_bbdev_perf.c | 2 +-
app/test-crypto-perf
.
* Moved to #if scheme instead of union.
* Updated release notes.
v3:
* Cleared use of buf_iova from cnxk PMD.
v2:
* Used RTE_IOVA_AS_VA instread of rte_is_iova_as_va_build().
* Moved mbuf next pointer to first cacheline if RTE_IOVA_AS_VA = 1.
Shijith Thotton (7):
mbuf: add API to get and set
>> Enabled the flag pmd_iova_as_va in cnxk driver build files as they work
>> with IOVA as VA. Updated cn9k and cn10k soc build configurations to
>> enable the IOVA as VA build by default.
>>
>> Signed-off-by: Shijith Thotton
>> ---
>> con
Hi Olivier,
>On Sat, Oct 08, 2022 at 01:00:23AM +0530, Shijith Thotton wrote:
>> Added APIs rte_mbuf_iova_set and rte_mbuf_iova_get to set and get the
>> physical address of an mbuf respectively. Updated applications and
>> library to use the same.
>>
>&g
>
>Fix stack overflow when event vector size is greater than
>MAX_BURST_SIZE.
>Add missing mac swap and rfc1812 stage.
>
>Fixes: e8adca1951d4 ("examples/l3fwd: support event vector")
>
>Signed-off-by: Pavan Nikhilesh
>---
> examples/l3fwd/l3fwd_fib.c | 123 ++---
> 1
>Use em vector path to process event vector.
>
>Signed-off-by: Pavan Nikhilesh
>---
> examples/l3fwd/l3fwd_em.c| 12 +++--
> examples/l3fwd/l3fwd_em.h| 29 +--
> examples/l3fwd/l3fwd_em_hlm.h| 72 +---
> examples/l3fwd/l3fwd_em_sequentia
>> This is a continuation of the discussions[1] to add mbuf physical address
>> field to
>dynamic field.
>> Previous version was to add PA field to dynamic field area based on the EAL
>IOVA mode option. It was
>> deemed unsafe as some components could still use the PA field without
>checking IOVA
Enabled software PMDs in IOVA as PA disabled build as they work with
IOVA as VA.
Signed-off-by: Shijith Thotton
---
drivers/crypto/armv8/meson.build| 1 +
drivers/crypto/ipsec_mb/meson.build | 1 +
drivers/crypto/null/meson.build | 1 +
drivers/crypto/openssl/meson.build | 1 +
drivers
Enabled the flag pmd_supports_disable_iova_as_pa in cnxk driver build
files as they work with IOVA as VA. Updated cn9k and cn10k soc build
configurations to disable the IOVA as PA build by default.
Signed-off-by: Shijith Thotton
---
config/arm/meson.build | 8 +++-
doc/guides
Swapped position of mbuf next pointer and second dynamic field (dynfield2)
if the build is configured to disable IOVA as PA. This is to move the
mbuf next pointer to first cache line.
Signed-off-by: Shijith Thotton
---
doc/guides/rel_notes/release_22_11.rst | 3 +++
lib/mbuf/rte_mbuf_core.h
If IOVA as PA is disabled during build, mbuf physical address field is
undefined. This space is used to add the second dynamic field.
Signed-off-by: Shijith Thotton
---
lib/mbuf/rte_mbuf_core.h | 6 +-
lib/mbuf/rte_mbuf_dyn.c | 3 +++
2 files changed, 8 insertions(+), 1 deletion(-)
diff
pmd_supports_disable_iova_as_pa in its build file.
mbuf structure holds the physical (PA) and virtual address (VA) of a
buffer. if IOVA as PA is disabled at compile time, PA field (buf_iova)
of mbuf is redundant as it is the same as VA and is replaced by a dummy
field.
Signed-off-by: Shijith Thotton
---
app/test
Used rte_mbuf_data_iova API to get the physical address of mbuf data.
