About maximum and recommended value of FB_MULTI_BUFFER

cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

About maximum and recommended value of FB_MULTI_BUFFER

2,717 Views
takayuki_ishii
Contributor IV

Hello community,

I have 3 question about envrionment valiable of FB_MULTI_BUFFER.

1) In iMX_Graphics_User's_Uide of Linux BSP document, FB_MULTI_BUFFER in Table

     of i.MX 6 FB API environment variables see that recommended values is 4 and maximum is 8.

      The other hand, in API description and syntax of fbGetDisplayByIndex, it say that

      "To use multiple buffers when rendering, set the environment variable FB_MULTI_BUFFER to an unsigned integer
      value, which indicates the number of buffers. Maximum is 3."

   That meen, if FB_MULTI_BUFFER > 3, fbGetDisplayByIndex() is failed and return NULL pointer,

   So in this case, we must use other API like a fbGetDisplay();

   Is it correct?

2)  What is a "hardware limitation of current IPU" in i.MX 6 FB API environment variables?

   Does It suggest that "i.MX6 X11 EGL is single buffered and there is no support for VSYNC."?

3) How to effect a FB_MULTI_BUFFER variable, if it is no IPU device like a i.MX6 SoloX?

Best regards,

Ishii.

Labels (4)
0 Kudos
5 Replies

1,656 Views
Bio_TICFSL
NXP TechSupport
NXP TechSupport

Hi Takayuki-san,

Sorry for the delay, now:

"No limit" mean that only 32bit unsigned integer value is a limit of FB_MULITBUFFER value.

Is it correct? --Correct!

"The hardware limitation is 4." mean that FB_MULTIBUFFER has no limit, but if set more than 4,

It will worked by software buffer handling. So performance are very slow.

To keep good benchmark, it must set less than equal 4 to use IPU hardware control.

Is it correct? --Correct!

So performance is not better than IPU device like Solo/DualLite/Dual/Quad.

Is it correct? --Correct!

If we use a X11 EGL, 

 1) It must reject a tearing on screen because X11 EGL not support VSYNC interrupt.

"Because of the hardware limitation of current IPU," comment in the "Table 13 i.MX 6 FB API

environment variables" in i.MX_Graphics_User's_Guide.pdf(IMXGRAPHICUG Rev. 0, 10/2016)

in the fsl_yocto-L4.1.15_2.0.0-ga.tar.gz mean that IPU have a double/triple buffer synchronization

mechanism with a video/graphics source.

Is it correct?

--- Correct

0 Kudos

1,656 Views
takayuki_ishii
Contributor IV

Hello Bio_TICFSL,

Thank you for your reply.

I will ask my customer it.

Best regards,

Ishii.

0 Kudos

1,656 Views
Bio_TICFSL
NXP TechSupport
NXP TechSupport

Hi Takayuki,

No limit, but we generally use FB_MULTIBUFFER =2

The hardware limitation is 4. The X11 EGL does not support VSYNC

0 Kudos

1,656 Views
takayuki_ishii
Contributor IV

Hello Bio

Thank you for your response.

But I can not understand your answer.

So please teach me more detail of your answer.

> No limit, but we generally use FB_MULTIBUFFER =2

> The hardware limitation is 4. The X11 EGL does not support VSYNC

"No limit" mean that only 32bit unsigned integer value is a limit of FB_MULITBUFFER value.

Is it correct?

"The hardware limitation is 4." mean that FB_MULTIBUFFER has no limit, but if set more than 4,

It will worked by software buffer handling. So performance are very slow.

To keep good benchmark, it must set less than equal 4 to use IPU hardware control.

Is it correct?

Like a SoloX, no IPU hardware device, worked by software control only.

So performance is not better than IPU device like Solo/DualLite/Dual/Quad.

Is it correct?

If we use a X11 EGL, 

 1) It must reject a tearing on screen because X11 EGL not support VSYNC interrupt.

 2) It must set to FB_MULTI_BUFFER=4 for no screen tearing appears. 

 3) other.

    i.e: For no screen tearing appears, we must select "Wayland" option build and 

          FB_MULTI_BUFFER=4 setting.

Which is a correct answer?

"Because of the hardware limitation of current IPU," comment in the "Table 13 i.MX 6 FB API

environment variables" in i.MX_Graphics_User's_Guide.pdf(IMXGRAPHICUG Rev. 0, 10/2016)

in the fsl_yocto-L4.1.15_2.0.0-ga.tar.gz mean that IPU have a double/triple buffer synchronization

mechanism with a video/graphics source.

Is it correct?

I am looking forward to hearing from you.

Best regards,

Ishii.

0 Kudos

1,656 Views
takayuki_ishii
Contributor IV

Hello Bio_TICFSL

Please give me your comment to me.

Should you need any further information, please do not hesitate to contact me.

Best regards,

Ishii.

0 Kudos