blob: 61c0531e044a75f68f4eea49f846d78d42f55194 [file] [log] [blame]
Linus Torvalds1da177e2005-04-16 15:20:36 -07001The tmscsim driver
2==================
3
41. Purpose and history
52. Installation
63. Features
74. Configuration via /proc/scsi/tmscsim/?
85. Configuration via boot/module params
96. Potential improvements
107. Bug reports, debugging and updates
118. Acknowledgements
129. Copyright
13
14
151. Purpose and history
16----------------------
17The tmscsim driver supports PCI SCSI Host Adapters based on the AM53C974
18chip. AM53C974 based SCSI adapters include:
19 Tekram DC390, DC390T
20 Dawicontrol 2974
21 QLogic Fast! PCI Basic
22 some on-board adapters
23(This is most probably not a complete list)
24
25It has originally written by C.L. Huang from the Tekram corp. to support the
26Tekram DC390(T) adapter. This is where the name comes from: tm = Tekram
27scsi = SCSI driver, m = AMD (?) as opposed to w for the DC390W/U/F
28(NCR53c8X5, X=2/7) driver. Yes, there was also a driver for the latter,
29tmscsiw, which supported DC390W/U/F adapters. It's not maintained any more,
Matt LaPlante3f6dee92006-10-03 22:45:33 +020030as the ncr53c8xx is perfectly supporting these adapters since some time.
Linus Torvalds1da177e2005-04-16 15:20:36 -070031
32The driver first appeared in April 1996, exclusively supported the DC390
33and has been enhanced since then in various steps. In May 1998 support for
34general AM53C974 based adapters and some possibilities to configure it were
35added. The non-DC390 support works by assuming some values for the data
36normally taken from the DC390 EEPROM. See below (chapter 5) for details.
37
38When using the DC390, the configuration is still be done using the DC390
39BIOS setup. The DC390 EEPROM is read and used by the driver, any boot or
40module parameters (chapter 5) are ignored! However, you can change settings
41dynamically, as described in chapter 4.
42
43For a more detailed description of the driver's history, see the first lines
44of tmscsim.c.
45The numbering scheme isn't consistent. The first versions went from 1.00 to
461.12, then 1.20a to 1.20t. Finally I decided to use the ncr53c8xx scheme. So
47the next revisions will be 2.0a to 2.0X (stable), 2.1a to 2.1X (experimental),
482.2a to 2.2X (stable, again) etc. (X = anything between a and z.) If I send
49fixes to people for testing, I create intermediate versions with a digit
50appended, e.g. 2.0c3.
51
52
532. Installation
54---------------
55If you got any recent kernel with this driver and document included in
56linux/drivers/scsi, you basically have to do nothing special to use this
57driver. Of course you have to choose to compile SCSI support and DC390(T)
58support into your kernel or as module when configuring your kernel for
59compiling.
60NEW: You may as well compile this module outside your kernel, using the
61supplied Makefile.
62
63 If you got an old kernel (pre 2.1.127, pre 2.0.37p1) with an old version of
64 this driver: Get dc390-21125-20b.diff.gz or dc390-2036p21-20b1.diff.gz from
65 my web page and apply the patch. Apply further patches to upgrade to the
66 latest version of the driver.
67
68 If you want to do it manually, you should copy the files (dc390.h,
69 tmscsim.h, tmscsim.c, scsiiom.c and README.tmscsim) from this directory to
70 linux/drivers/scsi. You have to recompile your kernel/module of course.
71
72 You should apply the three patches included in dc390-120-kernel.diff
73 (Applying them: cd /usr/src; patch -p0 <~/dc390-120-kernel.diff)
74 The patches are against 2.1.125, so you might have to manually resolve
75 rejections when applying to another kernel version.
76
77 The patches will update the kernel startup code to allow boot parameters to
78 be passed to the driver, update the Documentation and finally offer you the
79 possibility to omit the non-DC390 parts of the driver.
