mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Chancel Liu <chancel.liu@oss.nxp.com>
To: Mark Brown <broonie@kernel.org>,
	Liam Girdwood <lgirdwood@gmail.com>,
	Jaroslav Kysela <perex@perex.cz>, Takashi Iwai <tiwai@suse.com>,
	Charles Keepax <ckeepax@opensource.cirrus.com>,
	Oder Chiou <oder_chiou@realtek.com>,
	Cezary Rojewski <cezary.rojewski@intel.com>
Cc: Support Opensource <support.opensource@diasemi.com>,
	linux-sound@vger.kernel.org, linux-kernel@vger.kernel.org,
	patches@opensource.cirrus.com
Subject: [PATCH v2 18/21] ASoC: codecs: rt5677-spi: Free the DSP context on component remove
Date: Wed, 23 Sep 2026 15:13:02 +0900	[thread overview]
Message-ID: <20260923061305.4041197-19-chancel.liu@oss.nxp.com> (raw)
In-Reply-To: <20260923061305.4041197-1-chancel.liu@oss.nxp.com>

From: Chancel Liu <chancel.liu@nxp.com>

component->dev is the underlying SPI bus device, whose devres lifetime
follows the physical devices probe/remove rather than the ASoC cards
bind/unbind. The rt5677_dsp context allocated in the component probe with
devm_kzalloc(component->dev, ...) therefore leaks on every card
bind/unbind cycle, and the delayed work initialised there is never
cancelled on unbind.

Allocate the context with kzalloc() and add a component .remove callback
that cancels the copy work, destroys the mutex and frees the context.

The exported rt5677_spi_hotword_detected(), called from the rt5677 codec
interrupt path, fetches the context via dev_get_drvdata(&g_spi->dev) and
dereferences it. Freeing the context on remove would let that callback
run concurrently and touch freed memory. Add a dsp_lock that serializes
the fetch/use in rt5677_spi_hotword_detected() against clearing the
drvdata in the remove callback, so once remove has cleared the pointer
the callback either observes NULL and bails out or has already finished.

Signed-off-by: Chancel Liu <chancel.liu@nxp.com>
---
 sound/soc/codecs/rt5677-spi.c | 21 +++++++++++++++++++--
 1 file changed, 19 insertions(+), 2 deletions(-)

diff --git a/sound/soc/codecs/rt5677-spi.c b/sound/soc/codecs/rt5677-spi.c
index ebc527115ea5..0b2166082437 100644
--- a/sound/soc/codecs/rt5677-spi.c
+++ b/sound/soc/codecs/rt5677-spi.c
@@ -58,6 +58,8 @@
 
 static struct spi_device *g_spi;
 static DEFINE_MUTEX(spi_mutex);
+/* Serializes access to the DSP context against the exported hotword callback */
+static DEFINE_MUTEX(dsp_lock);
 
 struct rt5677_dsp {
 	struct device *dev;
@@ -380,8 +382,7 @@ static int rt5677_spi_pcm_probe(struct snd_soc_component *component)
 {
 	struct rt5677_dsp *rt5677_dsp;
 
-	rt5677_dsp = devm_kzalloc(component->dev, sizeof(*rt5677_dsp),
-			GFP_KERNEL);
+	rt5677_dsp = kzalloc_obj(*rt5677_dsp);
 	if (!rt5677_dsp)
 		return -ENOMEM;
 	rt5677_dsp->dev = &g_spi->dev;
@@ -392,9 +393,23 @@ static int rt5677_spi_pcm_probe(struct snd_soc_component *component)
 	return 0;
 }
 
+static void rt5677_spi_pcm_remove(struct snd_soc_component *component)
+{
+	struct rt5677_dsp *rt5677_dsp =
+			snd_soc_component_get_drvdata(component);
+
+	scoped_guard(mutex, &dsp_lock)
+		snd_soc_component_set_drvdata(component, NULL);
+
+	cancel_delayed_work_sync(&rt5677_dsp->copy_work);
+	mutex_destroy(&rt5677_dsp->dma_lock);
+	kfree(rt5677_dsp);
+}
+
 static const struct snd_soc_component_driver rt5677_spi_dai_component = {
 	.name			= DRV_NAME,
 	.probe			= rt5677_spi_pcm_probe,
+	.remove			= rt5677_spi_pcm_remove,
 	.open			= rt5677_spi_pcm_open,
 	.close			= rt5677_spi_pcm_close,
 	.hw_params		= rt5677_spi_hw_params,
@@ -579,6 +594,8 @@ void rt5677_spi_hotword_detected(void)
 	if (!g_spi)
 		return;
 
