Syslinux - Oct 2011 - [PATCH][GIT-PULL] lwIP undiif: Fixes for VMware platforms and general fixes

The following changes since commit 5b08704e284545c82b4c5d3ff3aaee815107f6ef:
  H. Peter Anvin (1):
        thread: allow marking invalid; allow for static mboxes

are available in the git repository at:

  git://github.com/geneC1/syslinux.git lwip-undiif-fix-for-hpa

Gene Cumm (4):
      lwip: undiif: Fix #error typo
      lwip: undiif: Fix debug option addressing in update_arp_entry()
      lwip:undiif: Use tcpip_input() rather than ethernet_input()/ip_input()
      lwip:undiif: Add interface with tcpip_input() callback

 core/lwip/src/netif/undiif.c |   12 ++++++------
 1 files changed, 6 insertions(+), 6 deletions(-)

-- 
-Gene

On 10/13/2011 07:29 PM, Gene Cumm wrote:
> The following changes since commit
5b08704e284545c82b4c5d3ff3aaee815107f6ef:
>    H. Peter Anvin (1):
>          thread: allow marking invalid; allow for static mboxes
>
> are available in the git repository at:
>
>    git://github.com/geneC1/syslinux.git lwip-undiif-fix-for-hpa
>
> Gene Cumm (4):
>        lwip: undiif: Fix #error typo
>        lwip: undiif: Fix debug option addressing in update_arp_entry()
>        lwip:undiif: Use tcpip_input() rather than
ethernet_input()/ip_input()
>        lwip:undiif: Add interface with tcpip_input() callback
>
>   core/lwip/src/netif/undiif.c |   12 ++++++------
>   1 files changed, 6 insertions(+), 6 deletions(-)
>
Hi Gene,

I have merged this... I'm still not convinced it is the right solution 
-- in particular I'm not convinced we should *ever* call 
ethernet_input() -- but it's a lot better than the previous (known 
broken) code.

Thanks!

	-hpa

On Fri, Oct 21, 2011 at 18:17, H. Peter Anvin <hpa at zytor.com>
wrote:
> On 10/20/2011 06:11 PM, Eric W. Biederman wrote:
>>
>> tcpip_input always calls ethernet_input if you are an ethernet device.
>> So the only practical difference is bounding through an extra thread.
>>
>> Maybe there is a good reason for bouncing through an extra thread.
>> I seem to have vague memories about it just being plain less efficient
>> and killing through put.
>>
>> I did not use tcpip_input deliberately. ?At the very least because it
>> was an apparent unnecessary complication.
>>
>
> ... except not doing that violates the locking constraints of the code.
Or probably requires complete replacement of tcpip_thread()

-- 
-Gene