aboutsummaryrefslogtreecommitdiffstats
path: root/Documentation/ide
diff options
context:
space:
mode:
Diffstat (limited to 'Documentation/ide')
-rw-r--r--Documentation/ide/ChangeLog.ide-tape.1995-2002257
-rw-r--r--Documentation/ide/ide-tape.txt146
2 files changed, 403 insertions, 0 deletions
diff --git a/Documentation/ide/ChangeLog.ide-tape.1995-2002 b/Documentation/ide/ChangeLog.ide-tape.1995-2002
new file mode 100644
index 000000000000..877fac8770b3
--- /dev/null
+++ b/Documentation/ide/ChangeLog.ide-tape.1995-2002
@@ -0,0 +1,257 @@
1/*
2 * Ver 0.1 Nov 1 95 Pre-working code :-)
3 * Ver 0.2 Nov 23 95 A short backup (few megabytes) and restore procedure
4 * was successful ! (Using tar cvf ... on the block
5 * device interface).
6 * A longer backup resulted in major swapping, bad
7 * overall Linux performance and eventually failed as
8 * we received non serial read-ahead requests from the
9 * buffer cache.
10 * Ver 0.3 Nov 28 95 Long backups are now possible, thanks to the
11 * character device interface. Linux's responsiveness
12 * and performance doesn't seem to be much affected
13 * from the background backup procedure.
14 * Some general mtio.h magnetic tape operations are
15 * now supported by our character device. As a result,
16 * popular tape utilities are starting to work with
17 * ide tapes :-)
18 * The following configurations were tested:
19 * 1. An IDE ATAPI TAPE shares the same interface
20 * and irq with an IDE ATAPI CDROM.
21 * 2. An IDE ATAPI TAPE shares the same interface
22 * and irq with a normal IDE disk.
23 * Both configurations seemed to work just fine !
24 * However, to be on the safe side, it is meanwhile
25 * recommended to give the IDE TAPE its own interface
26 * and irq.
27 * The one thing which needs to be done here is to
28 * add a "request postpone" feature to ide.c,
29 * so that we won't have to wait for the tape to finish
30 * performing a long media access (DSC) request (such
31 * as a rewind) before we can access the other device
32 * on the same interface. This effect doesn't disturb
33 * normal operation most of the time because read/write
34 * requests are relatively fast, and once we are
35 * performing one tape r/w request, a lot of requests
36 * from the other device can be queued and ide.c will
37 * service all of them after this single tape request.
38 * Ver 1.0 Dec 11 95 Integrated into Linux 1.3.46 development tree.
39 * On each read / write request, we now ask the drive
40 * if we can transfer a constant number of bytes
41 * (a parameter of the drive) only to its buffers,
42 * without causing actual media access. If we can't,
43 * we just wait until we can by polling the DSC bit.
44 * This ensures that while we are not transferring
45 * more bytes than the constant referred to above, the
46 * interrupt latency will not become too high and
47 * we won't cause an interrupt timeout, as happened
48 * occasionally in the previous version.
49 * While polling for DSC, the current request is
50 * postponed and ide.c is free to handle requests from
51 * the other device. This is handled transparently to
52 * ide.c. The hwgroup locking method which was used
53 * in the previous version was removed.
54 * Use of new general features which are provided by
55 * ide.c for use with atapi devices.
56 * (Programming done by Mark Lord)
57 * Few potential bug fixes (Again, suggested by Mark)
58 * Single character device data transfers are now
59 * not limited in size, as they were before.
60 * We are asking the tape about its recommended
61 * transfer unit and send a larger data transfer
62 * as several transfers of the above size.
63 * For best results, use an integral number of this
64 * basic unit (which is shown during driver
65 * initialization). I will soon add an ioctl to get
66 * this important parameter.
67 * Our data transfer buffer is allocated on startup,
68 * rather than before each data transfer. This should
69 * ensure that we will indeed have a data buffer.
70 * Ver 1.1 Dec 14 95 Fixed random problems which occurred when the tape
71 * shared an interface with another device.
72 * (poll_for_dsc was a complete mess).
