[dtaht/Cerowrt-3.3] 3aaa04: Reworked unaligned access patches

[dtaht/Cerowrt-3.3] 3aaa04: Reworked unaligned access patches

[GitHub]

[-- Attachment #1: Type: text/plain, Size: 1590 bytes --]

  Branch: refs/heads/master
  Home:   https://github.com/dtaht/Cerowrt-3.3
  Commit: 3aaa04f538d183bd7d420e5d67e8a88a938d14b0
      https://github.com/dtaht/Cerowrt-3.3/commit/3aaa04f538d183bd7d420e5d67e8a88a938d14b0
  Author: Dave Taht <dave.taht@bufferbloat.net>
  Date:   2012-04-12 (Thu, 12 Apr 2012)

  Changed paths:
    R target/linux/ar71xx/patches-3.3/0001-Export-DNSSL-RA-option-to-userspace.patch
    A target/linux/ar71xx/patches-3.3/0001-Treat-ND-option-31-as-userland-DNSSL-support.patch
    A target/linux/ar71xx/patches-3.3/0002-Unaligned-Access-Hacks-for-ar71xx-Architecture.patch
    A target/linux/ar71xx/patches-3.3/0003-Some-more-mostly-ipv6-related-alignment-patches.patch
    R target/linux/ar71xx/patches-3.3/902-unaligned_access_hacks.patch

  Log Message:
  -----------
  Reworked unaligned access patches

In a 60 second 93Mbit test of netperf directly to the router:

proto  traps
ipv4: 1529909
ipv6: 6080546

So throwing 100,000 unaligned traps/sec is rather bad.
25,000 is also bad.

I note that forwarding performance is currently over 200k for ipv4


  Commit: ecb2d9338153cdda5f2bfa63d0243c535cedcf33
      https://github.com/dtaht/Cerowrt-3.3/commit/ecb2d9338153cdda5f2bfa63d0243c535cedcf33
  Author: Dave Taht <dave.taht@bufferbloat.net>
  Date:   2012-04-12 (Thu, 12 Apr 2012)

  Changed paths:
    R package/hostapd/patches/500-random_pool_add_kernel.patch

  Log Message:
  -----------
  No point in the random patch from hostapd

we get some randomness now from other sources.


Compare: https://github.com/dtaht/Cerowrt-3.3/compare/91b4ca9...ecb2d93