blob: 486c699f4aea4722bb822a7af7ad12053f5c321c [file] [log] [blame]
Linus Torvalds1da177e2005-04-16 15:20:36 -07001
2 Information regarding the Enhanced IDE drive in Linux 2.6
3
4==============================================================================
5
Randy Dunlap331a5ad2008-03-07 21:53:50 +01006
Linus Torvalds1da177e2005-04-16 15:20:36 -07007 The hdparm utility can be used to control various IDE features on a
8 running system. It is packaged separately. Please Look for it on popular
9 linux FTP sites.
Randy Dunlap331a5ad2008-03-07 21:53:50 +010010
Linus Torvalds1da177e2005-04-16 15:20:36 -070011
12
13*** IMPORTANT NOTICES: BUGGY IDE CHIPSETS CAN CORRUPT DATA!!
14*** =================
15*** PCI versions of the CMD640 and RZ1000 interfaces are now detected
16*** automatically at startup when PCI BIOS support is configured.
17***
18*** Linux disables the "prefetch" ("readahead") mode of the RZ1000
19*** to prevent data corruption possible due to hardware design flaws.
20***
21*** For the CMD640, linux disables "IRQ unmasking" (hdparm -u1) on any
22*** drive for which the "prefetch" mode of the CMD640 is turned on.
23*** If "prefetch" is disabled (hdparm -p8), then "IRQ unmasking" can be
24*** used again.
25***
26*** For the CMD640, linux disables "32bit I/O" (hdparm -c1) on any drive
27*** for which the "prefetch" mode of the CMD640 is turned off.
28*** If "prefetch" is enabled (hdparm -p9), then "32bit I/O" can be
29*** used again.
30***
31*** The CMD640 is also used on some Vesa Local Bus (VLB) cards, and is *NOT*
32*** automatically detected by Linux. For safe, reliable operation with such
Bartlomiej Zolnierkiewiczade2daf2008-01-26 20:13:07 +010033*** interfaces, one *MUST* use the "cmd640.probe_vlb" kernel option.
Linus Torvalds1da177e2005-04-16 15:20:36 -070034***
35*** Use of the "serialize" option is no longer necessary.
36
37================================================================================
38Common pitfalls:
39
40- 40-conductor IDE cables are capable of transferring data in DMA modes up to
41 udma2, but no faster.
42
43- If possible devices should be attached to separate channels if they are
44 available. Typically the disk on the first and CD-ROM on the second.
45
46- If you mix devices on the same cable, please consider using similar devices
47 in respect of the data transfer mode they support.
48
49- Even better try to stick to the same vendor and device type on the same
50 cable.
51
52================================================================================
53
Randy Dunlap331a5ad2008-03-07 21:53:50 +010054This is the multiple IDE interface driver, as evolved from hd.c.
Linus Torvalds1da177e2005-04-16 15:20:36 -070055
56It supports up to 9 IDE interfaces per default, on one or more IRQs (usually
5714 & 15). There can be up to two drives per interface, as per the ATA-6 spec.
58
59Primary: ide0, port 0x1f0; major=3; hda is minor=0; hdb is minor=64
60Secondary: ide1, port 0x170; major=22; hdc is minor=0; hdd is minor=64
61Tertiary: ide2, port 0x1e8; major=33; hde is minor=0; hdf is minor=64
62Quaternary: ide3, port 0x168; major=34; hdg is minor=0; hdh is minor=64
63fifth.. ide4, usually PCI, probed
64sixth.. ide5, usually PCI, probed
65
66To access devices on interfaces > ide0, device entries please make sure that
67device files for them are present in /dev. If not, please create such
68entries, by using /dev/MAKEDEV.
69
70This driver automatically probes for most IDE interfaces (including all PCI
71ones), for the drives/geometries attached to those interfaces, and for the IRQ
72lines being used by the interfaces (normally 14, 15 for ide0/ide1).
73
Linus Torvalds1da177e2005-04-16 15:20:36 -070074Any number of interfaces may share a single IRQ if necessary, at a slight
75performance penalty, whether on separate cards or a single VLB card.
76The IDE driver automatically detects and handles this. However, this may
77or may not be harmful to your hardware.. two or more cards driving the same IRQ
78can potentially burn each other's bus driver, though in practice this
79seldom occurs. Be careful, and if in doubt, don't do it!
80
81Drives are normally found by auto-probing and/or examining the CMOS/BIOS data.
82For really weird situations, the apparent (fdisk) geometry can also be specified
83on the kernel "command line" using LILO. The format of such lines is:
84
Bartlomiej Zolnierkiewiczad1c53b2008-03-22 16:40:21 +010085 hdx=cyls,heads,sects
Linus Torvalds1da177e2005-04-16 15:20:36 -070086or hdx=cdrom
87
88where hdx can be any of hda through hdh, Three values are required
89(cyls,heads,sects). For example:
90
91 hdc=1050,32,64 hdd=cdrom
92
93either {hda,hdb} or {hdc,hdd}. The results of successful auto-probing may
94override the physical geometry/irq specified, though the "original" geometry
95may be retained as the "logical" geometry for partitioning purposes (fdisk).