80 (By selecting "Omit support for non DC390" you basically disable the
81 emulation of a DC390 EEPROM for non DC390 adapters. This saves a few bytes
82 of memory.)
83
84If you got a very old kernel without the tmscsim driver (pre 2.0.31)
85I recommend upgrading your kernel. However, if you don't want to, please
86contact me to get the appropriate patches.
87
88
89Upgrading a SCSI driver is always a delicate thing to do. The 2.0 driver has
90proven stable on many systems, but it's still a good idea to take some
91precautions. In an ideal world you would have a full backup of your disks.
92The world isn't ideal and most people don't have full backups (me neither).
93So take at least the following measures:
94* make your kernel remount the FS read-only on detecting an error:
95 tune2fs -e remount-ro /dev/sd??
96* have copies of your SCSI disk's partition tables on some safe location:
97 dd if=/dev/sda of=/mnt/floppy/sda bs=512 count=1
98 or just print it with:
99 fdisk -l | lpr
100* make sure you are able to boot Linux (e.g. from floppy disk using InitRD)
101 if your SCSI disk gets corrupted. You can use
102 ftp://student.physik.uni-dortmund.de/pub/linux/kernel/bootdisk.gz
103
104One more warning: I used to overclock my PCI bus to 41.67 MHz. My Tekram
105DC390F (Sym53c875) accepted this as well as my Millenium. But the Am53C974
106produced errors and started to corrupt my disks. So don't do that! A 37.50
107MHz PCI bus works for me, though, but I don't recommend using higher clocks
108than the 33.33 MHz being in the PCI spec.
109
110If you want to share the IRQ with another device and the driver refuses to
111do so, you might succeed with changing the DC390_IRQ type in tmscsim.c to
Thomas Gleixner6ce6c7f2006-07-01 19:29:47 -0700112IRQF_SHARED | IRQF_DISABLED.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700113
114
1153.Features
116----------
117- SCSI
118 * Tagged command queueing
119 * Sync speed up to 10 MHz
120 * Disconnection
121 * Multiple LUNs
122
123- General / Linux interface
124 * Support for up to 4 AM53C974 adapters.
125 * DC390 EEPROM usage or boot/module params
126 * Information via cat /proc/scsi/tmscsim/?
127 * Dynamically configurable by writing to /proc/scsi/tmscsim/?
128 * Dynamic allocation of resources
129 * SMP support: Locking on io_request lock (Linux 2.1/2.2) or adapter
130 specific locks (Linux 2.5?)
131 * Uniform source code for Linux-2.x.y
132 * Support for dyn. addition/removal of devices via add/remove-single-device
133 (Try: echo "scsi add-single-device C B T U" >/proc/scsi/scsi
134 C = Controller, B = Bus, T = Target SCSI ID, U = Unit SCSI LUN.)
135 Use with care!
136 * Try to use the partition table for the determination of the mapping
137
138
1394. Configuration via /proc/scsi/tmscsim/?
140-----------------------------------------
141First of all look at the output of /proc/scsi/tmscsim/? by typing
142 cat /proc/scsi/tmscsim/?
143The "?" should be replaced by the SCSI host number. (The shell might do this
144for you.)
145You will see some info regarding the adapter and, at the end, a listing of
146the attached devices and their settings.