Signed-off-by: Shijith Thotton
---
app/test/test_dmadev.c | 33 ++---
1 file changed, 14 insertions(+), 19 deletions(-)
diff --git a/app/test/test_dmadev.c b/app/test/test_dmadev.c
index
Added APIs rte_mbuf_iova_set and rte_mbuf_iova_get to set and get the
physical address of an mbuf respectively. Updated applications and
library to use the same.
Signed-off-by: Shijith Thotton
---
app/test-bbdev/test_bbdev_perf.c | 2 +-
app/test-crypto-perf/cperf_test_common.c | 5
.
* Updated release notes.
v3:
* Cleared use of buf_iova from cnxk PMD.
v2:
* Used RTE_IOVA_AS_VA instread of rte_is_iova_as_va_build().
* Moved mbuf next pointer to first cacheline if RTE_IOVA_AS_VA = 1.
Shijith Thotton (7):
mbuf: add API to get and set mbuf physical address
test/dma: use
>> >> Enabled the flag pmd_iova_as_va in cnxk driver build files as they work
>> >> with IOVA as VA. Updated cn9k and cn10k soc build configurations to
>> >> enable the IOVA as VA build by default.
>> >>
>> >> Signed-off-
>> Enabled the flag pmd_iova_as_va in cnxk driver build files as they work
>> with IOVA as VA. Updated cn9k and cn10k soc build configurations to
>> enable the IOVA as VA build by default.
>>
>> Signed-off-by: Shijith Thotton
>> ---
>> con
xt pointer.
>>
>> Signed-off-by: Shijith Thotton
>> ---
>> lib/mbuf/rte_mbuf_core.h | 29 +
>> lib/meson.build | 3 +++
>> 2 files changed, 24 insertions(+), 8 deletions(-)
>>
>> diff --git a/lib/mbuf/rte_mbuf_
>> > > mbuf physical address field is not used in builds which only uses VA. It
>> > > is used to expand the dynamic field area.
>> > >
>> > > Signed-off-by: Shijith Thotton
>> >
>> > We cannot condition the use of the dynamic f
d invalid if the build is
>> configured to use only VA.
>>
>> Signed-off-by: Shijith Thotton
>> ---
>> app/test-bbdev/test_bbdev_perf.c | 2 +-
>> app/test-crypto-perf/cperf_test_common.c | 5 +--
>> app/test/test_bpf.c |
Hi Olivier,
Thanks for the review.
>On Wed, Sep 21, 2022 at 07:26:16PM +0530, Shijith Thotton wrote:
>> This is a continuation of the discussions[1] to add mbuf physical address
>> field to
>dynamic field.
>> Previous version was to add PA field to dynamic field area bas
:
> * Cleared use of buf_iova from cnxk PMD.
>
>v2:
> * Used RTE_IOVA_AS_VA instread of rte_is_iova_as_va_build().
> * Moved mbuf next pointer to first cacheline if RTE_IOVA_AS_VA = 1.
>
>Shijith Thotton (5):
> build: add meson option to configure IOVA mode as VA
> mbuf
>>
>> This is a continuation of the discussions[1] to add mbuf physical address
>> field to
>dynamic field.
>> Previous version was to add PA field to dynamic field area based on the EAL
>IOVA mode option. It was
>> deemed unsafe as some components could still use the PA field without
>checking IO
Enabled software PMDs in IOVA as VA build as they work with IOVA as VA.
Signed-off-by: Shijith Thotton
---
drivers/crypto/armv8/meson.build| 1 +
drivers/crypto/ipsec_mb/meson.build | 1 +
drivers/crypto/null/meson.build | 1 +
drivers/crypto/openssl/meson.build | 1 +
drivers/dma
Enabled the flag pmd_iova_as_va in cnxk driver build files as they work
with IOVA as VA. Updated cn9k and cn10k soc build configurations to
enable the IOVA as VA build by default.
Signed-off-by: Shijith Thotton
---
config/arm/meson.build | 8 +++-
drivers/common/cnxk
Swapped position of mbuf next pointer and second dynamic field (dynfield2)
if the build is configured to use IOVA as VA. This is to move the mbuf
next pointer to first cache line. kni library is disabled for this
change as it depends on the offset value of next pointer.
