1 Thu May 11 12:30 2000 Pam Delaney (pam.delaney@lsil.com)
2 * revision 3.3b
3
4 Mon Apr 24 12:00 2000 Gerard Roudier (groudier@club-internet.fr)
5 * revision 3.2i
6 - Return value 1 (instead of 0) from the driver setup routine.
7 - Let the driver also attach controllers that have been set to
8 OFF in the NVRAM as it did prior to revision 3.2g.
9
10 Sat Apr 1 12:00 2000 Gerard Roudier (groudier@club-internet.fr)
11 * revision 3.2h
12 - Fix a compilation problem on Alpha introduced in version 3.2g.
13 (`port' changed to `base_io').
14 - Move from `sym' to this driver a tiny change for __sparc__ that
15 applies to cache line size (? Probably from David S Miller).
16 - Make sure no data transfer will happen for Scsi_Cmnd requests
17 that supply SCSI_DATA_NONE direction (this avoids some BUG()
18 statement in the PCI code when a data buffer is also supplied).
19
20 Thu Mar 16 9:30 2000 Pam Delaney (pam.delaney@lsil.com)
21 * revision 3.3b-3
22 - Added exclusion for the 53C1010 and 53C1010_66 chips
23 to the driver (change to sym53c8xx_comm.h).
24
25 Mon March 6 23:15 2000 Gerard Roudier (groudier@club-internet.fr)
26 * revision 3.2g
27 - Add the file sym53c8xx_comm.h that collects code that should
28 be shared by sym53c8xx and ncr53c8xx drivers. For now, it is
29 a header file that is only included by the ncr53c8xx driver,
30 but things will be cleaned up later. This code addresses
31 notably:
32 * Chip detection and PCI related initialisations
33 * NVRAM detection and reading
34 * DMA mapping
35 * Boot setup command
36 * And some other ...
37 - Add support for the new dynamic dma mapping kernel interface.
38 Requires Linux-2.3.47 (tested with pre-2.3.47-6).
39 - Get data transfer direction from the scsi command structure
40 (Scsi_Cmnd) when this information is available.
41
42 Mon March 6 23:15 2000 Gerard Roudier (groudier@club-internet.fr)
43 * revision 3.2g
44 - Add the file sym53c8xx_comm.h that collects code that should
45 be shared by sym53c8xx and ncr53c8xx drivers. For now, it is
46 a header file that is only included by the ncr53c8xx driver,
47 but things will be cleaned up later. This code addresses
48 notably:
49 * Chip detection and PCI related initialisations
50 * NVRAM detection and reading
51 * DMA mapping
52 * Boot setup command
53 * And some other ...
54 - Add support for the new dynamic dma mapping kernel interface.
55 Requires Linux-2.3.47 (tested with pre-2.3.47-6).
56 - Get data transfer direction from the scsi command structure
57 (Scsi_Cmnd) when this information is available.
58
59 Fri Jan 14 14:00 2000 Pam Delaney (pam.delaney@lsil.com)
60 * revision pre-3.3b-1
61 - Merge parallel driver series 3.31 and 3.2e
62
63 Tue Jan 11 14:00 2000 Pam Delaney (pam.delaney@lsil.com)
64 * revision 3.31
65 - Added support for mounting disks on wide-narrow-wide
66 scsi configurations.
67 - Built off of version 3.30
68
69 Mon Jan 10 13:30 2000 Pam Delaney (pam.delaney@lsil.com)
70 * revision 3.30
71 - Added capability to use the integrity checking code
72 in the kernel (optional).
73 - Disabled support for the 53C1010.
74 - Built off of version 3.2c
75
76 Sat Jan 8 22:00 2000 Gerard Roudier (groudier@club-internet.fr)
77 * revision 3.2e
78 - Add year 2000 copyright.
79 - Display correctly bus signals when bus is detected wrong.
80 - Remove the dead code that broke driver 3.2d.
