Hi Matti,
Thank you for your concern! I haven't used it in my projects yet. I came
across this issue while using smatch for code analysis. After reviewing
the code, I noticed a few minor issues. However, I did not realize that
the patch for the memory leak has already been submitted couple of days
ago. Apologies for not checking the patches thoroughly and submitting a
duplicate.
Yours,
-- Zicheng
On 2024/10/31 15:32, Matti Vaittinen wrote:
Hi Zicheng
On 31/10/2024 03:47, Zicheng Qu wrote:
In iio_gts_build_avail_scale_table(), the memory allocated for
per_time_gains is freed using kfree(per_time_gains) before return 0.
However, the type per_time_gains is 'int **', and the memory allocated
for its inner elements is not being freed, leading to a memory leak.
Cc: stable@xxxxxxxxxxxxxxx # v6.6+
Fixes: 38416c28e168 ("iio: light: Add gain-time-scale helpers")
Signed-off-by: Zicheng Qu <quzicheng@xxxxxxxxxx>
---
drivers/iio/industrialio-gts-helper.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/iio/industrialio-gts-helper.c
b/drivers/iio/industrialio-gts-helper.c
index 59d7615c0f56..f2450b2e740d 100644
--- a/drivers/iio/industrialio-gts-helper.c
+++ b/drivers/iio/industrialio-gts-helper.c
@@ -307,6 +307,8 @@ static int iio_gts_build_avail_scale_table(struct
iio_gts *gts)
if (ret)
goto err_free_out;
+ for (j = 0; j < gts->num_itime; j++)
+ kfree(per_time_gains[i]);
kfree(per_time_gains);
gts->per_time_avail_scale_tables = per_time_scales;
You're right, thanks!
This, however, was already fixed by:
https://lore.kernel.org/all/20241011095512.3667549-1-ruanjinjie@xxxxxxxxxx/
Out of the curiosity (and no need to respond if you don't feel like) -
are you using the gts helpers in some of your project(s)? I am glad
seeing these fixes coming in and just wondered if all these bugs are
found because these helpers are being used outside the ROHM drivers :)
Yours,
-- Matti
