Kernel patch enables gcc optimizations for additional CPUs.
Latest commit 7e4c6fb Jan 7, 2017 @graysky2 typo in comments

README.md

Kernel_gcc_patch

This kernel patch adds additional CPU options to the Linux kernel accessible under:

 Processor type and features  --->
 Processor family --->

Why a specific patch?

The kernel uses its own set of CFLAGS, KCFLAGS. For exmaple, see:

Expanded CPUs include

CPU Family GCC Optimization
Native optimizations autodetected by GCC -march=native
AMD Improved K8-family -march=k8-sse3
AMD K10-family -march=amdfam10
AMD Family 10h (Barcelona) -march=barcelona
AMD Family 14h (Bobcat) -march=btver1
AMD Family 16h (Jaguar) -march=btver2
AMD Family 15h (Bulldozer) -march=bdver1
AMD Family 15h (Piledriver) -march=bdver2
AMD Family 15h (Steamroller) -march=bdver3
AMD Family 15h (Excavator) -march=bdver4
AMD Family 17h (Zen) -march=znver1
Intel Bonnell family of low-power Atom processors (Bonnell) -march=bonnell
Intel Silvermont family of low-power Atom processors (Silvermont) -march=silvermont
Intel 1st Gen Core i3/i5/i7-family (Nehalem) -march=nehalem
Intel 1.5 Gen Core i3/i5/i7-family (Westmere) -march=westmere
Intel 2nd Gen Core i3/i5/i7-family (Sandybridge) -march=sandybridge
Intel 3rd Gen Core i3/i5/i7-family (Ivybridge) -march=ivybridge
Intel 4th Gen Core i3/i5/i7-family (Haswell) -march=haswell
Intel 5th Gen Core i3/i5/i7-family (Broadwell) -march=broadwell
Intel 6th Gen Core i3/i5/i7-family (Skylake) -march=skylake

Benchmarks

Intro

Three different machines running a generic x86-64 kernel and an otherwise identical kernel running with the optimized gcc options were tested using a make based endpoint.

Conclusion

There are small but real speed increases to running with this patch as judged by a make endpoint. The increases are on par with the speed increase that the upstream sanctioned core2 option gives users, so not including additional options seems somewhat arbitrary to me.

Details

  1. Three test machines: Intel Xeon X3360, Intel i7-2620M, Intel Core i7-3660K.
  2. All ran the make benchmark (linked below) 35 times while booted into a 'generic' kernel. Then all ran the same make benchmark 35 times after booting into an optimized kernel. Below are the optimizations chosen for each machine.
    • X3360 = core2
    • i7-2620M = sandybridge
    • i7-3660K = ivybridge
  3. Results were analyzed for statistical significance via ANOVA plots that clearly show statistically significant albeit small differences.

Discussion

  1. All the assumptions for ANOVA are met:
    • Data are normally distributed as show in the normal quantile plots.
    • The population variances are fairly equal (Levene and Barlett tests).
  2. The ANOVA plots clearly show significance.
    • Pair-wise analysis by Tukey-Kramer shows significance at the 0.05 level for all CPUs compared.

Below are the differences in median values:

CPU Difference in median value
core2 +87.5 ms
sandybridge +79.7 ms
ivybridge +257.2 ms

References

Credit

Legacy support

Find support for older version of the linux kernel and of gcc in the outdated_versions directory.

Data

Sandybridge vs. Generic

corei7avx

Ivybridge vs. Generic

coreavxi

Core2 vs. Generic

core2