[PATCH v2] drm/ttm: Fix compile error when CONFIG_SHMEM is not set

[PATCH v2] drm/ttm: Fix compile error when CONFIG_SHMEM is not set

[Steven Rostedt]

From: Steven Rostedt <rostedt@goodmis.org>

When CONFIG_SHMEM is not set, the following compiler error occurs:

ld: vmlinux.o: in function `ttm_backup_backup_page':
(.text+0x10363bc): undefined reference to `shmem_writeout'
make[3]: *** [/work/build/trace/nobackup/linux.git/scripts/Makefile.vmlinux:91: vmlinux.unstripped] Error 1
make[2]: *** [/work/build/trace/nobackup/linux.git/Makefile:1241: vmlinux] Error 2
make[1]: *** [/work/build/trace/nobackup/linux.git/Makefile:248: __sub-make] Error 2
make[1]: Leaving directory '/work/build/nobackup/tracetest'
make: *** [Makefile:248: __sub-make] Error 2

This is due to the replacement of writepage and calling swap_writeout()
and shmem_writeout() directly. The issue is that when CONFIG_SHMEM is not
defined, shmem_writeout() is also not defined.

The function ttm_backup_backup_page() called mapping->a_ops->writepage()
which was then changed to call shmem_writeout() directly.

Even before commit 84798514db50 ("mm: Remove swap_writepage() and
shmem_writepage()"), it didn't make sense to call anything other than
shmem_writeout() as the ttm_backup deals only with shmem folios.

Have DRM_TTM config option select SHMEM to guarantee that shmem_writeout()
is available.

Link: https://lore.kernel.org/all/20250602170500.48713a2b@gandalf.local.home/

Suggested-by: Hugh Dickins <hughd@google.com>
Fixes: 84798514db50 ("mm: Remove swap_writepage() and shmem_writepage()")
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
---
Changes since v1: https://lore.kernel.org/all/20250602170500.48713a2b@gandalf.local.home/

- Instead of adding a shmem_writeout() stub, just make CONFIG_DRM_TTM
  select CONFIG_SHMEM (Hugh Dickins)

 drivers/gpu/drm/Kconfig | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig
index f094797f3b2b..ded28c71d89c 100644
--- a/drivers/gpu/drm/Kconfig
+++ b/drivers/gpu/drm/Kconfig
@@ -188,6 +188,7 @@ source "drivers/gpu/drm/display/Kconfig"
 config DRM_TTM
 	tristate
 	depends on DRM && MMU
+	select SHMEM
 	help
 	  GPU memory management subsystem for devices with multiple
 	  GPU memory types. Will be enabled automatically if a device driver
-- 
2.47.2

Re: [PATCH v2] drm/ttm: Fix compile error when CONFIG_SHMEM is not set

[Thomas Hellström]

On Wed, 2025-06-04 at 08:51 -0400, Steven Rostedt wrote:
> From: Steven Rostedt <rostedt@goodmis.org>
> 
> When CONFIG_SHMEM is not set, the following compiler error occurs:
> 
> ld: vmlinux.o: in function `ttm_backup_backup_page':
> (.text+0x10363bc): undefined reference to `shmem_writeout'
> make[3]: ***
> [/work/build/trace/nobackup/linux.git/scripts/Makefile.vmlinux:91:
> vmlinux.unstripped] Error 1
> make[2]: *** [/work/build/trace/nobackup/linux.git/Makefile:1241:
> vmlinux] Error 2
> make[1]: *** [/work/build/trace/nobackup/linux.git/Makefile:248:
> __sub-make] Error 2
> make[1]: Leaving directory '/work/build/nobackup/tracetest'
> make: *** [Makefile:248: __sub-make] Error 2
> 
> This is due to the replacement of writepage and calling
> swap_writeout()
> and shmem_writeout() directly. The issue is that when CONFIG_SHMEM is
> not
> defined, shmem_writeout() is also not defined.
> 
> The function ttm_backup_backup_page() called mapping->a_ops-
> >writepage()
> which was then changed to call shmem_writeout() directly.
> 
> Even before commit 84798514db50 ("mm: Remove swap_writepage() and
> shmem_writepage()"), it didn't make sense to call anything other than
> shmem_writeout() as the ttm_backup deals only with shmem folios.
> 
> Have DRM_TTM config option select SHMEM to guarantee that
> shmem_writeout()
> is available.
> 
> Link:
> https://lore.kernel.org/all/20250602170500.48713a2b@gandalf.local.home/
> 
> Suggested-by: Hugh Dickins <hughd@google.com>
> Fixes: 84798514db50 ("mm: Remove swap_writepage() and
> shmem_writepage()")
> Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
> ---
> Changes since v1:
> https://lore.kernel.org/all/20250602170500.48713a2b@gandalf.local.home/
> 
> - Instead of adding a shmem_writeout() stub, just make CONFIG_DRM_TTM
>   select CONFIG_SHMEM (Hugh Dickins)
> 
>  drivers/gpu/drm/Kconfig | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig
> index f094797f3b2b..ded28c71d89c 100644
> --- a/drivers/gpu/drm/Kconfig
> +++ b/drivers/gpu/drm/Kconfig
> @@ -188,6 +188,7 @@ source "drivers/gpu/drm/display/Kconfig"
>  config DRM_TTM
>  	tristate
>  	depends on DRM && MMU
> +	select SHMEM
>  	help
>  	  GPU memory management subsystem for devices with multiple
>  	  GPU memory types. Will be enabled automatically if a
> device driver

Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>

Thanks,
Thomas

Re: [PATCH v2] drm/ttm: Fix compile error when CONFIG_SHMEM is not set

[Hugh Dickins]

[-- Attachment #1: Type: text/plain, Size: 2752 bytes --]

On Wed, 4 Jun 2025, Thomas Hellström wrote:
> On Wed, 2025-06-04 at 08:51 -0400, Steven Rostedt wrote:
> > From: Steven Rostedt <rostedt@goodmis.org>
> > 
> > When CONFIG_SHMEM is not set, the following compiler error occurs:
> > 
> > ld: vmlinux.o: in function `ttm_backup_backup_page':
> > (.text+0x10363bc): undefined reference to `shmem_writeout'
> > make[3]: ***
> > [/work/build/trace/nobackup/linux.git/scripts/Makefile.vmlinux:91:
> > vmlinux.unstripped] Error 1
> > make[2]: *** [/work/build/trace/nobackup/linux.git/Makefile:1241:
> > vmlinux] Error 2
> > make[1]: *** [/work/build/trace/nobackup/linux.git/Makefile:248:
> > __sub-make] Error 2
> > make[1]: Leaving directory '/work/build/nobackup/tracetest'
> > make: *** [Makefile:248: __sub-make] Error 2
> > 
> > This is due to the replacement of writepage and calling
> > swap_writeout()
> > and shmem_writeout() directly. The issue is that when CONFIG_SHMEM is
> > not
> > defined, shmem_writeout() is also not defined.
> > 
> > The function ttm_backup_backup_page() called mapping->a_ops-
> > >writepage()
> > which was then changed to call shmem_writeout() directly.
> > 
> > Even before commit 84798514db50 ("mm: Remove swap_writepage() and
> > shmem_writepage()"), it didn't make sense to call anything other than
> > shmem_writeout() as the ttm_backup deals only with shmem folios.
> > 
> > Have DRM_TTM config option select SHMEM to guarantee that
> > shmem_writeout()
> > is available.
> > 
> > Link:
> > https://lore.kernel.org/all/20250602170500.48713a2b@gandalf.local.home/
> > 
> > Suggested-by: Hugh Dickins <hughd@google.com>
> > Fixes: 84798514db50 ("mm: Remove swap_writepage() and
> > shmem_writepage()")
> > Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
> > ---
> > Changes since v1:
> > https://lore.kernel.org/all/20250602170500.48713a2b@gandalf.local.home/
> > 
> > - Instead of adding a shmem_writeout() stub, just make CONFIG_DRM_TTM
> >   select CONFIG_SHMEM (Hugh Dickins)
> > 
> >  drivers/gpu/drm/Kconfig | 1 +
> >  1 file changed, 1 insertion(+)
> > 
> > diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig
> > index f094797f3b2b..ded28c71d89c 100644
> > --- a/drivers/gpu/drm/Kconfig
> > +++ b/drivers/gpu/drm/Kconfig
> > @@ -188,6 +188,7 @@ source "drivers/gpu/drm/display/Kconfig"
> >  config DRM_TTM
> >  	tristate
> >  	depends on DRM && MMU
> > +	select SHMEM
> >  	help
> >  	  GPU memory management subsystem for devices with multiple
> >  	  GPU memory types. Will be enabled automatically if a
> > device driver
> 
> Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>

Acked-by: Hugh Dickins <hughd@google.com>

Thanks a lot!

Re: [PATCH v2] drm/ttm: Fix compile error when CONFIG_SHMEM is not set

[Steven Rostedt]

On Wed, 4 Jun 2025 09:26:21 -0700 (PDT)
Hugh Dickins <hughd@google.com> wrote:

> > Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>  
> 
> Acked-by: Hugh Dickins <hughd@google.com>

Thanks Thomas and Hugh,

Now the question is, who's gonna take it? ;-)

-- Steve

Re: [PATCH v2] drm/ttm: Fix compile error when CONFIG_SHMEM is not set

[Steven Rostedt]

On Wed, 4 Jun 2025 12:38:37 -0400
Steven Rostedt <rostedt@goodmis.org> wrote:

> On Wed, 4 Jun 2025 09:26:21 -0700 (PDT)
> Hugh Dickins <hughd@google.com> wrote:
> 
> > > Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>    
> > 
> > Acked-by: Hugh Dickins <hughd@google.com>  
> 
> Thanks Thomas and Hugh,
> 
> Now the question is, who's gonna take it? ;-)

Never mind, I see Linus took it.

-- Steve