+	guard(mutex)(&dsp_lock);
+
 	rt5677_dsp = dev_get_drvdata(&g_spi->dev);
 	if (!rt5677_dsp) {
 		dev_err(&g_spi->dev, "Can't get rt5677_dsp\n");
-- 
2.50.1


  parent reply	other threads:[~2026-09-23  6:15 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-09-23  6:12 [PATCH v2 00/21] ASoC: codecs: Fix resource leaks across card bind/unbind Chancel Liu
2026-09-23  6:12 ` [PATCH v2 01/21] ASoC: codecs: cpcap: Fix devm " Chancel Liu
2026-09-23  6:12 ` [PATCH v2 02/21] ASoC: codecs: da7218: " Chancel Liu
2026-09-23  6:12 ` [PATCH v2 03/21] ASoC: codecs: tlv320aic32x4: Fix clock leak from runtime callbacks Chancel Liu
2026-09-23  6:12 ` [PATCH v2 04/21] ASoC: codecs: twl4030: Acquire board params and hs_extmute GPIO in the platform probe Chancel Liu
2026-09-23  6:12 ` [PATCH v2 05/21] ASoC: codecs: es8316: Move mclk acquisition to the i2c probe Chancel Liu
2026-09-23  8:19   ` Cezary Rojewski
2026-09-23  6:12 ` [PATCH v2 06/21] ASoC: codecs: es8323: " Chancel Liu
2026-09-23  6:12 ` [PATCH v2 07/21] ASoC: codecs: es8311: " Chancel Liu
2026-09-23  6:12 ` [PATCH v2 08/21] ASoC: codecs: es8328: Move clk acquisition to the bus probe Chancel Liu
2026-09-23  6:12 ` [PATCH v2 09/21] ASoC: codecs: rt5640: Move mclk acquisition to the i2c probe Chancel Liu
2026-09-23  8:16   ` Cezary Rojewski
2026-09-23  8:22     ` Chancel Liu
2026-09-23  6:12 ` [PATCH v2 10/21] ASoC: codecs: rt5616: " Chancel Liu
2026-09-23  6:12 ` [PATCH v2 11/21] ASoC: codecs: rt5514: Move clk " Chancel Liu
2026-09-23  8:19   ` Cezary Rojewski
2026-09-23  6:12 ` [PATCH v2 12/21] ASoC: codecs: rt5682s: Move mclk " Chancel Liu
2026-09-23  6:12 ` [PATCH v2 13/21] ASoC: codecs: max98090: " Chancel Liu
2026-09-23  6:12 ` [PATCH v2 14/21] ASoC: codecs: max98095: " Chancel Liu
2026-09-23  6:12 ` [PATCH v2 15/21] ASoC: codecs: wm8985: Move regulator acquisition to the bus probe Chancel Liu
2026-09-23  8:56   ` Charles Keepax
2026-09-23  6:13 ` [PATCH v2 16/21] ASoC: codecs: wm8955: Move regulator acquisition to the i2c probe Chancel Liu
2026-09-23  6:13 ` [PATCH v2 17/21] ASoC: codecs: es8389: Move regulator and mclk " Chancel Liu
2026-09-23  6:13 ` Chancel Liu [this message]
2026-09-23  6:13 ` [PATCH v2 19/21] ASoC: codecs: rt1011: Free the bq/drc coefficient arrays on component remove Chancel Liu
2026-09-23  6:13 ` [PATCH v2 20/21] ASoC: codecs: rt5645: Free the hardware EQ parameters " Chancel Liu
2026-09-23  6:13 ` [PATCH v2 21/21] ASoC: codecs: rt5514-spi: Free the DSP context " Chancel Liu
2026-09-23 12:27 ` [PATCH v2 00/21] ASoC: codecs: Fix resource leaks across card bind/unbind Mark Brown

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=20260923061305.4041197-19-chancel.liu@oss.nxp.com \
    --to=chancel.liu@oss.nxp.com \
    --cc=broonie@kernel.org \
    --cc=cezary.rojewski@intel.com \
    --cc=ckeepax@opensource.cirrus.com \
    --cc=lgirdwood@gmail.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-sound@vger.kernel.org \
    --cc=oder_chiou@realtek.com \
    --cc=patches@opensource.cirrus.com \
    --cc=perex@perex.cz \
    --cc=support.opensource@diasemi.com \
    --cc=tiwai@suse.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

all inboxes | Powered by JetHome®