kmap_high/flush_tlb_all/smp_call_function problem

2000-10-09 Thread Rajagopal Ananthanarayanan
We are running some fairly intensive tests with XFS and with HIGHMEM on an X86 machine with 4G of memory and 4 cpus. The linux-kernel version is test8. The code also contains Stephen Tweedie's highmem-kiobuf. What we see is that after a while of running, smp_call_function getting stuck in the

kmap_high/flush_tlb_all/smp_call_function problem

2000-10-09 Thread Rajagopal Ananthanarayanan
We are running some fairly intensive tests with XFS and with HIGHMEM on an X86 machine with 4G of memory and 4 cpus. The linux-kernel version is test8. The code also contains Stephen Tweedie's highmem-kiobuf. What we see is that after a while of running, smp_call_function getting stuck in the