All of lore.kernel.org
 help / color / mirror / Atom feed
From: Bard Liao <yung-chuan.liao@linux.intel.com>
To: linux-sound@vger.kernel.org, vkoul@kernel.org
Cc: vinod.koul@linaro.org, linux-kernel@vger.kernel.org,
	pierre-louis.bossart@linux.intel.com, bard.liao@intel.com
Subject: [PATCH 1/2] soundwire: intel_ace2x: fix wakeup handling
Date: Mon,  8 Apr 2024 06:16:42 +0000	[thread overview]
Message-ID: <20240408061643.420916-2-yung-chuan.liao@linux.intel.com> (raw)
In-Reply-To: <20240408061643.420916-1-yung-chuan.liao@linux.intel.com>

From: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>

The initial programming sequence only worked in the case where the
OFLEN bit is set, i.e. the DSP handles the SoundWire interface. In the
Linux integration, the interface is owned by the host. This disconnect
leads to wake-ups being routed to the DSP and not to the host.

The suggested update is to rely on the global HDAudio WAKEEN/STATESTS
registers, with the SDI bits used to program the wakeups and check the
status.

Note that there is no way to know which peripheral generated a
wake-up. When the hardware detects a change, it sets all the bits
corresponding to LSDIIDx. The LSDIIDx information can be used to
figure out on which link the wakeup happened, but for further details
the software will have to check the status of each peripheral.

Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
---
 drivers/soundwire/intel_ace2x.c | 49 +++++++++++++++++++++++----------
 1 file changed, 35 insertions(+), 14 deletions(-)

diff --git a/drivers/soundwire/intel_ace2x.c b/drivers/soundwire/intel_ace2x.c
index 43a348db83bf..f26d4e5c2c7a 100644
--- a/drivers/soundwire/intel_ace2x.c
+++ b/drivers/soundwire/intel_ace2x.c
@@ -10,8 +10,10 @@
 #include <linux/soundwire/sdw_registers.h>
 #include <linux/soundwire/sdw.h>
 #include <linux/soundwire/sdw_intel.h>
-#include <sound/pcm_params.h>
+#include <sound/hdaudio.h>
 #include <sound/hda-mlink.h>
+#include <sound/hda_register.h>
+#include <sound/pcm_params.h>
 #include "cadence_master.h"
 #include "bus.h"
 #include "intel.h"
@@ -49,37 +51,56 @@ static void intel_shim_vs_set_clock_source(struct sdw_intel *sdw, u32 source)
 
 static int intel_shim_check_wake(struct sdw_intel *sdw)
 {
-	void __iomem *shim_vs;
+	u16 lsdiid = 0;
 	u16 wake_sts;
+	int ret;
 
-	shim_vs = sdw->link_res->shim_vs;
-	wake_sts = intel_readw(shim_vs, SDW_SHIM2_INTEL_VS_WAKESTS);
+	/* find out which bits are set in LSDIID for this sublink */
+	ret = hdac_bus_eml_sdw_get_lsdiid_unlocked(sdw->link_res->hbus, sdw->instance, &lsdiid);
+	if (ret < 0)
+		return ret;
 
-	return wake_sts & SDW_SHIM2_INTEL_VS_WAKEEN_PWS;
+	/*
+	 * we need to use the global HDaudio WAKEEN/STS to be able to detect
+	 * wakes in low-power modes
+	 */
+	wake_sts = snd_hdac_chip_readw(sdw->link_res->hbus, STATESTS);
+
+	return wake_sts & lsdiid;
 }
 
 static void intel_shim_wake(struct sdw_intel *sdw, bool wake_enable)
 {
-	void __iomem *shim_vs = sdw->link_res->shim_vs;
+	u16 lsdiid = 0;
 	u16 wake_en;
 	u16 wake_sts;
+	int ret;
 
-	wake_en = intel_readw(shim_vs, SDW_SHIM2_INTEL_VS_WAKEEN);
+	mutex_lock(sdw->link_res->shim_lock);
+
+	ret = hdac_bus_eml_sdw_get_lsdiid_unlocked(sdw->link_res->hbus, sdw->instance, &lsdiid);
+	if (ret < 0)
+		goto unlock;
+
+	wake_en = snd_hdac_chip_readw(sdw->link_res->hbus, WAKEEN);
 
 	if (wake_enable) {
 		/* Enable the wakeup */
-		wake_en |= SDW_SHIM2_INTEL_VS_WAKEEN_PWE;
-		intel_writew(shim_vs, SDW_SHIM2_INTEL_VS_WAKEEN, wake_en);
+		wake_en |= lsdiid;
+
+		snd_hdac_chip_writew(sdw->link_res->hbus, WAKEEN, wake_en);
 	} else {
 		/* Disable the wake up interrupt */
-		wake_en &= ~SDW_SHIM2_INTEL_VS_WAKEEN_PWE;
-		intel_writew(shim_vs, SDW_SHIM2_INTEL_VS_WAKEEN, wake_en);
+		wake_en &= ~lsdiid;
+		snd_hdac_chip_writew(sdw->link_res->hbus, WAKEEN, wake_en);
 
 		/* Clear wake status (W1C) */
-		wake_sts = intel_readw(shim_vs, SDW_SHIM2_INTEL_VS_WAKESTS);
-		wake_sts |= SDW_SHIM2_INTEL_VS_WAKEEN_PWS;
-		intel_writew(shim_vs, SDW_SHIM2_INTEL_VS_WAKESTS, wake_sts);
+		wake_sts = snd_hdac_chip_readw(sdw->link_res->hbus, STATESTS);
+		wake_sts |= lsdiid;
+		snd_hdac_chip_writew(sdw->link_res->hbus, STATESTS, wake_sts);
 	}
+unlock:
+	mutex_unlock(sdw->link_res->shim_lock);
 }
 
 static int intel_link_power_up(struct sdw_intel *sdw)
-- 
2.34.1


  reply	other threads:[~2024-04-08  6:17 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-04-08  6:16 [PATCH 0/2] soundwire: intel_ace2x: fix wakeup handling for LunarLake Bard Liao
2024-04-08  6:16 ` Bard Liao [this message]
2024-04-08  6:16 ` [PATCH 2/2] soundwire: intel_ace2x: simplify check_wake() Bard Liao

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=20240408061643.420916-2-yung-chuan.liao@linux.intel.com \
    --to=yung-chuan.liao@linux.intel.com \
    --cc=bard.liao@intel.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-sound@vger.kernel.org \
    --cc=pierre-louis.bossart@linux.intel.com \
    --cc=vinod.koul@linaro.org \
    --cc=vkoul@kernel.org \
    /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 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.