mbox series

[0/3] use a common eal device event for hot-unplug

Message ID 1544773540-89825-1-git-send-email-jia.guo@intel.com (mailing list archive)
Headers
Series use a common eal device event for hot-unplug |

Message

Guo, Jia Dec. 14, 2018, 7:45 a.m. UTC
  This patch set aim to use a common eal device event mechanism to manage
devices for hot-unplug whatever device types, so the previous ethdev
remove event will be deleted and the device detach will be remove from
app into eal framework.

Jeff Guo (3):
  eal: add --dev-hotplug option
  ethdev: remove ethdev rmv interrupt
  testpmd: remove device detach into eal

 app/test-pmd/parameters.c                  |  2 --
 app/test-pmd/testpmd.c                     | 41 +++---------------------------
 drivers/bus/pci/pci_common.c               |  7 +++++
 drivers/net/failsafe/failsafe_ether.c      | 12 ++++-----
 drivers/net/failsafe/failsafe_ops.c        |  3 +--
 drivers/net/failsafe/failsafe_private.h    |  6 ++---
 drivers/net/mlx4/mlx4_intr.c               |  1 -
 drivers/net/mlx5/mlx5_ethdev.c             |  7 ++---
 lib/librte_eal/bsdapp/eal/eal.c            | 16 ++++++++++++
 lib/librte_eal/common/eal_common_options.c |  5 ++++
 lib/librte_eal/common/eal_internal_cfg.h   |  1 +
 lib/librte_eal/common/eal_options.h        |  4 ++-
 lib/librte_eal/linuxapp/eal/eal.c          | 16 ++++++++++++
 lib/librte_eal/linuxapp/eal/eal_dev.c      | 16 +++++++++++-
 lib/librte_ethdev/rte_ethdev.h             |  1 -
 15 files changed, 79 insertions(+), 59 deletions(-)
  

Comments

Iremonger, Bernard April 8, 2019, 2:10 p.m. UTC | #1
Hi Jia

> -----Original Message-----
> From: Guo, Jia
> Sent: Friday, December 14, 2018 7:46 AM
> To: Iremonger, Bernard <bernard.iremonger@intel.com>; Lu, Wenzhuo
> <wenzhuo.lu@intel.com>; shahafs@mellanox.com; thomas@monjalon.net;
> matan@mellanox.com
> Cc: Yigit, Ferruh <ferruh.yigit@intel.com>; Ananyev, Konstantin
> <konstantin.ananyev@intel.com>; dev@dpdk.org; Guo, Jia
> <jia.guo@intel.com>; stephen@networkplumber.org; gaetan.rivet@6wind.com;
> Zhang, Qi Z <qi.z.zhang@intel.com>; arybchenko@solarflare.com; Richardson,
> Bruce <bruce.richardson@intel.com>; He, Shaopeng <shaopeng.he@intel.com>;
> Burakov, Anatoly <anatoly.burakov@intel.com>
> Subject: [PATCH 0/3] use a common eal device event for hot-unplug
> 
> This patch set aim to use a common eal device event mechanism to manage
> devices for hot-unplug whatever device types, so the previous ethdev remove
> event will be deleted and the device detach will be remove from app into eal
> framework.
> 
> Jeff Guo (3):
>   eal: add --dev-hotplug option
>   ethdev: remove ethdev rmv interrupt
>   testpmd: remove device detach into eal
> 
>  app/test-pmd/parameters.c                  |  2 --
>  app/test-pmd/testpmd.c                     | 41 +++---------------------------
>  drivers/bus/pci/pci_common.c               |  7 +++++
>  drivers/net/failsafe/failsafe_ether.c      | 12 ++++-----
>  drivers/net/failsafe/failsafe_ops.c        |  3 +--
>  drivers/net/failsafe/failsafe_private.h    |  6 ++---
>  drivers/net/mlx4/mlx4_intr.c               |  1 -
>  drivers/net/mlx5/mlx5_ethdev.c             |  7 ++---
>  lib/librte_eal/bsdapp/eal/eal.c            | 16 ++++++++++++
>  lib/librte_eal/common/eal_common_options.c |  5 ++++
>  lib/librte_eal/common/eal_internal_cfg.h   |  1 +
>  lib/librte_eal/common/eal_options.h        |  4 ++-
>  lib/librte_eal/linuxapp/eal/eal.c          | 16 ++++++++++++
>  lib/librte_eal/linuxapp/eal/eal_dev.c      | 16 +++++++++++-
>  lib/librte_ethdev/rte_ethdev.h             |  1 -
>  15 files changed, 79 insertions(+), 59 deletions(-)
> 
> --
> 2.7.4

This patch set fails to apply to DPDK 19.05.rc1 and probably needs to be rebased.

Regards,

Bernard.
  
Stephen Hemminger June 9, 2023, 5:18 p.m. UTC | #2
Unifying hotplug and device event support would be good idea
but this patch has gone stale. And supporting it would require
API changes (need to wait for next LTS) and a lot of testing.

Also, hotplug is a complex topic which has a lot of platform
dependencies.  If someone wants to do a more general solution
this patchset is worth looking at. But for now it needs to
be marked rejected in patchwork.