[arch-general] Nvidia/vesafb/GRUB2

[Don deJuan]

On 06/27/2012 07:42 AM, Arno Gaboury wrote:
> On 06/27/2012 04:18 PM, Don deJuan wrote:
>> On 06/27/2012 06:55 AM, Uroš Vampl wrote:
>>> Arno Gaboury <arnaud.gaboury <at> gmail.com> writes:
>>>
>>>> After lots of reading, especially *Nvidia* official readme, it seems
>>>> this card SUPPORTS indded *Vesafb*. So I think this error message has
>>>> nothing to do here, and I will keep my *grub* file as it was first.
>>>
>>> No, it supports the vesa standard. All cards do. But that's
>>> completely different
>>> from vesafb, a linux driver. When you use vesafb and the nvidia
>>> driver together,
>>> you effectively have two drivers poking at the card at the same time.
>>> That
>>> problems can arise from that makes full sense.
>>>
>>> You can keep using vesafb. After all, the message is just that - a
>>> message - and
>>> not an error. It's just that if something breaks, you get to keep the
>>> pieces.
>>>
>>>
>> Actually no Nvidia never supported VESA, it just happened to work.
>>
>> http://www.nvnews.net/vbulletin/showpost.php?p=2561806&postcount=39
>>
> Are you so sure?
>
> http://us.download.nvidia.com/XFree86/Linux-x86/173.14.09/README/chapter-19.html:
>
>
> First sentence: *The NVIDIA Accelerated Linux Graphics Driver supports
> all standard VGA and VESA modes*
Well the link I sent was from an Nvidia employee posting about these 
issues, so yes I am sure. Also why are you pointing to a readme for the 
legacy version?

It has been stated by Nvidia themselves in other places as well, that 
VESA just "happened" to work, was not "officially" supported. So take 
that as you will, but from what I have read and heard from Nvidia this 
is the case, even googling reveals the same information going all the 
way back to around 2003. But no point in arguing the semantics about 
this, I feel the information I found and gotten from Nvidia directly is 
more than sufficient, your choice if you want to listen to me and other 
Nvidia employees posting on this issue. You can easily get the same info 
regardless of what their README states. Or if you have any buddies in 
the Linux department at nvidia, shoot them an email and you will get the 
same response from them, it just happened to work.