96
97If the auto-probing during boot time confuses a drive (ie. the drive works
98with hd.c but not with ide.c), then an command line option may be specified
99for each drive for which you'd like the drive to skip the hardware
100probe/identification sequence. For example:
101
102 hdb=noprobe
103or
104 hdc=768,16,32
105 hdc=noprobe
106
107Note that when only one IDE device is attached to an interface, it should be
108jumpered as "single" or "master", *not* "slave". Many folks have had
109"trouble" with cdroms because of this requirement, so the driver now probes
110for both units, though success is more likely when the drive is jumpered
111correctly.
112
113Courtesy of Scott Snyder and others, the driver supports ATAPI cdrom drives
114such as the NEC-260 and the new MITSUMI triple/quad speed drives.
115Such drives will be identified at boot time, just like a hard disk.
116
117If for some reason your cdrom drive is *not* found at boot time, you can force
118the probe to look harder by supplying a kernel command line parameter
119via LILO, such as:
120
121 hdc=cdrom /* hdc = "master" on second interface */
122or
123 hdd=cdrom /* hdd = "slave" on second interface */
124
125For example, a GW2000 system might have a hard drive on the primary
126interface (/dev/hda) and an IDE cdrom drive on the secondary interface
127(/dev/hdc). To mount a CD in the cdrom drive, one would use something like:
128
129 ln -sf /dev/hdc /dev/cdrom
130 mkdir /mnt/cdrom
131 mount /dev/cdrom /mnt/cdrom -t iso9660 -o ro
132
133If, after doing all of the above, mount doesn't work and you see
134errors from the driver (with dmesg) complaining about `status=0xff',
135this means that the hardware is not responding to the driver's attempts
136to read it. One of the following is probably the problem:
137
138 - Your hardware is broken.
139
140 - You are using the wrong address for the device, or you have the
141 drive jumpered wrong. Review the configuration instructions above.
142
143 - Your IDE controller requires some nonstandard initialization sequence
144 before it will work properly. If this is the case, there will often
145 be a separate MS-DOS driver just for the controller. IDE interfaces
146 on sound cards usually fall into this category. Such configurations
147 can often be made to work by first booting MS-DOS, loading the
148 appropriate drivers, and then warm-booting linux (without powering
149 off). This can be automated using loadlin in the MS-DOS autoexec.
150
151If you always get timeout errors, interrupts from the drive are probably
152not making it to the host. Check how you have the hardware jumpered
153and make sure it matches what the driver expects (see the configuration
154instructions above). If you have a PCI system, also check the BIOS
155setup; I've had one report of a system which was shipped with IRQ 15
156disabled by the BIOS.
157
158The kernel is able to execute binaries directly off of the cdrom,
159provided it is mounted with the default block size of 1024 (as above).
160
161Please pass on any feedback on any of this stuff to the maintainer,
162whose address can be found in linux/MAINTAINERS.
163
Linus Torvalds1da177e2005-04-16 15:20:36 -0700164The IDE driver is modularized. The high level disk/CD-ROM/tape/floppy
165drivers can always be compiled as loadable modules, the chipset drivers
166can only be compiled into the kernel, and the core code (ide.c) can be
167compiled as a loadable module provided no chipset support is needed.
168
169When using ide.c as a module in combination with kmod, add:
170
171 alias block-major-3 ide-probe
172
173to /etc/modprobe.conf.
174
175When ide.c is used as a module, you can pass command line parameters to the
176driver using the "options=" keyword to insmod, while replacing any ',' with
177';'. For example:
178
Bartlomiej Zolnierkiewicz9a0e77f2008-04-18 00:46:24 +0200179 insmod ide.o options="hda=nodma hdb=nodma"
Linus Torvalds1da177e2005-04-16 15:20:36 -0700180
181
182================================================================================
183
184Summary of ide driver parameters for kernel command line
185--------------------------------------------------------
186
Bartlomiej Zolnierkiewiczad1c53b2008-03-22 16:40:21 +0100187 "hdx=" is recognized for all "x" from "a" to "u", such as "hdc".
Randy Dunlap331a5ad2008-03-07 21:53:50 +0100188
Bartlomiej Zolnierkiewiczad1c53b2008-03-22 16:40:21 +0100189 "idex=" is recognized for all "x" from "0" to "9", such as "ide1".
