atari_NCR5380.c

/* 
 * NCR 5380 generic driver routines.  These should make it *trivial*
 * 	to implement 5380 SCSI drivers under Linux with a non-trantor
 *	architecture.
 *
 *	Note that these routines also work with NR53c400 family chips.
 *
 * Copyright 1993, Drew Eckhardt
 *	Visionary Computing 
 *	(Unix and Linux consulting and custom programming)
 * 	drew@colorado.edu
 *	+1 (303) 666-5836
 *
 * DISTRIBUTION RELEASE 6. 
 *
 * For more information, please consult 
 *
 * NCR 5380 Family
 * SCSI Protocol Controller
 * Databook
 *
 * NCR Microelectronics
 * 1635 Aeroplaza Drive
 * Colorado Springs, CO 80916
 * 1+ (719) 578-3400
 * 1+ (800) 334-5454
 */

/*
 * ++roman: To port the 5380 driver to the Atari, I had to do some changes in
 * this file, too:
 *
 *  - Some of the debug statements were incorrect (undefined variables and the
 *    like). I fixed that.
 *
 *  - In information_transfer(), I think a #ifdef was wrong. Looking at the
 *    possible DMA transfer size should also happen for REAL_DMA. I added this
 *    in the #if statement.
 *
 *  - When using real DMA, information_transfer() should return in a DATAOUT
 *    phase after starting the DMA. It has nothing more to do.
 *
 *  - The interrupt service routine should run main after end of DMA, too (not
 *    only after RESELECTION interrupts). Additionally, it should _not_ test
 *    for more interrupts after running main, since a DMA process may have
 *    been started and interrupts are turned on now. The new int could happen
 *    inside the execution of NCR5380_intr(), leading to recursive
 *    calls.
 *
 *  - I've added a function merge_contiguous_buffers() that tries to
 *    merge scatter-gather buffers that are located at contiguous
 *    physical addresses and can be processed with the same DMA setup.
 *    Since most scatter-gather operations work on a page (4K) of
 *    4 buffers (1K), in more than 90% of all cases three interrupts and
 *    DMA setup actions are saved.
 *
 * - I've deleted all the stuff for AUTOPROBE_IRQ, REAL_DMA_POLL, PSEUDO_DMA
 *    and USLEEP, because these were messing up readability and will never be
 *    needed for Atari SCSI.
 * 
 * - I've revised the NCR5380_main() calling scheme (relax the 'main_running'
 *   stuff), and 'main' is executed in a bottom half if awoken by an
 *   interrupt.
 *
 * - The code was quite cluttered up by "#if (NDEBUG & NDEBUG_*) printk..."
 *   constructs. In my eyes, this made the source rather unreadable, so I
 *   finally replaced that by the *_PRINTK() macros.
 *
 */

/*
 * Further development / testing that should be done : 
 * 1.  Test linked command handling code after Eric is ready with 
 *     the high level code.
 */
#include <scsi/scsi_dbg.h>
#include <scsi/scsi_transport_spi.h>

#if (NDEBUG & NDEBUG_LISTS)
#define LIST(x,y) \
  { printk("LINE:%d   Adding %p to %p\n", __LINE__, (void*)(x), (void*)(y)); \
    if ((x)==(y)) udelay(5); }
#define REMOVE(w,x,y,z) \
  { printk("LINE:%d   Removing: %p->%p  %p->%p \n", __LINE__, \
	   (void*)(w), (void*)(x), (void*)(y), (void*)(z)); \
    if ((x)==(y)) udelay(5); }
#else
#define LIST(x,y)
#define REMOVE(w,x,y,z)
#endif

