Re: [PATCH] drm/vmwgfx: respect 'nomodeset'

[Thomas Hellstrom <thellstrom@vmware.com>]

On 10/15/2014 09:46 PM, Rob Clark wrote:
> On Wed, Oct 15, 2014 at 3:24 PM, Thomas Hellstrom <thellstrom@vmware.com> wrote:
>> On 10/15/2014 09:00 PM, Rob Clark wrote:
>>> Signed-off-by: Rob Clark <robdclark@gmail.com>
>>> ---
>>>  drivers/gpu/drm/vmwgfx/vmwgfx_drv.c | 7 +++++++
>>>  1 file changed, 7 insertions(+)
>>>
>>> diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
>>> index 18b54ac..f0267b8 100644
>>> --- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
>>> +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
>>> @@ -25,6 +25,7 @@
>>>   *
>>>   **************************************************************************/
>>>  #include <linux/module.h>
>>> +#include <linux/console.h>
>>>
>>>  #include <drm/drmP.h>
>>>  #include "vmwgfx_drv.h"
>>> @@ -1453,6 +1454,12 @@ static int vmw_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
>>>  static int __init vmwgfx_init(void)
>>>  {
>>>       int ret;
>>> +
>>> +#ifdef CONFIG_VGA_CONSOLE
>>> +     if (vgacon_text_force())
>>> +             return -EINVAL;
>>> +#endif
>>> +
>> Hmm,
>>
>> From the function name vgacon_text_force() it sounds like this should
>> just stop the driver from initializing fbcon? Not refuse to load?
> yeah, the function is badly named.. it perhaps should be
> vgacon_is_text_forced() or something like that.  But basically it
> returns whether we are forced to text mode.
>
> BR,
> -R

So then I guess it would be more correct to use the output of that
function when determining the value of dev_priv->enable_fb (vmwgfx can
enable the user-space modesetting API without turning on vmwgfx fbcon).

/Thomas