CHAPTER 6 Operations. Fujitsu MHV2040AH - Mobile - Hard Drive, MHV2100AH, MHV2080AH, MHV2040AH, MHV2060AH

Add to My manuals
280 Pages

advertisement

CHAPTER 6 Operations. Fujitsu MHV2040AH - Mobile - Hard Drive, MHV2100AH, MHV2080AH, MHV2040AH, MHV2060AH | Manualzz

CHAPTER 6 Operations

6.1 Device Response to the Reset

C141-E217 6-1

Operations

6.1 Device Response to the Reset

This section describes how the PDIAG- and DASP- signals responds when the power of the IDD is turned on or the IDD receives a reset or diagnostic command.

6.1.1 Response to power-on

After the master device (device 0) releases its own power-on reset state, the master device shall check a DASP- signal for least 500 ms to confirm presence of a slave device (device 1). The master device recognizes presence of the slave device when it confirms assertion of the DASP- signal. Then, the master device checks a

PDIAG- signal to see if the slave device has successfully completed the power-on diagnostics.

If the master device cannot confirm assertion of the DASP- signal within 500 ms, the master device recognizes that no slave device is connected.

After the slave device (device 1) releases its own power-on reset state, the slave device shall report its presence and the result of power-on diagnostics to the master device as described below:

DASP- signal: Asserted within 450 ms.

PDIAG- signal: Negated within 1 ms and asserted within 30 seconds.

The asserted PDIAG-signal is negated 30 seconds after it is asserted if the command is not received.

6-2 C141-E217

6.1 Device Response to the Reset

Power on

Master device

Power On Reset-

Status Reg.

BSY bit Max. 31 sec.

Checks DASP- for up to

500 ms.

If presence of a slave device is confirmed, PDIAG- is checked for up to 31 seconds.

Slave device

Power On Reset-

BSY bit

Max. 1 ms.

PDIAG-

DASP-

Max. 30 sec.

Max. 450 ms.

Figure 6.1 Response to power-on

Note: Figure 6.1 has an assumption that the device is kept on the power-off condition for more than

5 sec before the device power is turned on.

6.1.2 Response to hardware reset

Response to RESET- (hardware reset through the interface) is similar to the power-on reset.

Upon receipt of hardware reset, the master device checks a DASP- signal for up to

500 ms to confirm presence of a slave device. The master device recognizes the presence of the slave device when it confirms assertion of the DASP- signal.

Then the master device checks a PDIAG- signal to see if the slave device has successfully completed the self-diagnostics.

If the master device cannot confirm assertion of the DASP- signal within 450 ms, the master device recognizes that no slave device is connected.

C141-E217 6-3

Operations

Reset-

Master device

Status Reg.

BSY bit

After the slave device receives the hardware reset, the slave device shall report its presence and the result of the self-diagnostics to the master device as described below:

DASP- signal: Asserted within 450 ms.

PDIAG- signal: Negated within 1 ms and asserted within 30 seconds.

The asserted PDIAG-signal is negated 30 seconds after it is asserted if the command is not received.

Checks DASP- for up to

500 ms.

Max. 31 sec.

If presence of a slave device is confirmed, PDIAG- is checked for up to 31 seconds.

Slave device

BSY bit

Max. 1 ms.

PDIAG-

DASP-

Max. 30 sec.

Max. 450 ms.

.

Figure 6.2 Response to hardware reset

Note: Master Device does not check the DASP signal assertion for 2ms upon receipt of hardware reset.

6-4 C141-E217

6.1 Device Response to the Reset

6.1.3 Response to software reset

The master device does not check the DASP- signal for a software reset. If a slave device is present, the master device checks the PDIAG- signal for up to 15 seconds to see if the slave device has completed the self-diagnosis successfully.

X'3F6' Reg.

Master device

After the slave device receives the software reset, the slave device shall report its presence and the result of the self-diagnostics to the master device as described below:

PDIAG- signal: negated within 1 ms and asserted within 30 seconds

The asserted PDIAG-signal is negated 30 seconds after it is asserted if the command is not received.

When the IDD is set to a slave device, the IDD asserts the DASP- signal when negating the PDIAG- signal.

X"0C" or X"04"

X"00"

Status Reg.

BSY bit

Max. 31 sec.

If the slave device is preset, PDIAG- is checked for up to 31 seconds.

Slave device

BSY bit

PDIAG-

DASP-

Max. 1 ms.

Max. 30 sec.

