3 -=< The IBM Microchannel SCSI-Subsystem >=-
5 for the IBM PS/2 series
7 Low Level Software-Driver for Linux
9 Copyright (c) 1995 Strom Systems, Inc. under the terms of the GNU
10 General Public License. Originally written by Martin Kolinek, December 1995.
11 Officially maintained by Michael Lang since January 1999.
16 Last update: 29 July 2000
19 Authors of this Driver
20 ----------------------
21 - Chris Beauregard (improvement of the SCSI-device mapping by the driver)
22 - Martin Kolinek (origin, first release of this driver)
23 - Klaus Kudielka (multiple SCSI-host management/detection, adaption to
24 Linux Kernel 2.1.x, module support)
25 - Michael Lang (assigning original pun,lun mapping, dynamical ldn
26 assignment, this file, patch, official driver maintenance
27 and subsequent pains related with the driver :-))
33 2.1 IBM SCSI-Subsystem Detection
34 2.2 Physical Units, Logical Units, and Logical Devices
35 2.3 SCSI-Device Recognition and dynamical ldn Assignment
37 2.5 Regular SCSI-Command-Processing
38 2.6 Abort & Reset Commands
40 2.8 Kernel Boot Option
41 2.9 Driver Module Support
42 2.10 Multiple Hostadapter Support
43 2.11 /proc/scsi-Filesystem Information
44 2.12 /proc/mca-Filesystem Information
45 2.13 Supported IBM SCSI-Subsystems
46 2.14 Linux Kernel Versions
50 5.1 Commandline Parameters
57 7.2 Sponsors & Supporters
65 This README-file describes the IBM SCSI-subsystem low level driver for
66 Linux. The descriptions which were formerly kept in the source-code have
67 been taken out to this file to easify the codes' readability. The driver
68 description has been updated, as most of the former description was already
69 quite outdated. The history of the driver development is also kept inside
70 here. Multiple historical developments have been summarized to shorten the
71 textsize a bit. At the end of this file you can find a small manual for
72 this driver and hints to get it running on your machine.
76 2.1 IBM SCSI-Subsystem Detection
77 --------------------------------
78 This is done in the ibmmca_detect() function. It first checks, if the
79 Microchannel-bus support is enabled, as the IBM SCSI-subsystem needs the
80 Microchannel. In a next step, a free interrupt is chosen and the main
81 interrupt handler is connected to it to handle answers of the SCSI-
82 subsystem(s). If the F/W SCSI-adapter is forced by the BIOS to use IRQ11
83 instead of IRQ14, IRQ11 is used for the IBM SCSI-2 F/W adapter. In a
84 further step it is checked, if the adapter gets detected by force from
85 the kernel commandline, where the I/O port and the SCSI-subsystem id can
86 be specified. The next step checks if there is an integrated SCSI-subsystem
87 installed. This register area is fixed through all IBM PS/2 MCA-machines
88 and appears as something like a virtual slot 10 of the MCA-bus. On most
89 PS/2 machines, the POS registers of slot 10 are set to 0xff or 0x00 if not
90 integrated SCSI-controller is available. But on certain PS/2s, like model
91 9595, this slot 10 is used to store other information which at earlier
92 stage confused the driver and resulted in the detection of some ghost-SCSI.
93 If POS-register 2 and 3 are not 0x00 and not 0xff, but all other POS
94 registers are either 0xff or 0x00, there must be an integrated SCSI-
95 subsystem present and it will be registered as IBM Integrated SCSI-
96 Subsystem. The next step checks, if there is a slot-adapter installed on
97 the MCA-bus. To get this, the first two POS-registers, that represent the
98 adapter ID are checked. If they fit to one of the ids, stored in the
99 adapter list, a SCSI-subsystem is assumed to be found in a slot and will be
100 registered. This check is done through all possible MCA-bus slots to allow
101 more than one SCSI-adapter to be present in the PS/2-system and this is
102 already the first point of problems. Looking into the technical reference
103 manual for the IBM PS/2 common interfaces, the POS2 register must have
104 different interpretation of its single bits to avoid overlapping I/O
105 regions. While one can assume, that the integrated subsystem has a fix
106 I/O-address at 0x3540 - 0x3547, further installed IBM SCSI-adapters must
107 use a different I/O-address. This is expressed by bit 1 to 3 of POS2
108 (multiplied by 8 + 0x3540). Bits 2 and 3 are reserved for the integrated
109 subsystem, but not for the adapters! The following list shows, how the
110 bits of POS2 and POS3 should be interpreted.
112 The POS2-register of all PS/2 models' integrated SCSI-subsystems has the
113 following interpretation of bits:
114 Bit 7 - 4 : Chip Revision ID (Release)
116 Bit 1 : 8k NVRAM Disabled
117 Bit 0 : Chip Enable (EN-Signal)
118 The POS3-register is interpreted as follows (for most IBM SCSI-subsys.):
120 Bit 4 - 0 : Reserved = 0
121 The slot-adapters have different interpretation of these bits. The IBM SCSI
122 adapter (w/Cache) and the IBM SCSI-2 F/W adapter use the following
123 interpretation of the POS2 register:
124 Bit 7 - 4 : ROM Segment Address Select
125 Bit 3 - 1 : Adapter I/O Address Select (*8+0x3540)
126 Bit 0 : Adapter Enable (EN-Signal)
127 and for the POS3 register:
129 Bit 4 : Fairness Enable (SCSI ID3 f. F/W)
130 Bit 3 - 0 : Arbitration Level
131 The most modern product of the series is the IBM SCSI-2 F/W adapter, it
132 allows dual-bus SCSI and SCSI-wide addressing, which means, PUNs may be
133 between 0 and 15. Here, Bit 4 is the high-order bit of the 4-bit wide
134 adapter PUN expression. In short words, this means, that IBM PS/2 machines
135 can only support 1 single integrated subsystem by default. Additional
136 slot-adapters get ports assigned by the automatic configuration tool.
138 One day I found a patch in ibmmca_detect(), forcing the I/O-address to be
139 0x3540 for integrated SCSI-subsystems, there was a remark placed, that on
140 integrated IBM SCSI-subsystems of model 56, the POS2 register was showing 5.
141 This means, that really for these models, POS2 has to be interpreted
142 sticking to the technical reference guide. In this case, the bit 2 (4) is
143 a reserved bit and may not be interpreted. These differences between the
144 adapters and the integrated controllers are taken into account by the
145 detection routine of the driver on from version >3.0g.
147 Every time, a SCSI-subsystem is discovered, the ibmmca_register() function
148 is called. This function checks first, if the requested area for the I/O-
149 address of this SCSI-subsystem is still available and assigns this I/O-
150 area to the SCSI-subsystem. There are always 8 sequential I/O-addresses
151 taken for each individual SCSI-subsystem found, which are:
153 Offset Type Permissions
154 0 Command Interface Register 1 Read/Write
155 1 Command Interface Register 2 Read/Write
156 2 Command Interface Register 3 Read/Write
157 3 Command Interface Register 4 Read/Write
158 4 Attention Register Read/Write
159 5 Basic Control Register Read/Write
160 6 Interrupt Status Register Read
161 7 Basic Status Register Read
163 After the I/O-address range is assigned, the host-adapter is assigned
164 to a local structure which keeps all adapter information needed for the
165 driver itself and the mid- and higher-level SCSI-drivers. The SCSI pun/lun
166 and the adapters' ldn tables are initialized and get probed afterwards by
167 the check_devices() function. If no further adapters are found,
168 ibmmca_detect() quits.
170 2.2 Physical Units, Logical Units, and Logical Devices
171 ------------------------------------------------------
172 There can be up to 56 devices on the SCSI bus (besides the adapter):
173 there are up to 7 "physical units" (each identified by physical unit
174 number or pun, also called the scsi id, this is the number you select
175 with hardware jumpers), and each physical unit can have up to 8
176 "logical units" (each identified by logical unit number, or lun,
177 between 0 and 7). The IBM SCSI-2 F/W adapter offers this on up to two
178 busses and provides support for 30 logical devices at the same time, where
179 in wide-addressing mode you can have 16 puns with 32 luns on each device.