73 * Removed some old (non-active) code which had
74 * to do with supporting buffer cache originated
75 * requests.
76 * The block device interface can now be opened, so
77 * that general ide driver features like the unmask
78 * interrupts flag can be selected with an ioctl.
79 * This is the only use of the block device interface.
80 * New fast pipelined operation mode (currently only on
81 * writes). When using the pipelined mode, the
82 * throughput can potentially reach the maximum
83 * tape supported throughput, regardless of the
84 * user backup program. On my tape drive, it sometimes
85 * boosted performance by a factor of 2. Pipelined
86 * mode is enabled by default, but since it has a few
87 * downfalls as well, you may want to disable it.
88 * A short explanation of the pipelined operation mode
89 * is available below.
90 * Ver 1.2 Jan 1 96 Eliminated pipelined mode race condition.
91 * Added pipeline read mode. As a result, restores
92 * are now as fast as backups.
93 * Optimized shared interface behavior. The new behavior
94 * typically results in better IDE bus efficiency and
95 * higher tape throughput.
96 * Pre-calculation of the expected read/write request
97 * service time, based on the tape's parameters. In
98 * the pipelined operation mode, this allows us to
99 * adjust our polling frequency to a much lower value,
100 * and thus to dramatically reduce our load on Linux,
101 * without any decrease in performance.
102 * Implemented additional mtio.h operations.
103 * The recommended user block size is returned by
104 * the MTIOCGET ioctl.
105 * Additional minor changes.
106 * Ver 1.3 Feb 9 96 Fixed pipelined read mode bug which prevented the
107 * use of some block sizes during a restore procedure.
108 * The character device interface will now present a
109 * continuous view of the media - any mix of block sizes
110 * during a backup/restore procedure is supported. The
111 * driver will buffer the requests internally and
112 * convert them to the tape's recommended transfer
113 * unit, making performance almost independent of the
114 * chosen user block size.
115 * Some improvements in error recovery.
116 * By cooperating with ide-dma.c, bus mastering DMA can
117 * now sometimes be used with IDE tape drives as well.
118 * Bus mastering DMA has the potential to dramatically
119 * reduce the CPU's overhead when accessing the device,
120 * and can be enabled by using hdparm -d1 on the tape's
121 * block device interface. For more info, read the
122 * comments in ide-dma.c.
123 * Ver 1.4 Mar 13 96 Fixed serialize support.
124 * Ver 1.5 Apr 12 96 Fixed shared interface operation, broken in 1.3.85.
125 * Fixed pipelined read mode inefficiency.
126 * Fixed nasty null dereferencing bug.
127 * Ver 1.6 Aug 16 96 Fixed FPU usage in the driver.
128 * Fixed end of media bug.
129 * Ver 1.7 Sep 10 96 Minor changes for the CONNER CTT8000-A model.
130 * Ver 1.8 Sep 26 96 Attempt to find a better balance between good
131 * interactive response and high system throughput.
132 * Ver 1.9 Nov 5 96 Automatically cross encountered filemarks rather
133 * than requiring an explicit FSF command.
134 * Abort pending requests at end of media.
135 * MTTELL was sometimes returning incorrect results.
136 * Return the real block size in the MTIOCGET ioctl.
137 * Some error recovery bug fixes.
138 * Ver 1.10 Nov 5 96 Major reorganization.
139 * Reduced CPU overhead a bit by eliminating internal
140 * bounce buffers.
141 * Added module support.
142 * Added multiple tape drives support.
143 * Added partition support.
144 * Rewrote DSC handling.
145 * Some portability fixes.
146 * Removed ide-tape.h.
147 * Additional minor changes.
148 * Ver 1.11 Dec 2 96 Bug fix in previous DSC timeout handling.
149 * Use ide_stall_queue() for DSC overlap.
150 * Use the maximum speed rather than the current speed
151 * to compute the request service time.
152 * Ver 1.12 Dec 7 97 Fix random memory overwriting and/or last block data
153 * corruption, which could occur if the total number
154 * of bytes written to the tape was not an integral
155 * number of tape blocks.
156 * Add support for INTERRUPT DRQ devices.
