Re: 2.6.10-rc1-mm4 -1 EAGAIN after allocation failure was: Re: Kernel 2.6.9 Multiple Page Allocation Failures

On Wed, Nov 10, 2004 at 03:24:10PM +1100, Nick Piggin wrote:

Can you try the following patch, please? It is diffed against 2.6.10-rc1,
but I think it should apply to -mm kernels as well.

Basically 2.6.8 and earlier kernels had some quirks in the page allocator
that would allow for example, a large portion of "DMA" memory to be reserved
for network memory allocations (atomic allocations). After 'fixing' this
problem, 2.6.9 is effectively left with about a quarter the amount of memory
reserved for network allocations compared with 2.6.8.

The following patch roughly restores parity there. Thanks.

I applied the patch to 2.6.10-rc1-mm4 and the application froze again, but i
just remembered that i changed a kernel-option in mm4 and forgot about that
yesterday:
I unset CONFIG_PACKET_MMAP and i suppose this could have this kind of effect
on high connection rates.
I set it back to CONFIG_PACKET_MMAP=y and if the application does not freeze
for some hours at this load we can blame at least this issue (-1 EAGAIN) on
that parameter.

My variation of Harrisberger's Fourth Law of the Lab:
 Experience is directly proportional to the amount of braincells ruined.

*ouch*,
	Stefan