Peter Jeremy wrote: > I've just been tracking down a panic in 4.x and found that the problem > is still present in 6.x. > > According to bus_dma(9), bus_dmamem_alloc() can be invoked with a > flag BUS_DMA_NOWAIT to indicate that sleep()ing is not allowed. > > At least on the i386, if the requested size exceeds 1 page (or some > other cases), the requested memory will be allocated via contigmalloc(). > > bus_dmamem_alloc() maps BUS_DMA_NOWAIT to M_NOWAIT but contigmalloc() > does not support M_NOWAIT and will tsleep() under some conditions. > > Unfortunately, I don't know enough about the VM code to be able to > suggest a fix. > Will contigmalloc() actually sleep? If so, then this is something that needs to be addressed in contigmalloc. ScottReceived on Sun Jan 09 2005 - 04:25:22 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:25 UTC