Return to enet.Questions CVS log

Up to [Apple Darwin 0.x] / driverkit / notes

Darwin 0.2 Driver Kit

			Ethernet questions
			
1. Why does the 68k driver use IF_ENQUEUE and IF_DEQUEUE? These look like 
   private macros for the netif module...They'll prevent us from using 
   only netif.h! How to enqueue and dequeue from netif_t legally?
   
   -- Maybe don't have to enqueue mbufs on netif; 68k had to do that 
      because of finite-size DMA queue...We should just be able to 
      chan_dma_enqueue 'em.

2. How important is it to use netbuf's instead of mbufs? Is this actually 
   possible in current system (without a lot of hacking outside of enet
   driver)?
   
   -- Brad says: go for it, by the book. OK, let's try it.
   
3. Re: this comment at top of enoutput():

   Use trailer local net encapsulation if enough data in first
   packet leaves a multiple of 512 bytes of data in remainder.

   I don't see any code which does this.
   
   -- I think that the enet driver doesn't have to worry about this; 
      it's handled in if_venip.c.
   
4. engetbuf, I assume, is never called from interrupt level, since 
   it calls nb_alloc_wrapper, which the tech docs say can not be called
   from interrupt level. However, 
   engetbuf() {
   	en_get()
		enbuf_alloc()
			m_clalloc() ...
			
   m_clalloc() has a hack to allocate memory at interrupt level. 
   
   Will a network module's output or getbuf routines in fact be called
   at interrupt level?
   
   -- Brad and Joel G. both say NO. I'm going to go ahead and assume that
      I can do RPCs and kallocs from both routines. Brad agreed to help fix
      deadlock problems which may arise from this. 
		
5. Maybe we want the enetBuf's to always be wired. If so, kern_dev needs 
   one of two changes:
     -- at dma_enqueue() time, rememeber if current vm_region is pageable;
        if not, don't bother wiring it, and don't unwire it at dequeue time 
     -- att arg to chan_dma_enqueue_int() to specifiy whether or not to bother
        wiring and unwiring the memory.
	

This archive runs on limited infrastructure. Preserving old code on modern bandwidth. Automated agents are requested to crawl responsibly.