linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Tomasz Figa <tomasz.figa@gmail.com>
To: linux-samsung-soc@vger.kernel.org
Cc: linux-arm-kernel@lists.infradead.org,
	alsa-devel@alsa-project.org, linux-kernel@vger.kernel.org,
	linux-spi@vger.kernel.org, Dan Williams <djbw@fb.com>,
	Jaroslav Kysela <perex@perex.cz>,
	Kukjin Kim <kgene.kim@samsung.com>,
	Liam Girdwood <lgirdwood@gmail.com>,
	Linus Walleij <linus.walleij@linaro.org>,
	Mark Brown <broonie@kernel.org>,
	Mike Turquette <mturquette@linaro.org>,
	Padmavathi Venna <padma.v@samsung.com>,
	Russell King <linux@arm.linux.org.uk>,
	Sangbeom Kim <sbkim73@samsung.com>, Takashi Iwai <tiwai@suse.de>,
	Vinod Koul <vinod.koul@intel.com>,
	Tomasz Figa <tomasz.figa@gmail.com>
Subject: [PATCH 09/18] ASoC: Samsung: Do not queue cyclic buffers multiple times
Date: Sun, 11 Aug 2013 19:59:21 +0200	[thread overview]
Message-ID: <1376243970-6489-10-git-send-email-tomasz.figa@gmail.com> (raw)
In-Reply-To: <1376243970-6489-1-git-send-email-tomasz.figa@gmail.com>

The legacy S3C-DMA API required every period of a cyclic buffer to be
queued separately. After conversion of Samsung ASoC to Samsung DMA
wrappers somebody made an assumption that the same is needed for DMA
engine API, which is not true.

In effect, Samsung ASoC DMA code was queuing the whole cyclic buffer
multiple times with a shift of one period per iteration, leading to:
  a) severe memory waste - up to 13x times more DMA transfer descriptors
     are allocated than needed,
  b) possible memory corruption, because further cyclic buffers were out
     of the original buffers, due to the offset.

This patch fixes this problem by making the legacy S3C-DMA API use the
same semantics as DMA engine (the whole cyclic buffer is enqueued at
once) and modifying users of Samsung DMA wrappers in cyclic mode to
behave appropriately.

Signed-off-by: Tomasz Figa <tomasz.figa@gmail.com>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
---
 arch/arm/plat-samsung/s3c-dma-ops.c | 13 +++++++++++--
 sound/soc/samsung/dma.c             |  7 +++++++
 2 files changed, 18 insertions(+), 2 deletions(-)

diff --git a/arch/arm/plat-samsung/s3c-dma-ops.c b/arch/arm/plat-samsung/s3c-dma-ops.c
index 0cc40ae..98b10ba 100644
--- a/arch/arm/plat-samsung/s3c-dma-ops.c
+++ b/arch/arm/plat-samsung/s3c-dma-ops.c
@@ -82,7 +82,8 @@ static int s3c_dma_config(unsigned ch, struct samsung_dma_config *param)
 static int s3c_dma_prepare(unsigned ch, struct samsung_dma_prep *param)
 {
 	struct cb_data *data;
-	int len = (param->cap == DMA_CYCLIC) ? param->period : param->len;
+	dma_addr_t pos = param->buf;
+	dma_addr_t end = param->buf + param->len;
 
 	list_for_each_entry(data, &dma_list, node)
 		if (data->ch == ch)
@@ -94,7 +95,15 @@ static int s3c_dma_prepare(unsigned ch, struct samsung_dma_prep *param)
 		data->fp_param = param->fp_param;
 	}
 
-	s3c2410_dma_enqueue(ch, (void *)data, param->buf, len);
+	if (param->cap != DMA_CYCLIC) {
+		s3c2410_dma_enqueue(ch, (void *)data, param->buf, param->len);
+		return 0;
+	}
+
+	while (pos < end) {
+		s3c2410_dma_enqueue(ch, (void *)data, pos, param->period);
+		pos += param->period;
+	}
 
 	return 0;
 }
diff --git a/sound/soc/samsung/dma.c b/sound/soc/samsung/dma.c
index 21b7926..6e2b2b4 100644
--- a/sound/soc/samsung/dma.c
+++ b/sound/soc/samsung/dma.c
@@ -90,6 +90,13 @@ static void dma_enqueue(struct snd_pcm_substream *substream)
 	dma_info.period = prtd->dma_period;
 	dma_info.len = prtd->dma_period*limit;
 
+	if (dma_info.cap == DMA_CYCLIC) {
+		dma_info.buf = pos;
+		prtd->params->ops->prepare(prtd->params->ch, &dma_info);
+		prtd->dma_loaded += limit;
+		return;
+	}
+
 	while (prtd->dma_loaded < limit) {
 		pr_debug("dma_loaded: %d\n", prtd->dma_loaded);
 
-- 
1.8.3.2


  parent reply	other threads:[~2013-08-11 18:02 UTC|newest]

