From: "Martin Povišer" <povik+lin@cutebit.org> To: "Martin Povišer" <povik+lin@cutebit.org>, "Liam Girdwood" <lgirdwood@gmail.com>, "Mark Brown" <broonie@kernel.org>, "Rob Herring" <robh+dt@kernel.org>, "Krzysztof Kozlowski" <krzysztof.kozlowski+dt@linaro.org>, "Hector Martin" <marcan@marcan.st>, "Sven Peter" <sven@svenpeter.dev>, "Philipp Zabel" <p.zabel@pengutronix.de> Cc: Alyssa Rosenzweig <alyssa@rosenzweig.io>, asahi@lists.linux.dev, alsa-devel@alsa-project.org, devicetree@vger.kernel.org, linux-kernel@vger.kernel.org Subject: [PATCH v2 4/4] ASoC: apple: mca: Add locks on foreign cluster access Date: Fri, 19 Aug 2022 14:54:30 +0200 [thread overview] Message-ID: <20220819125430.4920-5-povik+lin@cutebit.org> (raw) In-Reply-To: <20220819125430.4920-1-povik+lin@cutebit.org> In DAI ops, accesses to the native cluster (of the DAI), and to data of clusters related to it by a DPCM frontend-backend link, should have been synchronized by the 'pcm_mutex' lock at ASoC level. What is not covered are the 'port_driver' accesses on foreign clusters to which the current cluster has no a priori relation, so fill in locking for that. (This should only matter in bizarre configurations of sharing one MCA peripheral between ASoC cards.) Signed-off-by: Martin Povišer <povik+lin@cutebit.org> --- sound/soc/apple/mca.c | 27 ++++++++++++++++++++++++--- 1 file changed, 24 insertions(+), 3 deletions(-) diff --git a/sound/soc/apple/mca.c b/sound/soc/apple/mca.c index 6eeb35d49986..659631772d38 100644 --- a/sound/soc/apple/mca.c +++ b/sound/soc/apple/mca.c @@ -159,6 +159,9 @@ struct mca_data { struct reset_control *rstc; struct device_link *pd_link; + /* Mutex for accessing port_driver of foreign clusters */ + struct mutex port_mutex; + int nclusters; struct mca_cluster clusters[]; }; @@ -297,16 +300,21 @@ static bool mca_fe_clocks_in_use(struct mca_cluster *cl) struct mca_cluster *be_cl; int stream, i; + mutex_lock(&mca->port_mutex); for (i = 0; i < mca->nclusters; i++) { be_cl = &mca->clusters[i]; if (be_cl->port_driver != cl->no) continue; - for_each_pcm_streams(stream) - if (be_cl->clocks_in_use[stream]) + for_each_pcm_streams(stream) { + if (be_cl->clocks_in_use[stream]) { + mutex_unlock(&mca->port_mutex); return true; + } + } } + mutex_unlock(&mca->port_mutex); return false; } @@ -331,8 +339,10 @@ static int mca_be_prepare(struct snd_pcm_substream *substream, */ if (!mca_fe_clocks_in_use(fe_cl)) { ret = mca_fe_enable_clocks(fe_cl); - if (ret < 0) + if (ret < 0) { + mutex_unlock(&mca->port_mutex); return ret; + } } cl->clocks_in_use[substream->stream] = true; @@ -350,6 +360,11 @@ static int mca_be_hw_free(struct snd_pcm_substream *substream, if (cl->port_driver < 0) return -EINVAL; + /* + * We are operating on a foreign cluster here, but since we + * belong to the same PCM, accesses should have been + * synchronized at ASoC level. + */ fe_cl = &mca->clusters[cl->port_driver]; if (!mca_fe_clocks_in_use(fe_cl)) return 0; /* Nothing to do */ @@ -722,7 +737,9 @@ static int mca_be_startup(struct snd_pcm_substream *substream, cl->base + REG_PORT_CLOCK_SEL); writel_relaxed(PORT_DATA_SEL_TXA(fe_cl->no), cl->base + REG_PORT_DATA_SEL); + mutex_lock(&mca->port_mutex); cl->port_driver = fe_cl->no; + mutex_unlock(&mca->port_mutex); cl->port_started[substream->stream] = true; return 0; @@ -732,6 +749,7 @@ static void mca_be_shutdown(struct snd_pcm_substream *substream, struct snd_soc_dai *dai) { struct mca_cluster *cl = mca_dai_to_cluster(dai); + struct mca_data *mca = cl->host; cl->port_started[substream->stream] = false; @@ -742,7 +760,9 @@ static void mca_be_shutdown(struct snd_pcm_substream *substream, */ writel_relaxed(0, cl->base + REG_PORT_ENABLES); writel_relaxed(0, cl->base + REG_PORT_DATA_SEL); + mutex_lock(&mca->port_mutex); cl->port_driver = -1; + mutex_unlock(&mca->port_mutex); } } @@ -963,6 +983,7 @@ static int apple_mca_probe(struct platform_device *pdev) return -ENOMEM; mca->dev = &pdev->dev; mca->nclusters = nclusters; + mutex_init(&mca->port_mutex); platform_set_drvdata(pdev, mca); clusters = mca->clusters; -- 2.33.0
WARNING: multiple messages have this Message-ID (diff)
From: "Martin Povišer" <povik+lin@cutebit.org> To: "Martin Povišer" <povik+lin@cutebit.org>, "Liam Girdwood" <lgirdwood@gmail.com>, "Mark Brown" <broonie@kernel.org>, "Rob Herring" <robh+dt@kernel.org>, "Krzysztof Kozlowski" <krzysztof.kozlowski+dt@linaro.org>, "Hector Martin" <marcan@marcan.st>, "Sven Peter" <sven@svenpeter.dev>, "Philipp Zabel" <p.zabel@pengutronix.de> Cc: devicetree@vger.kernel.org, alsa-devel@alsa-project.org, asahi@lists.linux.dev, Alyssa Rosenzweig <alyssa@rosenzweig.io>, linux-kernel@vger.kernel.org Subject: [PATCH v2 4/4] ASoC: apple: mca: Add locks on foreign cluster access Date: Fri, 19 Aug 2022 14:54:30 +0200 [thread overview] Message-ID: <20220819125430.4920-5-povik+lin@cutebit.org> (raw) In-Reply-To: <20220819125430.4920-1-povik+lin@cutebit.org> In DAI ops, accesses to the native cluster (of the DAI), and to data of clusters related to it by a DPCM frontend-backend link, should have been synchronized by the 'pcm_mutex' lock at ASoC level. What is not covered are the 'port_driver' accesses on foreign clusters to which the current cluster has no a priori relation, so fill in locking for that. (This should only matter in bizarre configurations of sharing one MCA peripheral between ASoC cards.) Signed-off-by: Martin Povišer <povik+lin@cutebit.org> --- sound/soc/apple/mca.c | 27 ++++++++++++++++++++++++--- 1 file changed, 24 insertions(+), 3 deletions(-) diff --git a/sound/soc/apple/mca.c b/sound/soc/apple/mca.c index 6eeb35d49986..659631772d38 100644 --- a/sound/soc/apple/mca.c +++ b/sound/soc/apple/mca.c @@ -159,6 +159,9 @@ struct mca_data { struct reset_control *rstc; struct device_link *pd_link; + /* Mutex for accessing port_driver of foreign clusters */ + struct mutex port_mutex; + int nclusters; struct mca_cluster clusters[]; }; @@ -297,16 +300,21 @@ static bool mca_fe_clocks_in_use(struct mca_cluster *cl) struct mca_cluster *be_cl; int stream, i; + mutex_lock(&mca->port_mutex); for (i = 0; i < mca->nclusters; i++) { be_cl = &mca->clusters[i]; if (be_cl->port_driver != cl->no) continue; - for_each_pcm_streams(stream) - if (be_cl->clocks_in_use[stream]) + for_each_pcm_streams(stream) { + if (be_cl->clocks_in_use[stream]) { + mutex_unlock(&mca->port_mutex); return true; + } + } } + mutex_unlock(&mca->port_mutex); return false; } @@ -331,8 +339,10 @@ static int mca_be_prepare(struct snd_pcm_substream *substream, */ if (!mca_fe_clocks_in_use(fe_cl)) { ret = mca_fe_enable_clocks(fe_cl); - if (ret < 0) + if (ret < 0) { + mutex_unlock(&mca->port_mutex); return ret; + } } cl->clocks_in_use[substream->stream] = true; @@ -350,6 +360,11 @@ static int mca_be_hw_free(struct snd_pcm_substream *substream, if (cl->port_driver < 0) return -EINVAL; + /* + * We are operating on a foreign cluster here, but since we + * belong to the same PCM, accesses should have been + * synchronized at ASoC level. + */ fe_cl = &mca->clusters[cl->port_driver]; if (!mca_fe_clocks_in_use(fe_cl)) return 0; /* Nothing to do */ @@ -722,7 +737,9 @@ static int mca_be_startup(struct snd_pcm_substream *substream, cl->base + REG_PORT_CLOCK_SEL); writel_relaxed(PORT_DATA_SEL_TXA(fe_cl->no), cl->base + REG_PORT_DATA_SEL); + mutex_lock(&mca->port_mutex); cl->port_driver = fe_cl->no; + mutex_unlock(&mca->port_mutex); cl->port_started[substream->stream] = true; return 0; @@ -732,6 +749,7 @@ static void mca_be_shutdown(struct snd_pcm_substream *substream, struct snd_soc_dai *dai) { struct mca_cluster *cl = mca_dai_to_cluster(dai); + struct mca_data *mca = cl->host; cl->port_started[substream->stream] = false; @@ -742,7 +760,9 @@ static void mca_be_shutdown(struct snd_pcm_substream *substream, */ writel_relaxed(0, cl->base + REG_PORT_ENABLES); writel_relaxed(0, cl->base + REG_PORT_DATA_SEL); + mutex_lock(&mca->port_mutex); cl->port_driver = -1; + mutex_unlock(&mca->port_mutex); } } @@ -963,6 +983,7 @@ static int apple_mca_probe(struct platform_device *pdev) return -ENOMEM; mca->dev = &pdev->dev; mca->nclusters = nclusters; + mutex_init(&mca->port_mutex); platform_set_drvdata(pdev, mca); clusters = mca->clusters; -- 2.33.0
next prev parent reply other threads:[~2022-08-19 12:55 UTC|newest] Thread overview: 40+ messages / expand[flat|nested] mbox.gz Atom feed top 2022-08-19 12:54 [PATCH v2 0/4] ASoC platform driver for Apple MCA Martin Povišer 2022-08-19 12:54 ` Martin Povišer 2022-08-19 12:54 ` [PATCH v2 1/4] dt-bindings: sound: Add Apple MCA I2S transceiver Martin Povišer 2022-08-19 12:54 ` Martin Povišer 2022-08-19 12:58 ` Martin Povišer 2022-08-19 12:58 ` Martin Povišer 2022-08-19 13:08 ` Krzysztof Kozlowski 2022-08-19 13:08 ` Krzysztof Kozlowski 2022-08-19 13:09 ` Krzysztof Kozlowski 2022-08-19 13:09 ` Krzysztof Kozlowski 2022-08-19 14:14 ` Martin Povišer 2022-08-19 14:14 ` Martin Povišer 2022-08-19 14:17 ` Krzysztof Kozlowski 2022-08-19 14:17 ` Krzysztof Kozlowski 2022-08-22 18:59 ` Rob Herring 2022-08-22 18:59 ` Rob Herring 2022-08-19 12:54 ` [PATCH v2 2/4] arm64: dts: apple: t8103: Add MCA and its support Martin Povišer 2022-08-19 12:54 ` Martin Povišer 2022-08-19 13:12 ` Krzysztof Kozlowski 2022-08-19 13:12 ` Krzysztof Kozlowski 2022-08-19 14:24 ` Martin Povišer 2022-08-19 14:24 ` Martin Povišer 2022-08-23 11:02 ` Krzysztof Kozlowski 2022-08-23 11:02 ` Krzysztof Kozlowski 2022-08-19 12:54 ` [PATCH v2 3/4] ASoC: apple: mca: Start new platform driver Martin Povišer 2022-08-19 12:54 ` Martin Povišer 2022-08-22 17:39 ` Mark Brown 2022-08-22 17:39 ` Mark Brown 2022-08-23 7:33 ` Martin Povišer 2022-08-23 7:33 ` Martin Povišer 2022-08-23 11:31 ` Mark Brown 2022-08-23 11:31 ` Mark Brown 2022-08-23 11:51 ` Martin Povišer 2022-08-23 11:51 ` Martin Povišer 2022-08-23 11:53 ` Martin Povišer 2022-08-23 11:53 ` Martin Povišer 2022-08-19 12:54 ` Martin Povišer [this message] 2022-08-19 12:54 ` [PATCH v2 4/4] ASoC: apple: mca: Add locks on foreign cluster access Martin Povišer 2022-08-22 17:45 ` Mark Brown 2022-08-22 17:45 ` 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=20220819125430.4920-5-povik+lin@cutebit.org \ --to=povik+lin@cutebit.org \ --cc=alsa-devel@alsa-project.org \ --cc=alyssa@rosenzweig.io \ --cc=asahi@lists.linux.dev \ --cc=broonie@kernel.org \ --cc=devicetree@vger.kernel.org \ --cc=krzysztof.kozlowski+dt@linaro.org \ --cc=lgirdwood@gmail.com \ --cc=linux-kernel@vger.kernel.org \ --cc=marcan@marcan.st \ --cc=p.zabel@pengutronix.de \ --cc=robh+dt@kernel.org \ --cc=sven@svenpeter.dev \ /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.