147
148Here's an example:
149garloff@kurt:/home/garloff > cat /proc/scsi/tmscsim/0
150Tekram DC390/AM53C974 PCI SCSI Host Adapter, Driver Version 2.0e7 2000-11-28
151SCSI Host Nr 1, AM53C974 Adapter Nr 0
152IOPortBase 0xb000, IRQ 10
153MaxID 8, MaxLUN 8, AdapterID 6, SelTimeout 250 ms, DelayReset 1 s
154TagMaxNum 16, Status 0x00, ACBFlag 0x00, GlitchEater 24 ns
155Statistics: Cmnds 1470165, Cmnds not sent directly 0, Out of SRB conds 0
156 Lost arbitrations 587, Sel. connected 0, Connected: No
157Nr of attached devices: 4, Nr of DCBs: 4
158Map of attached LUNs: 01 00 00 03 01 00 00 00
159Idx ID LUN Prty Sync DsCn SndS TagQ NegoPeriod SyncSpeed SyncOffs MaxCmd
16000 00 00 Yes Yes Yes Yes Yes 100 ns 10.0 M 15 16
16101 03 00 Yes Yes Yes Yes No 100 ns 10.0 M 15 01
16202 03 01 Yes Yes Yes Yes No 100 ns 10.0 M 15 01
16303 04 00 Yes Yes Yes Yes No 100 ns 10.0 M 15 01
164
165Note that the settings MaxID and MaxLUN are not zero- but one-based, which
166means that a setting MaxLUN=4, will result in the support of LUNs 0..3. This
167is somehow inconvenient, but the way the mid-level SCSI code expects it to be.
168
169ACB and DCB are acronyms for Adapter Control Block and Device Control Block.
170These are data structures of the driver containing information about the
171adapter and the connected SCSI devices respectively.
172
173Idx is the device index (just a consecutive number for the driver), ID and
174LUN are the SCSI ID and LUN, Prty means Parity checking, Sync synchronous
175negotiation, DsCn Disconnection, SndS Send Start command on startup (not
176used by the driver) and TagQ Tagged Command Queueing. NegoPeriod and
177SyncSpeed are somehow redundant, because they are reciprocal values
178(1 / 112 ns = 8.9 MHz). At least in theory. The driver is able to adjust the
179NegoPeriod more accurate (4ns) than the SyncSpeed (1 / 25ns). I don't know
180if certain devices will have problems with this discrepancy. Max. speed is
18110 MHz corresp. to a min. NegoPeriod of 100 ns.
182(The driver allows slightly higher speeds if the devices (Ultra SCSI) accept
183it, but that's out of adapter spec, on your own risk and unlikely to improve
184performance. You're likely to crash your disks.)
185SyncOffs is the offset used for synchronous negotiations; max. is 15.
186The last values are only shown, if Sync is enabled. (NegoPeriod is still
187displayed in brackets to show the values which will be used after enabling
188Sync.)
189MaxCmd ist the number of commands (=tags) which can be processed at the same
190time by the device.
191
192If you want to change a setting, you can do that by writing to
193/proc/scsi/tmscsim/?. Basically you have to imitate the output of driver.
194(Don't use the brackets for NegoPeriod on Sync disabled devices.)
195You don't have to care about capitalisation. The driver will accept space,
196tab, comma, = and : as separators.
197
198There are three kinds of changes:
199
200(1) Change driver settings:
201 You type the names of the parameters and the params following it.
202 Example:
203 echo "MaxLUN=8 seltimeout 200" >/proc/scsi/tmscsim/0
204
205 Note that you can only change MaxID, MaxLUN, AdapterID, SelTimeOut,
206 TagMaxNum, ACBFlag, GlitchEater and DelayReset. Don't change ACBFlag
207 unless you want to see what happens, if the driver hangs.
208
209(2) Change device settings: You write a config line to the driver. The Nr
210 must match the ID and LUN given. If you give "-" as parameter, it is
211 ignored and the corresponding setting won't be changed.
212 You can use "y" or "n" instead of "Yes" and "No" if you want to.
213 You don't need to specify a full line. The driver automatically performs
214 an INQUIRY on the device if necessary to check if it is capable to operate
215 with the given settings (Sync, TagQ).