Thread overview: 45+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-08-11 17:59 [PATCH 00/18] ARM: s3c64xx: Let amba-pl08x driver handle DMA Tomasz Figa
2013-08-11 17:59 ` [PATCH 01/18] dmaengine: PL08x: Refactor pl08x_getbytes_chan() to lower indentation Tomasz Figa
2013-08-11 17:59 ` [PATCH 02/18] dmaengine: PL08x: Add support for different offset of CONFIG register Tomasz Figa
2013-08-11 17:59 ` [PATCH 03/18] dmaengine: PL08x: Rework LLI handling to be less fragile Tomasz Figa
2013-08-28  6:49   ` Linus Walleij
2013-08-11 17:59 ` [PATCH 04/18] dmaengine: PL08x: Move LLI dumping code into separate function Tomasz Figa
2013-08-28  6:50   ` Linus Walleij
2013-08-11 17:59 ` [PATCH 05/18] dmaengine: PL08x: Add support for PL080S variant Tomasz Figa
2013-08-11 17:59 ` [PATCH 06/18] dmaengine: PL08x: Add support for different maximum transfer size Tomasz Figa
2013-08-11 17:59 ` [PATCH 07/18] dmaengine: PL08x: Fix reading the byte count in cctl Tomasz Figa
2013-08-11 17:59 ` [PATCH 08/18] dmaengine: PL08x: Add cyclic transfer support Tomasz Figa
2013-08-11 17:59 ` Tomasz Figa [this message]
2013-08-28  7:06   ` [PATCH 09/18] ASoC: Samsung: Do not queue cyclic buffers multiple times Linus Walleij
2013-08-28  8:23     ` Tomasz Figa
2013-08-28  9:53     ` Mark Brown
2013-08-29 16:36   ` [alsa-devel] " Mark Brown
2013-08-29 17:36     ` Linus Walleij
2013-08-11 17:59 ` [PATCH 10/18] spi: s3c64xx: Do not require legacy DMA API in case of S3C64XX Tomasz Figa
2013-08-12 23:36   ` Mark Brown
2013-08-12 23:49     ` Tomasz Figa
2013-08-13 12:02       ` Mark Brown
2013-08-13 18:55       ` Mark Brown
2013-08-13 20:18         ` Mark Brown
2013-08-13 20:20           ` Tomasz Figa
2013-08-13 20:29             ` Mark Brown
2013-08-11 17:59 ` [PATCH 11/18] ASoC: Samsung: " Tomasz Figa
2013-08-11 17:59 ` [PATCH 12/18] ARM: s3c64xx: Add support for DMA using generic amba-pl08x driver Tomasz Figa
2013-08-13 17:52   ` Mark Brown
2013-08-13 20:10     ` Tomasz Figa
2013-08-13 20:35       ` Mark Brown
2013-08-13 21:11         ` Tomasz Figa
2013-08-24  0:17   ` [PATCH v2 " Tomasz Figa
2013-08-11 17:59 ` [PATCH 13/18] ARM: s3c64xx: clock: Add clkdev lookup for DMA clocks Tomasz Figa
2013-08-11 17:59 ` [PATCH 14/18] clk: samsung: s3c64xx: Add aliases " Tomasz Figa
2013-08-12 19:45   ` Mike Turquette
2013-08-11 17:59 ` [PATCH 15/18] ARM: s3c64xx: Remove legacy DMA driver Tomasz Figa
2013-08-11 17:59 ` [PATCH 16/18] clk: samsung: s3c64xx: Remove clock aliases of old " Tomasz Figa
2013-08-12 19:45   ` Mike Turquette
2013-08-11 17:59 ` [PATCH 17/18] spi: s3c64xx: Always select S3C64XX_PL080 when ARCH_S3C64XX is enabled Tomasz Figa
2013-08-11 19:11   ` Mark Brown
2013-08-11 23:03     ` Tomasz Figa
2013-08-12 10:07       ` Mark Brown
2013-08-11 17:59 ` [PATCH 18/18] ASoC: Samsung: " Tomasz Figa
2013-08-15 11:50 ` [PATCH 00/18] ARM: s3c64xx: Let amba-pl08x driver handle DMA Tomasz Figa
2013-09-02  6:16 ` Vinod Koul

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=1376243970-6489-10-git-send-email-tomasz.figa@gmail.com \
    --to=tomasz.figa@gmail.com \
    --cc=alsa-devel@alsa-project.org \
    --cc=broonie@kernel.org \
    --cc=djbw@fb.com \
    --cc=kgene.kim@samsung.com \
    --cc=lgirdwood@gmail.com \
    --cc=linus.walleij@linaro.org \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-samsung-soc@vger.kernel.org \
    --cc=linux-spi@vger.kernel.org \
    --cc=linux@arm.linux.org.uk \
    --cc=mturquette@linaro.org \
    --cc=padma.v@samsung.com \
    --cc=perex@perex.cz \
    --cc=sbkim73@samsung.com \
    --cc=tiwai@suse.de \
    --cc=vinod.koul@intel.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: 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).