linux-kernel - [PATCH 2/9] ASoC: apple: mca: use readx_poll

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <20250518-mca-fixes-v1-2-ee1015a695f6@gmail.com>
Date: Sun, 18 May 2025 20:50:47 +1000
From: James Calligeros <jcalligeros99@...il.com>
To: Martin Povišer <povik+lin@...ebit.org>, 
 Liam Girdwood <lgirdwood@...il.com>, Mark Brown <broonie@...nel.org>, 
 Jaroslav Kysela <perex@...ex.cz>, Takashi Iwai <tiwai@...e.com>
Cc: asahi@...ts.linux.dev, linux-sound@...r.kernel.org, 
 linux-kernel@...r.kernel.org, James Calligeros <jcalligeros99@...il.com>
Subject: [PATCH 2/9] ASoC: apple: mca: use readx_poll_timeout to check for
 cluster reset

MCA clusters should take no longer than 1 microsecond to reset, however
it has been observed to take longer on very rare occasions.

Rather than just add an unreasonably long usleep(), use readx_poll_timeout
to poll the status register for a clear reset bit. This lets us have a
very safe maximum wait time, but continue early if we are ready to do so.

Signed-off-by: James Calligeros <jcalligeros99@...il.com>
---
 sound/soc/apple/mca.c | 14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)

diff --git a/sound/soc/apple/mca.c b/sound/soc/apple/mca.c
index 5dd24ab90d0f052bb48f451cf009dc2e9128014d..7113da4bdea7b687c0d44d2bbf3a511b8299056f 100644
--- a/sound/soc/apple/mca.c
+++ b/sound/soc/apple/mca.c
@@ -24,6 +24,7 @@
 #include <linux/clk.h>
 #include <linux/dma-mapping.h>
 #include <linux/init.h>
+#include <linux/iopoll.h>
 #include <linux/kernel.h>
 #include <linux/module.h>
 #include <linux/of.h>
@@ -197,6 +198,7 @@ static void mca_fe_early_trigger(struct snd_pcm_substream *substream, int cmd,
 	int serdes_unit = is_tx ? CLUSTER_TX_OFF : CLUSTER_RX_OFF;
 	int serdes_conf =
 		serdes_unit + (is_tx ? REG_TX_SERDES_CONF : REG_RX_SERDES_CONF);
+	int ret, status;
 
 	switch (cmd) {
 	case SNDRV_PCM_TRIGGER_START:
@@ -211,11 +213,15 @@ static void mca_fe_early_trigger(struct snd_pcm_substream *substream, int cmd,
 			   SERDES_STATUS_RST);
 		/*
 		 * Experiments suggest that it takes at most ~1 us
-		 * for the bit to clear, so wait 2 us for good measure.
+		 * for the bit to clear, however this has been seen to fail.
+		 * Wait up to 50 us for the reset bit to clear.
 		 */
-		udelay(2);
-		WARN_ON(readl_relaxed(cl->base + serdes_unit + REG_SERDES_STATUS) &
-			SERDES_STATUS_RST);
+		ret = readx_poll_timeout(readl_relaxed,
+					 cl->base + serdes_unit + REG_SERDES_STATUS,
+					 status, !(status & SERDES_STATUS_RST), 2, 50);
+		if (ret || (status & SERDES_STATUS_RST))
+			dev_warn(cl->host->dev, "MCA cluster failed to reset\n");
+
 		mca_modify(cl, serdes_conf, SERDES_CONF_SYNC_SEL,
 			   FIELD_PREP(SERDES_CONF_SYNC_SEL, 0));
 		mca_modify(cl, serdes_conf, SERDES_CONF_SYNC_SEL,

-- 
2.49.0