Re: [PATCH] net/af_packet: cache align Rx/Tx structs

On 2024-04-24 02:27, Honnappa Nagarahalli wrote: On Apr 23, 2024, at 3:56 PM, Mattias Rönnblom wrote: On 2024-04-23 13:15, Ferruh Yigit wrote: On 4/23/2024 10:08 AM, Mattias Rönnblom wrote: Cache align Rx and Tx queue struct to avoid false sharing. RX struct happens to be 64 bytes on x86_

RE: [PATCH v2] eal/linux: enhanced error handling for affinity

Hello, I've noticed some coding style issues eal/linux: enhanced error handling for affinity - Patchwork (dpdk.org) and have attempted to update the patch accordingly. Could you please help by updating

[PATCH] event: fix warning from useless snprintf

With Gcc-14, this warning is generated: ../drivers/event/sw/sw_evdev.c:263:3: warning: 'snprintf' will always be truncated; specified size is 12, but format string expands to at least 13 [-Wformat-truncation] 263 | snprintf(buf, sizeof(buf), "sw%d_iq_%d_rob", dev_id, i);

[RFC] ethdev: add VXLAN last reserved field

Add VxLAN last reserved bytes in rte flow packet fields. Signed-off-by: Rongwei Liu --- lib/ethdev/rte_flow.h | 1 + 1 file changed, 1 insertion(+) diff --git a/lib/ethdev/rte_flow.h b/lib/ethdev/rte_flow.h index 6e8ab1d4c7..5aaa0dc96f 100644 --- a/lib/ethdev/rte_flow.h +++ b/lib/ethdev/rte_flo

Re: [PATCH] net/af_packet: cache align Rx/Tx structs

> On Apr 23, 2024, at 3:56 PM, Mattias Rönnblom wrote: > > On 2024-04-23 13:15, Ferruh Yigit wrote: >> On 4/23/2024 10:08 AM, Mattias Rönnblom wrote: >>> Cache align Rx and Tx queue struct to avoid false sharing. >>> >>> RX struct happens to be 64 bytes on x86_64 already, so cache alignment >>

Re: [PATCH v4] lib/hash: add defer queue reclaim API

On Mon, 15 Apr 2024 11:26:02 + Abdullah Ömer Yamaç wrote: > + ret = rte_rcu_qsbr_dq_reclaim(h->dq, h->hash_rcu_cfg->max_reclaim_size, > + freed, > pending, available); Indention here is odd. I would expect "freed," to line up

Re: [PATCH] net/af_packet: cache align Rx/Tx structs

On 2024-04-23 13:15, Ferruh Yigit wrote: On 4/23/2024 10:08 AM, Mattias Rönnblom wrote: Cache align Rx and Tx queue struct to avoid false sharing. RX struct happens to be 64 bytes on x86_64 already, so cache alignment makes no change there, but it does on 32-bit ISAs. TX struct is 56 bytes on

Re: [PATCH v1 3/4] dts: unify super calls

Reviewed-by: Patrick Robb

Re: [PATCH v1 2/4] dts: unify class inheritance from object

Reviewed-by: Patrick Robb

DPDK 22.11.5 released

Hi all, Here is a new stable release: https://fast.dpdk.org/rel/dpdk-22.11.5.tar.xz The git tree is at: https://dpdk.org/browse/dpdk-stable/?h=22.11 Luca Boccassi --- .github/workflows/build.yml| 20 +- .mailmap

Re: [PATCH] lib/hash: setting the maximum reclamation size

Hello, Any comment on this patch? On Wed, Apr 17, 2024 at 4:39 PM Abdullah Ömer Yamaç wrote: > In the previous implementation, the maximum reclamation size was set > to RTE_HASH_RCU_DQ_RECLAIM_MAX and it was not configurable. This patch > uses the configuration argument to set the maximum reclam

Re: [PATCH v4] lib/hash: add defer queue reclaim API

Hello, is there any other comment on this patch? Thanks On Mon, Apr 15, 2024 at 2:26 PM Abdullah Ömer Yamaç wrote: > This patch adds a new feature to the hash library to allow the user to > reclaim the defer queue. This is useful when the user wants to force > reclaim resources that are not bein

[PATCH v1] dts: clean up close in remote session

The close method was split into two methods, one with common code and one that's subclass specific. There wasn't any common code so the split doesn't make sense. And if we ever need such a split, we can use super() in the future. There was also an unused argument, force and the order of methods was

Re: [PATCH v1] dts: rename execution to test run

On Tue, Apr 23, 2024 at 12:58 PM Luca Vizzarro wrote: > > Hi Juraj! > > I am not fussed about naming but I am certainly in favour of making > things less confusing, so I am happy with this change. > Yes, using execution in this context is very confusing. The main thing with this patch is to find

Re: [PATCH v1] dts: remove the OS UDP test suite

On Tue, Apr 23, 2024 at 12:39 PM Luca Vizzarro wrote: > > On 19/04/2024 14:52, Juraj Linkeš wrote: > > diff --git a/dts/framework/test_suite.py b/dts/framework/test_suite.py > > > > -def configure_testbed_ipv4(self, restore: bool = False) -> None: > > -"""Configure IPv4 addresses on a

Re: question about eth and vlan item in flow pattern

Hi, Please see below. On Tue, 23 Apr 2024, Jie Hai wrote: Hi, Ivan Malov, Sorry for the late reply. On 2024/4/10 21:19, Ivan Malov wrote: Hi Jie, Consider the following examples: 1) flow create 0 ingress pattern eth / ipv4 proto is 17 / udp / end \    actions queue index 1 / end 2) flow cre

