Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 1 | |
| 2 | Information regarding the Enhanced IDE drive in Linux 2.6 |
| 3 | |
| 4 | ============================================================================== |
| 5 | |
| 6 | |
| 7 | 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. |
| 10 | |
| 11 | |
| 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 |
| 33 | *** interfaces, one *MUST* use the "ide0=cmd640_vlb" kernel option. |
| 34 | *** |
| 35 | *** Use of the "serialize" option is no longer necessary. |
| 36 | |
| 37 | ================================================================================ |
| 38 | Common 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 | |
| 54 | This is the multiple IDE interface driver, as evolved from hd.c. |
| 55 | |
| 56 | It supports up to 9 IDE interfaces per default, on one or more IRQs (usually |
| 57 | 14 & 15). There can be up to two drives per interface, as per the ATA-6 spec. |
| 58 | |
| 59 | Primary: ide0, port 0x1f0; major=3; hda is minor=0; hdb is minor=64 |
| 60 | Secondary: ide1, port 0x170; major=22; hdc is minor=0; hdd is minor=64 |
| 61 | Tertiary: ide2, port 0x1e8; major=33; hde is minor=0; hdf is minor=64 |
| 62 | Quaternary: ide3, port 0x168; major=34; hdg is minor=0; hdh is minor=64 |
| 63 | fifth.. ide4, usually PCI, probed |
| 64 | sixth.. ide5, usually PCI, probed |
| 65 | |
| 66 | To access devices on interfaces > ide0, device entries please make sure that |
| 67 | device files for them are present in /dev. If not, please create such |
| 68 | entries, by using /dev/MAKEDEV. |
| 69 | |
| 70 | This driver automatically probes for most IDE interfaces (including all PCI |
| 71 | ones), for the drives/geometries attached to those interfaces, and for the IRQ |
| 72 | lines being used by the interfaces (normally 14, 15 for ide0/ide1). |
| 73 | |
| 74 | For special cases, interfaces may be specified using kernel "command line" |
| 75 | options. For example, |
| 76 | |
| 77 | ide3=0x168,0x36e,10 /* ioports 0x168-0x16f,0x36e, irq 10 */ |
| 78 | |
| 79 | Normally the irq number need not be specified, as ide.c will probe for it: |
| 80 | |
| 81 | ide3=0x168,0x36e /* ioports 0x168-0x16f,0x36e */ |
| 82 | |
| 83 | The standard port, and irq values are these: |
| 84 | |
| 85 | ide0=0x1f0,0x3f6,14 |
| 86 | ide1=0x170,0x376,15 |
| 87 | ide2=0x1e8,0x3ee,11 |
| 88 | ide3=0x168,0x36e,10 |
| 89 | |
| 90 | Note that the first parameter reserves 8 contiguous ioports, whereas the |
| 91 | second value denotes a single ioport. If in doubt, do a 'cat /proc/ioports'. |
| 92 | |
| 93 | In all probability the device uses these ports and IRQs if it is attached |
| 94 | to the appropriate ide channel. Pass the parameter for the correct ide |
| 95 | channel to the kernel, as explained above. |
| 96 | |
| 97 | Any number of interfaces may share a single IRQ if necessary, at a slight |
| 98 | performance penalty, whether on separate cards or a single VLB card. |
| 99 | The IDE driver automatically detects and handles this. However, this may |
| 100 | or may not be harmful to your hardware.. two or more cards driving the same IRQ |
| 101 | can potentially burn each other's bus driver, though in practice this |
| 102 | seldom occurs. Be careful, and if in doubt, don't do it! |
| 103 | |
| 104 | Drives are normally found by auto-probing and/or examining the CMOS/BIOS data. |
| 105 | For really weird situations, the apparent (fdisk) geometry can also be specified |
| 106 | on the kernel "command line" using LILO. The format of such lines is: |
| 107 | |
| 108 | hdx=cyls,heads,sects,wpcom,irq |
| 109 | or hdx=cdrom |
| 110 | |
| 111 | where hdx can be any of hda through hdh, Three values are required |
| 112 | (cyls,heads,sects). For example: |
| 113 | |
| 114 | hdc=1050,32,64 hdd=cdrom |
| 115 | |
| 116 | either {hda,hdb} or {hdc,hdd}. The results of successful auto-probing may |
| 117 | override the physical geometry/irq specified, though the "original" geometry |
| 118 | may be retained as the "logical" geometry for partitioning purposes (fdisk). |
| 119 | |
| 120 | If the auto-probing during boot time confuses a drive (ie. the drive works |
| 121 | with hd.c but not with ide.c), then an command line option may be specified |
| 122 | for each drive for which you'd like the drive to skip the hardware |
| 123 | probe/identification sequence. For example: |
