[Bug 60852] Random memory corruption reported after writing to /sys/module/drm_kms_helper/parameters/edid_firmware

[bugzilla-daemon at bugzilla.kernel.org]

https://bugzilla.kernel.org/show_bug.cgi?id=60852

Jani Nikula <jani.nikula at intel.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |NEEDINFO

--- Comment #3 from Jani Nikula <jani.nikula at intel.com> ---
(In reply to Chris Wilson from comment #2)
> You need to set that parameter on the command line for it to have effect for
> LVDS. I guess no one checked whether changing it on the fly was safe...

Hmm, it looked like a list corruption in the firmware loader to me. Which
shouldn't happen regardless of whether it's safe to set edid_firmware on the
fly.

Alex, please see if you can reproduce the problem with this patch:

diff --git a/drivers/gpu/drm/drm_edid_load.c b/drivers/gpu/drm/drm_edid_load.c
index 271b42b..0a54bdd7 100644
--- a/drivers/gpu/drm/drm_edid_load.c
+++ b/drivers/gpu/drm/drm_edid_load.c
@@ -152,7 +152,7 @@ static u8 *edid_load(struct drm_connector *connector, const
char *name,
         goto out;
     }

-    err = request_firmware(&fw, name, &pdev->dev);
+    err = 1;
     platform_device_unregister(pdev);

     if (err) {
@@ -231,7 +231,6 @@ static u8 *edid_load(struct drm_connector *connector, const
char *name,
         name, connector_name);

 relfw_out:
-    release_firmware(fw);

 out:
     if (err)

-- 
You are receiving this mail because:
You are on the CC list for the bug.
You are the assignee for the bug.