Figure 6.3 Response to software reset

C141-E217 6-5

Operations

6.1.4 Response to diagnostic command

When the master device receives an EXECUTE DEVICE DIAGNOSTIC command and the slave device is present, the master device checks the PDIAG- signal for up to 6 seconds to see if the slave device has completed the selfdiagnosis successfully.

The master device does not check the DASP- signal.

After the slave device receives the EXECUTE DEVICE DIAGNOSTIC command, it shall report the result of the self-diagnostics to the master device as described below:

PDIAG- signal: negated within 1 ms and asserted within 5 seconds

The asserted PDIAG-signal is negated 5 seconds after it is asserted if the command is not received. If the command is received, the PDIAG-signal is negated according to timing at which the command is received.

When the IDD is set to a slave device, the IDD asserts the DASP- signal when negating the PDIAG- signal.

X'1F7' Reg.

Write

Master device

Status Reg.

BSY bit

Max. 6 sec.

If the slave device is preset, PDIAG- signal is checked for up to6 seconds.

Slave device

BSY bit

PDIAG-

DASP-

Max. 1 ms.

Max. 5 sec.

Figure 6.4 Response to diagnostic command

6-6 C141-E217

6.2 Power Save

6.2 Power Save

The host can change the power consumption state of the device by issuing a power command to the device.

6.2.1 Power save mode

There are five types of power consumption state of the device including active mode where all circuits are active.

Active mode

Active idle mode

Low power idle mode

Standby mode

Sleep mode

The device enters the active idle mode by itself. The device also enters the idle mode in the same way after power-on sequence is completed. The subsequent mode transition changes depending on the APM setting.

(1) Active mode

In this mode, all the electric circuit in the device are active or the device is under seek, read or write operation.

A device enters the active mode under the following conditions:

• The media access system is received.

(2) Active idle mode

In this mode, circuits on the device are set to power save mode.

The device enters the Active idle mode under the following conditions:

• After completion of the command execution other than SLEEP and STANDBY commands.

(3) Low power idle mode

Sets circuits on the device to the power save mode. The heads are disabled in the safe state.

The device enters the low power mode under the following conditions:

After certain amount of time has elapsed in the active idle state (APM Mode

0, Mode 1 and Mode 2)

Upon completion of the power-on sequence

C141-E217 6-7

Operations

Upon receipt of a hard reset

Upon receipt of Idle/Idle Intermediate

(4) Standby mode

In this mode, the spindle motor has stopped from the low power idle state.

The device can receive commands through the interface. However if a command with disk access is issued, response time to the command under the standby mode takes longer than the active, active idle, or low power idle mode because the access to the disk medium cannot be made immediately.

The drive enters the standby mode under the following conditions:

A STANDBY or STANDBY IMMEDIATE command is issued.

A certain amount of time has elapsed in the low power idle state. (APM

Mode 2)

The time specified by the STANDBY or IDLE command has elapsed after completion of the command.

A reset is issued in the sleep mode. •

When one of following commands is issued, the command is executed normally and the device is still stayed in the standby mode.

Reset (hardware or software)

STANDBY command

STANDBY IMMEDIATE command

INITIALIZE DEVICE PARAMETERS command

CHECK POWER MODE command

(5) Sleep mode

The power consumption of the drive is minimal in this mode. The drive enters only the standby mode from the sleep mode. The only method to return from the standby mode is to execute a software or hardware reset.

The drive enters the sleep mode under the following condition:

• A SLEEP command is issued.

In this mode, the device does not accept the command. (It is ignored.)

6-8 C141-E217

6.3 Defect Processing

6.2.2 Power commands

The following commands are available as power commands.

IDLE

IDLE IMMEDIATE

STANDBY

STANDBY IMMEDIATE

SLEEP

CHECK POWER MODE

SET FEATURES (APM setting)

6.3 Defect Processing

This device performs alternating processing where the defective sector is alternated with the spare area depending on media defect location information.

The media defect location information is registered in the system space specified for the user area according to the format at shipment of the media from the plant.

6.3.1 Spare area

The following type of area is prepared as the spare area in user areas:

1) Spare cylinder for alternate assignment: This cylinder is used during automatic alternating processing for defective sector. More than 2000 sectors/drive.

C141-E217 6-9

Operations

6.3.2 Alternating processing for defective sectors

The following two types of technology are used for alternating processing:

(1) Sector slip processing

In this method, defective sectors are not used (thereby avoiding the effects of defects), and each defective sector is assigned to the next contiguous sector that is normal.