180 This section dexribes you the handling of devices on non-F/W adapters.
181 Just imagine, that you can have 16 * 32 = 512 devices on a F/W adapter
182 which means a lot of possible devices for such a small machine.
184 Typically the adapter has pun=7, so puns of other physical units
185 are between 0 and 6(15). On a wide-adapter a pun higher than 7 is
186 possible, but is normally not used. Almost all physical units have only
187 one logical unit, with lun=0. A CD-ROM jukebox would be an example of a
188 physical unit with more than one logical unit.
190 The embedded microprocessor of the IBM SCSI-subsystem hides the complex
191 two-dimensional (pun,lun) organization from the operating system.
192 When the machine is powered-up (or rebooted), the embedded microprocessor
193 checks, on its own, all 56 possible (pun,lun) combinations, and the first
194 15 devices found are assigned into a one-dimensional array of so-called
195 "logical devices", identified by "logical device numbers" or ldn. The last
196 ldn=15 is reserved for the subsystem itself. Wide adapters may have
197 to check up to 15 * 8 = 120 pun/lun combinations.
199 2.3 SCSI-Device Recognition and dynamical ldn Assignment
200 --------------------------------------------------------
201 One consequence of information hiding is that the real (pun,lun)
202 numbers are also hidden. The two possibilities to get around this problem
203 is to offer fake pun/lun combinations to the operating system or to
204 delete the whole mapping of the adapter and to reassign the ldns, using
205 the immediate assign command of the SCSI-subsystem for probing through
206 all possible pun/lun combinations. a ldn is a "logical device number"
207 which is used by IBM SCSI-subsystems to access some valid SCSI-device.
208 At the beginning of the development of this driver, the following approach
211 First, the driver checked the ldn's (0 to 6) to find out which ldn's
212 have devices assigned. This was done by the functions check_devices() and
213 device_exists(). The interrupt handler has a special paragraph of code
214 (see local_checking_phase_flag) to assist in the checking. Assume, for
215 example, that three logical devices were found assigned at ldn 0, 1, 2.
216 These are presented to the upper layer of Linux SCSI driver
217 as devices with bogus (pun, lun) equal to (0,0), (1,0), (2,0).
218 On the other hand, if the upper layer issues a command to device
219 say (4,0), this driver returns DID_NO_CONNECT error.
221 In a second step of the driver development, the following improvement has
222 been applied: The first approach limited the number of devices to 7, far
223 fewer than the 15 that it could usem then it just maped ldn ->
224 (ldn/8,ldn%8) for pun,lun. We ended up with a real mishmash of puns
225 and luns, but it all seemed to work.
227 The latest development, which is implemented from the driver version 3.0
228 and later, realizes the device recognition in the following way:
229 The physical SCSI-devices on the SCSI-bus are probed via immediate_assign-
230 and device_inquiry-commands, that is all implemented in a completely new
231 made check_devices() subroutine. This delivers an exact map of the physical
232 SCSI-world that is now stored in the get_scsi[][]-array. This means,
233 that the once hidden pun,lun assignment is now known to this driver.
234 It no longer believes in default-settings of the subsystem and maps all
235 ldns to existing pun,lun "by foot". This assures full control of the ldn
236 mapping and allows dynamical remapping of ldns to different pun,lun, if
237 there are more SCSI-devices installed than ldns available (n>15). The
238 ldns from 0 to 6 get 'hardwired' by this driver to puns 0 to 7 at lun=0,
239 excluding the pun of the subsystem. This assures, that at least simple
240 SCSI-installations have optimum access-speed and are not touched by
241 dynamical remapping. The ldns 7 to 14 are put to existing devices with
242 lun>0 or to non-existing devices, in order to satisfy the subsystem, if
243 there are less than 15 SCSI-devices connected. In the case of more than 15
244 devices, the dynamical mapping goes active. If the get_scsi[][] reports a
245 device to be existant, but it has no ldn assigned, it gets a ldn out of 7
246 to 14. The numbers are assigned in cyclic order. Therefore it takes 8
247 dynamical reassignments on the SCSI-devices, until a certain device
248 looses its ldn again. This assures, that dynamical remapping is avoided
249 during intense I/O between up to 15 SCSI-devices (means pun,lun
250 combinations). A further advantage of this method is, that people who
251 build their kernel without probing on all luns will get what they expect,
252 because the driver just won't assign everything with lun>0 when
253 multpile lun probing is inactive.
255 2.4 SCSI-Device Order
256 ---------------------
257 Because of the now correct recognition of physical pun,lun, and
258 their report to mid-level- and higher-level-drivers, the new reported puns
259 can be different from the old, faked puns. Therefore, Linux will eventually
260 change /dev/sdXXX assignments and prompt you for corrupted superblock
261 repair on boottime. In this case DO NOT PANIC, YOUR DISKS ARE STILL OK!!!
262 You have to reboot (CTRL-D) with an old kernel and set the /etc/fstab-file
263 entries right. After that, the system should come up as errorfree as before.
264 If your boot-partition is not coming up, also edit the /etc/lilo.conf-file
265 in a Linux session booted on old kernel and run lilo before reboot. Check
266 lilo.conf anyway to get boot on other partitions with foreign OSes right
267 again. But there exists a feature of this driver that allows you to change
268 the assignment order of the SCSI-devices by flipping the PUN-assignment.
269 See the next paragraph for a description.
271 The problem for this is, that Linux does not assign the SCSI-devices in the
272 way as described in the ANSI-SCSI-standard. Linux assigns /dev/sda to
273 the device with at minimum id 0. But the first drive should be at id 6,
274 because for historical reasons, drive at id 6 has, by hardware, the highest
275 priority and a drive at id 0 the lowest. IBM was one of the rare producers,
276 where the BIOS assigns drives belonging to the ANSI-SCSI-standard. Most
277 other producers' BIOS does not (I think even Adaptec-BIOS). The
278 IBMMCA_SCSI_ORDER_STANDARD flag, which you set while configuring the
279 kernel enables to choose the preferred way of SCSI-device-assignment.
280 Defining this flag would result in Linux determining the devices in the
281 same order as DOS and OS/2 does on your MCA-machine. This is also standard
282 on most industrial computers and OSes, like e.g. OS-9. Leaving this flag
283 undefined will get your devices ordered in the default way of Linux. See
284 also the remarks of Chris Beauregard from Dec 15, 1997 and the followups
287 2.5 Regular SCSI-Command-Processing
288 -----------------------------------
289 Only three functions get involved: ibmmca_queuecommand(), issue_cmd(),
290 and interrupt_handler().
292 The upper layer issues a scsi command by calling function
293 ibmmca_queuecommand(). This function fills a "subsystem control block"
294 (scb) and calls a local function issue_cmd(), which writes a scb
295 command into subsystem I/O ports. Once the scb command is carried out,
296 the interrupt_handler() is invoked. If a device is determined to be
297 existant and it has not assigned any ldn, it gets one dynamically.
298 For this, the whole stuff is done in ibmmca_queuecommand().
300 2.6 Abort & Reset Commands
301 --------------------------
302 These are implemented with busy waiting for interrupt to arrive.
303 ibmmca_reset() and ibmmca_abort() do not work sufficently well
304 up to now and need still a lot of development work. But, this seems
305 to be even a problem with other SCSI-low level drivers, too. However,
306 this should be no excuse.
310 The ibmmca_biosparams() function should return the same disk geometry
311 as the bios. This is needed for fdisk, etc. The returned geometry is
312 certainly correct for disks smaller than 1 gigabyte. In the meantime,
313 it has been proved, that this works fine even with disks larger than
316 2.8 Kernel Boot Option
317 ----------------------
318 The function ibmmca_scsi_setup() is called if option ibmmcascsi=n
319 is passed to the kernel. See file linux/init/main.c for details.