216 Examples:
217 echo "0 0 0 y y y - y - 10 " >/proc/scsi/tmscsim/0
218 echo "3 5 0 y n y " >/proc/scsi/tmscsim/0
219
220 To give a short explanation of the first example:
221 The first three numbers, "0 0 0" (Device index 0, SCSI ID 0, SCSI LUN 0),
222 select the device to which the following parameters apply. Note that it
223 would be sufficient to use the index or both SCSI ID and LUN, but I chose
224 to require all three to have a syntax similar to the output.
225 The following "y y y - y" enables Parity checking, enables Synchronous
226 transfers, Disconnection, leaves Send Start (not used) untouched and
227 enables Tagged Command Queueing for the selected device. The "-" skips
228 the Negotiation Period setting but the "10" sets the max sync. speed to
229 10 MHz. It's useless to specify both NegoPeriod and SyncSpeed as
230 discussed above. The values used in this example will result in maximum
231 performance.
232
233(3) Special commands: You can force a SCSI bus reset, an INQUIRY command, the
234 removal or the addition of a device's DCB and a SCSI register dump.
235 This is only used for debugging when you meet problems. The parameter of
236 the INQUIRY and REMOVE commands is the device index as shown by the
237 output of /proc/scsi/tmscsim/? in the device listing in the first column
238 (Idx). ADD takes the SCSI ID and LUN.
239 Examples:
240 echo "reset" >/proc/scsi/tmscsim/0
241 echo "inquiry 1" >/proc/scsi/tmscsim/0
242 echo "remove 2" >/proc/scsi/tmscsim/1
243 echo "add 2 3" >/proc/scsi/tmscsim/?
244 echo "dump" >/proc/scsi/tmscsim/0
245
246 Note that you will meet problems when you REMOVE a device's DCB with the
247 remove command if it contains partitions which are mounted. Only use it
248 after unmounting its partitions, telling the SCSI mid-level code to
249 remove it (scsi remove-single-device) and you really need a few bytes of
250 memory.
251 The ADD command allows you to configure a device before you tell the
252 mid-level code to try detection.
253
254
255I'd suggest reviewing the output of /proc/scsi/tmscsim/? after changing
256settings to see if everything changed as requested.
257
258
2595. Configuration via boot/module parameters
260-------------------------------------------
261With the DC390, the driver reads its EEPROM settings and tries to use them.
262But you may want to override the settings prior to being able to change the
263driver configuration via /proc/scsi/tmscsim/?.
264If you do have another AM53C974 based adapter, that's even the only
265possibility to adjust settings before you are able to write to the
266/proc/scsi/tmscsim/? pseudo-file, e.g. if you want to use another
267adapter ID than 7.
268(BTW, the log message "DC390: No EEPROM found!" is normal without a DC390.)
269For this purpose, you can pass options to the driver before it is initialised
270by using kernel or module parameters. See lilo(8) or modprobe(1) manual
271pages on how to pass params to the kernel or a module.
272[NOTE: Formerly, it was not possible to override the EEPROM supplied
273 settings of the DC390 with cmd line parameters. This has changed since
274 2.0e7]
275
276The syntax of the params is much shorter than the syntax of the /proc/...
277interface. This makes it a little bit more difficult to use. However, long
278parameter lines have the risk to be misinterpreted and the length of kernel
279parameters is limited.
280
281As the support for non-DC390 adapters works by simulating the values of the
282DC390 EEPROM, the settings are given in a DC390 BIOS' way.
283
284Here's the syntax:
285tmscsim=AdaptID,SpdIdx,DevMode,AdaptMode,TaggedCmnds,DelayReset
286
287Each of the parameters is a number, containing the described information:
288
289* AdaptID: The SCSI ID of the host adapter. Must be in the range 0..7
290 Default is 7.
291
292* SpdIdx: The index of the maximum speed as in the DC390 BIOS. The values
293 0..7 mean 10, 8.0, 6.7, 5.7, 5.0, 4.0, 3.1 and 2 MHz resp. Default is
294 0 (10.0 MHz).