Re: [PATCH v7 0/5] app/testpmd: support multiple process attach and detach port

Hi Ferruh and Thomas, It's been almost two years since this issue was reported. We have discussed a lot before, and also made some progress and consensus. Can you take a look at it again?  Looking forward to your reply. BR/ Huisong 在 2024/1/30 14:36, Huisong Li 写道: This patchset fix some bugs

Re: [PATCH] net/af_packet: cache align Rx/Tx structs

On 4/23/2024 10:08 AM, Mattias Rönnblom wrote: > Cache align Rx and Tx queue struct to avoid false sharing. > > RX struct happens to be 64 bytes on x86_64 already, so cache alignment > makes no change there, but it does on 32-bit ISAs. > > TX struct is 56 bytes on x86_64. > Hi Mattias, No obje

[DPDK/other Bug 1418] ice: no way to disable lsc-interrupts

https://bugs.dpdk.org/show_bug.cgi?id=1418 Bug ID: 1418 Summary: ice: no way to disable lsc-interrupts Product: DPDK Version: 24.03 Hardware: x86 OS: Linux Status: UNCONFIRMED Severity: normal Pr

Re: [PATCH v1] dts: rename execution to test run

Hi Juraj! I am not fussed about naming but I am certainly in favour of making things less confusing, so I am happy with this change. On 19/04/2024 09:51, Juraj Linkeš wrote: @staticmethod def from_dict( -d: ExecutionConfigDict, +d: TestRunConfigDict, node

[PATCH] common/cnxk: restore segregation of logs based on module

Originally the logs were segregated under various labels which could be selectively enabled. It was changed to use 'pmd.common.cnxk' while changing the macro used for registering logging. Address the same by restoring the segregation. Current logs: ... logtype3 pmd.common.c

Re: [PATCH v1] dts: remove the OS UDP test suite

On 19/04/2024 14:52, Juraj Linkeš wrote: diff --git a/dts/framework/test_suite.py b/dts/framework/test_suite.py -def configure_testbed_ipv4(self, restore: bool = False) -> None: -"""Configure IPv4 addresses on all testbed ports. - -The configured ports are: - -* SUT i

Re: [PATCH v1 0/4] node and inheritance improvements

On 23/04/2024 10:12, Juraj Linkeš wrote: There are two areas that are unified: The super() calls, where the arguments were removed as they're not needed in single inheritance. The classes derived from object don't need to explicitly state that. The tg node execution setup and teardown was omitte

Re: [PATCH v1 3/4] dts: unify super calls

Reviewed-by: Luca Vizzarro

Re: question about eth and vlan item in flow pattern

Hi, Ivan Malov, Sorry for the late reply. On 2024/4/10 21:19, Ivan Malov wrote: Hi Jie, Consider the following examples: 1) flow create 0 ingress pattern eth / ipv4 proto is 17 / udp / end \    actions queue index 1 / end 2) flow create 0 ingress pattern eth / ipv4 / udp / end \    actions que

Re: [PATCH v1 4/4] dts: refine pre-test setup and teardown steps

Reviewed-by: Luca Vizzarro

Re: [PATCH v1 2/4] dts: unify class inheritance from object

Reviewed-by: Luca Vizzarro

Re: [PATCH v1 1/4] dts: add tg node execution setup and teardown

Reviewed-by: Luca Vizzarro

[PATCH] net/af_packet: cache align Rx/Tx structs

Cache align Rx and Tx queue struct to avoid false sharing. RX struct happens to be 64 bytes on x86_64 already, so cache alignment makes no change there, but it does on 32-bit ISAs. TX struct is 56 bytes on x86_64. Both structs keep counters, and in the RX case they are updated even for empty pol

[PATCH v1 4/4] dts: refine pre-test setup and teardown steps

The major part is the removal of _set_up_execution() and _tear_down_execution() of Node in lieu of using super() in the superclasses, which simplifies the code while achieving the same thing. The minor changes are the move of virtual devices and build target setup/teardown into SutNode from Node s

[PATCH v1 3/4] dts: unify super calls

We have two ways of calling super() in the codebase. For single inheritance, there's no benefit in listing the arguments, as the function will do exactly what we need it to do. Signed-off-by: Juraj Linkeš --- dts/framework/test_result.py| 12 ++-- dts/framework/testbed_model/

[PATCH v1 2/4] dts: unify class inheritance from object

There are two ways we specify that a class inherits from object - implicitly and explicitly. There's no need to explicitly specify that a class inherits from object and is in fact mostly a remnant from Python2. Leaving it implicit is the standard in Python3 and offers a small bonus in cases where s

[PATCH v1 1/4] dts: add tg node execution setup and teardown

The setup and teardown for the execution stage was meant to be run on the tg node too, but was mistakenly omitted. Fixes: cecfe0aabf58 ("dts: add traffic generator abstractions") Signed-off-by: Juraj Linkeš --- dts/framework/runner.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/dts/fra

[PATCH v1 0/4] node and inheritance improvements

There are two areas that are unified: The super() calls, where the arguments were removed as they're not needed in single inheritance. The classes derived from object don't need to explicitly state that. The tg node execution setup and teardown was omitted and is fixed in this series. And probabl

Re: [PATCH v3] MAINTAINERS: remove maintainers that bounce

On Mon, Apr 22, 2024 at 8:40 PM Stephen Hemminger wrote: > > Recent patch set for rte_memcpy, discovered that the some > mail addresses could not be found. > > Signed-off-by: Stephen Hemminger > Acked-by: Bruce Richardson Acked-by: Juraj Linkeš