157 * Ver 1.13 Jan 2 98 Add "speed == 0" work-around for HP COLORADO 5GB
158 * Ver 1.14 Dec 30 98 Partial fixes for the Sony/AIWA tape drives.
159 * Replace cli()/sti() with hwgroup spinlocks.
160 * Ver 1.15 Mar 25 99 Fix SMP race condition by replacing hwgroup
161 * spinlock with private per-tape spinlock.
162 * Ver 1.16 Sep 1 99 Add OnStream tape support.
163 * Abort read pipeline on EOD.
164 * Wait for the tape to become ready in case it returns
165 * "in the process of becoming ready" on open().
166 * Fix zero padding of the last written block in
167 * case the tape block size is larger than PAGE_SIZE.
168 * Decrease the default disconnection time to tn.
169 * Ver 1.16e Oct 3 99 Minor fixes.
170 * Ver 1.16e1 Oct 13 99 Patches by Arnold Niessen,
171 * niessen@iae.nl / arnold.niessen@philips.com
172 * GO-1) Undefined code in idetape_read_position
173 * according to Gadi's email
174 * AJN-1) Minor fix asc == 11 should be asc == 0x11
175 * in idetape_issue_packet_command (did effect
176 * debugging output only)
177 * AJN-2) Added more debugging output, and
178 * added ide-tape: where missing. I would also
179 * like to add tape->name where possible
180 * AJN-3) Added different debug_level's
181 * via /proc/ide/hdc/settings
182 * "debug_level" determines amount of debugging output;
183 * can be changed using /proc/ide/hdx/settings
184 * 0 : almost no debugging output
185 * 1 : 0+output errors only
186 * 2 : 1+output all sensekey/asc
187 * 3 : 2+follow all chrdev related procedures
188 * 4 : 3+follow all procedures
189 * 5 : 4+include pc_stack rq_stack info
190 * 6 : 5+USE_COUNT updates
191 * AJN-4) Fixed timeout for retension in idetape_queue_pc_tail
192 * from 5 to 10 minutes
193 * AJN-5) Changed maximum number of blocks to skip when
194 * reading tapes with multiple consecutive write
195 * errors from 100 to 1000 in idetape_get_logical_blk
196 * Proposed changes to code:
197 * 1) output "logical_blk_num" via /proc
198 * 2) output "current_operation" via /proc
199 * 3) Either solve or document the fact that `mt rewind' is
200 * required after reading from /dev/nhtx to be
201 * able to rmmod the idetape module;
202 * Also, sometimes an application finishes but the
203 * device remains `busy' for some time. Same cause ?
204 * Proposed changes to release-notes:
205 * 4) write a simple `quickstart' section in the
206 * release notes; I volunteer if you don't want to
207 * 5) include a pointer to video4linux in the doc
208 * to stimulate video applications
209 * 6) release notes lines 331 and 362: explain what happens
210 * if the application data rate is higher than 1100 KB/s;
211 * similar approach to lower-than-500 kB/s ?
212 * 7) 6.6 Comparison; wouldn't it be better to allow different
213 * strategies for read and write ?
214 * Wouldn't it be better to control the tape buffer
215 * contents instead of the bandwidth ?
216 * 8) line 536: replace will by would (if I understand
217 * this section correctly, a hypothetical and unwanted situation
218 * is being described)
219 * Ver 1.16f Dec 15 99 Change place of the secondary OnStream header frames.
220 * Ver 1.17 Nov 2000 / Jan 2001 Marcel Mol, marcel@mesa.nl
221 * - Add idetape_onstream_mode_sense_tape_parameter_page
222 * function to get tape capacity in frames: tape->capacity.
223 * - Add support for DI-50 drives( or any DI- drive).
224 * - 'workaround' for read error/blank block around block 3000.
225 * - Implement Early warning for end of media for Onstream.
226 * - Cosmetic code changes for readability.
227 * - Idetape_position_tape should not use SKIP bit during
228 * Onstream read recovery.
229 * - Add capacity, logical_blk_num and first/last_frame_position
230 * to /proc/ide/hd?/settings.
231 * - Module use count was gone in the Linux 2.4 driver.
