DPDK patches and discussions
 help / color / mirror / Atom feed
From: Thomas Monjalon <thomas.monjalon@6wind.com>
To: Stephen Hemminger <stephen@networkplumber.org>
Cc: dev@dpdk.org
Subject: Re: [dpdk-dev] [PATCH 1/2] virtio: initialize iopl when device is initialized
Date: Wed, 29 Jul 2015 19:26:17 +0200	[thread overview]
Message-ID: <3368758.BASS8qfG8B@xps13> (raw)
In-Reply-To: <20150306082057.5c505f54@urahara>

2015-03-06 08:20, Stephen Hemminger:
> The issue is that virtio has no place it can do iopl() and have the IRQ thread
> work. It only shows up on real code where application is daemon, not in a toy
> demo or test application.
> 
> Right now:
>     gcc start
>        rte_virtio_pmd_init
>           iopl
>     main
>         daemon
>             fork
>                 fork 
>                   Process is now child of init not original process
> 
>                   rte_eal_init
>                      fork (pthread) for irq thread
>                                                        irq thread
>                                                         (no iopl permssion)
>                   program start
>                   rte_pmd_virtio_configure

This call tree is wrong.
Below is the right (and more complete) one:

gcc start
	driver constructor
		rte_eal_driver_register (if .a)
main
	rte_eal_init
		eal_parse_args
		rte_eal_pci_init
		rte_eal_memory_init
		rte_eal_intr_init
			pthread_create
				eal_intr_thread_main
					eal_intr_handle_interrupts
						eal_intr_process_interrupts
							virtio_interrupt_handler
		dlopen (if .so)
			driver constructor
				rte_eal_driver_register
		rte_eal_dev_init
			driver->init
				rte_virtio_pmd_init
					rte_eal_iopl_init
					rte_eth_driver_register
		pthread_create
		rte_eal_pci_probe
			driver->devinit
				rte_eth_dev_init
					rte_eth_dev_allocate
					eth_drv->eth_dev_init
						eth_virtio_dev_init
							virtio_resource_init

> So the only place where iopl() can be done in the proper context
> so that the IRQ (and other helper threads in future) have the correct
> permissions is in eal_init.

No there are 2 other possible solutions:
1) Move rte_eal_intr_init() after rte_eal_dev_init().
2) Move dlopen() before rte_eal_intr_init() and call iopl in the constructor.
With the second solution, we must also keep an iopl call in rte_virtio_pmd_init()
to return an error if iopl fails.

The second solution was proposed in the series sent by David:
	http://dpdk.org/ml/archives/dev/2015-March/014931.html
	http://dpdk.org/ml/archives/dev/2015-March/014932.html

  parent reply	other threads:[~2015-07-29 17:27 UTC|newest]

Thread overview: 43+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-03-06  0:45 [dpdk-dev] [PATCH 0/2] virtio: bugfixes Stephen Hemminger
2015-03-06  0:45 ` [dpdk-dev] [PATCH 1/2] virtio: initialize iopl when device is initialized Stephen Hemminger
2015-03-06  3:41   ` Ouyang, Changchun
2015-03-06 16:20     ` Stephen Hemminger
2015-03-06 16:33       ` David Marchand
2015-03-06 16:55         ` Stephen Hemminger
2015-03-06 22:04           ` David Marchand
2015-03-06 23:43             ` Stephen Hemminger
2015-03-07  6:53               ` David Marchand
2015-03-09 11:05       ` David Marchand
2015-03-09 14:56         ` [dpdk-dev] [PATCH 0/2] fix virtio interrupt handling David Marchand
2015-03-09 14:56           ` [dpdk-dev] [PATCH 1/2] eal/linux: move plugin load to very start of eal init David Marchand
2015-03-09 15:21             ` Neil Horman
     [not found]               ` <CALwxeUs4hPbYDPBUfz9u2AoiCoj_wwTsAyj=_1xxzuT6LLW6nw@mail.gmail.com>
2015-03-10 10:55                 ` Neil Horman
2015-10-14  0:05                   ` Stephen Hemminger
2015-10-14  9:55                     ` David Marchand
2015-03-09 14:56           ` [dpdk-dev] [PATCH 2/2] virtio: change io privilege level as early as possible David Marchand
2015-03-10 13:14             ` Neil Horman
2015-09-29 19:25               ` Stephen Hemminger
2015-09-30  8:28                 ` David Marchand
2015-09-30 14:52                   ` Neil Horman
2015-09-30 15:37                     ` Thomas Monjalon
2015-09-30 17:26                       ` Stephen Hemminger
2015-10-01 11:25                       ` Neil Horman
2015-10-12 20:08                         ` Stephen Hemminger
2015-10-14  0:07                         ` Stephen Hemminger
2015-10-14  8:00                           ` David Marchand
2015-10-14  9:49                             ` David Marchand
2015-10-14  9:50                               ` [dpdk-dev] [PATCH] eal: move interrupt init after device init David Marchand
2015-10-14 11:32                                 ` David Marchand
2015-10-20 21:22                                   ` Thomas Monjalon
2015-07-29 17:26       ` Thomas Monjalon [this message]
2015-03-06  0:45 ` [dpdk-dev] [PATCH 2/2] virtio: allow running w/o vlan filtering Stephen Hemminger
2015-03-06  3:39   ` Ouyang, Changchun
2015-03-06 16:24     ` Stephen Hemminger
2015-07-29 12:56       ` Thomas Monjalon
2015-07-30  1:23         ` Ouyang, Changchun
2015-08-04 12:51         ` Vincent JARDIN
2015-08-05  1:01           ` Ouyang, Changchun
2015-08-05  1:22             ` Stephen Hemminger
2015-08-05 10:49             ` Vincent JARDIN
2015-10-21 13:58               ` Thomas Monjalon
2017-02-15  8:38                 ` Thomas Monjalon

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=3368758.BASS8qfG8B@xps13 \
    --to=thomas.monjalon@6wind.com \
    --cc=dev@dpdk.org \
    --cc=stephen@networkplumber.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).