next up previous contents index
Next: 6. Database Definition Up: AppDevGuide Previous: 4. Build Facility   Contents   Index

Subsections

5. Database Locking, Scanning, And Processing

5.1 Overview

Before describing particular components of the IOC software, it is helpful to give an overview of three closely related topics: Database locking, scanning, and processing. Locking is done to prevent two different tasks from simultaneously modifying related database records. Database scanning is the mechanism for deciding when records should be processed. The basics of record processing involves obtaining the current value of input fields and outputting the current value of output fields. As records become more complex so does the record processing.

One powerful feature of the DATABASE is that records can contain links to other records. This feature also causes considerable complication. Thus, before discussing locking, scanning, and processing, record links are described.

5.2 Record Links

A database record may contain links to other records. Each link is one of the following types:

Links are defined in file link.h.

NOTE: This chapter discusses mainly database links.

5.3 Database Links

Database links are referenced by calling one of the following routines:

A forward link only makes sense when it refers to a passive record that should be processed after the record that contains the link.

For input and output links, two additional attributes can be specified by the application developer: process passive, and maximize severity.

5.3.1 Process Passive

The Process Passive attribute takes the value NPP (Non-Process Passive) or PP (Process Passive). It determines if the linked record should be processed before getting a value from an input link or after writing a value to an output link. The linked record will be processed only if link's Process Passive attribute is PP and the target record's SCAN field is Passive.

NOTE: Three other options may also be specified: CA, CP, and CPP. These options force the link to be handled like a Channel Access Link. See last section of this chapter for details.

5.3.2 Maximize Severity

The Maximize Severity attribute is one of NMS (Non-Maximize Severity), MS (Maximize Severity), MSS (Maximize Status and Severity) or MSI (Maximize Severity if Invalid). It determines whether alarm severity is propagated across links. If the attribute is MSI only a severity of INVALID_ALARM is propagated; settings of MS or MSS propagate all alarms that are more severe than the record's current severity. For input links the alarm severity of the record referred to by the link is propagated to the record containing the link. For output links the alarm severity of the record containing the link is propagated to the record referred to by the link. If the severity is changed the associated alarm status is set to LINK_ALARM, except if the attribute is MSS when the alarm status will be copied along with the severity.

The method of determining if the alarm status and severity should be changed is called ``maximize severity". In addition to its actual status and severity, each record also has a new status and severity. The new status and severity are initially 0, which means NO_ALARM. Every time a software component wants to modify the status and severity, it first checks the new severity and only makes a change if the severity it wants to set is greater than the current new severity. If it does make a change, it changes the new status and new severity, not the current status and severity. When database monitors are checked, which is normally done by a record processing routine, the current status and severity are set equal to the new values and the new values reset to zero. The end result is that the current alarm status and severity reflect the highest severity outstanding alarm. If multiple alarms of the same severity are present the alarm status reflects the first one detected.

5.4 Database Locking

The purpose of database locking is to prevent a record from being processed simultaneously by two different tasks. In addition, it prevents ``outside" tasks from changing any field while the record is being processed.

The following routines are provided for database locking.

dbScanLock(precord);
dbScanUnlock(precord);

The basic idea is to call dbScanLock before accessing database records and calling dbScanUnlock afterwords. Because of database links (Input, Output, and Forward) a modification to one record can cause modification to other records. Records linked together with database links are placed in the same lock set. dbScanLock locks the entire lock set, not just the record requested. dbScanUnlock unlocks the entire set.

The following rules determine when the lock routines must be called:

  1. The periodic, I/O event, and event tasks lock before and unlock after processing:

  2. dbPutField locks before modifying a record and unlocks afterwards.

  3. dbGetField locks before reading and unlocks afterwards.

  4. Any asynchronous record support completion routine must lock before modifying a record and unlock afterwards.

All records connected by any kind of database link are placed in the same lock set. Versions of EPICS Base prior to R3.14 allowed an NPP NMS input link to span two different lock sets, but this was not safe where the read and write operations on the field value were not atomic in nature and is no longer available to break a lockset.

5.5 Database Scanning

Database scanning refers to requests that database records be processed. Four types of scanning are possible:

  1. Periodic - Records are scanned at regular intervals.

  2. I/O event - A record is scanned as the result of an I/O interrupt.

  3. Event - A record is scanned as the result of any task issuing a post_event request.

  4. Passive - A record is scanned as a result of a call to dbScanPassive. dbScanPassive will issue a record processing request if and only if the record is passive and is not already being processed.

A dbScanPassive request results from a task calling one of the following routines:

All non-record processing tasks (Channel Access, Sequence Programs, etc.) call dbGetField to obtain database values. dbGetField just reads values without asking that a record be processed.

5.6 Record Processing

A record is processed as a result of a call to dbProcess. Each record support module must supply a routine process. This routine does most of the work related to record processing. Since the details of record processing are record type specific this topic is discussed in greater detail in the Chapter ``Record Support".

5.7 Guidelines for Creating Database Links

The ability to link records together is an extremely powerful feature of the IOC software. In order to use links properly it is important that the Application Developer understand how they are processed. As an introduction consider the following example:

\includegraphics{lockScanProcess_6}

Assume that A, B, and C are all passive records. The notation states that A has a forward link to B and B to C. C has an input link obtaining a value from A. Assume, for some reason, A gets processed. The following sequence of events occurs:

  1. A begins processing. While processing a request is made to process B.

  2. B starts processing. While processing a request is made to process C.

  3. C starts processing. One of the first steps is to get a value from A via the input link.

  4. At this point a question occurs. Note that the input link specifies process passive (signified by the PP after InLink). But process passive states that A should be processed before the value is retrieved. Are we in an infinite loop? The answer is no. Every record contains a field PACT (processing active), which is set TRUE when record processing begins and is not set FALSE until all processing completes. When C is processed A still has PACT TRUE and will not be processed again.