321 2.9 Driver Module Support
322 -------------------------
323 Is implemented and tested by K. Kudielka. This could probably not work
326 2.10 Multiple Hostadapter Support
327 ---------------------------------
328 This driver supports up to eight interfaces of type IBM-SCSI-Subsystem.
329 Integrated-, and MCA-adapters are automatically recognized. Unrecognizable
330 IBM-SCSI-Subsystem interfaces can be specified as kernel-parameters.
332 2.11 /proc/scsi-Filesystem Information
333 --------------------------------------
334 Information about the driver condition is given in
335 /proc/scsi/ibmmca/<host_no>. ibmmca_proc_info() provides this information.
337 This table is quite informative for interested users. It shows the load
338 of commands on the subsystem and wether you are running the bypassed
339 (software) or integrated (hardware) SCSI-command set (see below). The
340 amount of accesses is shown. Read, write, modeselect is shown seperately
341 in order to help debugging problems with CD-ROMs or tapedrives.
343 The following table shows the list of 15 logical device numbers, that are
344 used by the SCSI-subsystem. The load on each ldn is shown in the table,
345 again, read and write commands are split. The last column shows the amount
346 of reassignments, that have been applied to the ldns, if you have more than
347 15 pun/lun combinations available on the SCSI-bus.
349 The last two tables show the pun/lun map and the positions of the ldns
350 on this pun/lun map. This may change during operation, when a ldn is
351 reassigned to another pun/lun combination. If the necessity for dynamical
352 assignments is set to 'no', the ldn structure keeps static.
354 2.12 /proc/mca-Filesystem Information
355 -------------------------------------
356 The slot-file contains all default entries and in addition chip and I/O-
357 address information of the SCSI-subsystem. This information is provided
360 2.13 Supported IBM SCSI-Subsystems
361 ----------------------------------
362 The following IBM SCSI-subsystems are supported by this driver:
364 - IBM Fast/Wide SCSI-2 Adapter
365 - IBM 7568 Industrial Computer SCSI Adapter w/cache
366 - IBM Expansion Unit SCSI Controller
367 - IBM SCSI Adapter w/Cache
369 - IBM Integrated SCSI Controller
370 - All clones, 100% compatible with the chipset and subsystem command
371 system of IBM SCSI-adapters (forced detection)
373 2.14 Linux Kernel Versions
374 --------------------------
375 The IBM SCSI-subsystem low level driver is prepared to be used with
376 all versions of Linux between 2.0.x and 2.4.x. The compatibility checks
377 are fully implemented up from version 3.1e of the driver. This means, that
378 you just need the latest ibmmca.h and ibmmca.c file and copy it in the
379 linux/drivers/scsi directory. The code is automatically adapted during
384 Jan 15 1996: First public release.
387 Jan 23 1996: Scrapped code which reassigned scsi devices to logical
388 device numbers. Instead, the existing assignment (created
389 when the machine is powered-up or rebooted) is used.
390 A side effect is that the upper layer of Linux SCSI
391 device driver gets bogus scsi ids (this is benign),
392 and also the hard disks are ordered under Linux the
393 same way as they are under dos (i.e., C: disk is sda,
394 D: disk is sdb, etc.).
397 I think that the CD-ROM is now detected only if a CD is
398 inside CD_ROM while Linux boots. This can be fixed later,
399 once the driver works on all types of PS/2's.
402 Feb 7 1996: Modified biosparam function. Fixed the CD-ROM detection.
403 For now, devices other than harddisk and CD_ROM are
404 ignored. Temporarily modified abort() function
405 to behave like reset().
408 Mar 31 1996: The integrated scsi subsystem is correctly found
409 in PS/2 models 56,57, but not in model 76. Therefore
410 the ibmmca_scsi_setup() function has been added today.
411 This function allows the user to force detection of
412 scsi subsystem. The kernel option has format
414 where n is the scsi_id (pun) of the subsystem. Most likely, n is 7.
417 Aug 21 1996: Modified the code which maps ldns to (pun,0). It was
418 insufficient for those of us with CD-ROM changers.
421 Dec 14 1996: More improvements to the ldn mapping. See check_devices
422 for details. Did more fiddling with the integrated SCSI detection,
423 but I think it's ultimately hopeless without actually testing the
424 model of the machine. The 56, 57, 76 and 95 (ultimedia) all have
425 different integrated SCSI register configurations. However, the 56
426 and 57 are the only ones that have problems with forced detection.
429 Mar 8-16 1997: Modified driver to run as a module and to support
430 multiple adapters. A structure, called ibmmca_hostdata, is now
431 present, containing all the variables, that were once only
432 available for one single adapter. The find_subsystem-routine has vanished.
433 The hardware recognition is now done in ibmmca_detect directly.
434 This routine checks for presence of MCA-bus, checks the interrupt
435 level and continues with checking the installed hardware.
436 Certain PS/2-models do not recognize a SCSI-subsystem automatically.
437 Hence, the setup defined by command-line-parameters is checked first.
438 Thereafter, the routine probes for an integrated SCSI-subsystem.
439 Finally, adapters are checked. This method has the advantage to cover all
440 possible combinations of multiple SCSI-subsystems on one MCA-board. Up to
441 eight SCSI-subsystems can be recognized and announced to the upper-level
442 drivers with this improvement. A set of defines made changes to other
443 routines as small as possible.
447 1) SCSI-command capability enlarged by the recognition of MODE_SELECT.
448 This needs the RD-Bit to be disabled on IM_OTHER_SCSI_CMD_CMD which
449 allows data to be written from the system to the device. It is a
450 necessary step to be allowed to set blocksize of SCSI-tape-drives and
451 the tape-speed, whithout confusing the SCSI-Subsystem.
452 2) The recognition of a tape is included in the check_devices routine.
453 This is done by checking for TYPE_TAPE, that is already defined in
454 the kernel-scsi-environment. The markup of a tape is done in the
455 global ldn_is_tape[] array. If the entry on index ldn
456 is 1, there is a tapedrive connected.
457 3) The ldn_is_tape[] array is necessary to distinguish between tape- and
458 other devices. Fixed blocklength devices should not cause a problem
459 with the SCB-command for read and write in the ibmmca_queuecommand
460 subroutine. Therefore, I only derivate the READ_XX, WRITE_XX for
461 the tape-devices, as recommended by IBM in this Technical Reference,
462 mentioned below. (IBM recommends to avoid using the read/write of the
463 subsystem, but the fact was, that read/write causes a command error from
464 the subsystem and this causes kernel-panic.)
465 4) In addition, I propose to use the ldn instead of a fix char for the
466 display of PS2_DISK_LED_ON(). On 95, one can distinguish between the
467 devices that are accessed. It shows activity and easyfies debugging.
468 The tape-support has been tested with a SONY SDT-5200 and a HP DDS-2
469 (I do not know yet the type). Optimization and CD-ROM audio-support,
473 June 19 1997: (v1.6b)
474 1) Submitting the extra-array ldn_is_tape[] -> to the local ld[]
476 2) CD-ROM Audio-Play seems to work now.
477 3) When using DDS-2 (120M) DAT-Tapes, mtst shows still density-code
478 0x13 for ordinary DDS (61000 BPM) instead 0x24 for DDS-2. This appears
479 also on Adaptec 2940 adaptor in a PCI-System. Therefore, I assume that
480 the problem is independent of the low-level-driver/bus-architecture.
481 4) Hexadecimal ldn on PS/2-95 LED-display.
482 5) Fixing of the PS/2-LED on/off that it works right with tapedrives and
483 does not confuse the disk_rw_in_progress counter.
486 June 21 1997: (v1.7b)
487 1) Adding of a proc_info routine to inform in /proc/scsi/ibmmca/<host> the
488 outer-world about operational load statistics on the different ldns,
489 seen by the driver. Everybody that has more than one IBM-SCSI should
490 test this, because I only have one and cannot see what happens with more
491 than one IBM-SCSI hosts.
492 2) Definition of a driver version-number to have a better recognition of
493 the source when there are existing too much releases that may confuse
494 the user, when reading about release-specific problems. Up to know,
495 I calculated the version-number to be 1.7. Because we are in BETA-test
496 yet, it is today 1.7b.
