From: Sailesh Nandanavanam <saileshnandanavanam@gmail.com>
To: andersson@kernel.org, mathieu.poirier@linaro.org
Cc: linux-arm-msm@vger.kernel.org, linux-remoteproc@vger.kernel.org,
linux-kernel@vger.kernel.org,
Sailesh Nandanavanam <saileshnandanavanam@gmail.com>
Subject: [PATCH] remoteproc: qcom_q6v5_pas: Release dtb_firmware on stop
Date: Mon, 22 Jun 2026 09:31:14 +0530 [thread overview]
Message-ID: <20260622040114.17128-1-saileshnandanavanam@gmail.com> (raw)
qcom_pas_load() acquires a reference to pas->dtb_firmware via
request_firmware() whenever the platform defines dtb_pas_id. This
reference is only released on the error path inside qcom_pas_load()
itself, when qcom_mdt_pas_load() fails.
On the success path, pas->dtb_firmware is never released by
qcom_pas_start(), qcom_pas_stop(), qcom_pas_unprepare(), or
qcom_pas_remove(). Since qcom_pas_load() runs on every load/reload
cycle, including subsystem-restart recovery, this leaks a firmware
reference on every successful boot of remoteprocs that use a DTB
co-firmware image (e.g. ADSP/CDSP/MPSS on SM8550, SM8650, SM8750,
X1E80100).
A second, narrower leak window exists if qcom_pas_start() itself
fails after the DTB carveout has been mapped: the remoteproc core
does not invoke .stop for a remoteproc that failed to start, so
qcom_pas_stop() never runs to release the reference in that case.
Release pas->dtb_firmware in both qcom_pas_stop() and at the
unmap_dtb_carveout label in qcom_pas_start(), alongside the existing
teardown of the DTB carveout mapping, and clear the pointer to avoid
holding a stale reference into the next load cycle.
Fixes: 29814986b82e ("remoteproc: qcom_q6v5_pas: add support for dtb co-firmware loading")
Signed-off-by: Sailesh Nandanavanam <saileshnandanavanam@gmail.com>
---
drivers/remoteproc/qcom_q6v5_pas.c | 7 ++++++-
1 file changed, 6 insertions(+), 1 deletion(-)
diff --git a/drivers/remoteproc/qcom_q6v5_pas.c b/drivers/remoteproc/qcom_q6v5_pas.c
index da27d1d3c9da..040aabbe5860 100644
--- a/drivers/remoteproc/qcom_q6v5_pas.c
+++ b/drivers/remoteproc/qcom_q6v5_pas.c
@@ -360,8 +360,11 @@ static int qcom_pas_start(struct rproc *rproc)
qcom_scm_pas_metadata_release(pas->dtb_pas_ctx);
unmap_dtb_carveout:
- if (pas->dtb_pas_id)
+ if (pas->dtb_pas_id) {
qcom_pas_unmap_carveout(rproc, pas->dtb_mem_phys, pas->dtb_mem_size);
+ release_firmware(pas->dtb_firmware);
+ pas->dtb_firmware = NULL;
+ }
disable_px_supply:
if (pas->px_supply)
regulator_disable(pas->px_supply);
@@ -419,6 +422,8 @@ static int qcom_pas_stop(struct rproc *rproc)
dev_err(pas->dev, "failed to shutdown dtb: %d\n", ret);
qcom_pas_unmap_carveout(rproc, pas->dtb_mem_phys, pas->dtb_mem_size);
+ release_firmware(pas->dtb_firmware);
+ pas->dtb_firmware = NULL;
}
qcom_pas_unmap_carveout(rproc, pas->mem_phys, pas->mem_size);
--
2.34.1
next reply other threads:[~2026-06-22 4:01 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-06-22 4:01 Sailesh Nandanavanam [this message]
2026-06-22 6:41 ` [PATCH] remoteproc: qcom_q6v5_pas: Release dtb_firmware on stop Mukesh Ojha
2026-06-24 17:53 ` [PATCH v2] remoteproc: qcom_q6v5_pas: Fix dtb firmware lifecycle and leak Sailesh Nandanavanam
2026-06-30 7:55 ` Mukesh Ojha
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=20260622040114.17128-1-saileshnandanavanam@gmail.com \
--to=saileshnandanavanam@gmail.com \
--cc=andersson@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 \
/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