All of lore.kernel.org
 help / color / mirror / Atom feed
From: Ezequiel Garcia <ezequiel@collabora.com>
To: linux-media@vger.kernel.org, linux-usb@vger.kernel.org,
	linux-arm-kernel@lists.infradead.org,
	linux-kernel@vger.kernel.org
Cc: Laurent Pinchart <laurent.pinchart@ideasonboard.com>,
	Tomasz Figa <tfiga@chromium.org>,
	"Matwey V . Kornilov" <matwey@sai.msu.ru>,
	Alan Stern <stern@rowland.harvard.edu>,
	kernel@collabora.com, Keiichi Watanabe <keiichiw@chromium.org>,
	Ezequiel Garcia <ezequiel@collabora.com>
Subject: [RFC 0/3] Introduce usb_{alloc,free}_noncoherent API
Date: Thu, 30 Aug 2018 14:20:27 -0300	[thread overview]
Message-ID: <20180830172030.23344-1-ezequiel@collabora.com> (raw)

Following the discussion on PWC [1] and UVC [2] drivers, where
use non-consistent mappings for the URB transfer buffers was
shown to improve transfer speed significantly, here's a proposal
for a non-coherent USB helpers.

With this pachset, it's possible to get 360x288 raw analog video
using stk1160 and a AM335x Beaglebone Black board. This isn't
possible in mainline, for the same reasons Matwey has explained [1].

First patch is a hack, obviously incomplete, to add support for
non-consistent mappings on ARM.

The second patch introduces the usb_{alloc,free}_noncoherent API,
while the third patch is an example on stk1160.

I'm sending this patchset as RFC, just to get the ball rolling.

[1] https://lkml.org/lkml/2018/8/21/663
[2] https://lkml.org/lkml/2018/6/27/188

Ezequiel Garcia (3):
  HACK: ARM: dma-mapping: Get writeback memory for non-consistent
    mappings
  USB: core: Add non-coherent buffer allocation helpers
  stk1160: Use non-coherent buffers for USB transfers

 arch/arm/include/asm/pgtable.h            |  3 ++
 arch/arm/mm/dma-mapping.c                 |  9 ++--
 drivers/media/usb/stk1160/stk1160-video.c | 22 +++------
 drivers/usb/core/buffer.c                 | 29 +++++++-----
 drivers/usb/core/hcd.c                    |  5 +-
 drivers/usb/core/usb.c                    | 56 ++++++++++++++++++++++-
 include/linux/usb.h                       |  5 ++
 include/linux/usb/hcd.h                   |  4 +-
 8 files changed, 97 insertions(+), 36 deletions(-)

-- 
2.18.0


WARNING: multiple messages have this Message-ID (diff)
From: ezequiel@collabora.com (Ezequiel Garcia)
To: linux-arm-kernel@lists.infradead.org
Subject: [RFC 0/3] Introduce usb_{alloc,free}_noncoherent API
Date: Thu, 30 Aug 2018 14:20:27 -0300	[thread overview]
Message-ID: <20180830172030.23344-1-ezequiel@collabora.com> (raw)

Following the discussion on PWC [1] and UVC [2] drivers, where
use non-consistent mappings for the URB transfer buffers was
shown to improve transfer speed significantly, here's a proposal
for a non-coherent USB helpers.

With this pachset, it's possible to get 360x288 raw analog video
using stk1160 and a AM335x Beaglebone Black board. This isn't
possible in mainline, for the same reasons Matwey has explained [1].

First patch is a hack, obviously incomplete, to add support for
non-consistent mappings on ARM.

The second patch introduces the usb_{alloc,free}_noncoherent API,
while the third patch is an example on stk1160.

I'm sending this patchset as RFC, just to get the ball rolling.

