From: Mark Brown <broonie@kernel.org> To: Trent Piepho <tpiepho@gmail.com> Cc: Peter Korsgaard <peter@korsgaard.com>, balbi@ti.com, "linux-mtd@lists.infradead.org" <linux-mtd@lists.infradead.org>, "spi-devel-general@lists.sourceforge.net" <spi-devel-general@lists.sourceforge.net>, Sourav Poddar <sourav.poddar@ti.com>, Brian Norris <computersforpeace@gmail.com>, David Woodhouse <dwmw2@infradead.org> Subject: Re: [PATCH 1/3] spi/qspi: Add memory mapped read support. Date: Fri, 11 Oct 2013 10:59:48 +0100 [thread overview] Message-ID: <20131011095948.GZ21581@sirena.org.uk> (raw) In-Reply-To: <CA+7tXigijA6ASOiHkRDz45PTruv2vwYJu0UUWz0hvKr0=zwu5w@mail.gmail.com> [-- Attachment #1.1: Type: text/plain, Size: 1532 bytes --] On Thu, Oct 10, 2013 at 04:53:46PM -0700, Trent Piepho wrote: > On Thu, Oct 10, 2013 at 3:10 AM, Mark Brown <broonie@kernel.org> wrote: > > Yeah, though of course at the minute the implementation of that thread > > is pretty much up to the individual drivers which isn't triumph - and > > the quality of implementation does vary rather a lot. I'm currently > > working on trying to factor more of this out, hopefully then it'll be > > easier to push out improvements. It may be nice to be able to kick off > > the first DMA transfer from within the caller for example. > I did testing with the mxs driver, which uses transfer_one > _message and the spi core queue pumping code. For small messages the > overhead of queuing work to the pump_messages queue and waiting for > completion is rather more than the time the actual transfer takes. Which > makes using a kthread rather pointless. Part of the problem could be the > high context switch cost for ARMv5. Indeed - that's what I'd expect to happen. Sufficiently small PIO transfers and DMA initiation should both be faster if done from the calling context. The thread starts to get more useful once you start building up work (at which point it should reduce context switches, especially if multiple threads are using the bus) and if the system is very heavily loaded when you can raise the thread priority so it can drive data in more quickly. That said it's also helpful in terms of making the code simpler to just have the one data path. [-- Attachment #1.2: Digital signature --] [-- Type: application/pgp-signature, Size: 836 bytes --] [-- Attachment #2: Type: text/plain, Size: 144 bytes --] ______________________________________________________ Linux MTD discussion mailing list http://lists.infradead.org/mailman/listinfo/linux-mtd/
WARNING: multiple messages have this Message-ID (diff)
From: Mark Brown <broonie@kernel.org> To: Trent Piepho <tpiepho@gmail.com> Cc: Peter Korsgaard <peter@korsgaard.com>, balbi@ti.com, "linux-mtd@lists.infradead.org" <linux-mtd@lists.infradead.org>, "spi-devel-general@lists.sourceforge.net" <spi-devel-general@lists.sourceforge.net>, Sourav Poddar <sourav.poddar@ti.com>, Brian Norris <computersforpeace@gmail.com>, David Woodhouse <dwmw2@infradead.org> Subject: Re: [PATCH 1/3] spi/qspi: Add memory mapped read support. Date: Fri, 11 Oct 2013 10:59:48 +0100 [thread overview] Message-ID: <20131011095948.GZ21581@sirena.org.uk> (raw) In-Reply-To: <CA+7tXigijA6ASOiHkRDz45PTruv2vwYJu0UUWz0hvKr0=zwu5w@mail.gmail.com> [-- Attachment #1: Type: text/plain, Size: 1532 bytes --] On Thu, Oct 10, 2013 at 04:53:46PM -0700, Trent Piepho wrote: > On Thu, Oct 10, 2013 at 3:10 AM, Mark Brown <broonie@kernel.org> wrote: > > Yeah, though of course at the minute the implementation of that thread > > is pretty much up to the individual drivers which isn't triumph - and > > the quality of implementation does vary rather a lot. I'm currently > > working on trying to factor more of this out, hopefully then it'll be > > easier to push out improvements. It may be nice to be able to kick off > > the first DMA transfer from within the caller for example. > I did testing with the mxs driver, which uses transfer_one > _message and the spi core queue pumping code. For small messages the > overhead of queuing work to the pump_messages queue and waiting for > completion is rather more than the time the actual transfer takes. Which > makes using a kthread rather pointless. Part of the problem could be the > high context switch cost for ARMv5. Indeed - that's what I'd expect to happen. Sufficiently small PIO transfers and DMA initiation should both be faster if done from the calling context. The thread starts to get more useful once you start building up work (at which point it should reduce context switches, especially if multiple threads are using the bus) and if the system is very heavily loaded when you can raise the thread priority so it can drive data in more quickly. That said it's also helpful in terms of making the code simpler to just have the one data path. [-- Attachment #2: Digital signature --] [-- Type: application/pgp-signature, Size: 836 bytes --]
next prev parent reply other threads:[~2013-10-11 9:59 UTC|newest] Thread overview: 104+ messages / expand[flat|nested] mbox.gz Atom feed top 2013-10-09 15:24 [PATCH 0/3]Add quad/memory mapped support for SPI flash Sourav Poddar 2013-10-09 15:24 ` Sourav Poddar 2013-10-09 15:24 ` [PATCH 1/3] spi/qspi: Add memory mapped read support Sourav Poddar 2013-10-09 15:24 ` Sourav Poddar 2013-10-09 16:07 ` Mark Brown 2013-10-09 16:07 ` Mark Brown [not found] ` <20131009160759.GQ21581-GFdadSzt00ze9xe1eoZjHA@public.gmane.org> 2013-10-09 16:54 ` Sourav Poddar 2013-10-09 16:54 ` Sourav Poddar 2013-10-09 17:40 ` Mark Brown 2013-10-09 17:40 ` Mark Brown 2013-10-09 18:15 ` Sourav Poddar 2013-10-09 18:15 ` Sourav Poddar 2013-10-09 18:41 ` Mark Brown 2013-10-09 18:41 ` Mark Brown [not found] ` <20131009174027.GS21581-GFdadSzt00ze9xe1eoZjHA@public.gmane.org> 2013-10-09 19:01 ` Peter Korsgaard 2013-10-09 19:01 ` Peter Korsgaard 2013-10-09 19:36 ` Mark Brown 2013-10-09 19:36 ` Mark Brown [not found] ` <87hacq1d5k.fsf-D6SC8u56vOOJDPpyT6T3/w@public.gmane.org> 2013-10-10 2:27 ` Trent Piepho 2013-10-10 2:27 ` Trent Piepho 2013-10-10 8:52 ` Sourav Poddar 2013-10-10 8:52 ` Sourav Poddar 2013-10-10 10:14 ` Mark Brown 2013-10-10 10:14 ` Mark Brown 2013-10-10 10:17 ` Sourav Poddar 2013-10-10 10:17 ` Sourav Poddar 2013-10-10 11:08 ` Sourav Poddar 2013-10-10 11:08 ` Sourav Poddar 2013-10-11 10:08 ` Mark Brown 2013-10-11 10:08 ` Mark Brown 2013-10-15 6:06 ` Sourav Poddar 2013-10-15 6:06 ` Sourav Poddar 2013-10-15 11:16 ` Mark Brown 2013-10-15 11:16 ` Mark Brown 2013-10-15 11:49 ` Sourav Poddar 2013-10-15 11:49 ` Sourav Poddar 2013-10-15 12:46 ` Mark Brown 2013-10-15 12:46 ` Mark Brown [not found] ` <20131015124656.GM2443-GFdadSzt00ze9xe1eoZjHA@public.gmane.org> 2013-10-15 13:23 ` Sourav Poddar 2013-10-15 13:23 ` Sourav Poddar 