[ofa-general] [PATCH] RDMA/nes: application hang during large cluster test
Roland Dreier
rdreier at cisco.com
Mon Apr 20 13:59:56 PDT 2009
This patch seems to be conflating multiple fixes -- or am I
misunderstanding? For example:
> * Under heavy load, sometimes it takes longer to receive the response from
> application to the MPA request. The rexmit timeout value is too low.
so bumping the rexmit timeout would be one independent fix.
> * check_seq(), does not check for condition if the seq# is wrapped.
and this seems completely independent.
etc etc.
can you split this up into a series of patches that each fix one bug only?
More information about the general
mailing list