mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Mukesh Ojha <mukesh.ojha@oss.qualcomm.com>
To: Sumit Garg <sumit.garg@oss.qualcomm.com>,
	Bjorn Andersson <andersson@kernel.org>,
	Konrad Dybcio <konradybcio@kernel.org>,
	Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: Abel Vesa <abelvesa@kernel.org>,
	linux-arm-msm@vger.kernel.org, linux-kernel@vger.kernel.org,
	linux-remoteproc@vger.kernel.org,
	Mukesh Ojha <mukesh.ojha@oss.qualcomm.com>
Subject: [PATCH v8] firmware: qcom: pas: Make MDT buffer retention explicit
Date: Fri, 25 Sep 2026 22:22:31 +0530	[thread overview]
Message-ID: <20260925165231.10722-1-mukesh.ojha@oss.qualcomm.com> (raw)

The PAS image initialization path retains metadata whenever a valid
qcom_pas_context is supplied, although callers may not need the metadata
after initialization. This leaves ownership implicit and causes unused
metadata buffers to remain allocated.

Add keep_mdt_buf to struct qcom_pas_context and have the generic PAS
wrapper release metadata unless the caller requests retention. This applies
uniformly to SCM and OP-TEE backends, while qcom_q6v5_pas keeps metadata
for subsequent PAS operations.

Signed-off-by: Mukesh Ojha <mukesh.ojha@oss.qualcomm.com>
---
Changes in v8:
 - Commit rephrased a bit.
 - Make the commit aware of all present backends(scm and tee).
 - Link to v7: https://lore.kernel.org/lkml/20260920074919.3338358-1-mukesh.ojha@oss.qualcomm.com/

Changes in v7:
- Since qcom_scm_pas_context is removed, v6 needs to be rebased removing
  the changes related to it.
- Link to v6: https://lore.kernel.org/lkml/20260805133759.2790755-1-mukesh.ojha@oss.qualcomm.com/

Changes in v6:
   - Other patches from the series is merged.
   - Removed dependency on below series by adding keep_mdt_buf
     into qcom_scm_pas_context structure as well.
     https://lore.kernel.org/lkml/20260702115835.167602-1-sumit.garg@kernel.org/
   - Addressed minor comment on the documentationi.
    - Link to v5: 6/6 of https://lore.kernel.org/lkml/20260724182858.1868271-7-mukesh.ojha@oss.qualcomm.com/



 drivers/firmware/qcom/qcom_pas.c       | 12 +++++++++---
 drivers/firmware/qcom/qcom_scm.c       |  2 +-
 drivers/remoteproc/qcom_q6v5_pas.c     |  3 +++
 include/linux/firmware/qcom/qcom_pas.h |  1 +
 4 files changed, 14 insertions(+), 4 deletions(-)

diff --git a/drivers/firmware/qcom/qcom_pas.c b/drivers/firmware/qcom/qcom_pas.c
index 24485dd0fa10..93d78a5c1201 100644
--- a/drivers/firmware/qcom/qcom_pas.c
+++ b/drivers/firmware/qcom/qcom_pas.c
@@ -60,17 +60,23 @@ EXPORT_SYMBOL_GPL(devm_qcom_pas_context_alloc);
  *
  * Return: 0 on success.
  *
- * Upon successful return, the PAS metadata context (@ctx) will be used to
- * track the metadata allocation, this needs to be released by invoking
+ * If @ctx requests metadata retention, the PAS metadata context will track
+ * the allocation, which needs to be released by invoking
  * qcom_pas_metadata_release() by the caller.
  */
 int qcom_pas_init_image(u32 pas_id, const void *metadata, size_t size,
 			struct qcom_pas_context *ctx)
 {
+	int ret;
+
 	if (!ops_ptr)
 		return -ENODEV;
 
-	return ops_ptr->init_image(ops_ptr->dev, pas_id, metadata, size, ctx);
+	ret = ops_ptr->init_image(ops_ptr->dev, pas_id, metadata, size, ctx);
+	if (!ret && ctx && !ctx->keep_mdt_buf)
+		qcom_pas_metadata_release(ctx);
+
+	return ret;
 }
 EXPORT_SYMBOL_GPL(qcom_pas_init_image);
 
diff --git a/drivers/firmware/qcom/qcom_scm.c b/drivers/firmware/qcom/qcom_scm.c
index df20773dd754..dc57c87870cf 100644
--- a/drivers/firmware/qcom/qcom_scm.c
+++ b/drivers/firmware/qcom/qcom_scm.c
@@ -666,7 +666,7 @@ static int qcom_scm_pas_init_image(struct device *dev, u32 pas_id,
 	ret = __qcom_scm_pas_init_image(dev, pas_id, mdata_phys, &res);
 	if (ret < 0 || !ctx) {
 		dma_free_coherent(dev, size, mdata_buf, mdata_phys);
-	} else if (ctx) {
+	} else {
 		ctx->ptr = mdata_buf;
 		ctx->phys = mdata_phys;
 		ctx->size = size;
diff --git a/drivers/remoteproc/qcom_q6v5_pas.c b/drivers/remoteproc/qcom_q6v5_pas.c
index 879c465c4a7a..ae93f693358c 100644
--- a/drivers/remoteproc/qcom_q6v5_pas.c
+++ b/drivers/remoteproc/qcom_q6v5_pas.c
@@ -733,6 +733,7 @@ static int qcom_pas_alloc_memory_region(struct qcom_pas *pas)
 	if (IS_ERR(pas->pas_ctx))
 		return PTR_ERR(pas->pas_ctx);
 
+	pas->pas_ctx->keep_mdt_buf = true;
 	if (!pas->dtb_pas_id)
 		return 0;
 
@@ -751,6 +752,8 @@ static int qcom_pas_alloc_memory_region(struct qcom_pas *pas)
 	if (IS_ERR(pas->dtb_pas_ctx))
 		return PTR_ERR(pas->dtb_pas_ctx);
 
+	pas->dtb_pas_ctx->keep_mdt_buf = true;
+
 	return 0;
 }
 
diff --git a/include/linux/firmware/qcom/qcom_pas.h b/include/linux/firmware/qcom/qcom_pas.h
index fb2ec3be6a16..1d132e89536e 100644
--- a/include/linux/firmware/qcom/qcom_pas.h
+++ b/include/linux/firmware/qcom/qcom_pas.h
@@ -22,6 +22,7 @@ struct qcom_pas_context {
 	dma_addr_t phys;
 	ssize_t size;
 	bool use_tzmem;
+	bool keep_mdt_buf;
 };
 
 static inline void __iomem *qcom_pas_ctx_map(struct qcom_pas_context *ctx)
-- 
2.55.0


             reply	other threads:[~2026-09-25 16:52 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-09-25 16:52 Mukesh Ojha [this message]
2026-10-21 12:56 ` Konrad Dybcio

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=20260925165231.10722-1-mukesh.ojha@oss.qualcomm.com \
    --to=mukesh.ojha@oss.qualcomm.com \
    --cc=abelvesa@kernel.org \
    --cc=andersson@kernel.org \
    --cc=konradybcio@kernel.org \
    --cc=linux-arm-msm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-remoteproc@vger.kernel.org \
    --cc=mathieu.poirier@linaro.org \
    --cc=sumit.garg@oss.qualcomm.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®