2013-10-15 15:53 ` Mark Brown 2013-10-15 15:53 ` Mark Brown [not found] ` <525D41E2.30206-l0cyMroinI0@public.gmane.org> 2013-10-15 15:33 ` Gupta, Pekon 2013-10-15 15:33 ` Gupta, Pekon 2013-10-15 16:01 ` Mark Brown 2013-10-15 16:01 ` Mark Brown 2013-10-15 16:54 ` Gupta, Pekon 2013-10-15 16:54 ` Gupta, Pekon 2013-10-15 18:01 ` Brian Norris 2013-10-15 18:01 ` Brian Norris 2013-10-15 18:10 ` Sourav Poddar 2013-10-15 18:10 ` Sourav Poddar [not found] ` <20131015180142.GS23337-bU/DPfM3abD4WzifrMjOTkcViWtcw2C0@public.gmane.org> 2013-10-15 18:13 ` Trent Piepho 2013-10-15 18:13 ` Trent Piepho 2013-10-15 18:33 ` Gupta, Pekon 2013-10-15 18:33 ` Gupta, Pekon 2013-10-15 20:52 ` Mark Brown 2013-10-15 20:52 ` Mark Brown [not found] ` <20131015205254.GX2443-GFdadSzt00ze9xe1eoZjHA@public.gmane.org> 2013-10-15 21:03 ` Trent Piepho 2013-10-15 21:03 ` Trent Piepho 2013-10-15 22:10 ` Mark Brown 2013-10-15 22:10 ` Mark Brown [not found] ` <20980858CB6D3A4BAE95CA194937D5E73EA23640-yXqyApvAXouIQmiDNMet8wC/G2K4zDHf@public.gmane.org> 2013-10-17 12:24 ` Sourav Poddar 2013-10-17 12:24 ` Sourav Poddar 2013-10-17 12:38 ` Mark Brown 2013-10-17 12:38 ` Mark Brown 2013-10-17 13:03 ` Gupta, Pekon 2013-10-17 13:03 ` Gupta, Pekon 2013-10-17 23:42 ` Mark Brown 2013-10-18 4:06 ` Sourav Poddar 2013-10-18 5:56 ` Trent Piepho 2013-10-18 6:10 ` Sourav Poddar 2013-10-18 7:27 ` Sourav Poddar 2013-10-18 10:31 ` Mark Brown 2013-10-18 11:48 ` Sourav Poddar 2013-10-18 13:08 ` Mark Brown 2013-10-18 14:47 ` Sourav Poddar 2013-10-15 20:59 ` Mark Brown 2013-10-15 20:59 ` Mark Brown [not found] ` <52566ACC.1080100-l0cyMroinI0@public.gmane.org> 2013-10-11 9:30 ` Gupta, Pekon 2013-10-11 9:30 ` Gupta, Pekon 2013-10-10 10:10 ` Mark Brown 2013-10-10 10:10 ` Mark Brown [not found] ` <20131010101052.GF21581-GFdadSzt00ze9xe1eoZjHA@public.gmane.org> 2013-10-10 23:53 ` Trent Piepho 2013-10-11 9:59 ` Mark Brown [this message] 2013-10-11 9:59 ` Mark Brown 2013-10-09 15:24 ` [PATCHv3 2/3] drivers: mtd: devices: Add quad " Sourav Poddar 2013-10-09 15:24 ` Sourav Poddar [not found] ` <1381332284-21822-3-git-send-email-sourav.poddar-l0cyMroinI0@public.gmane.org> 2013-10-09 18:15 ` Jagan Teki 2013-10-09 18:15 ` Jagan Teki [not found] ` <CAD6G_RShZMkSpVzvXWEE0+sDX=pcnf7ndChndgDG5_T4EVL2vQ-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org> 2013-10-11 7:10 ` Gupta, Pekon 2013-10-11 7:10 ` Gupta, Pekon 2013-10-24 1:06 ` Brian Norris 2013-10-24 5:44 ` Sourav Poddar 2013-10-24 7:34 ` Brian Norris 2013-10-24 8:44 ` Sourav Poddar 2013-10-24 17:07 ` Brian Norris 2013-10-24 17:55 ` Sourav Poddar 2013-10-09 15:24 ` [RFC/PATCH 3/3] drivers: mtd: devices: Add memory mapped " Sourav Poddar 2013-10-09 15:24 ` Sourav Poddar 2013-10-09 15:45 ` Mark Brown 2013-10-09 15:45 ` Mark Brown 2013-10-24 0:22 ` [PATCH 0/3]Add quad/memory mapped support for SPI flash Brian Norris 2013-10-24 4:51 ` Sourav Poddar
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=20131011095948.GZ21581@sirena.org.uk \ --to=broonie@kernel.org \ --cc=balbi@ti.com \ --cc=computersforpeace@gmail.com \ --cc=dwmw2@infradead.org \ --cc=linux-mtd@lists.infradead.org \ --cc=peter@korsgaard.com \ --cc=sourav.poddar@ti.com \ --cc=spi-devel-general@lists.sourceforge.net \ --cc=tpiepho@gmail.com \ /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: linkBe 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.