497 3) Sorry for the heavy bug I programmed on June 19 1997! After that, the
498 CD-ROM did not work any more! The C7-command was a fake impression
499 I got while programming. Now, the READ and WRITE commands for CD-ROM are
500 no longer running over the subsystem, but just over
501 IM_OTHER_SCSI_CMD_CMD. On my observations (PS/2-95), now CD-ROM mounts
502 much faster(!) and hopefully all fancy multimedia-functions, like direct
503 digital recording from audio-CDs also work. (I tried it with cdda2wav
504 from the cdwtools-package and it filled up the harddisk immediately :-).)
505 To easify boolean logics, a further local device-type in ld[], called
506 is_cdrom has been included.
507 4) If one uses a SCSI-device of unsupported type/commands, one
508 immediately runs into a kernel-panic caused by Command Error. To better
509 understand which SCSI-command caused the problem, I extended this
510 specific panic-message slightly.
513 June 25 1997: (v1.8b)
514 1) Some cosmetical changes for the handling of SCSI-device-types.
515 Now, also CD-Burners / WORMs and SCSI-scanners should work. For
516 MO-drives I have no experience, therefore not yet supported.
517 In logical_devices I changed from different type-variables to one
518 called 'device_type' where the values, corresponding to scsi.h,
519 of a SCSI-device are stored.
520 2) There existed a small bug, that maps a device, coming after a SCSI-tape
521 wrong. Therefore, e.g. a CD-ROM changer would have been mapped wrong
523 3) Extension of the logical_device structure. Now it contains also device,
524 vendor and revision-level of a SCSI-device for internal usage.
527 June 26-29 1997: (v2.0b)
528 1) The release number 2.0b is necessary because of the completely new done
529 recognition and handling of SCSI-devices with the adapter. As I got
530 from Chris the hint, that the subsystem can reassign ldns dynamically,
531 I remembered this immediate_assign-command, I found once in the handbook.
532 Now, the driver first kills all ldn assignments that are set by default
533 on the SCSI-subsystem. After that, it probes on all puns and luns for
534 devices by going through all combinations with immediate_assign and
535 probing for devices, using device_inquiry. The found physical(!) pun,lun
536 structure is stored in get_scsi[][] as device types. This is followed
537 by the assignment of all ldns to existing SCSI-devices. If more ldns
538 than devices are available, they are assigned to non existing pun,lun
539 combinations to satisfy the adapter. With this, the dynamical mapping
540 was possible to implement. (For further info see the text in the
541 source-code and in the description below. Read the description
542 below BEFORE installing this driver on your system!)
543 2) Changed the name IBMMCA_DRIVER_VERSION to IBMMCA_SCSI_DRIVER_VERSION.
544 3) The LED-display shows on PS/2-95 no longer the ldn, but the SCSI-ID
545 (pun) of the accessed SCSI-device. This is now senseful, because the
546 pun known within the driver is exactly the pun of the physical device
547 and no longer a fake one.
548 4) The /proc/scsi/ibmmca/<host_no> consists now of the first part, where
549 hit-statistics of ldns is shown and a second part, where the maps of
550 physical and logical SCSI-devices are displayed. This could be very
551 interesting, when one is using more than 15 SCSI-devices in order to
552 follow the dynamical remapping of ldns.
555 June 26-29 1997: (v2.0b-1)
556 1) I forgot to switch the local_checking_phase_flag to 1 and back to 0
557 in the dynamical remapping part in ibmmca_queuecommand for the
558 device_exist routine. Sorry.
561 July 1-13 1997: (v3.0b,c)
562 1) Merging of the driver-developments of Klaus Kudielka and Michael Lang
563 in order to get a optimum and unified driver-release for the
564 IBM-SCSI-Subsystem-Adapter(s).
565 For people, using the Kernel-release >=2.1.0, module-support should
566 be no problem. For users, running under <2.1.0, module-support may not
567 work, because the methods have changed between 2.0.x and 2.1.x.
568 2) Added some more effective statistics for /proc-output.
569 3) Change typecasting at necessary points from (unsigned long) to
571 4) Included #if... at special points to have specific adaption of the
572 driver to kernel 2.0.x and 2.1.x. It should therefore also run with
574 5) Magneto-Optical drives and medium-changers are also recognized, now.
575 Therefore, we have a completely gapfree recognition of all SCSI-
576 device-types, that are known by Linux up to kernel 2.1.31.
577 6) The flag SCSI_IBMMCA_DEV_RESET has been inserted. If it is set within
578 the configuration, each connected SCSI-device will get a reset command
579 during boottime. This can be necessary for some special SCSI-devices.
580 This flag should be included in Config.in.
581 (See also the new Config.in file.)
582 Probable next improvement: bad disk handler.
585 Sept 14 1997: (v3.0c)
586 1) Some debugging and speed optimization applied.
590 - chrisb@truespectra.com
591 - made the front panel display thingy optional, specified from the
592 command-line via ibmmcascsi=display. Along the lines of the /LED
593 option for the OS/2 driver.
594 - fixed small bug in the LED display that would hang some machines.
595 - reversed ordering of the drives (using the
596 IBMMCA_SCSI_ORDER_STANDARD define). This is necessary for two main
598 - users who've already installed Linux won't be screwed. Keep
599 in mind that not everyone is a kernel hacker.
600 - be consistent with the BIOS ordering of the drives. In the
601 BIOS, id 6 is C:, id 0 might be D:. With this scheme, they'd be
602 backwards. This confuses the crap out of those heathens who've
603 got a impure Linux installation (which, <wince>, I'm one of).
604 This whole problem arises because IBM is actually non-standard with
605 the id to BIOS mappings. You'll find, in fdomain.c, a similar
606 comment about a few FD BIOS revisions. The Linux (and apparently
607 industry) standard is that C: maps to scsi id (0,0). Let's stick
609 - Since this is technically a branch of my own, I changed the
610 version number to 3.0e-cpb.
612 Jan 17, 1998: (v3.0f)
613 1) Addition of some statistical info for /proc in proc_info.
614 2) Taking care of the SCSI-assignment problem, dealed by Chris at Dec 15
615 1997. In fact, IBM is right, concerning the assignment of SCSI-devices
616 to driveletters. It is conform to the ANSI-definition of the SCSI-
617 standard to assign drive C: to SCSI-id 6, because it is the highest
618 hardware priority after the hostadapter (that has still today by
619 default everywhere id 7). Also realtime-operating systems that I use,
620 like LynxOS and OS9, which are quite industrial systems use top-down
621 numbering of the harddisks, that is also starting at id 6. Now, one
622 sits a bit between two chairs. On one hand side, using the define
623 IBMMCA_SCSI_ORDER_STANDARD makes Linux assigning disks conform to
624 the IBM- and ANSI-SCSI-standard and keeps this driver downward
625 compatible to older releases, on the other hand side, people is quite
626 habituated in believing that C: is assigned to (0,0) and much other
627 SCSI-BIOS do so. Therefore, I moved the IBMMCA_SCSI_ORDER_STANDARD
628 define out of the driver and put it into Config.in as subitem of
629 'IBM SCSI support'. A help, added to Documentation/Configure.help
630 explains the differences between saying 'y' or 'n' to the user, when
631 IBMMCA_SCSI_ORDER_STANDARD prompts, so the ordinary user is enabled to
632 choose the way of assignment, depending on his own situation and gusto.
633 3) Adapted SCSI_IBMMCA_DEV_RESET to the local naming convention, so it is
634 now called IBMMCA_SCSI_DEV_RESET.
635 4) Optimization of proc_info and its subroutines.
636 5) Added more in-source-comments and extended the driver description by
637 some explanation about the SCSI-device-assignment problem.
640 Jan 18, 1998: (v3.0g)
641 1) Correcting names to be absolutely conform to the later 2.1.x releases.
642 This is necessary for
643 IBMMCA_SCSI_DEV_RESET -> CONFIG_IBMMCA_SCSI_DEV_RESET
644 IBMMCA_SCSI_ORDER_STANDARD -> CONFIG_IBMMCA_SCSI_ORDER_STANDARD
647 Jan 18, 1999: (v3.1 MCA-team internal)
648 1) The multiple hosts structure is accessed from every subroutine, so there
649 is no longer the address of the device structure passed from function
650 to function, but only the hostindex. A call by value, nothing more. This
651 should really be understood by the compiler and the subsystem should get
652 the right values and addresses.