295
296* DevMode is a bit mapped value describing the per-device features. It
297 applies to all devices. (Sync, Disc and TagQ will only apply, if the
298 device supports it.) The meaning of the bits (* = default):
299
300 Bit Val(hex) Val(dec) Meaning
301 *0 0x01 1 Parity check
302 *1 0x02 2 Synchronous Negotiation
303 *2 0x04 4 Disconnection
304 *3 0x08 8 Send Start command on startup. (Not used)
305 *4 0x10 16 Tagged Command Queueing
306
307 As usual, the desired value is obtained by adding the wanted values. If
308 you want to enable all values, e.g., you would use 31(0x1f). Default is 31.
309
310* AdaptMode is a bit mapped value describing the enabled adapter features.
311
312 Bit Val(hex) Val(dec) Meaning
313 *0 0x01 1 Support more than two drives. (Not used)
314 *1 0x02 2 Use DOS compatible mapping for HDs greater than 1GB.
315 *2 0x04 4 Reset SCSI Bus on startup.
316 *3 0x08 8 Active Negation: Improves SCSI Bus noise immunity.
317 4 0x10 16 Immediate return on BIOS seek command. (Not used)
318 (*)5 0x20 32 Check for LUNs >= 1.
319
320 The default for LUN Check depends on CONFIG_SCSI_MULTI_LUN.
321
322* TaggedCmnds is a number indicating the maximum number of Tagged Commands.
323 It is the binary logarithm - 1 of the actual number. Max is 4 (32).
324 Value Number of Tagged Commands
325 0 2
326 1 4
327 2 8
328 *3 16
329 4 32
330
331* DelayReset is the time in seconds (minus 0.5s), the adapter waits, after a
332 bus reset. Default is 1 (corresp. to 1.5s).
333
334Example:
335 modprobe tmscsim tmscsim=6,2,31
336would set the adapter ID to 6, max. speed to 6.7 MHz, enable all device
337features and leave the adapter features, the number of Tagged Commands
338and the Delay after a reset to the defaults.
339
340As you can see, you don't need to specify all of the six params.
341If you want values to be ignored (i.e. the EEprom settings or the defaults
342will be used), you may pass -2 (not 0!) at the corresponding position.
343
344The defaults (7,0,31,15,3,1) are aggressive to allow good performance. You
345can use tmscsim=7,0,31,63,4,0 for maximum performance, if your SCSI chain
346allows it. If you meet problems, you can use tmscsim=-1 which is a shortcut
347for tmscsim=7,4,9,15,2,10.
348
349
3506. Potential improvements
351-------------------------
352Most of the intended work on the driver has been done. Here are a few ideas
353to further improve its usability:
354
355* Cleanly separate per-Target and per-LUN properties (DCB)
356* More intelligent abort() routine
357* Use new_eh code (Linux-2.1+)
358* Have the mid-level (ML) code (and not the driver) handle more of the
359 various conditions.
360* Command queueing in the driver: Eliminate Query list and use ML instead.
361* More user friendly boot/module param syntax
362
363Further investigation on these problems:
364
365* Driver hangs with sync readcdda (xcdroast) (most probably VIA PCI error)
366
367Known problems:
368Please see http://www.garloff.de/kurt/linux/dc390/problems.html
369
370* Changing the parameters of multi-lun by the tmscsim/? interface will
371 cause problems, cause these settings are mostly per Target and not per LUN
372 and should be updated accordingly. To be fixed for 2.0d24.
373* CDRs (eg Yam CRW4416) not recognized, because some buggy devices don't
374 recover from a SCSI reset in time. Use a higher delay or don't issue
375 a SCSI bus reset on driver initialization. See problems page.
376 For the CRW4416S, this seems to be solved with firmware 1.0g (reported by
377 Jean-Yves Barbier).
378* TEAC CD-532S not being recognized. (Works with 1.11).
379* Scanners (eg. Astra UMAX 1220S) don't work: Disable Sync Negotiation.
