From: slash.tmp@free•fr (Mason)
To: linux-arm-kernel@lists•infradead.org
Subject: Tearing down DMA transfer setup after DMA client has finished
Date: Thu, 24 Nov 2016 16:20:16 +0100 [thread overview]
Message-ID: <58370530.5080605@free.fr> (raw)
In-Reply-To: <yw1xr3619d3l.fsf@unicorn.mansr.com>
On 24/11/2016 15:17, M?ns Rullg?rd wrote:
> Mason wrote:
>
>> [ 35.085854] SETUP DMA
>> [ 35.088272] START NAND TRANSFER
>> [ 35.091670] tangox_dma_pchan_start from tangox_dma_irq
>> [ 35.096882] tango_dma_callback from vchan_complete
>> [ 45.102513] DONE FAKE SPINNING
>>
>> So the IRQ rolls in, the ISR calls tangox_dma_pchan_start,
>> which calls tangox_dma_pchan_detach to tear down the sbox
>> setup; and only sometime later does the DMA framework call
>> my callback function.
>
> Yes, I realised this soon after I said it. The dma driver could be
> rearranged to make it work though.
There is a way to make the tasklet run and invoke the callback
before the interrupt service routine proceeds? Can you say more
about this?
>> So far, the work-arounds I've tested are:
>>
>> 1) delay sbox tear-down by 10 ?s in tangox_dma_pchan_detach.
>> 2) statically setup sbox in probe, and never touch it henceforth.
>>
>> WA1 is fragile, it might break for devices other than NFC.
>> WA2 is what I used when I wrote the NFC driver.
>>
>> Can tangox_dma_irq() be changed to have the framework call
>> the client's callback *before* tangox_dma_pchan_start?
>>
>> (Thinking out loud) The DMA_PREP_INTERRUPT requests that the
>> DMA framework invoke the callback from tasklet context,
>> maybe a different flag DMA_PREP_INTERRUPT_EX can request
>> calling the call-back directly from within the ISR?
>>
>> (Looking at existing flags) Could I use DMA_CTRL_ACK?
>> Description sounds like some kind hand-shake between
>> client and dmaengine.
>>
>> Grepping for DMA_PREP_INTERRUPT, I don't see where the framework
>> checks that flag to spawn the tasklet? Or is that up to each
>> driver individually?
>
> Those flags all have defined meanings and abusing them for other things
> is a bad idea. As far as possible, device drivers should work with any
> dma driver.
I was asking about introducing a new flag, not abusing existing
flags. (I don't understand the semantics of DMA_CTRL_ACK.)
(FWIW, both the NFC and the MBUS agent are custom designs,
not third-party IP blocks.)
Regards.
next prev parent reply other threads:[~2016-11-24 15:20 UTC|newest]
Thread overview: 82+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-11-23 10:25 Tearing down DMA transfer setup after DMA client has finished Mason
2016-11-23 12:13 ` Måns Rullgård
2016-11-23 12:41 ` Mason
2016-11-23 17:21 ` Måns Rullgård
2016-11-24 10:53 ` Mason
2016-11-24 14:17 ` Måns Rullgård
2016-11-24 15:20 ` Mason [this message]
2016-11-24 16:37 ` Måns Rullgård
2016-11-25 4:55 ` Vinod Koul
2016-11-25 11:57 ` Måns Rullgård
2016-11-25 14:05 ` Mason
2016-11-25 14:12 ` Måns Rullgård
2016-11-25 14:28 ` Mason
2016-11-25 14:42 ` Måns Rullgård
2016-11-25 12:45 ` Russell King - ARM Linux
2016-11-25 13:07 ` Måns Rullgård
2016-11-25 13:34 ` Russell King - ARM Linux
2016-11-25 13:50 ` Måns Rullgård
2016-11-25 13:58 ` Russell King - ARM Linux
2016-11-25 14:03 ` Måns Rullgård
2016-11-25 14:17 ` Russell King - ARM Linux
2016-11-25 14:40 ` Måns Rullgård
2016-11-25 14:56 ` Russell King - ARM Linux
2016-11-25 15:08 ` Måns Rullgård
2016-11-25 15:02 ` Mason
2016-11-25 15:12 ` Måns Rullgård
2016-11-25 15:21 ` Mason
2016-11-25 15:28 ` Måns Rullgård
2016-11-25 12:46 ` Mason
2016-11-25 13:11 ` Måns Rullgård
2016-11-25 14:21 ` Mason
2016-11-25 14:37 ` Måns Rullgård
2016-11-25 15:35 ` Mason
2016-11-29 18:25 ` Mason
2016-12-06 5:12 ` Vinod Koul
2016-12-06 12:42 ` Mason
2016-12-06 13:14 ` Måns Rullgård
2016-12-06 15:24 ` Mason
2016-12-06 15:34 ` Måns Rullgård
2016-12-06 22:55 ` Mason
2016-12-07 16:43 ` Vinod Koul
2016-12-07 16:45 ` Måns Rullgård
2016-12-08 10:39 ` Vinod Koul
2016-12-08 10:54 ` Mason
2016-12-08 11:18 ` Geert Uytterhoeven
2016-12-08 11:47 ` Måns Rullgård
2016-12-08 12:03 ` Geert Uytterhoeven
2016-12-08 12:17 ` Mason
2016-12-08 12:21 ` Måns Rullgård
2016-12-08 16:37 ` Vinod Koul
2016-12-08 16:48 ` Måns Rullgård
2016-12-09 6:59 ` Vinod Koul
2016-12-09 10:25 ` Sebastian Frias
2016-12-09 11:34 ` Måns Rullgård
2016-12-09 11:35 ` 1Måns Rullgård
2016-12-09 17:17 ` Vinod Koul
2016-12-09 17:28 ` Måns Rullgård
2016-12-09 17:53 ` Vinod Koul
2016-12-09 17:34 ` Mason
2016-12-09 17:56 ` Vinod Koul
2016-12-09 18:17 ` Vinod Koul
2016-12-09 18:23 ` Mason
2016-12-12 5:01 ` Vinod Koul
2016-12-15 11:17 ` Mark Brown
2016-12-08 11:44 ` Måns Rullgård
2016-12-08 11:59 ` Geert Uytterhoeven
2016-12-08 12:20 ` Måns Rullgård
2016-12-08 12:31 ` Geert Uytterhoeven
2016-12-08 12:41 ` Mason
2016-12-08 12:44 ` Måns Rullgård
2016-12-08 13:29 ` Mason
2016-12-08 13:39 ` Måns Rullgård
2016-12-08 15:50 ` Vinod Koul
2016-12-08 16:36 ` Måns Rullgård
2016-12-08 15:40 ` Vinod Koul
2016-12-08 15:43 ` Mason
2016-12-08 16:21 ` Vinod Koul
2016-12-08 16:46 ` Måns Rullgård
2016-12-07 16:41 ` Vinod Koul
2016-12-07 16:44 ` Måns Rullgård
2016-12-08 10:37 ` Vinod Koul
2016-12-08 11:44 ` Måns Rullgård
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=58370530.5080605@free.fr \
--to=slash.tmp@free$(echo .)fr \
--cc=linux-arm-kernel@lists$(echo .)infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
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