RE: Regression caused by using node_to_bdi()

[Zhao Lei]

Hi, Christoph

> -----Original Message-----
> From: 'Christoph Hellwig' [mailto:hch@xxxxxx]
> Sent: Tuesday, April 14, 2015 1:33 AM
> To: Zhao Lei
> Cc: 'Christoph Hellwig'; linux-fsdevel@xxxxxxxxxxxxxxx; 'Jan Kara'; 'Jens Axboe';
> 'LKML'
> Subject: Re: Regression caused by using node_to_bdi()
> 
> Tejun posted a series last week that moves node_to_bdi back inline to -fsdevel.
> 
> Can you apply patches 1-18 of the series
> 
> "[PATCHSET 1/3 v3 block/for-4.1/core] writeback: cgroup writeback support"
> 
> and see if that helps your benchmark?
I done performance test for this patchset(1~18).

But the result is not changed, performace is not increased:

v3.19-rc1              : io_speed: valcnt=10 avg=214.688 range=[211.460,216.190] diff=  2.24% stdev=1.417 cv=0.66%
v3.19                  : io_speed: valcnt=10 avg=212.529 range=[206.160,216.980] diff=  5.25% stdev=3.588 cv=1.69%

# This is 3 patch before
v3.19_04754_1b30e6     : io_speed: valcnt=10 avg=214.072 range=[210.690,217.440] diff=  3.20% stdev=2.222 cv=1.04%
v3.19_04765_495a27     : io_speed: valcnt=10 avg=210.163 range=[205.130,215.630] diff=  5.12% stdev=3.140 cv=1.49%
v3.19_04766_26ff13     : io_speed: valcnt=10 avg=210.045 range=[207.590,212.550] diff=  2.39% stdev=1.860 cv=0.89%

# This is patch of using node_to_bdi()
v3.19_04767_de1414     : io_speed: valcnt=10 avg=205.216 range=[202.870,206.620] diff=  1.85% stdev=0.977 cv=0.48%

# This is 3 patch after
v3.19_04768_b83ae6     : io_speed: valcnt=10 avg=205.265 range=[203.720,206.170] diff=  1.20% stdev=0.721 cv=0.35%
v3.19_04769_e4d275     : io_speed: valcnt=10 avg=205.787 range=[205.250,206.150] diff=  0.44% stdev=0.306 cv=0.15%
v3.19_04770_7b14a2     : io_speed: valcnt=10 avg=203.558 range=[203.110,204.020] diff=  0.45% stdev=0.308 cv=0.15%

v4.0-rc1               : io_speed: valcnt=10 avg=204.917 range=[203.370,205.890] diff=  1.24% stdev=0.663 cv=0.32%

# This is 1~18 of above patchset on top of v4.0-rc1
v4.0-rc1_00018_579465  : io_speed: valcnt=10 avg=205.348 range=[204.430,206.120] diff=  0.83% stdev=0.572 cv=0.28%

v4.0                   : io_speed: valcnt=10 avg=205.442 range=[204.680,207.290] diff=  1.28% stdev=0.831 cv=0.40%

(Above performance data is changed with my first mail,
because the old harddisk was broken, then I replaced to a
new one and run all performance test again)

Thanks
Zhaolei


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/