| 124 | |
| 125 | hdb=noprobe |
| 126 | or |
| 127 | hdc=768,16,32 |
| 128 | hdc=noprobe |
| 129 | |
| 130 | Note that when only one IDE device is attached to an interface, it should be |
| 131 | jumpered as "single" or "master", *not* "slave". Many folks have had |
| 132 | "trouble" with cdroms because of this requirement, so the driver now probes |
| 133 | for both units, though success is more likely when the drive is jumpered |
| 134 | correctly. |
| 135 | |
| 136 | Courtesy of Scott Snyder and others, the driver supports ATAPI cdrom drives |
| 137 | such as the NEC-260 and the new MITSUMI triple/quad speed drives. |
| 138 | Such drives will be identified at boot time, just like a hard disk. |
| 139 | |
| 140 | If for some reason your cdrom drive is *not* found at boot time, you can force |
| 141 | the probe to look harder by supplying a kernel command line parameter |
| 142 | via LILO, such as: |
| 143 | |
| 144 | hdc=cdrom /* hdc = "master" on second interface */ |
| 145 | or |
| 146 | hdd=cdrom /* hdd = "slave" on second interface */ |
| 147 | |
| 148 | For example, a GW2000 system might have a hard drive on the primary |
| 149 | interface (/dev/hda) and an IDE cdrom drive on the secondary interface |
| 150 | (/dev/hdc). To mount a CD in the cdrom drive, one would use something like: |
| 151 | |
| 152 | ln -sf /dev/hdc /dev/cdrom |
| 153 | mkdir /mnt/cdrom |
| 154 | mount /dev/cdrom /mnt/cdrom -t iso9660 -o ro |
| 155 | |
| 156 | If, after doing all of the above, mount doesn't work and you see |
| 157 | errors from the driver (with dmesg) complaining about `status=0xff', |
| 158 | this means that the hardware is not responding to the driver's attempts |
| 159 | to read it. One of the following is probably the problem: |
| 160 | |
| 161 | - Your hardware is broken. |
| 162 | |
| 163 | - You are using the wrong address for the device, or you have the |
| 164 | drive jumpered wrong. Review the configuration instructions above. |
| 165 | |
| 166 | - Your IDE controller requires some nonstandard initialization sequence |
| 167 | before it will work properly. If this is the case, there will often |
| 168 | be a separate MS-DOS driver just for the controller. IDE interfaces |
| 169 | on sound cards usually fall into this category. Such configurations |
| 170 | can often be made to work by first booting MS-DOS, loading the |
| 171 | appropriate drivers, and then warm-booting linux (without powering |
| 172 | off). This can be automated using loadlin in the MS-DOS autoexec. |
| 173 | |
| 174 | If you always get timeout errors, interrupts from the drive are probably |
| 175 | not making it to the host. Check how you have the hardware jumpered |
| 176 | and make sure it matches what the driver expects (see the configuration |
| 177 | instructions above). If you have a PCI system, also check the BIOS |
| 178 | setup; I've had one report of a system which was shipped with IRQ 15 |
| 179 | disabled by the BIOS. |
| 180 | |
| 181 | The kernel is able to execute binaries directly off of the cdrom, |
| 182 | provided it is mounted with the default block size of 1024 (as above). |
| 183 | |
| 184 | Please pass on any feedback on any of this stuff to the maintainer, |
| 185 | whose address can be found in linux/MAINTAINERS. |
| 186 | |
| 187 | Note that if BOTH hd.c and ide.c are configured into the kernel, |
| 188 | hd.c will normally be allowed to control the primary IDE interface. |
| 189 | This is useful for older hardware that may be incompatible with ide.c, |
| 190 | and still allows newer hardware to run on the 2nd/3rd/4th IDE ports |
| 191 | under control of ide.c. To have ide.c also "take over" the primary |
| 192 | IDE port in this situation, use the "command line" parameter: ide0=0x1f0 |
| 193 | |
| 194 | The IDE driver is modularized. The high level disk/CD-ROM/tape/floppy |
| 195 | drivers can always be compiled as loadable modules, the chipset drivers |
| 196 | can only be compiled into the kernel, and the core code (ide.c) can be |
| 197 | compiled as a loadable module provided no chipset support is needed. |
| 198 | |
| 199 | When using ide.c as a module in combination with kmod, add: |
| 200 | |
| 201 | alias block-major-3 ide-probe |
| 202 | |
| 203 | to /etc/modprobe.conf. |
| 204 | |
| 205 | When ide.c is used as a module, you can pass command line parameters to the |
