Borislav Petkov | 1056971 | 2008-04-27 15:38:28 +0200 | [diff] [blame] | 1 | IDE ATAPI streaming tape driver. |
| 2 | |
| 3 | This driver is a part of the Linux ide driver. |
| 4 | |
| 5 | The driver, in co-operation with ide.c, basically traverses the |
| 6 | request-list for the block device interface. The character device |
| 7 | interface, on the other hand, creates new requests, adds them |
| 8 | to the request-list of the block device, and waits for their completion. |
| 9 | |
| 10 | The block device major and minor numbers are determined from the |
| 11 | tape's relative position in the ide interfaces, as explained in ide.c. |
| 12 | |
| 13 | The character device interface consists of the following devices: |
| 14 | |
| 15 | ht0 major 37, minor 0 first IDE tape, rewind on close. |
| 16 | ht1 major 37, minor 1 second IDE tape, rewind on close. |
| 17 | ... |
| 18 | nht0 major 37, minor 128 first IDE tape, no rewind on close. |
| 19 | nht1 major 37, minor 129 second IDE tape, no rewind on close. |
| 20 | ... |
| 21 | |
| 22 | The general magnetic tape commands compatible interface, as defined by |
| 23 | include/linux/mtio.h, is accessible through the character device. |
| 24 | |
| 25 | General ide driver configuration options, such as the interrupt-unmask |
| 26 | flag, can be configured by issuing an ioctl to the block device interface, |
| 27 | as any other ide device. |
| 28 | |
| 29 | Our own ide-tape ioctl's can be issued to either the block device or |
| 30 | the character device interface. |
| 31 | |
| 32 | Maximal throughput with minimal bus load will usually be achieved in the |
| 33 | following scenario: |
| 34 | |
| 35 | 1. ide-tape is operating in the pipelined operation mode. |
| 36 | 2. No buffering is performed by the user backup program. |
| 37 | |
| 38 | Testing was done with a 2 GB CONNER CTMA 4000 IDE ATAPI Streaming Tape Drive. |
| 39 | |
| 40 | Here are some words from the first releases of hd.c, which are quoted |
| 41 | in ide.c and apply here as well: |
| 42 | |
| 43 | | Special care is recommended. Have Fun! |
| 44 | |
| 45 | Possible improvements: |
| 46 | |
| 47 | 1. Support for the ATAPI overlap protocol. |
| 48 | |
| 49 | In order to maximize bus throughput, we currently use the DSC |
| 50 | overlap method which enables ide.c to service requests from the |
| 51 | other device while the tape is busy executing a command. The |
| 52 | DSC overlap method involves polling the tape's status register |
| 53 | for the DSC bit, and servicing the other device while the tape |
| 54 | isn't ready. |
| 55 | |
| 56 | In the current QIC development standard (December 1995), |
| 57 | it is recommended that new tape drives will *in addition* |
| 58 | implement the ATAPI overlap protocol, which is used for the |
| 59 | same purpose - efficient use of the IDE bus, but is interrupt |
| 60 | driven and thus has much less CPU overhead. |
| 61 | |
| 62 | ATAPI overlap is likely to be supported in most new ATAPI |
| 63 | devices, including new ATAPI cdroms, and thus provides us |
| 64 | a method by which we can achieve higher throughput when |
| 65 | sharing a (fast) ATA-2 disk with any (slow) new ATAPI device. |