Re: [PATCH] fbdev: udlfb: make CONFIG_FB_DEVICE optional

Re: [PATCH] fbdev: udlfb: make CONFIG_FB_DEVICE optional

[kernel test robot]

Hi Sukrut,

kernel test robot noticed the following build errors:

[auto build test ERROR on drm-misc/drm-misc-next]
[also build test ERROR on linus/master v6.17-rc7 next-20250926]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Sukrut-Heroorkar/fbdev-udlfb-make-CONFIG_FB_DEVICE-optional/20250925-015939
base:   git://anongit.freedesktop.org/drm/drm-misc drm-misc-next
patch link:    https://lore.kernel.org/r/20250924175743.6790-1-hsukrut3%40gmail.com
patch subject: [PATCH] fbdev: udlfb: make CONFIG_FB_DEVICE optional
config: x86_64-randconfig-001-20250927 (https://download.01.org/0day-ci/archive/20250927/202509272320.3K8kdDCw-lkp@intel.com/config)
compiler: clang version 20.1.8 (https://github.com/llvm/llvm-project 87f0227cb60147a26a1eeb4fb06e3b505e9c7261)
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20250927/202509272320.3K8kdDCw-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202509272320.3K8kdDCw-lkp@intel.com/

All errors (new ones prefixed by >>):

>> drivers/video/fbdev/udlfb.c:345:16: error: no member named 'dev' in 'struct fb_info'
     345 |         dev_dbg(info->dev, "mmap() framebuffer addr:%lu size:%lu\n",
         |                 ~~~~  ^
   include/linux/dev_printk.h:171:28: note: expanded from macro 'dev_dbg'
     171 |         dev_no_printk(KERN_DEBUG, dev, dev_fmt(fmt), ##__VA_ARGS__)
         |                                   ^~~
   include/linux/dev_printk.h:139:23: note: expanded from macro 'dev_no_printk'
     139 |                         _dev_printk(level, dev, fmt, ##__VA_ARGS__);    \
         |                                            ^~~
   drivers/video/fbdev/udlfb.c:933:16: error: no member named 'dev' in 'struct fb_info'
     933 |         dev_dbg(info->dev, "open, user=%d fb_info=%p count=%d\n",
         |                 ~~~~  ^
   include/linux/dev_printk.h:171:28: note: expanded from macro 'dev_dbg'
     171 |         dev_no_printk(KERN_DEBUG, dev, dev_fmt(fmt), ##__VA_ARGS__)
         |                                   ^~~
   include/linux/dev_printk.h:139:23: note: expanded from macro 'dev_no_printk'
     139 |                         _dev_printk(level, dev, fmt, ##__VA_ARGS__);    \
         |                                            ^~~
   drivers/video/fbdev/udlfb.c:986:16: error: no member named 'dev' in 'struct fb_info'
     986 |         dev_dbg(info->dev, "release, user=%d count=%d\n", user, dlfb->fb_count);
         |                 ~~~~  ^
   include/linux/dev_printk.h:171:28: note: expanded from macro 'dev_dbg'
     171 |         dev_no_printk(KERN_DEBUG, dev, dev_fmt(fmt), ##__VA_ARGS__)
         |                                   ^~~
   include/linux/dev_printk.h:139:23: note: expanded from macro 'dev_no_printk'
     139 |                         _dev_printk(level, dev, fmt, ##__VA_ARGS__);    \
         |                                            ^~~
   drivers/video/fbdev/udlfb.c:1099:16: error: no member named 'dev' in 'struct fb_info'
    1099 |         dev_dbg(info->dev, "blank, mode %d --> %d\n",
         |                 ~~~~  ^
   include/linux/dev_printk.h:171:28: note: expanded from macro 'dev_dbg'