| 206 | driver using the "options=" keyword to insmod, while replacing any ',' with |
| 207 | ';'. For example: |
| 208 | |
| 209 | insmod ide.o options="ide0=serialize ide1=serialize ide2=0x1e8;0x3ee;11" |
| 210 | |
| 211 | |
| 212 | ================================================================================ |
| 213 | |
| 214 | Summary of ide driver parameters for kernel command line |
| 215 | -------------------------------------------------------- |
| 216 | |
| 217 | "hdx=" is recognized for all "x" from "a" to "h", such as "hdc". |
| 218 | |
| 219 | "idex=" is recognized for all "x" from "0" to "3", such as "ide1". |
| 220 | |
| 221 | "hdx=noprobe" : drive may be present, but do not probe for it |
| 222 | |
| 223 | "hdx=none" : drive is NOT present, ignore cmos and do not probe |
| 224 | |
| 225 | "hdx=nowerr" : ignore the WRERR_STAT bit on this drive |
| 226 | |
| 227 | "hdx=cdrom" : drive is present, and is a cdrom drive |
| 228 | |
| 229 | "hdx=cyl,head,sect" : disk drive is present, with specified geometry |
| 230 | |
| 231 | "hdx=remap" : remap access of sector 0 to sector 1 (for EZDrive) |
| 232 | |
| 233 | "hdx=remap63" : remap the drive: add 63 to all sector numbers |
| 234 | (for DM OnTrack) |
| 235 | |
| 236 | "hdx=autotune" : driver will attempt to tune interface speed |
| 237 | to the fastest PIO mode supported, |
| 238 | if possible for this drive only. |
| 239 | Not fully supported by all chipset types, |
| 240 | and quite likely to cause trouble with |
| 241 | older/odd IDE drives. |
| 242 | |
| 243 | "hdx=swapdata" : when the drive is a disk, byte swap all data |
| 244 | |
| 245 | "hdx=bswap" : same as above.......... |
| 246 | |
| 247 | "hdx=scsi" : the return of the ide-scsi flag, this is useful for |
| 248 | allowing ide-floppy, ide-tape, and ide-cdrom|writers |
| 249 | to use ide-scsi emulation on a device specific option. |
| 250 | |
| 251 | "idebus=xx" : inform IDE driver of VESA/PCI bus speed in MHz, |
| 252 | where "xx" is between 20 and 66 inclusive, |
| 253 | used when tuning chipset PIO modes. |
| 254 | For PCI bus, 25 is correct for a P75 system, |
| 255 | 30 is correct for P90,P120,P180 systems, |
| 256 | and 33 is used for P100,P133,P166 systems. |
| 257 | If in doubt, use idebus=33 for PCI. |
| 258 | As for VLB, it is safest to not specify it. |
| 259 | Bigger values are safer than smaller ones. |
| 260 | |
| 261 | "idex=noprobe" : do not attempt to access/use this interface |
| 262 | |
| 263 | "idex=base" : probe for an interface at the addr specified, |
| 264 | where "base" is usually 0x1f0 or 0x170 |
| 265 | and "ctl" is assumed to be "base"+0x206 |
| 266 | |
| 267 | "idex=base,ctl" : specify both base and ctl |
| 268 | |
| 269 | "idex=base,ctl,irq" : specify base, ctl, and irq number |
| 270 | |
| 271 | "idex=autotune" : driver will attempt to tune interface speed |
| 272 | to the fastest PIO mode supported, |
| 273 | for all drives on this interface. |
| 274 | Not fully supported by all chipset types, |
| 275 | and quite likely to cause trouble with |
| 276 | older/odd IDE drives. |
| 277 | |
| 278 | "idex=noautotune" : driver will NOT attempt to tune interface speed |
| 279 | This is the default for most chipsets, |
| 280 | except the cmd640. |
| 281 | |
| 282 | "idex=serialize" : do not overlap operations on idex. Please note |
| 283 | that you will have to specify this option for |
Matt LaPlante | d6bc8ac | 2006-10-03 22:54:15 +0200 | [diff] [blame] | 284 | both the respective primary and secondary channel |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 285 | to take effect. |
| 286 | |
| 287 | "idex=four" : four drives on idex and ide(x^1) share same ports |
| 288 | |
| 289 | "idex=reset" : reset interface after probe |
| 290 | |
| 291 | "idex=dma" : automatically configure/use DMA if possible. |
| 292 | |
| 293 | "idex=ata66" : informs the interface that it has an 80c cable |
| 294 | for chipsets that are ATA-66 capable, but the |
| 295 | ability to bit test for detection is currently |
| 296 | unknown. |
| 297 | |
| 298 | "ide=reverse" : formerly called to pci sub-system, but now local. |
| 299 | |
| 300 | "ide=nodma" : disable DMA globally for the IDE subsystem. |
| 301 | |
| 302 | The following are valid ONLY on ide0, which usually corresponds |
| 303 | to the first ATA interface found on the particular host, and the defaults for |
| 304 | the base,ctl ports must not be altered. |