[1] https://lkml.org/lkml/2018/8/21/663
[2] https://lkml.org/lkml/2018/6/27/188

Ezequiel Garcia (3):
  HACK: ARM: dma-mapping: Get writeback memory for non-consistent
    mappings
  USB: core: Add non-coherent buffer allocation helpers
  stk1160: Use non-coherent buffers for USB transfers

 arch/arm/include/asm/pgtable.h            |  3 ++
 arch/arm/mm/dma-mapping.c                 |  9 ++--
 drivers/media/usb/stk1160/stk1160-video.c | 22 +++------
 drivers/usb/core/buffer.c                 | 29 +++++++-----
 drivers/usb/core/hcd.c                    |  5 +-
 drivers/usb/core/usb.c                    | 56 ++++++++++++++++++++++-
 include/linux/usb.h                       |  5 ++
 include/linux/usb/hcd.h                   |  4 +-
 8 files changed, 97 insertions(+), 36 deletions(-)

-- 
2.18.0

             reply	other threads:[~2018-08-30 17:21 UTC|newest]

Thread overview: 32+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-08-30 17:20 Ezequiel Garcia [this message]
2018-08-30 17:20 ` [RFC 0/3] Introduce usb_{alloc,free}_noncoherent API Ezequiel Garcia
2018-08-30 17:20 ` [RFC 1/3] HACK: ARM: dma-mapping: Get writeback memory for non-consistent mappings Ezequiel Garcia
2018-08-30 17:20   ` Ezequiel Garcia
2018-08-30 17:20   ` [RFC,1/3] " Ezequiel Garcia
2018-08-30 17:20 ` [RFC 2/3] USB: core: Add non-coherent buffer allocation helpers Ezequiel Garcia
2018-08-30 17:20   ` Ezequiel Garcia
2018-08-30 17:20   ` [RFC,2/3] " Ezequiel Garcia
2018-08-30 17:58   ` [RFC 2/3] " Christoph Hellwig
2018-08-30 17:58     ` Christoph Hellwig
2018-08-30 17:58     ` [RFC,2/3] " Christoph Hellwig
2018-08-30 22:11     ` [RFC 2/3] " Ezequiel Garcia
2018-08-30 22:11       ` Ezequiel Garcia
2018-08-30 22:11       ` [RFC,2/3] " Ezequiel Garcia
2018-08-31  5:50       ` [RFC 2/3] " Christoph Hellwig
2018-08-31  5:50         ` Christoph Hellwig
2018-08-31  5:50         ` [RFC,2/3] " Christoph Hellwig
2018-08-31  6:51         ` [RFC 2/3] " Tomasz Figa
2018-08-31  6:51           ` Tomasz Figa
2018-08-31  6:51           ` [RFC,2/3] " Tomasz Figa
2018-10-31  1:55           ` [RFC 2/3] " Tomasz Figa
2018-10-31  1:55             ` Tomasz Figa
2018-10-31  1:55             ` [RFC,2/3] " Tomasz Figa
2018-08-30 17:20 ` [RFC 3/3] stk1160: Use non-coherent buffers for USB transfers Ezequiel Garcia
2018-08-30 17:20   ` Ezequiel Garcia
2018-08-30 17:20   ` [RFC,3/3] " Ezequiel Garcia
2018-08-30 17:59   ` [RFC 3/3] " Christoph Hellwig
2018-08-30 17:59     ` Christoph Hellwig
2018-08-30 17:59     ` [RFC,3/3] " Christoph Hellwig
2018-09-07  8:54     ` [RFC 3/3] " Tomasz Figa
2018-09-07  8:54       ` Tomasz Figa
2018-09-07  8:54       ` [RFC,3/3] " Tomasz Figa

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=20180830172030.23344-1-ezequiel@collabora.com \
    --to=ezequiel@collabora.com \
    --cc=keiichiw@chromium.org \
    --cc=kernel@collabora.com \
    --cc=laurent.pinchart@ideasonboard.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-media@vger.kernel.org \
    --cc=linux-usb@vger.kernel.org \
    --cc=matwey@sai.msu.ru \
    --cc=stern@rowland.harvard.edu \
    --cc=tfiga@chromium.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.