Re: [PATCH] drm/i915: set IDICOS to medium uncore resources

[Ben Widawsky <ben@bwidawsk.net>]

On Sun, 06 May 2012 23:51:26 -0700
Eric Anholt <eric@anholt.net> wrote:

> On Fri,  4 May 2012 18:58:59 -0700, Ben Widawsky <ben@bwidawsk.net> wrote:
> > I'm seeing about a 5% FPS improvement across various benchmarks on my
> > IVB i3. Rumor has it that the higher end parts show even more benefit.
> > 
> > This derives from a patch originally given to me by Bernard. The docs
> > are  confusing about the definition names (ie. medium really seems like
> > max), but it would seem it gives more cache to the GT at the expense of
> > uncore. This configuration makes the split most in favor of the GT. I've
> > not tried the other IDICOS values.
> 
> My IVB system started with an SNBPCR value of 0x1b40.  To test the effect of
> this patch, I just smashed the value of the reg at runtime:
> 
> Results:
> x intel_reg_write 0x900c 0x001b40
> + intel_reg_write 0x900c 0x201b40
> 
> citybench:
> +------------------------------------------------------------------------------+
> |                   xxx                          +                             |
> |   x               xxx          +               + +                           |
> |   x               xxx          +               +++                           |
> |   x               xxx          +            + ++++                           |
> |   xx             xxxx          +            + ++++                           |
> |x xxx           x xxxx        + +            ++++++ ++                        |
> |xxxxx          xxxxxxxx       + + + +        +*++++ ++                        |
> |xxxxxxxxx      xxxxxxxx x     +++++ ++++     +*+++++++                        |
> |xxxxxxxxx    xxxxxxxxxx x     ++++++++++   ++**+++++++ +          +           |
> |xxxxxxxxx  xxxxxxxxxxxxxx    *++++++++++++ ****+*++++++++      +  ++++++     +|
> |    |___________AM_________|       |_________A_M_______|                      |
> +------------------------------------------------------------------------------+
>     N           Min           Max        Median           Avg        Stddev
> x 118       335.779       387.456       354.306     352.63131     12.462796
> + 118       367.324        419.02        386.14     384.47139     10.970918
> Difference at 95.0% confidence
> 	31.8401 +/- 2.99585
> 	9.02928% +/- 0.849569%
> 	(Student's t, pooled s = 11.7406)
> 
> openarena:
> +------------------------------------------------------------------------------+
> |         +                ++                                xx xx x  x xx     |
> |         +      +         ++ + +                            xxxxxxx xxxxx     |
> |        +++  ++++++++++ ++++++ ++ +                    x  xxxxxxxxxxxxxxx   xx|
> |++ + ++++++++++++++++++++++++++++ + + +             x  x xxxxxxxxxxxxxxxxxx xx|
> |          |_________A________|                              |_____A____|      |
> +------------------------------------------------------------------------------+
>     N           Min           Max        Median           Avg        Stddev
> x  65         273.5         292.2           284     283.73385     4.2561822
> +  65         234.8         263.2         249.9         249.6     6.9873323
> Difference at 95.0% confidence
> 	-34.1338 +/- 1.989
> 	-12.0302% +/- 0.70101%
> 
> nexuiz:
> +------------------------------------------------------------------------------+
> |                                                                 +            |
> |                                                   x        +    +            |
> |                                             +     x   x xx *x+  +x           |
> |+     +         +        x   *x+x xxx    + x++ + x+*xxxx+x***x+++*x**+x ++  *+|
> |                                   |____|____________AAM____M_____|______|    |
> +------------------------------------------------------------------------------+
>     N           Min           Max        Median           Avg        Stddev
> x  33     74.314888     76.177197     75.424764     75.332161    0.48205891
> +  33     73.397254     76.219868     75.584749     75.363209    0.69408523
> 
> Using these 3 tests, this patch does not look like a win.  Note that OA
> is the most CPU-bound of these, while nexuiz is usually on the edge
> between being CPU and GPU-bound, and citybench is definitely GPU
> limited.

Strange. It would be helpful to start to investigate these systems I
suppose. This is basically what I see on my SNB (though I use xonotic
instead of nexuiz). My SNB is an i7, and my IVB is an i3.
Below are the interesting ones from IVB (I've lost OA somehow).

x xonotic_dinq_fps
+ xonotic_poland_fps
+----------------------------------------------------------------------+
|        x       xx                                             +      |
|x x     xx  +*x *x+                                +      +   ++  +  +|
|     |_____A_M___|  |_______________________A_____________M__________||
+----------------------------------------------------------------------+
    N           Min           Max        Median           Avg        Stddev
x  11     98.398296     101.24008     100.58156     100.19283    0.97987835
+  11     100.33134     109.63333     107.76804      105.6379      3.939517
Difference at 95.0% confidence
        5.44507 +/- 2.55327
        5.43459% +/- 2.54835%
        (Student's t, pooled s = 2.87054)


Not sure why my citybench numbers are so low

x citybench_dinq_fps
+ citybench_poland_fps
+----------------------------------------------------------------------+
|    xxxxxx    x          +               +     +                      |
|+x xxxxxxx xx x          + + +  +        +  ++ +   +         ++  ++  +|
|    |_MA__|              |_________________AM________________|        |
+----------------------------------------------------------------------+
    N           Min           Max        Median           Avg        Stddev
x  26       196.857       200.568      198.4245     198.56062     0.8468172
+  18        196.69       215.299      208.6485     208.31394     4.8783414
Difference at 95.0% confidence
        9.75333 +/- 1.96251
        4.91202% +/- 0.988368%
        (Student's t, pooled s = 3.17166)


-- 
Ben Widawsky, Intel Open Source Technology Center