81
82 Mon Dec 6 22:00 1999 Gerard Roudier (groudier@club-internet.fr)
83 * revision 3.2d
84 - Change messages written by the driver at initialisation and
85 through the /proc FS (rather cosmetic changes that consist in
86 printing out the PCI bus number and device/function).
87 - Get rid of the old PCI bios interface, but preserve kernel 2.0
88 compatibility from a simple wrapper.
89 - Remove the compilation condition about having to acquire the
90 io_request_lock since it seems to be a definite feature now.:)
91 - proc_dir structure no longer needed for kernel >= 2.3.27.
92 - Change the driver detection code by the sym53c8xx one, modulo
93 some minor changes. The driver can now attach any number of
94 controllers (>40) and does no longer hoger stack space at
95 initialisation.
96 - Definitely disable overlapped PCI arbitration for all dual
97 function chips, since I cannot make sure for what chip revisions
98 it is actually safe.
99 - Add support for the SYM53C1510D.
100 - Update the poor Tekram sync factor table.
101 - Remove the compilation condition about having to acquire the
102 io_request_lock since it seems to be a definite feature now.:)
103 - proc_dir structure no longer needed for kernel >= 2.3.27.
104
105 Sat Sep 11 18:00 1999 Gerard Roudier (groudier@club-internet.fr)
106 * revision 3.2c
107 - Handle correctly (hopefully) jiffies wrap-around.
108 - Restore the entry used to detect 875 until revision 0xff.
109 (I removed it inadvertently, it seems :) )
110 - Replace __initfunc() which is deprecated stuff by __init which
111 is not yet so. ;-)
112 - Add support of some 'resource handling' for linux-2.3.13.
113 Basically the BARs have been changed to something more complex
114 in the pci_dev structure.
115 - Remove some deprecated code.
116
117 Sat May 10 11:00 1999 Gerard Roudier (groudier@club-internet.fr)
118 * revision pre-3.2b-1
119 - Support for the 53C895A by Pamela Delaney <pam.delaney@lsil.com>
120 The 53C895A contains all of the features of the 896 but has only
121 one channel and has a 32 bit PCI bus. It does 64 bit PCI addressing
122 using dual cycle PCI data transfers.
123 - Miscellaneous minor fixes.
124 - Some additions to the README.ncr53c8xx file.
125
126 Sun Apr 11 10:00 1999 Gerard Roudier (groudier@club-internet.fr)
127 * revision 3.2a
128 - Add 'hostid:#id' boot option. This option allows to change the
129 default SCSI id the driver uses for controllers.
130 - Remove nvram layouts and driver set-up structures from the C source,
131 and use the one defined in sym53c8xx_defs.h file.
132 (shared by both drivers).
133 - Set for now MAX LUNS to 16 (instead of 8).
134
135 Thu Mar 11 23:00 1999 Gerard Roudier (groudier@club-internet.fr)
136 * revision 3.2 (8xx-896 driver bundle)
137 - Only define the host template in ncr53c8xx.h and include the
138 sym53c8xx_defs.h file.
139 - Declare static all symbols that do not need to be visible from
140 outside the driver code.
141 - Add 'excl' boot command option that allows to pass to the driver
142 io address of devices not to attach.
143 - Add info() function called from the host template to print
144 driver/host information.
145 - Minor documentation additions.
146
147 Sat Mar 6 11:00 1999 Gerard Roudier (groudier@club-internet.fr)
148 * revision 3.1h
149 - Fix some oooold bug that hangs the bus if a device rejects a
150 negotiation. Btw, the corresponding stuff also needed some cleanup
151 and thus the change is a bit larger than it could have been.
152 - Still some typo that made compilation fail for 64 bit (trivial fix).
153
154 Sun Feb 14:00 1999 Gerard Roudier (groudier@club-internet.fr)
155 * revision 3.1g
156 - Deal correctly with 64 bit PCI address registers on Linux 2.2.