     171 |         dev_no_printk(KERN_DEBUG, dev, dev_fmt(fmt), ##__VA_ARGS__)
         |                                   ^~~
   include/linux/dev_printk.h:139:23: note: expanded from macro 'dev_no_printk'
     139 |                         _dev_printk(level, dev, fmt, ##__VA_ARGS__);    \
         |                                            ^~~
   drivers/video/fbdev/udlfb.c:1193:18: error: no member named 'dev' in 'struct fb_info'
    1193 |                         dev_err(info->dev, "Virtual framebuffer alloc failed\n");
         |                                 ~~~~  ^
   include/linux/dev_printk.h:154:44: note: expanded from macro 'dev_err'
     154 |         dev_printk_index_wrap(_dev_err, KERN_ERR, dev, dev_fmt(fmt), ##__VA_ARGS__)
         |                                                   ^~~
   include/linux/dev_printk.h:110:11: note: expanded from macro 'dev_printk_index_wrap'
     110 |                 _p_func(dev, fmt, ##__VA_ARGS__);                       \
         |                         ^~~
   drivers/video/fbdev/udlfb.c:1217:19: error: no member named 'dev' in 'struct fb_info'
    1217 |                         dev_info(info->dev,
         |                                  ~~~~  ^
   include/linux/dev_printk.h:160:46: note: expanded from macro 'dev_info'
     160 |         dev_printk_index_wrap(_dev_info, KERN_INFO, dev, dev_fmt(fmt), ##__VA_ARGS__)
         |                                                     ^~~
   include/linux/dev_printk.h:110:11: note: expanded from macro 'dev_printk_index_wrap'
     110 |                 _p_func(dev, fmt, ##__VA_ARGS__);                       \
         |                         ^~~
   drivers/video/fbdev/udlfb.c:1251:12: error: no member named 'dev' in 'struct fb_info'
    1251 |         if (info->dev) {
         |             ~~~~  ^
   drivers/video/fbdev/udlfb.c:1255:15: error: no member named 'dev' in 'struct fb_info'
    1255 |                 dev = info->dev;
         |                       ~~~~  ^
   drivers/video/fbdev/udlfb.c:1379:12: error: no member named 'dev' in 'struct fb_info'
    1379 |         if (info->dev)
         |             ~~~~  ^
   drivers/video/fbdev/udlfb.c:1720:40: error: no member named 'dev' in 'struct fb_info'
    1720 |         retval = device_create_bin_file(info->dev, &edid_attr);
         |                                         ~~~~  ^
   drivers/video/fbdev/udlfb.c:1727:19: error: no member named 'dev' in 'struct fb_info'
    1727 |                  dev_name(info->dev), info->var.xres, info->var.yres,
         |                           ~~~~  ^
   include/linux/dev_printk.h:160:67: note: expanded from macro 'dev_info'
     160 |         dev_printk_index_wrap(_dev_info, KERN_INFO, dev, dev_fmt(fmt), ##__VA_ARGS__)
         |                                                                          ^~~~~~~~~~~
   include/linux/dev_printk.h:110:23: note: expanded from macro 'dev_printk_index_wrap'
     110 |                 _p_func(dev, fmt, ##__VA_ARGS__);                       \
         |                                     ^~~~~~~~~~~
   drivers/video/fbdev/udlfb.c:1767:31: error: no member named 'dev' in 'struct fb_info'
    1767 |         device_remove_bin_file(info->dev, &edid_attr);
         |                                ~~~~  ^
   12 errors generated.