653 2) The SCSI-subsystem detection was not complete and quite hugely buggy up
654 to now, compared to the technical manual. The interpretation of the pos2
655 register is not as assumed by people before, therefore, I dropped a note
656 in the ibmmca_detect function to show the registers' interpretation.
657 The pos-registers of integrated SCSI-subsystems do not contain any
658 information concerning the IO-port offset, really. Instead, they contain
659 some info about the adapter, the chip, the NVRAM .... The I/O-port is
660 fixed to 0x3540 - 0x3547. There can be more than one adapters in the
661 slots and they get an offset for the I/O area in order to get their own
662 I/O-address area. See chapter 2 for detailed description. At least, the
663 detection should now work right, even on models other than 95. The 95ers
664 came happily around the bug, as their pos2 register contains always 0
665 in the critical area. Reserved bits are not allowed to be interpreted,
666 therefore, IBM is allowed to set those bits as they like and they may
667 really vary between different PS/2 models. So, now, no interpretation
668 of reserved bits - hopefully no trouble here anymore.
669 3) The command error, which you may get on models 55, 56, 57, 70, 77 and
670 P70 may have been caused by the fact, that adapters of older design do
671 not like sending commands to non-existing SCSI-devices and will react
672 with a command error as a sign of protest. While this error is not
673 present on IBM SCSI Adapter w/cache, it appears on IBM Integrated SCSI
674 Adapters. Therefore, I implemented a workarround to forgive those
675 adapters their protests, but it is marked up in the statisctis, so
676 after a successful boot, you can see in /proc/scsi/ibmmca/<host_number>
677 how often the command errors have been forgiven to the SCSI-subsystem.
678 If the number is bigger than 0, you have a SCSI subsystem of older
679 design, what should no longer matter.
680 4) ibmmca_getinfo() has been adapted very carefully, so it shows in the
681 slotn file really, what is senseful to be presented.
682 5) ibmmca_register() has been extended in its parameter list in order to
683 pass the right name of the SCSI-adapter to Linux.
687 1) Finally, after some 3.1Beta-releases, the 3.1 release. Sorry, for
688 the delayed release, but it was not finished with the release of
693 1) Added a new commandline parameter called 'bypass' in order to bypass
694 every integrated subsystem SCSI-command consequently in case of
696 2) Concatenated read_capacity requests to the harddisks. It gave a lot
697 of troubles with some controllers and after I wanted to apply some
698 extensions, it jumped out in the same situation, on my w/cache, as like
699 on D. Weinehalls' Model 56, having integrated SCSI. This gave me the
700 descissive hint to move the code-part out and declare it global. Now,
701 it seems to work by far much better an more stable. Let us see, what
702 the world thinks of it...
703 3) By the way, only Sony DAT-drives seem to show density code 0x13. A
704 test with a HP drive gave right results, so the problem is vendor-
705 specific and not a problem of the OS or the driver.
709 1) The abort command and the reset function have been checked for
710 inconsistencies. From the logical point of thinking, they work
711 at their optimum, now, but as the subsystem does not answer with an
712 interrupt, abort never finishes, sigh...
713 2) Everything, that is accessed by a busmaster request from the adapter
714 is now declared as global variable, even the return-buffer in the
715 local checking phase. This assures, that no accesses to undefined memory
717 3) In ibmmca.h, the line unchecked_isa_dma is added with 1 in order to
718 avoid memory-pointers for the areas higher than 16MByte in order to
719 be sure, it also works on 16-Bit Microchannel bus systems.
720 4) A lot of small things have been found, but nothing that endangered the
721 driver operations. Just it should be more stable, now.
725 1) I took the warning from the Linux Kernel Hackers Guide serious and
726 checked the cmd->result return value to the done-function very carefuly.
727 It is obvious, that the IBM SCSI only delivers the tsb.dev_status, if
728 some error appeared, else it is undefined. Now, this is fixed. Before
729 any SCB command gets queued, the tsb.dev_status is set to 0, so the
730 cmd->result won't screw up Linux higher level drivers.
731 2) The reset-function has slightly improved. This is still planed for
732 abort. During the abort and the reset function, no interrupts are
733 allowed. This is however quite hard to cope with, so the INT-status
734 register is read. When the interrupt gets queued, one can find its
735 status immediately on that register and is enabled to continue in the
736 reset function. I had no chance to test this really, only in a bogus
737 situation, I got this function running, but the situation was too much
738 worse for Linux :-(, so tests will continue.
739 3) Buffers got now consistent. No open address mapping, as before and
740 therefore no further troubles with the unassigned memory segmentation
741 faults that scrambled probes on 95XX series and even on 85XX series,
742 when the kernel is done in a not so perfectly fitting way.
743 4) Spontaneous interrupts from the subsystem, appearing without any
744 command previously queued are answered with a DID_BAD_INTR result.
745 5) Taken into account ZP Gus' proposals to reverse the SCSI-device
746 scan order. As it does not work on Kernel 2.1.x or 2.2.x, as proposed
747 by him, I implemented it in a slightly derived way, which offers in
748 addition more flexibility.
751 Apr 23, 2000 (v3.2pre1)
752 1) During a very long time, I collected a huge amount of bugreports from
753 various people, trying really quite different things on their SCSI-
754 PS/2s. Today, all these bugreports are taken into account and should be
755 mostly solved. The major topics were:
756 - Driver crashes during boottime by no obvious reason.
757 - Driver panics while the midlevel-SCSI-driver is trying to inquire
758 the SCSI-device properties, even though hardware is in perfect state.
759 - Displayed info for the various slot-cards is interpreted wrong.
760 The main reasons for the crashes were two:
761 1) The commands to check for device information like INQUIRY,
762 TEST_UNIT_READY, REQUEST_SENSE and MODE_SENSE cause the devices
763 to deliver information of up to 255 bytes. Midlevel drivers offer
764 1024 bytes of space for the answer, but the IBM-SCSI-adapters do
765 not accept this, as they stick quite near to ANSI-SCSI and report
766 a COMMAND_ERROR message which causes the driver to panic. The main
767 problem was located around the INQUIRY command. Now, for all the
768 mentioned commands, the buffersize, sent to the adapter is at
769 maximum 255 which seems to be a quite reasonable solution.
770 TEST_UNIT_READY gets a buffersize of 0 to make sure, that no
771 data is transferred in order to avoid any possible command failure.
772 2) On unsuccessful TEST_UNIT_READY, the midlevel-driver has to send
773 a REQUEST_SENSE in order to see, where the problem is located. This
774 REQUEST_SENSE may have various length in its answer-buffer. IBM
775 SCSI-subsystems report a command failure, if the returned buffersize
776 is different from the sent buffersize, but this can be supressed by
777 a special bit, which is now done and problems seem to be solved.
778 2) Code adaption to all kernel-releases. Now, the 3.2 code compiles on
779 2.0.x, 2.1.x, 2.2.x and 2.3.x kernel releases without any code-changes.
780 3) Commandline-parameters are recognized again, even under Kernel 2.3.x or
784 April 27, 2000 (v3.2pre2)
785 1) Bypassed commands get read by the adapter by one cycle instead of two.
786 This increases SCSI-performance.
787 2) Synchronous datatransfer is provided for sure to be 5 MHz on older
788 SCSI and 10 MHz on internal F/W SCSI-adapter.
789 3) New commandline parameters allow to force the adapter to slow down while
790 in synchronous transfer. Could be helpful for very old devices.
793 June 2, 2000 (v3.2pre5)
794 1) Added Jim Shorney's contribution to make the activity indicator
795 flashing in addition to the LED-alphanumeric display-panel on
796 models 95A. To be enabled to choose this feature freely, a new
797 commandline parameter is added, called 'activity'.
798 2) Added the READ_CONTROL bit for test_unit_ready SCSI-command.
799 3) Added some suppress_exception bits to read_device_capacity and
800 all device_inquiry occurences in the driver code.
