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>] [day] [month] [year] [list]
Message-ID: <20251121091125.6289-1-ansuelsmth@gmail.com>
Date: Fri, 21 Nov 2025 10:11:20 +0100
From: Christian Marangi <ansuelsmth@...il.com>
To: Bjorn Andersson <andersson@...nel.org>,
	Konrad Dybcio <konradybcio@...nel.org>,
	Christian Marangi <ansuelsmth@...il.com>,
	linux-arm-msm@...r.kernel.org,
	linux-kernel@...r.kernel.org
Cc: Dan Carpenter <dan.carpenter@...aro.org>
Subject: [PATCH] soc: qcom: smem: fix qcom_smem_is_available and check if __smem is valid

Commit 7a94d5f31b54 ("soc: qcom: smem: better track SMEM uninitialized
state") changed the usage of __smem and init now as an error pointer
instead of NULL.

qcom_smem_is_available() wasn't updated to reflect this change and also
.qcom_smem_remove doesn't reset it on module exit.

While at it also proced other expoert symbol if __smem is not set to a
correct pointer.

The changed symbol are unlikely to cause a kernel panic (as it's
expected to always check qcom_smem_is_available() before any SMEM usage)
but it's still worth to protect them by any wrong usage.

Fixes: 7a94d5f31b54 ("soc: qcom: smem: better track SMEM uninitialized state")
Reported-by: Dan Carpenter <dan.carpenter@...aro.org>
Signed-off-by: Christian Marangi <ansuelsmth@...il.com>
---
 drivers/soc/qcom/smem.c | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)

diff --git a/drivers/soc/qcom/smem.c b/drivers/soc/qcom/smem.c
index fef840b54574..c8db29c54ddb 100644
--- a/drivers/soc/qcom/smem.c
+++ b/drivers/soc/qcom/smem.c
@@ -381,6 +381,9 @@ static struct qcom_smem *__smem = INIT_ERR_PTR(-EPROBE_DEFER);
  */
 int qcom_smem_bust_hwspin_lock_by_host(unsigned int host)
 {
+	if (IS_ERR(__smem))
+		return PTR_ERR(__smem);
+
 	/* This function is for remote procs, so ignore SMEM_HOST_APPS */
 	if (host == SMEM_HOST_APPS || host >= SMEM_HOST_COUNT)
 		return -EINVAL;
@@ -396,7 +399,7 @@ EXPORT_SYMBOL_GPL(qcom_smem_bust_hwspin_lock_by_host);
  */
 bool qcom_smem_is_available(void)
 {
-	return !!__smem;
+	return !IS_ERR(__smem);
 }
 EXPORT_SYMBOL_GPL(qcom_smem_is_available);
 
@@ -777,6 +780,9 @@ phys_addr_t qcom_smem_virt_to_phys(void *p)
 	u64 offset;
 	u32 i;
 
+	if (IS_ERR(__smem))
+		return 0;
+
 	for (i = 0; i < SMEM_HOST_COUNT; i++) {
 		part = &__smem->partitions[i];
 
@@ -1247,7 +1253,8 @@ static void qcom_smem_remove(struct platform_device *pdev)
 {
 	platform_device_unregister(__smem->socinfo);
 
-	__smem = NULL;
+	/* Set to -EPROBE_DEFER to signal unprobed state */
+	__smem = ERR_PTR(-EPROBE_DEFER);
 }
 
 static const struct of_device_id qcom_smem_of_match[] = {
-- 
2.51.0


Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