Signed-off-by: Shijith
mbuf physical address field is not used in builds which only uses VA. It
is used to expand the dynamic field area.
Signed-off-by: Shijith Thotton
---
lib/mbuf/rte_mbuf_core.h | 26 +-
lib/mbuf/rte_mbuf_dyn.c | 2 ++
2 files changed, 19 insertions(+), 9 deletions
file to enable build.
mbuf structure holds the physical (PA) and virtual address (VA) of a
buffer. if IOVA mode is set to VA, PA is redundant as it is the same as
VA. So PA field need not be updated and marked invalid if the build is
configured to use only VA.
Signed-off-by: Shijith Thotton
rte_is_iova_as_va_build().
* Moved mbuf next pointer to first cacheline if RTE_IOVA_AS_VA = 1.
Shijith Thotton (5):
build: add meson option to configure IOVA mode as VA
mbuf: add second dynamic field member for VA only build
lib: move mbuf next pointer to first cache line
drivers: mark Marvell cnxk PMDs
Removed support to limit XAQ from devargs. If XAQ is limited, new add
works could run out of XAQ entries and disable the queue.
Signed-off-by: Shijith Thotton
---
v2:
* Removed used function parameter.
doc/guides/eventdevs/cnxk.rst | 5 ++---
drivers/common/cnxk/roc_mbox.h | 2
Removed support to limit XAQ from devargs. If XAQ is limited, new add
works could run out of XAQ entries and disable the queue.
Signed-off-by: Shijith Thotton
---
doc/guides/eventdevs/cnxk.rst | 5 ++---
drivers/common/cnxk/roc_mbox.h | 2 +-
drivers/common/cnxk/roc_sso.c | 2
>On 7/1/2022 1:24 PM, Shijith Thotton wrote:
>>>> If all devices are configured to run in IOVA mode as VA, physical
>>>> address field of mbuf (buf_iova) won't be used. In such cases, buf_iova
>>>> space is free to use as a dynamic field. So a new dynam
>> >
>> > --
>> > On Mon, Aug 29, 2022 at 08:32:20PM +0200, Morten Brørup wrote:
>> > >
>> > > > From: Shijith Thotton [mailto:sthot...@marvell.com]
>> > > &g
Enabled software PMDs in IOVA as VA build as they work with IOVA as VA.
Signed-off-by: Shijith Thotton
---
drivers/crypto/armv8/meson.build| 1 +
drivers/crypto/ipsec_mb/meson.build | 1 +
drivers/crypto/null/meson.build | 1 +
drivers/crypto/openssl/meson.build | 1 +
drivers/dma
Enabled the flag pmd_iova_as_va in cnxk driver build files as they work
with IOVA as VA. Updated cn9k and cn10k soc build configurations to
enable the IOVA as VA build by default.
Signed-off-by: Shijith Thotton
---
config/arm/meson.build | 8 ++--
drivers/common/cnxk
Swapped position of mbuf next pointer and second dynamic field (dynfield2)
if the build is configured to use IOVA as VA. This is to move the mbuf
next pointer to first cache line. kni library is disabled for this
change as it depends on the offset value of next pointer.
Signed-off-by: Shijith
file to enable build.
mbuf structure holds the physical (PA) and virtual address (VA) of a
buffer. if IOVA mode is set to VA, PA is redundant as it is the same as
VA. So PA field need not be updated and marked invalid if the build is
configured to use only VA.
Signed-off-by: Shijith Thotton
mbuf physical address field is not used in builds which only uses VA. It
is used to expand the dynamic field area.
Signed-off-by: Shijith Thotton
---
lib/mbuf/rte_mbuf_core.h | 26 +-
lib/mbuf/rte_mbuf_dyn.c | 2 ++
2 files changed, 19 insertions(+), 9 deletions
1 - 100 of 531 matches
Mail list logo