380 If this does not help, try echo "INQUIRY t" >/proc/scsi/tmscsim/? (t
381 replaced by the dev index of your scanner). You may try to reset your SCSI
382 bus afterwards (echo "RESET" >/proc/scsi/tmscsim/?).
383 The problem seems to be solved as of 2.0d18, thanks to Andreas Rick.
Matt LaPlantefff92892006-10-03 22:47:42 +0200384* If there is a valid partition table, the driver will use it for determining
Linus Torvalds1da177e2005-04-16 15:20:36 -0700385 the mapping. If there's none, a reasonable mapping (Symbios-like) will be
386 assumed. Other operating systems may not like this mapping, though
387 it's consistent with the BIOS' behaviour. Old DC390 drivers ignored the
388 partition table and used a H/S = 64/32 or 255/63 translation. So if you
389 want to be compatible to those, use this old mapping when creating
390 partition tables. Even worse, on bootup the DC390 might complain if other
391 mappings are found, so auto rebooting may fail.
392* In some situations, the driver will get stuck in an abort loop. This is a
393 bad interaction between the Mid-Layer of Linux' SCSI code and the driver.
394 Try to disable DsCn, if you meet this problem. Please contact me for
395 further debugging.
396
397
3987. Bug reports, debugging and updates
399-------------------------------------
400Whenever you have problems with the driver, you are invited to ask the
401author for help. However, I'd suggest reading the docs and trying to solve
402the problem yourself, first.
403If you find something, which you believe to be a bug, please report it to me.
404Please append the output of /proc/scsi/scsi, /proc/scsi/tmscsim/? and
405maybe the DC390 log messages to the report.
406
407Bug reports should be send to me (Kurt Garloff <dc390@garloff.de>) as well
408as to the linux-scsi list (<linux-scsi@vger.kernel.org>), as sometimes bugs
409are caused by the SCSI mid-level code.
410
411I will ask you for some more details and probably I will also ask you to
412enable some of the DEBUG options in the driver (tmscsim.c:DC390_DEBUGXXX
413defines). The driver will produce some data for the syslog facility then.
414Beware: If your syslog gets written to a SCSI disk connected to your
415AM53C974, the logging might produce log output again, and you might end
416having your box spending most of its time doing the logging.
417
418The latest version of the driver can be found at:
419 http://www.garloff.de/kurt/linux/dc390/
420 ftp://ftp.suse.com/pub/people/garloff/linux/dc390/
421
422
4238. Acknowledgements
424-------------------
425Thanks to Linus Torvalds, Alan Cox, the FSF people, the XFree86 team and
426all the others for the wonderful OS and software.
427Thanks to C.L. Huang and Philip Giang (Tekram) for the initial driver
428release and support.
John Anthony Kazos Jrbe2a6082007-05-09 08:50:42 +0200429Thanks to Doug Ledford, GĂ©rard Roudier for support with SCSI coding.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700430Thanks to a lot of people (espec. Chiaki Ishikawa, Andreas Haumer, Hubert
431Tonneau) for intensively testing the driver (and even risking data loss
432doing this during early revisions).
433Recently, SuSE GmbH, Nuernberg, FRG, has been paying me for the driver
434development and maintenance. Special thanks!
435
436
4379. Copyright
438------------
439 This driver is free software; you can redistribute it and/or modify
440 it under the terms of the GNU General Public License as published by
441 the Free Software Foundation; version 2 of the License.
442 If you want to use any later version of the GNU GPL, you will probably
443 be allowed to, but you have to ask me and Tekram <erich@tekram.com.tw>
444 before.
445
446-------------------------------------------------------------------------
447Written by Kurt Garloff <kurt@garloff.de> 1998/06/11
448Last updated 2000/11/28, driver revision 2.0e7
449$Id: README.tmscsim,v 2.25.2.7 2000/12/20 01:07:12 garloff Exp $