vim +345 drivers/video/fbdev/udlfb.c

59277b679f8b5c drivers/staging/udlfb/udlfb.c Bernie Thompson    2009-11-24  323  
4574203f45eb26 drivers/staging/udlfb/udlfb.c Bernie Thompson    2010-02-15  324  static int dlfb_ops_mmap(struct fb_info *info, struct vm_area_struct *vma)
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  325  {
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  326  	unsigned long start = vma->vm_start;
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  327  	unsigned long size = vma->vm_end - vma->vm_start;
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  328  	unsigned long offset = vma->vm_pgoff << PAGE_SHIFT;
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  329  	unsigned long page, pos;
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  330  
5905585103276b drivers/video/fbdev/udlfb.c   Thomas Zimmermann  2022-04-29  331  	if (info->fbdefio)
5905585103276b drivers/video/fbdev/udlfb.c   Thomas Zimmermann  2022-04-29  332  		return fb_deferred_io_mmap(info, vma);
5905585103276b drivers/video/fbdev/udlfb.c   Thomas Zimmermann  2022-04-29  333  
76f92201b821dd drivers/video/fbdev/udlfb.c   Thomas Zimmermann  2023-11-27  334  	vma->vm_page_prot = pgprot_decrypted(vma->vm_page_prot);
76f92201b821dd drivers/video/fbdev/udlfb.c   Thomas Zimmermann  2023-11-27  335  
04f8afbec37f63 drivers/video/udlfb.c         Tomi Valkeinen     2013-04-18  336  	if (vma->vm_pgoff > (~0UL >> PAGE_SHIFT))
04f8afbec37f63 drivers/video/udlfb.c         Tomi Valkeinen     2013-04-18  337  		return -EINVAL;
04f8afbec37f63 drivers/video/udlfb.c         Tomi Valkeinen     2013-04-18  338  	if (size > info->fix.smem_len)
04f8afbec37f63 drivers/video/udlfb.c         Tomi Valkeinen     2013-04-18  339  		return -EINVAL;
04f8afbec37f63 drivers/video/udlfb.c         Tomi Valkeinen     2013-04-18  340  	if (offset > info->fix.smem_len - size)
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  341  		return -EINVAL;
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  342  
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  343  	pos = (unsigned long)info->fix.smem_start + offset;
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  344  
5865889fe43194 drivers/video/fbdev/udlfb.c   Ladislav Michl     2018-01-16 @345  	dev_dbg(info->dev, "mmap() framebuffer addr:%lu size:%lu\n",
2685cffa9e7992 drivers/staging/udlfb/udlfb.c Bernie Thompson    2010-09-05  346  		pos, size);
2685cffa9e7992 drivers/staging/udlfb/udlfb.c Bernie Thompson    2010-09-05  347  
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  348  	while (size > 0) {
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  349  		page = vmalloc_to_pfn((void *)pos);
f05e0575ed334a drivers/staging/udlfb/udlfb.c Greg Kroah-Hartman 2009-06-03  350  		if (remap_pfn_range(vma, start, page, PAGE_SIZE, PAGE_SHARED))
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  351  			return -EAGAIN;
f05e0575ed334a drivers/staging/udlfb/udlfb.c Greg Kroah-Hartman 2009-06-03  352  
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  353  		start += PAGE_SIZE;
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  354  		pos += PAGE_SIZE;
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  355  		if (size > PAGE_SIZE)
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  356  			size -= PAGE_SIZE;
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  357  		else
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  358  			size = 0;
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  359  	}
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  360  
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  361  	return 0;
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  362  }
88e58b1a42f8c1 drivers/staging/udlfb/udlfb.c Roberto De Ioris   2009-06-03  363  

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki

Re: [PATCH] fbdev: udlfb: make CONFIG_FB_DEVICE optional

[David Hunter]

On 9/27/25 12:12, kernel test robot wrote:
> Hi Sukrut,
> 
> kernel test robot noticed the following build errors:
> 

Did you compile and test this code before submitting this patch?

Re: [PATCH] fbdev: udlfb: make CONFIG_FB_DEVICE optional

[sukrut heroorkar]

Hi David,
Apologies for the late reply,

On Mon, Sep 29, 2025 at 1:29 AM David Hunter
<david.hunter.linux@gmail.com> wrote:
>
> On 9/27/25 12:12, kernel test robot wrote:
> > Hi Sukrut,
> >
> > kernel test robot noticed the following build errors:
> >
>
> Did you compile and test this code before submitting this patch?
>
>

Yes, I had compiled & loaded the udlfb module with no errors. Please
let me know how to proceed
in this case.

Re: [PATCH] fbdev: udlfb: make CONFIG_FB_DEVICE optional

[Helge Deller]

>>> kernel test robot noticed the following build errors:
>>
>> Did you compile and test this code before submitting this patch?
> 
> Yes, I had compiled & loaded the udlfb module with no errors. Please
> let me know how to proceed in this case.

Look at the reported build error, which seems to happen in dev_dbg().
So, maybe in your testing you did not have debugging enabled?
The report contains the .config file with which you can test.

Helge

Re: [PATCH] fbdev: udlfb: make CONFIG_FB_DEVICE optional

[Thomas Zimmermann]

Hi

Am 02.10.25 um 08:41 schrieb Helge Deller:
>>>> kernel test robot noticed the following build errors:
>>>
>>> Did you compile and test this code before submitting this patch?
>>
>> Yes, I had compiled & loaded the udlfb module with no errors. Please
>> let me know how to proceed in this case.
>
> Look at the reported build error, which seems to happen in dev_dbg().
> So, maybe in your testing you did not have debugging enabled?
> The report contains the .config file with which you can test.

Can we rather make an effort to remove the udlfb driver entirely? A few 
years back, there was one user who was still using it because of some 
problems with the DRM udl driver. But I think we've addressed them. The 
discussion is at [1].