157 Pointed out by Leonard Zubkoff.
158 - Allow to tune request_irq() flags from the boot command line using
159 ncr53c8xx=irqm:??, as follows:
160 a) If bit 0x10 is set in irqm, SA_SHIRQ flag is not used.
161 b) If bit 0x20 is set in irqm, SA_INTERRUPT flag is not used.
162 By default the driver uses both SA_SHIRQ and SA_INTERRUPT.
163 Option 'ncr53c8xx=irqm:0x20' may be used when an IRQ is shared by
164 a 53C8XX adapter and a network board.
165 - Tiny mispelling fixed (ABORT instead of ABRT). Was fortunately
166 harmless.
167 - Negotiate SYNC data transfers with CCS devices.
168
169 Sat Jan 16 17:30 1999 Gerard Roudier (groudier@club-internet.fr)
170 * revision 3.1f
171 - Some PCI fix-ups not needed any more for PPC (from Cort).
172 - Cache line size set to 16 DWORDS for Sparc (from DSM).
173 - Waiting list look-up didn't work for the first command of the list.
174 - Remove 2 useless lines of code.
175
176 Sun Dec 13 18:00 1998 Gerard Roudier (groudier@club-internet.fr)
177 * revision 3.1e
178 - Same work-around as for the 53c876 rev <= 0x15 for 53c896 rev 1:
179 Disable overlapped arbitration. This will not make difference
180 since the chip has on-chip RAM.
181
182 Thu Nov 26 22:00 1998 Gerard Roudier (groudier@club-internet.fr)
183 * revision 3.1d
184 - The SISL RAID change requires now remap_pci_mem() stuff to be
185 compiled for __i386__ when normal IOs are used.
186 - Minor spelling fixes in doc files.
187
188 Sat Nov 21 18:00 1998 Gerard Roudier (groudier@club-internet.fr)
189 * revision 3.1c
190 - Ignore chips that are driven by SISL RAID (DAC 960).
191 Change sent by Leonard Zubkoff and slightly reworked.
192 - Still a buglet in the tags initial settings that needed to be fixed.
193 It was not possible to disable TGQ at system startup for devices
194 that claim TGQ support. The driver used at least 2 for the queue
195 depth but did'nt keep track of user settings for tags depth lower
196 than 2.
197
198 Wed Nov 11 10:00 1998 Gerard Roudier (groudier@club-internet.fr)
199 * revision 3.1b
200 - The driver was unhappy when configured with default_tags > MAX_TAGS
201 Hopefully doubly-fixed.
202 - Update the Configure.help driver section that speaks of TAGS.
203
204 Wed Oct 21 21:00 1998 Gerard Roudier (groudier@club-internet.fr)
205 * revision 3.1a
206 - Changes from Eddie Dost for Sparc and Alpha:
207 ioremap/iounmap support for Sparc.
208 pcivtophys changed to bus_dvma_to_phys.
209 - Add the 53c876 description to the chip table. This is only usefull
210 for printing the right name of the controller.
211 - DEL-441 Item 2 work-around for the 53c876 rev <= 5 (0x15).
212 - Add additionnal checking of INQUIRY data:
213 Check INQUIRY data received length is at least 7. Byte 7 of
214 inquiry data contains device features bits and the driver might
215 be confused by garbage. Also check peripheral qualifier.
216 - Cleanup of the SCSI tasks management:
217 Remove the special case for 32 tags. Now the driver only uses the
218 scheme that allows up to 64 tags per LUN.
219 Merge some code from the 896 driver.
220 Use a 1,3,5,...MAXTAGS*2+1 tag numbering. Previous driver could
221 use any tag number from 1 to 253 and some non conformant devices
222 might have problems with large tag numbers.
223 - 'no_sync' changed to 'no_disc' in the README file. This is an old
224 and trivial mistake that seems to demonstrate the README file is
225 not often read. :)
226
227 Sun Oct 4 14:00 1998 Gerard Roudier (groudier@club-internet.fr)
228 * revision 3.0i
229 - Cosmetic changes for sparc (but not for the driver) that needs
230 __irq_itoa() to be used for printed IRQ value to be understandable.
