Linux 6.4.12 Breaks Virtualbox 6.1.46 DKMS Build
All, This shouldn't happen, but the linux-6.4.11 to linux-6.4.12 broke the dkms build of virtualbox 6.1.46 modules. Specifically from the log: DKMS make.log for vboxhost-6.1.46_non_OSE for kernel 6.4.12-arch1-1 (x86_64) Fri Aug 25 05:35:45 PM CDT 2023 CC [M] /var/lib/dkms/vboxhost/6.1.46_non_OSE/build/vboxnetflt/linux/VBoxNetFlt-linux.o CC [M] /var/lib/dkms/vboxhost/6.1.46_non_OSE/build/vboxdrv/linux/SUPDrv-linux.o /var/lib/dkms/vboxhost/6.1.46_non_OSE/build/vboxnetflt/linux/VBoxNetFlt-linux.c: In function ‘vboxNetFltLinuxForwardToIntNetInner’: /var/lib/dkms/vboxhost/6.1.46_non_OSE/build/vboxnetflt/linux/VBoxNetFlt-linux.c:1561:40: error: implicit declaration of function ‘skb_gso_segment’; did you mean ‘skb_gso_reset’? [-Werror=implicit-function-declaration] 1561 | struct sk_buff *pSegment = skb_gso_segment(pBuf, 0 /*supported features*/); | ^~~~~~~~~~~~~~~ | skb_gso_reset /var/lib/dkms/vboxhost/6.1.46_non_OSE/build/vboxnetflt/linux/VBoxNetFlt-linux.c:1561:40: warning: initialization of ‘struct sk_buff *’ from ‘int’ makes pointer from integer without a cast [-Wint-conversion] cc1: some warnings being treated as errors make[2]: *** [scripts/Makefile.build:252: /var/lib/dkms/vboxhost/6.1.46_non_OSE/build/vboxnetflt/linux/VBoxNetFlt-linux.o] Error 1 make[1]: *** [scripts/Makefile.build:497: /var/lib/dkms/vboxhost/6.1.46_non_OSE/build/vboxnetflt] Error 2 make[1]: *** Waiting for unfinished jobs.... CC [M] /var/lib/dkms/vboxhost/6.1.46_non_OSE/build/vboxdrv/SUPDrv.o CC [M] /var/lib/dkms/vboxhost/6.1.46_non_OSE/build/vboxdrv/SUPDrvGip.o In file included from ./include/linux/string.h:254, from /var/lib/dkms/vboxhost/6.1.46_non_OSE/build/vboxdrv/include/iprt/string.h:39, from /var/lib/dkms/vboxhost/6.1.46_non_OSE/build/vboxdrv/SUPDrvInternal.h:46, from /var/lib/dkms/vboxhost/6.1.46_non_OSE/build/vboxdrv/SUPDrv.c:33: In function ‘fortify_memcpy_chk’, inlined from ‘supdrvIOCtlInnerUnrestricted’ at /var/lib/dkms/vboxhost/6.1.46_non_OSE/build/vboxdrv/SUPDrv.c:1746:13, inlined from ‘supdrvIOCtl’ at /var/lib/dkms/vboxhost/6.1.46_non_OSE/build/vboxdrv/SUPDrv.c:2698:14: ./include/linux/fortify-string.h:520:25: warning: call to ‘__write_overflow_field’ declared with attribute warning: detected write beyond size of field (1st parameter); maybe use struct_group()? [-Wattribute-warning] 520 | __write_overflow_field(p_size_field, size); | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Should I open a bug with Arch? -- David C. Rankin, J.D.,P.E.
On Fri, 25 Aug 2023 17:49:03 -0500 "David C. Rankin" <drankinatty@gmail.com> wrote:
All,
This shouldn't happen, but the linux-6.4.11 to linux-6.4.12 broke the dkms build of virtualbox 6.1.46 modules. Specifically from the log:
Pretty normal, actually. If you want to run an old version, it's your responsibility to make it work. Kernels breaking compatibility happens all of the time.
Should I open a bug with Arch?
No, this is not an Arch issue, as Arch doesn't ship virtualbox 6.1.46 modules. Doug
On 8/25/23 18:30, Doug Newgard wrote:
On Fri, 25 Aug 2023 17:49:03 -0500 "David C. Rankin" <drankinatty@gmail.com> wrote:
All,
This shouldn't happen, but the linux-6.4.11 to linux-6.4.12 broke the dkms build of virtualbox 6.1.46 modules. Specifically from the log:
Pretty normal, actually. If you want to run an old version, it's your responsibility to make it work. Kernels breaking compatibility happens all of the time.
