On Sun, Sep 25, 2011 at 10:33:40PM +0800, Sepherosa Ziehau wrote: > > Hmm, even though the above patch is correct, but we probably should > set POLLERR if errors happened on the connection, please test the > latest master: > aa622c3d0448dbcf9cad62ea40234b9b01173663 > > I believe even without the patching firefox, the bug should have been fixed.
It is. With the new kernel, a few seconds after loading web pages, a non-patched seamonkey goes back to using 0% cpu. Thank you! -- Francois Tigeot