Depending on the format defined at shipment from the plant, this processing is performed for defective sectors.

Figure 6.5 shows an example where sector (physical) 5 with cylinder 0 and head 0 is defective.

Cylinder 0

Head 0

Sector (physical)

778 779 780

Defective sector

(Not used)

777 778 779

Note: When an access request for sector 5 is issued, physical sector 6 must be accessed instead of physical sector 5.

Figure 6.5 Sector slip processing

(2) Track slip processing

In this method, defective tracks not used (there by avoiding the effects of defects), and each defective track is assigned to the next contiguous track that is normal.

Depending on the format defined at shipment from the plant, this processing is performed for defective tracks.

6-10 C141-E217

6.3 Defect Processing

(3) Automatic alternating processing

This technology assigns a defective sector to a spare sector of a spare cylinder for alternate assignment.

This device performs automatic alternating processing in the event of any of the following errors.

• Automatic alternating processing is attempted for read error recovery by reaching the specified retry cycle while a read error retry is in progress.

Before attempting automatic alternating processing, writing and reading of already corrected data is repeated for the sector in which an error occurred. If a read error does not occur during this reading operation, automatic alternating processing is not performed.

• If error recovery is not successful even if a write fault error retry is executed, automatic alternating processing is performed.

Figure 6.6 shows an example where automatic alternating processing is applied to sector (physical) 5 with cylinder 0 and head 0.

Sector (physical)

779 780

Cylinder 0

Head 0

Defective sector

(Not used)

779 780

Alternate cylinder 0

Head 0

Already assigned

This is assigned to an unassigned sector.

Notes:

1. The alternate cylinder is assigned to an inner cylinder in each zone.

2. When an access request for sector 5 is issued, the sector assigned for alternating processing of the alternate cylinder must be accessed instead of physical sector 5.

If an access request for sectors after sector 5 is issued, seek is executed to cylinder 0, head 0 in order to continue processing.

Figure 6.6 Automatic alternating processing

C141-E217 6-11

Operations

6.4 Read-ahead Cache

Read-ahead Cache is the function for automatically reading data blocks upon completion of the read command in order to read data from disk media and save data block on a data buffer.

If a subsequent command requests reading of the read-ahead data, data on the data buffer can be transferred without accessing the disk media. As the result, faster data access becomes possible for the host.

6.4.1 DATA buffer structure

This device contains a data buffer. This buffer is divided into two areas: one area is used for MPU work, and the other is used as a read cache for another command.

(See Figure 6.7) a) 8MB buffer (8,388,608 bytes)

For MPU work

557,056 bytes

8,388,608 bytes

For R/W command

7,831,552 bytes

Figure 6.7 Data buffer structure

The read-ahead operation is done by the following commands.

READ SECTOR(S) (EXT)

READ MULTIPLE (EXT)

READ DMA (EXT)

READ SECTOR(S) EXT

READ DMA EXT

READ MULTIPLE EXT

6-12 C141-E217

6.4 Read-ahead Cache

6.4.2 Caching operation

The caching operation is performed only when the commands listed below are received. If any of the following data are stored on the data buffer, the data is sent to the host system.

All of the sector data that this command processes.

A part of the sector data including the start sector, that this command processes.

If part of the data to be processed is stored on the data buffer, the remaining data is read from disk media and sent to the host system.

(1) Commands that are targets of caching

The commands that are targets of caching are as follows:

READ SECTOR(s)

READ MULTIPLE

READ DMA

READ SECTOR(s) EXT

READ DMA EXT

READ MULTIPLE EXT

However, if the caching function is prohibited by the SET FEATURES command, the caching operation is not performed.

(2) Data that is a target of caching

The data that is a target of caching are as follows:

1) Read-ahead data that is read from disk media and saved to the data buffer upon completion of execution of a command that is a target of caching.

2) Pre-read data that is read from disk media and saved to the data buffer before execution of a command that is a target of caching.

3) Data required by a command that is a target of caching and has been sent to the host system one. If the sector data requested by the host has not been completely stored in the read cache portion of the buffer, this data does not become a target of caching. Also, if sequential hits occur continuously, the caching-target data required by the host becomes invalid because that data is overwrited by new data.

(3) Invalidating caching-target data

Data that is a target of caching on the data buffer is invalidated under the following conditions:

C141-E217 6-13

Operations

1)-1 Any command other than the following commands is issued. (All cachingtarget data is invalidated.)

