On Thu, Apr 09, 2026 at 10:43:58AM +0200, Greg Kroah-Hartman wrote:

On Thu, Apr 09, 2026 at 10:26:55AM +0200, Greg Kroah-Hartman wrote:

quoted

On Thu, Apr 09, 2026 at 10:09:59AM +0200, Jiri Slaby wrote:

quoted

Hi,

On 08. 04. 26, 20:04, Greg Kroah-Hartman wrote:

quoted

6.19-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Dmitry Torokhov <dmitry.torokhov@gmail.com>

commit b981e9e94c687b7b19ae8820963f005b842cb2f2 upstream.

...

quoted

--- a/arch/x86/platform/geode/geode-common.c
+++ b/arch/x86/platform/geode/geode-common.c

...

quoted

@@ -127,6 +129,12 @@ int __init geode_create_leds(const char
  		goto err_free_swnodes;
  	}
+	gpio_refs = kzalloc_objs(*gpio_refs, n_leds);

On x86_32, this fails to build:
arch/x86/platform/geode/geode-common.c:132:21: error: implicit declaration
of function ‘kzalloc_objs’; did you mean ‘kzalloc_node’?
[-Wimplicit-function-declaration]
arch/x86/platform/geode/geode-common.c:132:19: error: assignment to ‘struct
software_node_ref_args *’ from ‘int’ makes pointer from integer without a
cast [-Wint-conversion]
  132 |         gpio_refs = kzalloc_objs(*gpio_refs, n_leds);


6.19 does not have kzalloc_objs() yet.

Ugh, yeah, Sasha caught this too, let me go fix this up, my fault.  I
guess I never build x32 kernels anymore :(

Ok, now fixed up, I'll go push out new -rc2 releases with this fix in it
to make sure I didn't get it wrong.

Ah, Sasha just dropped the commit, that was easier to do :)