231 - Some problems with the driver that didn't occur using driver 2.5f
232 were due to a SCSI selection problem triggered by a clearly
233 documented feature that in fact seems not to work: (53C8XX chips
234 are claimed by the manuals to be able to execute SCSI scripts just
235 after abitration while the SCSI core is performing SCSI selection).
236 This optimization is broken and has been removed.
237 - Some broken scsi devices are confused when a negotiation is started
238 on a LUN that does not correspond to a real device. According to
239 SCSI specs, this is a device firmware bug. This has been worked
240 around by only starting negotiation if the LUN has previously be
241 used for at least 1 successful SCSI command.
242 - The 'last message sent' printed out on M_REJECT message reception
243 was read from the SFBR i/o register after the previous message had
244 been sent.
245 This was not correct and affects all previous driver versions and
246 the original FreeBSD one as well. The SCSI scripts has been fixed
247 so that it now provides the right information to the C code.
248
249 Sat Jul 18 13:00 1998 Gerard Roudier (groudier@club-internet.fr)
250 * revision 3.0g
251 - Preliminary fixes for Big Endian (sent by Eddie C. Dost).
252 Big Endian architectures should work again with the driver.
253 Eddie's patch has been partially applied since current 2.1.109
254 does not have all the Sparc changes of the vger tree.
255 - Use of BITS_PER_LONG instead of (~0UL == 0xffffffffUL) has fixed
256 the problem observed when the driver was compiled using EGCS or
257 PGCC.
258
259 Mon Jul 13 20:00 1998 Gerard Roudier (groudier@club-internet.fr)
260 * revision 3.0f
261 - Some spelling fixes.
262 - linux/config.h misplaced in ncr53c8xx.h
263 - MODULE_PARM stuff added for linux 2.1.
264 - check INQUIRY response data format is exactly 2.
265 - use BITS_PER_LONG if defined.
266
267 Sun Jun 28 12:00 1998 Gerard Roudier (groudier@club-internet.fr)
268 * revision 3.0e
269 - Some cleanup, spelling fixes, version checks, documentations
270 changes, etc ...
271
272 Sat Jun 20 20:00 1998 Gerard Roudier (groudier@club-internet.fr)
273 * revision 3.0c
274 - Add a boot setup option that allows to set up device queue depths
275 at boot-up. This option is very usefull since Linux does not
276 allow to change scsi device queue depth once the system has been
277 booted up.
278
279 Sun Jun 15 23:00 1998 Gerard Roudier (groudier@club-internet.fr)
280 * revision 3.0a
281 - Support for up to 64 TAGS per LUN.
282 - Rewrite the TARGET vs LUN capabilities management.
283 CmdQueue is now handled as a LUN capability as it shall be.
284 This also fixes a bug triggered when disabling tagged command
285 queuing for a device that had this feature enabled.
286 - Remove the ncr_opennings() stuff that was useless under Linux
287 and hard to understand to me.
288 - Add "setverbose" procfs driver command. It allows to tune
289 verbose level after boot-up. Setting this level to zero, for
290 example avoid flooding the syslog file.
291 - Add KERN_XXX to some printk's.
292
293 Tue Jun 10 23:00 1998 Gerard Roudier (groudier@club-internet.fr)
294 * revision 3.0
295 - Linux config changes for 2.0.34:
296 Remove NVRAM detection config option. This option is now enabled
297 by default but can be disabled by editing the driver header file.
298 Add a PROFILE config option.
299 - Update Configure.help
300 - Add calls to new function mdelay() for milli-seconds delay if
301 kernel version >= 2.1.105.
302 - Replace all printf(s) by printk(s). After all, the ncr53c8xx is
303 a driver for Linux.