RECALIBRATE

IDLE IMMEDIATE

DOWNLOAD MICROCODE

DEVICE CONFIGURATION

READ BUFFER

WRITE BUFFER

SET FEATURES

SECURITY ERASE UNIT

READ LOG EXT

WRITE LOG EXT

UNSUPPORT COMMAND (INVALID COMMAND)

1)-2 Commands that partially invalidate caching data

(When data in the buffer or on media is overwritten, the overwritten data is invalidated.)

READ SECTOR (s) / READ MULTIPLE / READ DMA

READ SECTOR (s) EXT / READ DMA EXT / READ MULTIPLE EXT

WRITE SECTOR(s) / WRITE MULTIPLE / WRITE DMA

WRITE SECTOR (s) EXT / WRITE DMA EXT / WRITE MULTIPLE EXT

WRITE DMA FUA EXT / WRITE MULTIPLE FUA EXT

SMART

2) A hard reset is issued or the power is turned off.

3) When ICRC (Interface CRC) Error has occurred.

6-14 C141-E217

6.4 Read-ahead Cache

6.4.3 Using the read segment buffer

Methods of using the read segment buffer are explained for following situations.

6.4.3.1 Miss-hit

In this situations, the top block of read requested data is not stored at all in the data buffer. As a result, all of the read requested data is read from disk media.

1) HAP (host address pointer) and DAP (disk address pointer) are defined in the head of the segment allocated from Buffer. (If pre-read is executed, HAP is set at the requested data reading position.)

HAP (host address pointer)

Read segment

DAP (disk address pointer)

2) During reading of read requested data, the request data that has already been read is sent to the host system.

Read requested data is stored until this point

HAP

Read requested data

Free space

DAP

3) When reading of read requested data is completed and transfer of the read requested data to the host system is completed, reading of the disk continues until a certain amount of data is stored.

HAP (stop)

Read requested data Read-ahead data

DAP

4) The following cache valid data is for the read command that is executed next:

Cache valid data

LAST LBA START LBA

C141-E217 6-15

Operations

6.4.3.2 Sequential hit

When the read command that is targeted at a sequential address is received after execution of the read commands is completed, the read command transmits the

Read requested data to the host system continuing read-ahead without newly allocating the buffer for read.

1) When the sequential read command is received, HAP is set in the sequential address of the last read command, and DAP is set at a present read position as it is.

Cache valid data

HAP (host address pointer)

Read requested data

Read-ahead data

Free space

DAP (disk address pointer)

2) During reading of read requested data, the request data that has already been read is sent to the host system.

Cache valid data

HAP (host address pointer)

Read requested data

Free space

DAP (disk address pointer)

3) When reading of read requested data is completed and transfer of the read requested data to the host system is completed, the read-ahead operation continues until a certain amount of data is stored.

Cache valid data

Read requested data

HAP (host address pointer)

Free space

DAP (disk address pointer)

4) The following cache valid data is for the read command that is executed next:

Cache valid data

LAST LBA START LBA

6-16 C141-E217

6.4 Read-ahead Cache

6.4.3.3 Full hit

In this situation, all read requested data is stored in the data buffer. Transfer of the read requested data is started from the location where hit data is stored. For data that is a target of caching and remains before a full hit, the data is retained when execution of the command is completed. This is done so that a new read-ahead operation is not performed. If the full hit command is received during the readahead operation, a transfer of the read requested data starts while the read-ahead operation is in progress.

1) An example is the state shown below where the previous read command is executing sequential reading. First, HAP is set at the location where hit data is stored.

HAP

HAP end location of the previous read command

HAP (It is reset to the hit data location for transfers.)

Cache data Full hit data Cache data

DAP

DAP end location of the previous read command

2) The read requested data is transferred, and a new read-ahead operation is not performed.

Cache data Full hit data

HAP

(stop)

Cache data

C141-E217 6-17

Operations

6.4.3.4 Partial hit

In this situation, a part of read requested data including the top sector is stored in the data buffer. A transfer of the read requested data starts from the address where the data that is hit is stored until the top sector of the read requested data.

Remaining part of insufficient data is read then.

An example is a case where a partial hit occurs in cache data, as shown below.

Cache valid data

START LBA

1) HAP is set at the address where partial hit data is stored, and Transfer is started.

LAST LBA

Cache valid data

HAP (host address pointer)

Partial hit data

2) DAP and HAP are set at the head of Buffer newly allocated, and insufficient data is read.