#ifndef notyet
#undef LINKED
#endif

/*
 * Design
 * Issues :
 *
 * The other Linux SCSI drivers were written when Linux was Intel PC-only,
 * and specifically for each board rather than each chip.  This makes their
 * adaptation to platforms like the Mac (Some of which use NCR5380's)
 * more difficult than it has to be.
 *
 * Also, many of the SCSI drivers were written before the command queuing
 * routines were implemented, meaning their implementations of queued 
 * commands were hacked on rather than designed in from the start.
 *
 * When I designed the Linux SCSI drivers I figured that 
 * while having two different SCSI boards in a system might be useful
 * for debugging things, two of the same type wouldn't be used.
 * Well, I was wrong and a number of users have mailed me about running
 * multiple high-performance SCSI boards in a server.
 *
 * Finally, when I get questions from users, I have no idea what 
 * revision of my driver they are running.
 *
 * This driver attempts to address these problems :
 * This is a generic 5380 driver.  To use it on a different platform, 
 * one simply writes appropriate system specific macros (ie, data
 * transfer - some PC's will use the I/O bus, 68K's must use 
 * memory mapped) and drops this file in their 'C' wrapper.
 *
 * As far as command queueing, two queues are maintained for 
 * each 5380 in the system - commands that haven't been issued yet,
 * and commands that are currently executing.  This means that an 
 * unlimited number of commands may be queued, letting 
 * more commands propagate from the higher driver levels giving higher 
 * throughput.  Note that both I_T_L and I_T_L_Q nexuses are supported, 
 * allowing multiple commands to propagate all the way to a SCSI-II device 
 * while a command is already executing.
 *
 * To solve the multiple-boards-in-the-same-system problem, 
 * there is a separate instance structure for each instance
 * of a 5380 in the system.  So, multiple NCR5380 drivers will
 * be able to coexist with appropriate changes to the high level
 * SCSI code.  
 *
 * A NCR5380_PUBLIC_REVISION macro is provided, with the release
 * number (updated for each public release) printed by the 
 * NCR5380_print_options command, which should be called from the 
 * wrapper detect function, so that I know what release of the driver
 * users are using.
 *
 * Issues specific to the NCR5380 : 
 *
 * When used in a PIO or pseudo-dma mode, the NCR5380 is a braindead 
 * piece of hardware that requires you to sit in a loop polling for 
 * the REQ signal as long as you are connected.  Some devices are 
 * brain dead (ie, many TEXEL CD ROM drives) and won't disconnect 
 * while doing long seek operations.
 * 
 * The workaround for this is to keep track of devices that have
 * disconnected.  If the device hasn't disconnected, for commands that
 * should disconnect, we do something like 
 *
 * while (!REQ is asserted) { sleep for N usecs; poll for M usecs }
 * 
 * Some tweaking of N and M needs to be done.  An algorithm based 
 * on "time to data" would give the best results as long as short time
 * to datas (ie, on the same track) were considered, however these 
 * broken devices are the exception rather than the rule and I'd rather
 * spend my time optimizing for the normal case.
 *
 * Architecture :
 *
 * At the heart of the design is a coroutine, NCR5380_main,
 * which is started when not running by the interrupt handler,
 * timer, and queue command function.  It attempts to establish
 * I_T_L or I_T_L_Q nexuses by removing the commands from the 
 * issue queue and calling NCR5380_select() if a nexus 
 * is not established. 
 *
 * Once a nexus is established, the NCR5380_information_transfer()
 * phase goes through the various phases as instructed by the target.
 * if the target goes into MSG IN and sends a DISCONNECT message,
 * the command structure is placed into the per instance disconnected
 * queue, and NCR5380_main tries to find more work.  If USLEEP
 * was defined, and the target is idle for too long, the system
 * will try to sleep.
 *
 * If a command has disconnected, eventually an interrupt will trigger,
 * calling NCR5380_intr()  which will in turn call NCR5380_reselect
 * to reestablish a nexus.  This will run main if necessary.
 *
 * On command termination, the done function will be called as 
 * appropriate.
 *
 * SCSI pointers are maintained in the SCp field of SCSI command 
 * structures, being initialized after the command is connected
 * in NCR5380_select, and set as appropriate in NCR5380_information_transfer.
 * Note that in violation of the standard, an implicit SAVE POINTERS operation
 * is done, since some BROKEN disks fail to issue an explicit SAVE POINTERS.
 */

/*
 * Using this file :
 * This file a skeleton Linux SCSI driver for the NCR 5380 series
 * of chips.  To use it, you write an architecture specific functions 
 * and macros and include this file in your driver.