801 4) Complaints about the various KERNEL_VERSION implementations are
802 taken into account. Every local_LinuxKernelVersion occurence is
803 now replaced by KERNEL_VERSION, defined in linux/version.h.
804 Corresponding changes were applied to ibmmca.h, too. This was a
805 contribution to all kernel-parts by Philipp Hahn.
808 July 17, 2000 (v3.2pre8)
809 A long period of collecting bugreports from all corners of the world
810 now lead to the following corrections to the code:
811 1) SCSI-2 F/W support crashed with a COMMAND ERROR. The reason for this
812 was, that it is possible to disbale Fast-SCSI for the external bus.
813 The feature-control command, where this crash appeared regularly tried
814 to set the maximum speed of 10MHz synchronous transfer speed and that
815 reports a COMMAND ERROR, if external bus Fast-SCSI is disabled. Now,
816 the feature-command probes down from maximum speed until the adapter
817 stops to complain, which is at the same time the maximum possible
818 speed selected in the reference program. So, F/W external can run at
819 5 MHz (slow-) or 10 MHz (fast-SCSI). During feature probing, the
820 COMMAND ERROR message is used to detect if the adapter does not complain.
821 2) Up to now, only combined busmode is supported, if you use external
822 SCSI-devices, attached to the F/W-controller. If dual bus is selected,
823 only the internal SCSI-devices get accessed by Linux. For most
824 applications, this should do fine.
825 3) Wide-SCSI-addressing (16-Bit) is now possible for the internal F/W
826 bus on the F/W adapter. If F/W adapter is detected, the driver
827 automatically uses the extended PUN/LUN <-> LDN mapping tables, which
828 are now new from 3.2pre8. This allows PUNs between 0 and 15 and should
829 provide more fun with the F/W adapter.
830 4) Several machines use the SCSI: POS registers for internal/undocumented
831 storage of system relevant info. This confused the driver, mainly on
832 models 9595, as it expected no onboard SCSI only, if all POS in
833 the integrated SCSI-area are set to 0x00 or 0xff. Now, the mechanism
834 to check for integrated SCSI is much more restrictive and these problems
838 July 18, 2000 (v3.2pre9)
839 This develop rather quickly at the moment. Two major things were still
841 1) The adapter PUN for F/W adapters has 4-bits, while all other adapters
842 have 3-bits. This is now taken into account for F/W.
843 2) When you select CONFIG_IBMMCA_SCSI_ORDER_STANDARD, you should
844 normally get the inverse probing order of your devices on the SCSI-bus.
845 The ANSI device order gets scrambled in version 3.2pre8!! Now, a new
846 and tested algorithm inverts the device-order on the SCSI-bus and
847 automatically avoids accidental access to whatever SCSI PUN the adapter
848 is set and works with SCSI- and Wide-SCSI-addressing.
851 July 23, 2000 (v3.2pre10 unpublished)
852 1) LED panel display supports wide-addressing in ibmmca=display mode.
853 2) Adapter-information and autoadaption to address-space is done.
854 3) Auto-probing for maximum synchronous SCSI transfer rate is working.
855 4) Optimization to some embedded function calls is applied.
856 5) Added some comment for the user to wait for SCSI-devices beeing probed.
857 6) Finished version 3.2 for Kernel 2.4.0. It least, I thought it is but...
860 July 26, 2000 (v3.2pre11)
861 1) I passed a horrible weekend getting mad with NMIs on kernel 2.2.14 and
862 a model 9595. Asking around in the community, nobody except of me has
863 seen such errors. Weired, but I am trying to recompile everything on
864 the model 9595. Maybe, as I use a specially modified gcc, that could
865 cause problems. But, it was not the reason. The true background was,
866 that the kernel was compiled for i386 and the 9595 has a 486DX-2.
867 Normally, no troubles should appear, but for this special machine,
868 only the right processor support is working fine!
869 2) Previous problems with synchronous speed, slowing down from one adapter
870 to the next during probing are corrected. Now, local variables store
871 the synchronous bitmask for every single adapter found on the MCA bus.
872 3) LED alphanumeric panel support for XX95 systems is now showing some
873 alive rotator during boottime. This makes sense, when no monitor is
874 connected to the system. You can get rid of all display activity, if
875 you do not use any parameter or just ibmmcascsi=activity, for the
876 harddrive activity LED, existant on all PS/2, except models 8595-XXX.
877 If no monitor is available, please use ibmmcascsi=display, which works
878 fine together with the linuxinfo utility for the LED-panel.
882 1) Submission of this driver for kernel 2.4test-XX and 2.2.17.
887 - IBM SCSI-2 F/W external SCSI bus support in seperate mode.
888 - It seems that the handling of bad disks is really bad -
889 non-existent, in fact.
890 - More testing of the full driver-controlled dynamical ldn
891 (re)mapping for up to 56 SCSI-devices. I guess, it won't work
892 at the moment, but nobody ever really tried it.
893 - Abort and Reset functions still slightly buggy.
897 5.1 Commandline Parameters
898 --------------------------
899 There exist several features for the IBM SCSI-subsystem driver.
900 The commandline parameter format is:
902 ibmmcascsi=<command1>,<command2>,<command3>,...
904 where commandN can be one of the following:
906 display Owners of a model 95 or other PS/2 systems with an
907 alphanumeric LED display may set this to have their
908 display showing the following output of the 8 digits:
912 where '-' stays dark, 'D' shows the SCSI-device id
913 and 'A' shows the SCSI hostindex, beeing currently
914 accessed. During boottime, this will give the message
918 on the LED-panel, where the * represents a rotator,
919 showing the activity during the probing phase of the
920 driver which can take up to two minutes per SCSI-adapter.
921 adisplay This works like display, but gives more optical overview
922 of the activities on the SCSI-bus. The display will have
923 the following output:
927 where the numbers 0 to 6 light up at the shown position,
928 when the SCSI-device is accessed. A shows again the SCSI
929 hostindex. If display nor adisplay is set, the internal
930 PS/2 harddisk LED is used for media-activities. So, if
931 you really do not have a system with a LED-display, you
932 should not set display or adisplay. Keep in mind, that
933 display and adisplay can only be used alternatively. It
934 is not recommended to use this option, if you have some
935 wide-addressed devices e.g. at the SCSI-2 F/W adapter in
936 your system. In addition, the usage of the display for
937 other tasks in parallel, like the linuxinfo-utility makes
938 no sense with this option.
939 activity This enables the PS/2 harddisk LED activity indicator.
940 Most PS/2 have no alphanumeric LED display, but some
941 indicator. So you should use this parameter to activate it.
942 If you own model 9595 (Server95), you can have both, the
943 LED panel and the activity indicator in parallel. However,
944 some PS/2s, like the 8595 do not have any harddisk LED
945 activity indicator, which means, that you must use the
946 alphanumeric LED display if you want to monitor SCSI-
948 bypass This commandline parameter forces the driver never to use
949 SCSI-subsystems' integrated SCSI-command set. Except of
950 the immediate assign, which is of vital importance for
951 every IBM SCSI-subsystem to set its ldns right. Instead,
952 the ordinary ANSI-SCSI-commands are used and passed by the
953 controller to the SCSI-devices, therefore 'bypass'. The
954 effort, done by the subsystem is quite bogus and at a
955 minimum and therefore it should work everywhere. This
956 could maybe solve troubles with old or integrated SCSI-
957 controllers and nasty harddisks. Keep in mind, that using
958 this flag will slow-down SCSI-accesses slightly, as the
959 software generated commands are always slower than the
960 hardware. Non-harddisk devices always get read/write-
961 commands in bypass mode. On the most recent releases of
962 the Linux IBM-SCSI-driver, the bypass command should be
963 no longer a necessary thing, if you are sure about your
965 normal This is the parameter, introduced on the 2.0.x development
966 rail by ZP Gu. This parameter defines the SCSI-device
967 scan order in the new industry standard. This means, that
968 the first SCSI-device is the one with the lowest pun.