304 - Perform auto-sense on COMMAND TERMINATED. Not sure it is usefull.
305 - Some other minor changes.
306
307 Tue Jun 4 23:00 1998 Gerard Roudier (groudier@club-internet.fr)
308 * revision 2.6n
309 - Code cleanup and simplification:
310 Remove kernel 1.2.X and 1.3.X support.
311 Remove the _old_ target capabilities table.
312 Remove the error recovery code that have'nt been really usefull.
313 Use a single alignment boundary (CACHE_LINE_SIZE) for data
314 structures.
315 - Several aggressive SCRIPTS optimizations and changes:
316 Reselect SCRIPTS code rewritten.
317 Support for selection/reselection without ATN.
318 And some others.
319 - Miscallaneous changes in the C code:
320 Count actual number of CCB queued to the controller (future use).
321 Lots of other minor changes.
322
323 Wed May 13 20:00 1998 Gerard Roudier (groudier@club-internet.fr)
324 * revision 2.6m
325 - Problem of missed SCSI bus reset with the 53C895 fixed by
326 Richard Waltham. The 53C895 needs about 650 us for the bus
327 mode to settle. Delays used while resetting the controller
328 and the bus have been adjusted. Thanks Richard!
329 - Some simplification for 64 bit arch done ccb address testing.
330 - Add a check of the MSG_OUT phase after Selection with ATN.
331 - The new tagged queue stuff seems ok, so some informationnal
332 message have been conditionned by verbose >= 3.
333 - Donnot reset if a SBMC interrupt reports the same bus mode.
334 - Print out the whole driver set-up. Some options were missing and
335 the print statement was misplaced for modules.
336 - Ignore a SCSI parity interrupt if the chip is not connected to
337 the SCSI bus.
338
339 Sat May 1 16:00 1998 Gerard Roudier (groudier@club-internet.fr)
340 * revision 2.6l
341 - Add CCB done queue support for Alpha and perhaps some other
342 architectures.
343 - Add some barriers to enforce memory ordering for x86 and
344 Alpha architectures.
345 - Fix something that looks like an old bug in the nego SIR
346 interrupt code in case of negotiation failure.
347
348 Sat Apr 25 21:00 1998 Gerard Roudier (groudier@club-internet.fr)
349 * revision 2.6k
350 - Remove all accesses to the on-chip RAM from the C code:
351 Use SCRIPTS to load the on-chip RAM.
352 Use SCRIPTS to repair the start queue on selection timeout.
353 Use the copy of script in main memory to calculate the chip
354 context on phase mismatch.
355 - The above allows now to use the on-chip RAM without requiring
356 to get access to the on-chip RAM from the C code. This makes
357 on-chip RAM useable for linux-1.2.13 and for Linux-Alpha for
358 instance.
359 - Some simplifications and cleanups in the SCRIPTS and C code.
360 - Buglet fixed in parity error recovery SCRIPTS (never tested).
361 - Minor updates in README.ncr53c8xx.
362
363 Wed Apr 15 21:00 1998 Gerard Roudier (groudier@club-internet.fr)
364 * revision 2.6j
365 - Incorporate changes from linux-2.1.95 ncr53c8xx driver version.
366 - Add SMP support for linux-2.1.95 and above.
367 - Fix a bug when QUEUE FULL is returned and no commands are
368 disconnected. This happens with Atlas I / L912 and may happen
369 with Atlas II / LXY4.
370 - Nail another one on CHECK condition when requeuing the command
371 for auto-sense.
372 - Call scsi_done() for all completed commands after interrupt
373 handling.
374 - Increase the done queue to 24 entries.
375
376 Sat Apr 4 20:00 1998 Gerard Roudier (groudier@club-internet.fr)
377 * revision 2.6i
378 - CTEST0 is used by the 53C885 for Power Management and
379 priority setting between the 2 functions.
380 Use SDID instead as actual target number. Just have had to
381 overwrite it with SSID on reselection.