232 * Ver 1.17a Apr 2001 Willem Riede osst@riede.org
233 * - Get drive's actual block size from mode sense block descriptor
234 * - Limit size of pipeline
235 * Ver 1.17b Oct 2002 Alan Stern <stern@rowland.harvard.edu>
236 * Changed IDETAPE_MIN_PIPELINE_STAGES to 1 and actually used
237 * it in the code!
238 * Actually removed aborted stages in idetape_abort_pipeline
239 * instead of just changing the command code.
240 * Made the transfer byte count for Request Sense equal to the
241 * actual length of the data transfer.
242 * Changed handling of partial data transfers: they do not
243 * cause DMA errors.
244 * Moved initiation of DMA transfers to the correct place.
245 * Removed reference to unallocated memory.
246 * Made __idetape_discard_read_pipeline return the number of
247 * sectors skipped, not the number of stages.
248 * Replaced errant kfree() calls with __idetape_kfree_stage().
249 * Fixed off-by-one error in testing the pipeline length.
250 * Fixed handling of filemarks in the read pipeline.
251 * Small code optimization for MTBSF and MTBSFM ioctls.
252 * Don't try to unlock the door during device close if is
253 * already unlocked!
254 * Cosmetic fixes to miscellaneous debugging output messages.
255 * Set the minimum /proc/ide/hd?/settings values for "pipeline",
256 * "pipeline_min", and "pipeline_max" to 1.
257 */
diff --git a/Documentation/ide/ide-tape.txt b/Documentation/ide/ide-tape.txt
new file mode 100644
index 000000000000..658f271a373f
--- /dev/null
+++ b/Documentation/ide/ide-tape.txt
@@ -0,0 +1,146 @@
1/*
2 * IDE ATAPI streaming tape driver.
3 *
4 * This driver is a part of the Linux ide driver.
5 *
6 * The driver, in co-operation with ide.c, basically traverses the
7 * request-list for the block device interface. The character device
8 * interface, on the other hand, creates new requests, adds them
9 * to the request-list of the block device, and waits for their completion.
10 *
11 * Pipelined operation mode is now supported on both reads and writes.
12 *
13 * The block device major and minor numbers are determined from the
14 * tape's relative position in the ide interfaces, as explained in ide.c.
15 *
16 * The character device interface consists of the following devices:
17 *
18 * ht0 major 37, minor 0 first IDE tape, rewind on close.
19 * ht1 major 37, minor 1 second IDE tape, rewind on close.
20 * ...
21 * nht0 major 37, minor 128 first IDE tape, no rewind on close.
22 * nht1 major 37, minor 129 second IDE tape, no rewind on close.
23 * ...
24 *
25 * The general magnetic tape commands compatible interface, as defined by
26 * include/linux/mtio.h, is accessible through the character device.
27 *
28 * General ide driver configuration options, such as the interrupt-unmask
29 * flag, can be configured by issuing an ioctl to the block device interface,
30 * as any other ide device.
31 *
32 * Our own ide-tape ioctl's can be issued to either the block device or
33 * the character device interface.
34 *
35 * Maximal throughput with minimal bus load will usually be achieved in the
36 * following scenario:
37 *
38 * 1. ide-tape is operating in the pipelined operation mode.
39 * 2. No buffering is performed by the user backup program.
40 *
41 * Testing was done with a 2 GB CONNER CTMA 4000 IDE ATAPI Streaming Tape Drive.
42 *
43 * Here are some words from the first releases of hd.c, which are quoted
44 * in ide.c and apply here as well:
45 *
46 * | Special care is recommended. Have Fun!
47 *
48 *
49 * An overview of the pipelined operation mode.
50 *
51 * In the pipelined write mode, we will usually just add requests to our
52 * pipeline and return immediately, before we even start to service them. The
53 * user program will then have enough time to prepare the next request while
54 * we are still busy servicing previous requests. In the pipelined read mode,
55 * the situation is similar - we add read-ahead requests into the pipeline,
56 * before the user even requested them.