Linus Torvalds1da177e2005-04-16 15:20:36 -0700190
191 "hdx=noprobe" : drive may be present, but do not probe for it
Randy Dunlap331a5ad2008-03-07 21:53:50 +0100192
Linus Torvalds1da177e2005-04-16 15:20:36 -0700193 "hdx=none" : drive is NOT present, ignore cmos and do not probe
Randy Dunlap331a5ad2008-03-07 21:53:50 +0100194
Linus Torvalds1da177e2005-04-16 15:20:36 -0700195 "hdx=nowerr" : ignore the WRERR_STAT bit on this drive
Randy Dunlap331a5ad2008-03-07 21:53:50 +0100196
Linus Torvalds1da177e2005-04-16 15:20:36 -0700197 "hdx=cdrom" : drive is present, and is a cdrom drive
Randy Dunlap331a5ad2008-03-07 21:53:50 +0100198
Linus Torvalds1da177e2005-04-16 15:20:36 -0700199 "hdx=cyl,head,sect" : disk drive is present, with specified geometry
200
Linus Torvalds1da177e2005-04-16 15:20:36 -0700201 "hdx=autotune" : driver will attempt to tune interface speed
202 to the fastest PIO mode supported,
203 if possible for this drive only.
204 Not fully supported by all chipset types,
205 and quite likely to cause trouble with
206 older/odd IDE drives.
207
Bartlomiej Zolnierkiewiczc2237012007-10-16 22:29:58 +0200208 "hdx=nodma" : disallow DMA
209
Linus Torvalds1da177e2005-04-16 15:20:36 -0700210 "idebus=xx" : inform IDE driver of VESA/PCI bus speed in MHz,
211 where "xx" is between 20 and 66 inclusive,
212 used when tuning chipset PIO modes.
213 For PCI bus, 25 is correct for a P75 system,
214 30 is correct for P90,P120,P180 systems,
215 and 33 is used for P100,P133,P166 systems.
216 If in doubt, use idebus=33 for PCI.
217 As for VLB, it is safest to not specify it.
218 Bigger values are safer than smaller ones.
219
Linus Torvalds1da177e2005-04-16 15:20:36 -0700220 "idex=serialize" : do not overlap operations on idex. Please note
221 that you will have to specify this option for
Matt LaPlanted6bc8ac2006-10-03 22:54:15 +0200222 both the respective primary and secondary channel
Linus Torvalds1da177e2005-04-16 15:20:36 -0700223 to take effect.
224
Linus Torvalds1da177e2005-04-16 15:20:36 -0700225 "idex=reset" : reset interface after probe
Linus Torvalds1da177e2005-04-16 15:20:36 -0700226
227 "idex=ata66" : informs the interface that it has an 80c cable
228 for chipsets that are ATA-66 capable, but the
229 ability to bit test for detection is currently
230 unknown.
231
Linus Torvalds1da177e2005-04-16 15:20:36 -0700232 "ide=doubler" : probe/support IDE doublers on Amiga
233
234There may be more options than shown -- use the source, Luke!
235
236Everything else is rejected with a "BAD OPTION" message.
237
Bartlomiej Zolnierkiewicz84913882007-03-03 17:48:55 +0100238For legacy IDE VLB host drivers (ali14xx/dtc2278/ht6560b/qd65xx/umc8672)
239you need to explicitly enable probing by using "probe" kernel parameter,
240i.e. to enable probing for ALI M14xx chipsets (ali14xx host driver) use:
241
242* "ali14xx.probe" boot option when ali14xx driver is built-in the kernel
243
244* "probe" module parameter when ali14xx driver is compiled as module
245 ("modprobe ali14xx probe")
246
Bartlomiej Zolnierkiewiczade2daf2008-01-26 20:13:07 +0100247Also for legacy CMD640 host driver (cmd640) you need to use "probe_vlb"
248kernel paremeter to enable probing for VLB version of the chipset (PCI ones
249are detected automatically).
250
Bartlomiej Zolnierkiewiczffd4f6f2008-04-18 00:46:34 +0200251You also need to use "probe" kernel parameter for ide-4drives driver
252(support for IDE generic chipset with four drives on one port).
253
Linus Torvalds1da177e2005-04-16 15:20:36 -0700254================================================================================
255
Linus Torvalds1da177e2005-04-16 15:20:36 -0700256Some Terminology
257----------------
258IDE = Integrated Drive Electronics, meaning that each drive has a built-in
259controller, which is why an "IDE interface card" is not a "controller card".
260
261ATA = AT (the old IBM 286 computer) Attachment Interface, a draft American
262National Standard for connecting hard drives to PCs. This is the official
263name for "IDE".
264
265The latest standards define some enhancements, known as the ATA-6 spec,
266which grew out of vendor-specific "Enhanced IDE" (EIDE) implementations.
267
268ATAPI = ATA Packet Interface, a new protocol for controlling the drives,
269similar to SCSI protocols, created at the same time as the ATA2 standard.
270ATAPI is currently used for controlling CDROM, TAPE and FLOPPY (ZIP or
271LS120/240) devices, removable R/W cartridges, and for high capacity hard disk
272drives.
273
274mlord@pobox.com
275--
276
277Wed Apr 17 22:52:44 CEST 2002 edited by Marcin Dalecki, the current
278maintainer.
279
Matt LaPlante4ae0edc2006-11-30 04:58:40 +0100280Wed Aug 20 22:31:29 CEST 2003 updated ide boot options to current ide.c
Linus Torvalds1da177e2005-04-16 15:20:36 -0700281comments at 2.6.0-test4 time. Maciej Soltysiak <solt@dns.toxicfilms.tv>