Should I open a bug with Arch?
No, this is not an Arch issue, as Arch doesn't ship virtualbox 6.1.46 modules.
Doug
True but was that a kernel change or an Arch change to the kernel that broke it? -- David C. Rankin, J.D.,P.E.
On Fri, 25 Aug 2023 18:53:07 -0500 "David C. Rankin" <drankinatty@gmail.com> wrote:
On 8/25/23 18:30, Doug Newgard wrote:
On Fri, 25 Aug 2023 17:49:03 -0500 "David C. Rankin" <drankinatty@gmail.com> wrote:
All,
This shouldn't happen, but the linux-6.4.11 to linux-6.4.12 broke the dkms build of virtualbox 6.1.46 modules. Specifically from the log:
Pretty normal, actually. If you want to run an old version, it's your responsibility to make it work. Kernels breaking compatibility happens all of the time.
Should I open a bug with Arch?
No, this is not an Arch issue, as Arch doesn't ship virtualbox 6.1.46 modules.
Doug
True but was that a kernel change or an Arch change to the kernel that broke it?
What do you think Arch does to the kernel? But really, does it even matter? Arch keeps the modules in the repo up to date, including patching them when necessary for newer kernels. Again, if you want to run an old version that Arch doesn't maintain, that part is up to you. Maybe check what Arch does? Or see if upstream has patches?
Additionally,
This is not an old version. 6.1.46 was released July 31, 2023, less than 30 days ago.
Nobody cares if its lts/still supported, 6.1 is an old version; end of story. Even upstream lists it under "Old Builds". Arch is on 7.0 and maintains that. If you don't want to use that, it's on you. And 6.1.46 was July 18th, not 31st, and is only supported for a few more months. Doug
Doug Newgard <dnewgard@outlook.com> wrote:
On Fri, 25 Aug 2023 18:53:07 -0500 "David C. Rankin" <drankinatty@gmail.com> wrote:
On 8/25/23 18:30, Doug Newgard wrote:
On Fri, 25 Aug 2023 17:49:03 -0500 "David C. Rankin" <drankinatty@gmail.com> wrote:
All,
This shouldn't happen, but the linux-6.4.11 to linux-6.4.12 broke the dkms build of virtualbox 6.1.46 modules. Specifically from the log:
Pretty normal, actually. If you want to run an old version, it's your responsibility to make it work. Kernels breaking compatibility happens all of the time.
Should I open a bug with Arch?
No, this is not an Arch issue, as Arch doesn't ship virtualbox 6.1.46 modules.
Doug
True but was that a kernel change or an Arch change to the kernel that broke it?
What do you think Arch does to the kernel?
Doesn't usually everything arch does to the kernel are the few lines in between `Arch Linux kernel vA.B.CC-arch1' and `Linux A.B.CC'? As in $ git log --oneline -n6 477b5e926a64 Arch Linux kernel v6.4.12-arch1 879f0f7fc0f9 sphinx: kfigure.py: Convert outdir to str before using len e5f6f3e36b5e tpm: Don't make vendor check required for probe 3720e1b0eb7d drivers/firmware: skip simpledrm if nvidia-drm.modeset=1 is set 093ed5b87ed2 ZEN: Add sysctl and CONFIG to disallow unprivileged CLONE_NEWUSER 05d8970cca01 Linux 6.4.12 ? -- u34
But really, does it even matter? Arch keeps the modules in the repo up to date, including patching them when necessary for newer kernels. Again, if you want to run an old version that Arch doesn't maintain, that part is up to you. Maybe check what Arch does? Or see if upstream has patches?
Additionally,
This is not an old version. 6.1.46 was released July 31, 2023, less than 30 days ago.
Nobody cares if its lts/still supported, 6.1 is an old version; end of story. Even upstream lists it under "Old Builds". Arch is on 7.0 and maintains that. If you don't want to use that, it's on you.
And 6.1.46 was July 18th, not 31st, and is only supported for a few more months.
Doug
On 8/25/23 18:30, Doug Newgard wrote:
Pretty normal, actually. If you want to run an old version, it's your responsibility to make it work. Kernels breaking compatibility happens all of the time.
Additionally, This is not an old version. 6.1.46 was released July 31, 2023, less than 30 days ago. -- David C. Rankin, J.D.,P.E.
participants (3)
-
David C. Rankin
-
Doug Newgard
-
u34＠net9.cf