[1] 
https://lore.kernel.org/dri-devel/20201130125200.10416-1-tzimmermann@suse.de/

Best regards
Thomas

>
> Helge

-- 
--
Thomas Zimmermann
Graphics Driver Developer
SUSE Software Solutions Germany GmbH
Frankenstrasse 146, 90461 Nuernberg, Germany
GF: Ivo Totev, Andrew Myers, Andrew McDonald, Boudien Moerman
HRB 36809 (AG Nuernberg)

Re: [PATCH] fbdev: udlfb: make CONFIG_FB_DEVICE optional

[sukrut heroorkar]

Hi

On Thu, Oct 2, 2025 at 8:52 AM Thomas Zimmermann <tzimmermann@suse.de> wrote:
>
> Hi
>
> Am 02.10.25 um 08:41 schrieb Helge Deller:
> >>>> kernel test robot noticed the following build errors:
> >>>
> >>> Did you compile and test this code before submitting this patch?
> >>
> >> Yes, I had compiled & loaded the udlfb module with no errors. Please
> >> let me know how to proceed in this case.
> >
> > Look at the reported build error, which seems to happen in dev_dbg().
> > So, maybe in your testing you did not have debugging enabled?
> > The report contains the .config file with which you can test.
>
> Can we rather make an effort to remove the udlfb driver entirely? A few
> years back, there was one user who was still using it because of some
> problems with the DRM udl driver. But I think we've addressed them. The
> discussion is at [1].
Should I send a patch series to completely remove udlfb, since [1] echoed that
DRM udl driver is good enough?
>
> [1]
> https://lore.kernel.org/dri-devel/20201130125200.10416-1-tzimmermann@suse.de/
>
> Best regards
> Thomas
>
> >
> > Helge
>
> --
> --
> Thomas Zimmermann
> Graphics Driver Developer
> SUSE Software Solutions Germany GmbH
> Frankenstrasse 146, 90461 Nuernberg, Germany
> GF: Ivo Totev, Andrew Myers, Andrew McDonald, Boudien Moerman
> HRB 36809 (AG Nuernberg)
>
>

Re: [PATCH] fbdev: udlfb: make CONFIG_FB_DEVICE optional

[Helge Deller]

On 10/3/25 20:43, sukrut heroorkar wrote:
> On Thu, Oct 2, 2025 at 8:52 AM Thomas Zimmermann <tzimmermann@suse.de> wrote:
>> Am 02.10.25 um 08:41 schrieb Helge Deller:
>>>>>> kernel test robot noticed the following build errors:
>>>>>
>>>>> Did you compile and test this code before submitting this patch?
>>>>
>>>> Yes, I had compiled & loaded the udlfb module with no errors. Please
>>>> let me know how to proceed in this case.
>>>
>>> Look at the reported build error, which seems to happen in dev_dbg().
>>> So, maybe in your testing you did not have debugging enabled?
>>> The report contains the .config file with which you can test.
>>
>> Can we rather make an effort to remove the udlfb driver entirely? A few
>> years back, there was one user who was still using it because of some
>> problems with the DRM udl driver. But I think we've addressed them. The
>> discussion is at [1].

Would be good to know if they issues/crashes really have been solved.
In [1] it seems the crashes still happened with DRM.

> Should I send a patch series to completely remove udlfb, 

No. (at least not yet)

> since [1] echoed that DRM udl driver is good enough?
>> [1] https://lore.kernel.org/dri-devel/20201130125200.10416-1-tzimmermann@suse.de/

Well, some people who do *NOT* actively use fbdev with the old
cards say the DRM replacements are "good enough".
For tThose people who really depend on fbdev and the speed it has
over DRM, the DRM "basic-drivers" are simply a nice-to-have-but-not-really-useable
type of drivers.
So, unless the really affected people say the DRM replacement
is fully usable, we need to keep the fbdev driver.

Helge

Re: [PATCH] fbdev: udlfb: make CONFIG_FB_DEVICE optional

[Thomas Zimmermann]

Hi

Am 03.10.25 um 21:50 schrieb Helge Deller:
> On 10/3/25 20:43, sukrut heroorkar wrote:
>> On Thu, Oct 2, 2025 at 8:52 AM Thomas Zimmermann 
>> <tzimmermann@suse.de> wrote:
>>> Am 02.10.25 um 08:41 schrieb Helge Deller:
>>>>>>> kernel test robot noticed the following build errors:
>>>>>>
>>>>>> Did you compile and test this code before submitting this patch?
>>>>>
>>>>> Yes, I had compiled & loaded the udlfb module with no errors. Please
>>>>> let me know how to proceed in this case.
>>>>
>>>> Look at the reported build error, which seems to happen in dev_dbg().
>>>> So, maybe in your testing you did not have debugging enabled?
>>>> The report contains the .config file with which you can test.
>>>
>>> Can we rather make an effort to remove the udlfb driver entirely? A few
>>> years back, there was one user who was still using it because of some
>>> problems with the DRM udl driver. But I think we've addressed them. The
>>> discussion is at [1].
>
> Would be good to know if they issues/crashes really have been solved.
> In [1] it seems the crashes still happened with DRM.

