[PATCH] remoteproc: qcom_wcnss: Fix error check after devm_ioremap_wc() switch

[PATCH] remoteproc: qcom_wcnss: Fix error check after devm_ioremap_wc() switch

[Lucas Sproule]

Commit f9b888599418 ("remoteproc: qcom_wcnss: Fix reserved region
mapping failure") switched back from devm_ioremap_resource_wc() to
devm_ioremap_wc(), but kept the IS_ERR() check that had been added to
match the resource variant.

devm_ioremap_wc() returns NULL on failure rather than an error pointer,
so IS_ERR() never fires and wcnss_alloc_memory_region() returns success
with wcnss->mem_region left NULL.  Probe completes, and every subsequent
firmware boot then fails with a bare -EINVAL from qcom_mdt_load_no_init(),
masking the real mapping failure.

Check for NULL instead, and return -ENOMEM to match the other remoteproc
drivers.

Fixes: f9b888599418 ("remoteproc: qcom_wcnss: Fix reserved region mapping failure")
Cc: stable@vger.kernel.org # v7.0+
Signed-off-by: Lucas Sproule <lucas.sproule.42@gmail.com>
---
 drivers/remoteproc/qcom_wcnss.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/remoteproc/qcom_wcnss.c b/drivers/remoteproc/qcom_wcnss.c
index c856a92af43c..90747af237f7 100644
--- a/drivers/remoteproc/qcom_wcnss.c
+++ b/drivers/remoteproc/qcom_wcnss.c
@@ -542,9 +542,9 @@ static int wcnss_alloc_memory_region(struct qcom_wcnss *wcnss)
  wcnss->mem_phys = wcnss->mem_reloc = res.start;
  wcnss->mem_size = resource_size(&res);
  wcnss->mem_region = devm_ioremap_wc(wcnss->dev, wcnss->mem_phys, wcnss->mem_size);
- if (IS_ERR(wcnss->mem_region)) {
+ if (!wcnss->mem_region) {
    dev_err(wcnss->dev, "unable to map memory region: %pR\n", &res);
-   return PTR_ERR(wcnss->mem_region);
+   return -ENOMEM;
  }
 
  return 0;
-- 
2.53.0

Re: [PATCH] remoteproc: qcom_wcnss: Fix error check after devm_ioremap_wc() switch

[Konrad Dybcio]

On 8/27/26 9:27 AM, Lucas Sproule wrote:
> Commit f9b888599418 ("remoteproc: qcom_wcnss: Fix reserved region
> mapping failure") switched back from devm_ioremap_resource_wc() to
> devm_ioremap_wc(), but kept the IS_ERR() check that had been added to
> match the resource variant.
> 
> devm_ioremap_wc() returns NULL on failure rather than an error pointer,
> so IS_ERR() never fires and wcnss_alloc_memory_region() returns success
> with wcnss->mem_region left NULL.  Probe completes, and every subsequent
> firmware boot then fails with a bare -EINVAL from qcom_mdt_load_no_init(),
> masking the real mapping failure.
> 
> Check for NULL instead, and return -ENOMEM to match the other remoteproc
> drivers.
> 
> Fixes: f9b888599418 ("remoteproc: qcom_wcnss: Fix reserved region mapping failure")
> Cc: stable@vger.kernel.org # v7.0+
> Signed-off-by: Lucas Sproule <lucas.sproule.42@gmail.com>
> ---
>  drivers/remoteproc/qcom_wcnss.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/remoteproc/qcom_wcnss.c b/drivers/remoteproc/qcom_wcnss.c
> index c856a92af43c..90747af237f7 100644
> --- a/drivers/remoteproc/qcom_wcnss.c
> +++ b/drivers/remoteproc/qcom_wcnss.c
> @@ -542,9 +542,9 @@ static int wcnss_alloc_memory_region(struct qcom_wcnss *wcnss)
>   wcnss->mem_phys = wcnss->mem_reloc = res.start;
>   wcnss->mem_size = resource_size(&res);
>   wcnss->mem_region = devm_ioremap_wc(wcnss->dev, wcnss->mem_phys, wcnss->mem_size);
> - if (IS_ERR(wcnss->mem_region)) {
> + if (!wcnss->mem_region) {
>     dev_err(wcnss->dev, "unable to map memory region: %pR\n", &res);
> -   return PTR_ERR(wcnss->mem_region);
> +   return -ENOMEM;
>   }

It seems like whatever program you used to send this patch messed up its
contents by squashing the whitespace - I would recommend using the b4
tool:

https://b4.docs.kernel.org/en/latest/index.html

for the patch itself, please carry my:

Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>

Konrad