57 *
58 * The pipeline can be viewed as a "safety net" which will be activated when
59 * the system load is high and prevents the user backup program from keeping up
60 * with the current tape speed. At this point, the pipeline will get
61 * shorter and shorter but the tape will still be streaming at the same speed.
62 * Assuming we have enough pipeline stages, the system load will hopefully
63 * decrease before the pipeline is completely empty, and the backup program
64 * will be able to "catch up" and refill the pipeline again.
65 *
66 * When using the pipelined mode, it would be best to disable any type of
67 * buffering done by the user program, as ide-tape already provides all the
68 * benefits in the kernel, where it can be done in a more efficient way.
69 * As we will usually not block the user program on a request, the most
70 * efficient user code will then be a simple read-write-read-... cycle.
71 * Any additional logic will usually just slow down the backup process.
72 *
73 * Using the pipelined mode, I get a constant over 400 KBps throughput,
74 * which seems to be the maximum throughput supported by my tape.
75 *
76 * However, there are some downfalls:
77 *
78 * 1. We use memory (for data buffers) in proportional to the number
79 * of pipeline stages (each stage is about 26 KB with my tape).
80 * 2. In the pipelined write mode, we cheat and postpone error codes
81 * to the user task. In read mode, the actual tape position
82 * will be a bit further than the last requested block.
83 *
84 * Concerning (1):
85 *
86 * 1. We allocate stages dynamically only when we need them. When
87 * we don't need them, we don't consume additional memory. In
88 * case we can't allocate stages, we just manage without them
89 * (at the expense of decreased throughput) so when Linux is
90 * tight in memory, we will not pose additional difficulties.
91 *
92 * 2. The maximum number of stages (which is, in fact, the maximum
93 * amount of memory) which we allocate is limited by the compile
94 * time parameter IDETAPE_MAX_PIPELINE_STAGES.
95 *
96 * 3. The maximum number of stages is a controlled parameter - We
97 * don't start from the user defined maximum number of stages
98 * but from the lower IDETAPE_MIN_PIPELINE_STAGES (again, we
99 * will not even allocate this amount of stages if the user
100 * program can't handle the speed). We then implement a feedback
101 * loop which checks if the pipeline is empty, and if it is, we
102 * increase the maximum number of stages as necessary until we
103 * reach the optimum value which just manages to keep the tape
104 * busy with minimum allocated memory or until we reach
105 * IDETAPE_MAX_PIPELINE_STAGES.
106 *
107 * Concerning (2):
108 *
109 * In pipelined write mode, ide-tape can not return accurate error codes
110 * to the user program since we usually just add the request to the
111 * pipeline without waiting for it to be serviced. In case an error
112 * occurs, I will report it on the next user request.
113 *
114 * In the pipelined read mode, subsequent read requests or forward
115 * filemark spacing will perform correctly, as we preserve all blocks
116 * and filemarks which we encountered during our excess read-ahead.
117 *
118 * For accurate tape positioning and error reporting, disabling
119 * pipelined mode might be the best option.
120 *
121 * You can enable/disable/tune the pipelined operation mode by adjusting
122 * the compile time parameters below.
123 *
124 *
125 * Possible improvements.
126 *
127 * 1. Support for the ATAPI overlap protocol.
128 *
129 * In order to maximize bus throughput, we currently use the DSC
130 * overlap method which enables ide.c to service requests from the
131 * other device while the tape is busy executing a command. The
132 * DSC overlap method involves polling the tape's status register
133 * for the DSC bit, and servicing the other device while the tape
134 * isn't ready.
135 *
136 * In the current QIC development standard (December 1995),
137 * it is recommended that new tape drives will *in addition*
138 * implement the ATAPI overlap protocol, which is used for the
139 * same purpose - efficient use of the IDE bus, but is interrupt
140 * driven and thus has much less CPU overhead.
141 *
142 * ATAPI overlap is likely to be supported in most new ATAPI
143 * devices, including new ATAPI cdroms, and thus provides us
144 * a method by which we can achieve higher throughput when
145 * sharing a (fast) ATA-2 disk with any (slow) new ATAPI device.
146 */
generated by cgit v1.2.2 (git 2.25.0) at 2024-12-03 13:08:30 -0500