The thread at [1] was the original removal attempt. And that was 5 years 
ago. I think we could retry and take the reporter (Mikulas) into cc.

Best regards
Thomas

>
>> Should I send a patch series to completely remove udlfb, 
>
> No. (at least not yet)
>
>> since [1] echoed that DRM udl driver is good enough?
>>> [1] 
>>> https://lore.kernel.org/dri-devel/20201130125200.10416-1-tzimmermann@suse.de/
>
> Well, some people who do *NOT* actively use fbdev with the old
> cards say the DRM replacements are "good enough".
> For tThose people who really depend on fbdev and the speed it has
> over DRM, the DRM "basic-drivers" are simply a 
> nice-to-have-but-not-really-useable
> type of drivers.
> So, unless the really affected people say the DRM replacement
> is fully usable, we need to keep the fbdev driver.
>
> Helge

-- 
--
Thomas Zimmermann
Graphics Driver Developer
SUSE Software Solutions Germany GmbH
Frankenstrasse 146, 90461 Nuernberg, Germany
GF: Ivo Totev, Andrew Myers, Andrew McDonald, Boudien Moerman
HRB 36809 (AG Nuernberg)

Re: [PATCH] fbdev: udlfb: make CONFIG_FB_DEVICE optional

[Mikulas Patocka]

[-- Attachment #1: Type: text/plain, Size: 1666 bytes --]



On Tue, 7 Oct 2025, Thomas Zimmermann wrote:

> Hi
> 
> Am 03.10.25 um 21:50 schrieb Helge Deller:
> > On 10/3/25 20:43, sukrut heroorkar wrote:
> > > On Thu, Oct 2, 2025 at 8:52 AM Thomas Zimmermann <tzimmermann@suse.de>
> > > wrote:
> > > > Am 02.10.25 um 08:41 schrieb Helge Deller:
> > > > > > > > kernel test robot noticed the following build errors:
> > > > > > > 
> > > > > > > Did you compile and test this code before submitting this patch?
> > > > > > 
> > > > > > Yes, I had compiled & loaded the udlfb module with no errors. Please
> > > > > > let me know how to proceed in this case.
> > > > > 
> > > > > Look at the reported build error, which seems to happen in dev_dbg().
> > > > > So, maybe in your testing you did not have debugging enabled?
> > > > > The report contains the .config file with which you can test.
> > > > 
> > > > Can we rather make an effort to remove the udlfb driver entirely? A few
> > > > years back, there was one user who was still using it because of some
> > > > problems with the DRM udl driver. But I think we've addressed them. The
> > > > discussion is at [1].

It was me - and I am still using it on an ARM64 MacchiatoBIN board because 
the board doesn't have graphics output.

The problems with the UDL DRM driver were:

* crashes with full-screen framebuffer applications, such as "links2 -g", 
"fbi" or "fbgs". On UDLFB, there are no crashes.

* worse performance - the UDL DRM driver updates everything in a given 
rectangle, while the UDLFB driver keeps back-buffer and front-buffer and 
updates only differences between them.

* crash when you unplug the card while Xorg was running (already fixed)

Mikulas

Re: [PATCH] fbdev: udlfb: make CONFIG_FB_DEVICE optional

[David Hunter]

On 10/2/25 02:35, sukrut heroorkar wrote:
>>> kernel test robot noticed the following build errors:
>>>
>> Did you compile and test this code before submitting this patch?
>>
>>
> Yes, I had compiled & loaded the udlfb module with no errors. Please
> let me know how to proceed
> in this case.

Hey Sukrut,

When you make code that deletes something from the kernel, you need to
make sure that all other code that references that code will still
function properly.

When you surround things in the #ifdev, depending on the config file,
the compiler strips those things out and compiles without them. That
means that you actually need to compile and test under two conditions.
Once when CONFIG_FB_DEVICE=y and another time when CONFIG_FB_DEVICE=n.

The test robot gave you a sample config file that you can use. Please
ensure that you have the proper version of Clang on your Host machine
for that particular config file, if you choose to use it.

Also, I strongly recommend that you run the code on a sufficient
hardware or emulator after you test it so that you can verify that the
code does what you think it does.

Thanks,
David Hunter