969 E.g. harddisk at pun=0 is scanned before harddisk at
970 pun=6, which means, that harddisk at pun=0 gets sda
971 and the one at pun=6 gets sdb.
972 ansi The ANSI-standard for the right scan order, as done by
973 IBM, Microware and Microsoft, scans SCSI-devices starting
974 at the highest pun, which means, that e.g. harddisk at
975 pun=6 gets sda and a harddisk at pun=0 gets sdb. If you
976 like to have the same SCSI-device order, as in DOS, OS-9
977 or OS/2, just use this parameter.
978 fast SCSI-I/O in synchronous mode is done at 5 MHz for IBM-
979 SCSI-devices. SCSI-2 Fast/Wide Adapter/A external bus
980 should then run at 10 MHz if Fast-SCSI is enabled,
981 and at 5 MHz if Fast-SCSI is disabled on the external
982 bus. This is the default setting when nothing is
984 medium Synchronous rate is at 50% approximately, which means
985 2.5 MHz for IBM SCSI-adapters and 5.0 MHz for F/W ext.
986 SCSI-bus (when Fast-SCSI speed enabled on external bus).
987 slow The slowest possible synchronous transfer rate is set.
988 This means 1.82 MHz for IBM SCSI-adapters and 2.0 MHz
989 for F/W external bus at Fast-SCSI speed on the external
992 A further option is that you can force the SCSI-driver to accept a SCSI-
993 subsystem at a certain I/O-address with a predefined adapter PUN. This
997 commandN+1 = adapter PUN
999 e.g. ibmmcascsi=0x3540,7 will force the driver to detect a SCSI-subsystem
1000 at I/O-address 0x3540 with adapter PUN 7.
1004 ibmmcascsi=adisplay,bypass
1006 This will use the advanced display mode for the model 95 LED display and
1007 every SCSI-command passed to an attached device will get bypassed in order
1008 not to use any of the subsystem built-in commands.
1010 ibmmcascsi=display,0x3558,7
1012 This will activate the default display mode for the model 95 LED display
1013 and will force the driver to accept a SCSI-subsystem at I/O-base 0x3558
1018 The following FAQs should help you to solve some major problems with this
1021 Q: "Reset SCSI-devices at boottime" halts the system at boottime, why?
1022 A: This is only tested with the IBM SCSI Adapter w/cache. It is not
1023 yet prooved to run on other adapters, however you may be lucky.
1024 In version 3.1d this has been hugely improved and should work better,
1025 now. Normally you really won't need to activate this flag in the
1026 kernel configuration, as all post 1989 SCSI-devices should accept
1027 the reset-signal, when the computer is switched on. The SCSI-
1028 subsystem generates this reset while beeing initialized. This flag
1029 is really reserved for users with very old, very strange or self-made
1031 Q: Why is the SCSI-order of my drives mirrored to the device-order
1032 seen from OS/2 or DOS ?
1033 A: It depends on the operating system, if it looks at the devices in
1034 ANSI-SCSI-standard (starting from pun 6 and going down to pun 0) or
1035 if it just starts at pun 0 and counts up. If you want to be conform
1036 with OS/2 and DOS, you have to activate this flag in the kernel
1037 configuration or you should set 'ansi' as parameter for the kernel.
1038 The parameter 'normal' sets the new industry standard, starting
1039 from pun 0, scanning up to pun 6. This allows you to change your
1040 opinion still after having already compiled the kernel.
1041 Q: Why I cannot find the IBM MCA SCSI support in the config menue?
1042 A: You have to activate MCA bus support, first.
1043 Q: Where can I find the latest info about this driver?
1044 A: See the file MAINTAINERS for the current WWW-address, which offers
1045 updates, info and Q/A lists. At this files' origin, the webaddress
1046 was: http://www.uni-mainz.de/~langm000/linux.html
1047 Q: My SCSI-adapter is not recognized by the driver, what can I do?
1048 A: Just force it to be recognized by kernel parameters. See section 5.1.
1049 Q: The driver screws up, if it starts to probe SCSI-devices, is there
1051 A: Yes, that was some recognition problem of the correct SCSI-adapter
1052 and its I/O base addresses. Upgrade your driver to the latest release
1053 and it should be fine again.
1054 Q: I get a message: panic IBM MCA SCSI: command error .... , what can
1056 A: Previously, I followed the way by ignoring command errors by using
1057 ibmmcascsi=forgiveall, but this command no longer exists and is
1058 obsolete. If such a problem appears, it is caused by some segmentation
1059 fault of the driver, which maps to some unallowed area. The latest
1060 version of the driver should be ok, as most bugs have been solved.
1061 Q: There are still kernel panics, even after having set
1062 ibmmcascsi=forgiveall. Are there other possibilities to prevent
1064 A: No, get just the latest release of the driver and it should work
1065 better and better with increasing version number. Forget about this
1066 ibmmcascsi=forgiveall, as also ignorecmd are obsolete.!
1067 Q: Linux panics or stops without any comment, but it is probable, that my
1068 harddisk(s) have bad blocks.
1069 A: Sorry, the bad-block handling is still a feeble point of this driver,
1070 but is on the schedule for development in the near future.
1071 Q: Linux panics while dynamically assigning SCSI-ids or ldns.
1072 A: If you disconnect a SCSI-device from the machine, while Linux is up
1073 and the driver uses dynamical reassignment of logical device numbers
1074 (ldn), it really gets "angry" if it won't find devices, that were still
1075 present at boottime and stops Linux.
1076 Q: The system does not recover after an abort-command has been generated.
1077 A: This is regrettably true, as it is not yet understood, why the
1078 SCSI-adapter does really NOT generate any interrupt at the end of
1079 the abort-command. As no interrupt is generated, the abort command
1080 cannot get finished and the system hangs, sorry, but checks are
1081 running to hunt down this problem. If there is a real pending command,
1082 the interrupt MUST get generated after abort. In this case, it
1084 Q: The system gets in bad shape after a SCSI-reset, is this known?
1085 A: Yes, as there are a lot of prescriptions (see the Linux Hackers'
1086 Guide) what has to be done for reset, we still share the bad shape of
1087 the reset functions with all other low level SCSI-drivers.
1088 Astonishingly, reset works in most cases quite ok, but the harddisks
1089 won't run in synchonous mode anymore after a reset, until you reboot.
1090 Q: Why does my XXX w/Cache adapter not use read-prefetch?
1091 A: Ok, that is not completely possible. If a cache is present, the
1092 adapter tries to use it internally. Explicitly, one can use the cache
1093 with a read prefetch command, maybe in future, but this requires
1094 some major overhead of SCSI-commands that risks the performance to
1095 go down more than it gets improved. Tests with that are running.
1096 Q: I have a IBM SCSI-2 Fast/Wide adapter, it boots in some way and hangs.
1097 A: Yes, that is understood, as for sure, your SCSI-2 Fast/Wide adapter
1098 was in such a case recognized as integrated SCSI-adapter or something
1099 else, but not as the correct adapter. As the I/O-ports get assigned
1100 wrongly by that reason, the system should crash in most cases. You
1101 should upgrade to the latest release of the SCSI-driver. The
1102 recommended version is 3.2 or later. Here, the F/W support is in
1103 a stable and reliable condition. Wide-addressing is in addition
1105 Q: I get a Ooops message and something like "killing interrupt".
1106 A: The reason for this is that the IBM SCSI-subsystem only sends a
1107 termination status back, if some error appeared. In former releases
1108 of the driver, it was not checked, if the termination status block
1109 is NULL. From version 3.2, it is taken care of this.
1110 Q: I have a F/W adapter and the driver sees my internal SCSI-devices,
1111 but ignores the external ones.
1112 A: Select combined busmode in the config-program and check for that
1113 no SCSI-id on the external devices appears on internal devices.
1114 Reboot afterwards. Dual busmode is supported, but works only for the
1115 internal bus, yet. External bus is still ignored. Take care for your
1116 SCSI-ids. If combined bus-mode is activated, on some adapters,
1117 the wide-addressing is not possible, so devices with ids between 8
1118 and 15 get ignored by the driver & adapter!