| 305 | |
| 306 | "ide0=dtc2278" : probe/support DTC2278 interface |
| 307 | "ide0=ht6560b" : probe/support HT6560B interface |
| 308 | "ide0=cmd640_vlb" : *REQUIRED* for VLB cards with the CMD640 chip |
| 309 | (not for PCI -- automatically detected) |
| 310 | "ide0=qd65xx" : probe/support qd65xx interface |
| 311 | "ide0=ali14xx" : probe/support ali14xx chipsets (ALI M1439/M1443/M1445) |
| 312 | "ide0=umc8672" : probe/support umc8672 chipsets |
| 313 | |
| 314 | "ide=doubler" : probe/support IDE doublers on Amiga |
| 315 | |
| 316 | There may be more options than shown -- use the source, Luke! |
| 317 | |
| 318 | Everything else is rejected with a "BAD OPTION" message. |
| 319 | |
| 320 | ================================================================================ |
| 321 | |
| 322 | IDE ATAPI streaming tape driver |
| 323 | ------------------------------- |
| 324 | |
| 325 | This driver is a part of the Linux ide driver and works in co-operation |
| 326 | with linux/drivers/block/ide.c. |
| 327 | |
| 328 | The driver, in co-operation with ide.c, basically traverses the |
| 329 | request-list for the block device interface. The character device |
| 330 | interface, on the other hand, creates new requests, adds them |
| 331 | to the request-list of the block device, and waits for their completion. |
| 332 | |
| 333 | Pipelined operation mode is now supported on both reads and writes. |
| 334 | |
| 335 | The block device major and minor numbers are determined from the |
| 336 | tape's relative position in the ide interfaces, as explained in ide.c. |
| 337 | |
| 338 | The character device interface consists of the following devices: |
| 339 | |
| 340 | ht0 major 37, minor 0 first IDE tape, rewind on close. |
| 341 | ht1 major 37, minor 1 second IDE tape, rewind on close. |
| 342 | ... |
| 343 | nht0 major 37, minor 128 first IDE tape, no rewind on close. |
| 344 | nht1 major 37, minor 129 second IDE tape, no rewind on close. |
| 345 | ... |
| 346 | |
| 347 | Run /dev/MAKEDEV to create the above entries. |
| 348 | |
| 349 | The general magnetic tape commands compatible interface, as defined by |
| 350 | include/linux/mtio.h, is accessible through the character device. |
| 351 | |
| 352 | General ide driver configuration options, such as the interrupt-unmask |
| 353 | flag, can be configured by issuing an ioctl to the block device interface, |
| 354 | as any other ide device. |
| 355 | |
| 356 | Our own ide-tape ioctl's can be issued to either the block device or |
| 357 | the character device interface. |
| 358 | |
| 359 | Maximal throughput with minimal bus load will usually be achieved in the |
| 360 | following scenario: |
| 361 | |
| 362 | 1. ide-tape is operating in the pipelined operation mode. |
| 363 | 2. No buffering is performed by the user backup program. |
| 364 | |
| 365 | |
| 366 | |
| 367 | ================================================================================ |
| 368 | |
| 369 | Some Terminology |
| 370 | ---------------- |
| 371 | IDE = Integrated Drive Electronics, meaning that each drive has a built-in |
| 372 | controller, which is why an "IDE interface card" is not a "controller card". |
| 373 | |
| 374 | ATA = AT (the old IBM 286 computer) Attachment Interface, a draft American |
| 375 | National Standard for connecting hard drives to PCs. This is the official |
| 376 | name for "IDE". |
| 377 | |
| 378 | The latest standards define some enhancements, known as the ATA-6 spec, |
| 379 | which grew out of vendor-specific "Enhanced IDE" (EIDE) implementations. |
| 380 | |
| 381 | ATAPI = ATA Packet Interface, a new protocol for controlling the drives, |
| 382 | similar to SCSI protocols, created at the same time as the ATA2 standard. |
| 383 | ATAPI is currently used for controlling CDROM, TAPE and FLOPPY (ZIP or |
| 384 | LS120/240) devices, removable R/W cartridges, and for high capacity hard disk |
| 385 | drives. |
| 386 | |
| 387 | mlord@pobox.com |
| 388 | -- |
| 389 | |
| 390 | Wed Apr 17 22:52:44 CEST 2002 edited by Marcin Dalecki, the current |
| 391 | maintainer. |
| 392 | |
Matt LaPlante | 4ae0edc | 2006-11-30 04:58:40 +0100 | [diff] [blame] | 393 | Wed Aug 20 22:31:29 CEST 2003 updated ide boot options to current ide.c |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 394 | comments at 2.6.0-test4 time. Maciej Soltysiak <solt@dns.toxicfilms.tv> |