Your message dated Tue, 21 Sep 2021 11:37:14 +0200
with message-id <[email protected]>
and subject line Re: Bug#994814: nvidia-legacy-390xx-kernel-dkms: build
failures on Linux 5.14
has caused the Debian Bug report #994814,
regarding nvidia-legacy-390xx-kernel-dkms: build failures on Linux 5.14
to be marked as done.
This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.
(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact [email protected]
immediately.)
--
994814: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=994814
Debian Bug Tracking System
Contact [email protected] with problems
--- Begin Message ---
Package: nvidia-legacy-390xx-kernel-dkms
Version: 390.144-1
Severity: normal
Tags: patch
Hi,
today I've upgraded to linux-image-5.14.0-1-amd64-unsigned, and
nvidia-legacy-390xx-kernel-dkms failed to build with the following
error:
/var/lib/dkms/nvidia-legacy-390xx/390.144/build/nvidia/os-interface.c:583:18:
error: ‘struct task_struct’ has no member named ‘state’; did you mean ‘__state’?
583 | current->state = TASK_INTERRUPTIBLE;
| ^~~~~
| __state
Other than the changes to struct task_struct, the other relevant news in
Linux 5.14 is that the pdev field is now gone from struct drm_device
(see https://gist.github.com/joanbm/144a965c36fc1dc0d1f1b9be3438a368).
/var/lib/dkms/nvidia-legacy-390xx/390.144/build/nvidia-drm/nvidia-drm-drv.c:814:10:
error: ‘struct drm_device’ has no member named ‘pdev’; did you mean ‘dev’?
814 | dev->pdev = pdev;
| ^~~~
| dev
The attached changes to /usr/src/nvidia-legacy-390xx-390.144 fix the
driver with 5.14 on my workstation.
Thanks,
Emanuele
--- /usr/src/nvidia-legacy-390xx-390.144/nvidia/os-interface.c.orig 2021-09-21 10:03:57.965273627 +0200
+++ /usr/src/nvidia-legacy-390xx-390.144/nvidia/os-interface.c 2021-09-21 10:10:30.581119711 +0200
@@ -8,6 +8,8 @@
* _NVRM_COPYRIGHT_END_
*/
+#include <linux/version.h>
+
#define __NO_VERSION__
#include "nv-misc.h"
@@ -580,7 +582,11 @@
// the requested timeout has expired, loop until less
// than a jiffie of the desired delay remains.
//
+#if (LINUX_VERSION_CODE < KERNEL_VERSION(5, 14, 0))
current->state = TASK_INTERRUPTIBLE;
+#else
+ WRITE_ONCE(current->__state, TASK_INTERRUPTIBLE);
+#endif
do
{
schedule_timeout(jiffies);
--- /usr/src/nvidia-legacy-390xx-390.144/nvidia/nvlink_linux.c.orig 2021-09-21 10:15:06.578262557 +0200
+++ /usr/src/nvidia-legacy-390xx-390.144/nvidia/nvlink_linux.c 2021-09-21 10:14:02.386927630 +0200
@@ -21,6 +21,7 @@
*******************************************************************************/
+#include <linux/version.h>
#include "conftest.h"
#include "nvlink_common.h"
@@ -597,7 +598,11 @@
// the requested timeout has expired, loop until less
// than a jiffie of the desired delay remains.
//
+#if (LINUX_VERSION_CODE < KERNEL_VERSION(5, 14, 0))
current->state = TASK_INTERRUPTIBLE;
+#else
+ WRITE_ONCE(current->__state, TASK_INTERRUPTIBLE);
+#endif
do
{
schedule_timeout(jiffies);
--- /usr/src/nvidia-legacy-390xx-390.144/nvidia-drm/nvidia-drm-drv.c.orig 2021-09-21 10:18:02.980433147 +0200
+++ /usr/src/nvidia-legacy-390xx-390.144/nvidia-drm/nvidia-drm-drv.c 2021-09-21 10:21:21.454372046 +0200
@@ -59,6 +59,7 @@
#include <drm/drm_ioctl.h>
#endif
+#include <linux/version.h>
#include <linux/pci.h>
/*
@@ -811,7 +812,10 @@
dev->dev_private = nv_dev;
nv_dev->dev = dev;
+#if (LINUX_VERSION_CODE < KERNEL_VERSION(5, 14, 0))
+ // Rel. commit "drm: Remove pdev field from struct drm_device" (Thomas Zimmermann, 3 May 2021)
dev->pdev = pdev;
+#endif
/* Register DRM device to DRM sub-system */
--- End Message ---
--- Begin Message ---
Version: 390.144-2
On 21/09/2021 11.28, Emanuele Rocca wrote:
Package: nvidia-legacy-390xx-kernel-dkms
Version: 390.144-1
today I've upgraded to linux-image-5.14.0-1-amd64-unsigned, and
nvidia-legacy-390xx-kernel-dkms failed to build with the following
error:
I already uploaded the fix yesterday ;-)
Andreas
--- End Message ---