HAP (host address pointer)

Read segment

DAP (disk address pointer)

3) When reading the read requested data ends and the transmission of the read requested data to the host system ends, the read-ahead operation continues until a certain amount of data is stored.

The method of storing the read-ahead data at Partial hit is the same as the

Miss hit.

Cache valid data

LAST LBA START LBA

6-18 C141-E217

6.5 Write Cache

6.5 Write Cache

Write Cache is the function for reducing the command processing time by separating command control to disk media from write control to disk media.

When Write Cache is permitted, the write command can be keep receiving as long as the space available for data transfers remains free on the data buffer. Because of this function, command processing appears to be completed swiftly from the viewpoint of the host. It improves system throughput.

6.5.1 Cache operation

(1) Command that are targets of caching

The Commands that are targets of caching are as follows:

WRITE SECTOR (S)

WRITE MULTIPLE

WRITE DMA

WRITE SECTOR (S) EXT

WRITE MULTIPLE EXT

• WRITE DMA EXT

However, the caching operation is not performed when the caching function is prohibited by the SET FEATURES command.

(2) Invalidation of cached data

If an error occurs during writing onto media, write processing is repeated up to as many times as specified for retry processing. If retry fails for a sector because the retry limit is reached, automatic alternate sector processing is executed for the sector. If the automatic alternate sector processing fails, the data in the sector for which automatic alternate sector processing failed is invalidated without being guaranteed.

If data remains in sectors following a sector for which automatic alternate sector processing failed, the data is invalidated without being guaranteed.

Moreover, when the command (clause 6.4.2(3)) is accepted and HOST CRC Error is generated, the caching data is invalidated.

<Exception>

• If a Reset or command is received while a transfer of one sector of data is in progress, data is not written in the sector of the media where the interruption occurred, and sectors accepted before interruption occurred is written in the medium.

C141-E217 6-19

Operations

(3) Status report in the event of an error

The status report concerning an error occurring during writing onto media is created when the next command is issued. Where the command reporting the error status is not executed, only the error status is reported. Only the status of an error that occurs during write processing is reported.

<Exceptions>

The error status is not reported in the following case:

The reset command is received after an error has occurred during writing to media.

Reset processing is performed as usual. The error status that has occurred during writing to media is not reported.

(4) Enabling and disabling

Enabling and disabling of the Write Cache function can be set only with the SET

FEATURES command. The setting does not changed even when the error status is reported.

The initial setting is stored in the system area of media. System area information is loaded whenever the power is turned on.

(5) Reset response

When a reset is received while cached data is stored on the data buffer, data of the data buffer is written on the media, and reset processing is then performed. This is true for both a hard reset and soft reset.

(6) Caching function when power supply is turned on.

The caching function is invalid until Calibration is done after the power supply is turned on (about 7 or 8 sec). It is effective in Default after that as long as the caching function is not invalidly set by the SET FEATURES command.

6-20 C141-E217

6.5 Write Cache

If Write Cache is enabled, there is a possibility that data transferred from the host with the Write Cache enable command is not completely written on disk media before the normal end interrupt is issued.

If an unrecoverable error occurs while multiple commands that are targets of write caching are received, the host has difficulty determining which command caused the error. (An error report is not issued to the host if automatic alternating processing for the error is performed normally.) Therefore, the host cannot execute a retry for the unrecoverable error while Write Cache is enabled. Be very careful on this point when using this function.

If a write error occurs, an abort response is sent to all subsequent commands.

C141-E217 6-21

This page is intentionally left blank.

advertisement

Key Features

  • 60 GB 2.5" 5400 RPM Ultra-ATA/100
  • HDD
  • Storage drive buffer size: 8 MB
  • 101 g

Related manuals

Frequently Answers and Questions

What is the interface type of the Fujitsu MHV2060AH?
The Fujitsu MHV2060AH utilizes an ATA interface, ensuring compatibility with a wide range of systems and motherboards.
Is the Fujitsu MHV2060AH suitable for use in NAS (Network Attached Storage) devices?
While the Fujitsu MHV2060AH can be used in NAS devices, it is primarily designed for use in laptops, desktops, and external storage applications. For optimal performance and reliability in NAS environments, consider using hard disk drives specifically designed for NAS applications.
What is the rotational speed of the Fujitsu MHV2060AH?
The rotational speed of the Fujitsu MHV2060AH is not explicitly mentioned in the provided manual.

advertisement

Table of contents