1119 Q: I have a 9595 and I get a NMI during heavy SCSI I/O e.g. during fsck.
1120 A COMMAND ERROR is reported and characters on the screen are missing.
1121 Warm reboot is not possible. Things look like quite weired.
1122 A: Check the processor type of your 9595. If you have an 80486 or 486DX-2
1123 processor complex on your mainboard and you compiled a kernel that
1124 supports 80386 processors, it is possible, that the kernel cannot
1125 keep track of the PS/2 interrupt handling and stops on an NMI. Just
1126 compile a kernel for the correct processor type of your PS/2 and
1127 everything should be fine. This is necessary even if one assumes,
1128 that some 80486 system should be downward compatible to 80386
1133 If you really find bugs in the sourcecode or the driver will successfully
1134 refuse to work on your machine, you should send a bug report to me. The
1135 best for this is to follow the instructions on the WWW-page for this
1136 driver. Fill out the bug-report form, placed on the WWW-page and ship it,
1137 so the bugs can be taken into account with maximum efforts. But, please
1138 do not send bug reports about this driver to Linus Torvalds or Leonard
1139 Zubkoff, as Linus is burried in E-Mail and Leonard is supervising all
1140 SCSI-drivers and won't have the time left to look inside every single
1141 driver to fix a bug and especially DO NOT send modified code to Linus
1142 Torvalds or Alan J. Cox which has not been checked here!!! They are both
1143 quite burried in E-mail (as me, sometimes, too) and one should first check
1144 for problems on my local teststand. Recently, I got a lot of
1145 bugreports for errors in the ibmmca.c code, which I could not imagine, but
1146 a look inside some Linux-distribution showed me quite often some modified
1147 code, which did no longer work on most other machines than the one of the
1148 modifier. Ok, so now that there is maintenance service available for this
1149 driver, please use this address first in order to keep the level of
1150 confusion low. Thank you!
1152 When you get a SCSI-error message that panics your system, a list of
1153 register-entries of the SCSI-subsystem is shown (from Version 3.1d). With
1154 this list, it is very easy for the maintainer to localize the problem in
1155 the driver or in the configuration of the user. Please write down all the
1156 values from this report and send them to the maintainer. This would really
1157 help a lot and makes life easier concerning misunderstandings.
1159 Use the bug-report form (see 5.4 for its address) to send all the bug-
1160 stuff to the maintainer or write e-mail with the values from the table.
1162 5.4 Support WWW-page
1163 --------------------
1164 The address of the IBM SCSI-subsystem supporting WWW-page is:
1166 http://www.uni-mainz.de/~langm000/linux.html
1168 Here you can find info about the background of this driver, patches,
1169 troubleshooting support, news and a bugreport form. Please check that
1170 WWW-page regularly for latest hints.
1172 For the bugreport, please fill out the formular on the corresponding
1173 WWW-page. Read the dedicated instructions and write as much as you
1174 know about your problem. If you do not like such formulars, please send
1175 some e-mail directly, but at least with the same information as required by
1178 If you have extensive bugreports, including Ooops messages and
1179 screen-shots, please feel free to send it directly to the address
1180 of the maintainer, too. The current address of the maintainer is:
1182 Michael Lang <langa2@kph.uni-mainz.de>
1186 IBM Corp., "Update for the PS/2 Hardware Interface Technical Reference,
1187 Common Interfaces", Armonk, September 1991, PN 04G3281,
1188 (available in the U.S. for $21.75 at 1-800-IBM-PCTB or in Germany for
1189 around 40,-DM at "Hallo IBM").
1191 IBM Corp., "Personal System/2 Micro Channel SCSI
1192 Adapter with Cache Technical Reference", Armonk, March 1990, PN 68X2365.
1194 IBM Corp., "Personal System/2 Micro Channel SCSI
1195 Adapter Technical Reference", Armonk, March 1990, PN 68X2397.
1197 IBM Corp., "SCSI-2 Fast/Wide Adapter/A Technical Reference - Dual Bus",
1198 Armonk, March 1994, PN 83G7545.
1200 Friedhelm Schmidt, "SCSI-Bus und IDE-Schnittstelle - Moderne Peripherie-
1201 Schnittstellen: Hardware, Protokollbeschreibung und Anwendung", 2. Aufl.
1202 Addison Wesley, 1996.
1204 Michael K. Johnson, "The Linux Kernel Hackers' Guide", Version 0.6, Chapel
1205 Hill - North Carolina, 1995
1207 Andreas Kaiser, "SCSI TAPE BACKUP for OS/2 2.0", Version 2.12, Stuttgart
1210 Helmut Rompel, "IBM Computerwelt GUIDE", What is what bei IBM., Systeme *
1211 Programme * Begriffe, IWT-Verlag GmbH - Muenchen, 1988
1218 who already a long time ago gave me the old code from the
1219 SCSI-driver in order to get it running for some old machine
1222 who wrote the first release of the IBM SCSI-subsystem driver.
1224 who for a long time maintained MCA-Linux and the SCSI-driver
1225 in the beginning. Chris, wherever you are: Cheers to you!
1227 with whom in the 2.1.x times, I had a quite fruitful
1228 cooperation to get the driver running as a module and to get
1229 it running with multiple SCSI-adapters.
1231 for his excellent maintenance of the MCA-stuff and the quite
1232 detailed bug reports and ideas for this driver (and his
1235 for his bugreports and his bold activities in cross-checking
1236 the driver-code with his teststand.
1238 7.2 Sponsors & Supporters
1239 -------------------------
1241 IBM-Deutschland GmbH
1242 the service of IBM-Deutschland for customers. Their E-Mail
1243 service is unbeatable. Whatever old stuff I asked for, I
1244 always got some helpful answers.
1246 IBM Klub - Sparte IBM Geschichte, Sindelfingen
1247 for sending me a copy of the w/Cache manual from the
1248 IBM-Deutschland archives.
1250 for his extensive hardware donations which allows me today
1251 still to test the driver in various constellations.
1253 for his very kind sponsoring.
1256 for support by shipping me an IBM SCSI-2 Fast/Wide manual.
1257 In addition, the contribution of various hardware is quite
1258 decessive and will make it possible to add FWSR (RAID)
1259 adapter support to the driver in the near future! So,
1260 complaints about no RAID support won't remain forever.
1261 Yes, folks, that is no joke, RAID support is going to rise!
1263 for the great deal we made about a model 9595 and the nice
1264 surrounding equipment and the cool trip to Mannheim
1265 second-hand computer market.
1267 for his direct shipment of a SCSI F/W adapter, which allowed
1268 me immediately on the first stage to try it on model 8557
1269 together with onboard SCSI adapter and some SCSI w/Cache.
1271 for his support by memory and an IBM SCSI-adapter. Collecting
1272 all this together now allows me to try really things with
1273 the driver at maximum load and variety on various models in
1274 a very quick and efficient way.
1276 for his model 30, which serves me as part of my teststand
1277 and his cool remark about how you make an ordinary diskette
1278 drive working and how to connect it to an IBM-diskette port.
1279 Johannes Gutenberg-University, Mainz &
1280 Institut fuer Kernphysik, MAMI
1281 for the offered space, the link, placed on the central
1282 homepage and the space to store and offer the driver and
1283 related material and the free working times, which allow
1284 me to answer all your e-mail.
1288 IBM, PS/2, OS/2, Microchannel are registered trademarks of International
1289 Business Machines Corporation
1291 MS-DOS is a registered trademark of Microsoft Corporation
1293 Microware, OS-9 are registered trademarks of Microware Systems
1297 Beside the GNU public license and the dependant disclaimers and disclaimers
1298 concerning the Linux-kernel in special, this SCSI-driver comes without any
1299 warranty. Its functionality is tested as good as possible on certain
1300 machines and combinations of computer hardware, which does not exclude,
1301 that dataloss or severe damage of hardware is possible while using this
1302 part of software on some arbitrary computer hardware or in combination
1303 with other software packages. It is highly recommended to make backup
1304 copies of your data before using this software.
1306 This driver supports hardware, produced by International Business Machines
1311 (langa2@kph.uni-mainz.de)