382 - Split DATA_IN and DATA_OUT scripts into 2 sub-scripts.
383 64 segments are moved from on-chip RAM scripts.
384 If more segments, a script in main memory is used for the
385 additionnal segments.
386 - Since the SCRIPTS processor continues SCRIPTS execution after
387 having won arbitration, do some stuff prior to testing any SCSI
388 phase on reselection. This should have the vertue to process
389 scripts in parallel with the SCSI core performing selection.
390 - Increase the done queue to 12 entries.
391
392 Sun Mar 29 12:00 1998 Gerard Roudier (groudier@club-internet.fr)
393 * revision 2.6h
394 - Some fixes.
395
396 Tue Mar 26 23:00 1998 Gerard Roudier (groudier@club-internet.fr)
397 * revision 2.6g
398 - New done queue. 8 entries by default (6 always useable).
399 Can be increased if needed.
400 - Resources management using doubly linked queues.
401 - New auto-sense and QUEUE FULL handling that does not need to
402 stall the NCR queue any more.
403 - New CCB starvation avoiding algorithm.
404 - Prepare CCBs for SCSI commands that cannot be queued, instead of
405 inserting these commands into the waiting list. The waiting list
406 is now only used while resetting and when memory for CCBs is not
407 yet available?
408
409 Sun Feb 8 22:00 1998 Gerard Roudier (groudier@club-internet.fr)
410 * revision 2.6f
411 - Some fixes in order to really support the 53C895, at least with
412 FAST-20 devices.
413 - Heavy changes in the target/lun resources management to allow
414 the scripts to jump directly to the CCB on reselection instead
415 of walking on the lun CCBs list. Up to 32 tags per lun are now
416 supported without script processor and PCI traffic overhead.
417
418 Sun Jan 11 22:00 1998 Gerard Roudier (groudier@club-internet.fr)
419 * revision 2.6d
420 - new (different ?) implementation of the start queue:
421 Use a simple CALL to a launch script in the CCB.
422 - implement a minimal done queue (1 entry :-) ).
423 this avoid scanning all CCBs on INT FLY (Only scan all CCBs, on
424 overflow). Hit ratio is better than 99.9 % on my system, so no
425 need to have a larger done queue.
426 - generalization of the restart of CCB on special condition as
427 Abort, QUEUE FULL, CHECK CONDITION.
428 This has been called 'silly scheduler'.
429 - make all the profiling code conditionned by a config option.
430 This spare some PCI traffic and C code when this feature is not
431 needed.
432 - handle more cleanly the situation where direction is unknown.
433 The pointers patching is now performed by the SCRIPTS processor.
434 - remove some useless scripts instructions.
435
436 Ported from driver 2.5 series:
437 ------------------------------
438 - Use FAST-5 instead of SLOW for slow scsi devices according to
439 new SPI-2 draft.
440 - Make some changes in order to accomodate with 875 rev <= 3
441 device errata listing 397. Minor consequences are:
442 . Leave use of PCI Write and Invalidate under user control.
443 Now, by default the driver does not enable PCI MWI and option
444 'specf:y' is required in order to enable this feature.
445 . Memory Read Line is not enabled for 875 and 875-like chips.
446 . Programmed burst length set to 64 DWORDS (instead of 128).
447 (Note: SYMBIOS uses 32 DWORDS for the SDMS BIOS)
448 - Add 'buschk' boot option.
449 This option enables checking of SCSI BUS data lines after SCSI
450 RESET (set by default). (Submitted by Richard Waltham).
451 - Update the README file.
452 - Dispatch CONDITION MET and RESERVATION CONFLICT scsi status
453 as OK driver status.
454 - Update the README file and the Symbios NVRAM format definition
455 with removable media flags values (available with SDMS 4.09).
456 - Several PCI configuration registers fix-ups for powerpc.
457 (Patch sent by Cort).
This page was automatically generated by the
LXR engine.
Visit the LXR main site for more
information.