  5. C obtains the value from A and completes its processing. Control returns to B.

  6. B completes returning control to A

  7. A completes processing.

This brief example demonstrates that database links need more discussion.

5.7.1 Rules Relating to Database Links

5.7.1.1 Processing Order

The processing order follows the following rules:

  1. Forward links are processed in order from left to right and top to bottom. For example the following records are processed in the order FLNK1, FLNK2, FLNK3, FLNK4 .

    \includegraphics{lockScanProcess_9}

  2. If a record has multiple input links (such as the calculation or select records) the input values are nornally fetched in the natural order. For example for link fields named INPA, INPB, ..., INPL, the links would be read in the order A, B, C etc. Thus if obtaining an input results in a record being processed, the processing order is guaranteed. Some record types may not follow this rule however.

  3. All input and output links are processed before the forward link.

5.7.1.2 Lock Sets

All records, except for the conditions listed in the next paragraph, linked together directly or indirectly are placed in the same lock set. When dbScanLock is called the entire set, not just the specified record, is locked. This prevents two different tasks from simultaneously modifying records in the same lock set.

5.7.1.3 PACT - Process Active

Every record contains a field PACT. This field is set TRUE at the beginning of record processing and is not set FALSE until the record is completely processed. To prevent infinite processing loops, whenever a record gets processed through a forward link, or a database link with the PP link option, the linking record's PACT field is saved and set to TRUE, then restored again afterwards. The example given at the beginning of this section gives an example. It will be seen in the next two sections that PACT has other uses.

5.7.1.4 Process Passive: Link option

Input and output links have an option called process passive. For each such link the application developer can specify process passive TRUE (PP) or process passive FALSE (NPP). Consider the following example:

\includegraphics{lockScanProcess_16}

Assume that all records except fanout are passive. When the fanout record is processed the following sequence of events occur:

  1. Fanout starts processing and asks that B be processed.

  2. B begins processing. It calls dbGetLink to obtain data from A.

  3. Because the input link has process passive true, a request is made to process A.

  4. A is processed, the data value fetched, and control is returned to B

  5. B completes processing and control is returned to fanout. Fanout asks that C be processed.

  6. C begins processing. It calls dbGetLink to obtain data from A.

  7. Because the input link has process passive TRUE, a request is made to process A.

  8. A is processed, the data value fetched, and control is returned to C.

  9. C completes processing and returns to fanout

  10. The fanout completes

Note that A was processed twice. This is unnecessary. If the input link to C were declared No Process Passive then A would only be processed once. Thus a better solution would be:

\includegraphics{lockScanProcess_26}

5.7.1.5 Process Passive: Field attribute

All record type field definitions have an attribute called process_passive which is specified in the record definition file. It cannot be changed by an IOC application developer. This attribute is used only by dbPutField. It determines if a passive record will be processed after dbPutField sets a field in the record. Consult the record specific information in the record reference manual for the setting of individual fields.

5.7.1.6 Maximize Severity: Link option

Input and output links have an option called maximize severity. For each such link the application developer can specify the option as MS (Maximize Severity), NMS (Non-Maximize Severity), MSS (Maximize Status and Severity) or MSI (Maximize Severity if Invalid).

When database input or output links are defined, the application developer can use this option to specify whether and how alarm severities should be propagated across links with the data. The alarm severity is transferred only if the new severity will be greater than the current severity of the destination record. If the severity is propagated the alarm status is set equal to LINK_ALARM (unless the link option is MSS when the alarm status will also be copied from the source record).

5.8 Guidelines for Synchronous Records

A synchronous record is a record that can be completely processed without waiting. Thus the application developer never needs to consider the possibility of delays when he defines a set of related records. The only consideration is deciding when records should be processed and in what order a set of records should be processed.

The following reviews the methods available to the application programmer for deciding when to process a record and for enforcing the order of record processing.

  1. A record can be scanned periodically (at one of several rates), via I/O event, or via Event.

  2. For each periodic group and for each Event group the PHAS field can be used to specify processing order.

  3. The application programmer has no control over the record processing order of records in different groups.

  4. The disable fields (SDIS, DISA, and DISV) can be used to disable records from being processed. By letting the SDIS field of an entire set of records refer to the same input record, the entire set can be enabled or disabled simultaneously. See the Record Reference Manual for details.

  5. A record (periodic or other) can be the root of a set of passive records that will all be processed whenever the root record is processed. The set is formed by input, output, and forward links.

  6. The process_passive attribute of each record field determines if a passive record will be processed when a dbPutField is directed to the field. The application developer must be aware of the possibility of record processing being triggered by external sources using this mechanism.

  7. The process_passive option for input and output links provides the application developer control over how a set of records are scanned.

  8. General link structures can be defined. The application programmer should be wary, however, of defining arbitrary structures without carefully analyzing the processing order.

5.9 Guidelines for Asynchronous Records

The previous discussion does not cover asynchronous device support. An example might be a GPIB input record. When the record is processed the GPIB request is started and the processing routine returns. Processing, however, is not really complete until the GPIB request completes. This is handled via an asynchronous completion routine. Let's state a few attributes of asynchronous record processing.

During the initial processing for all asynchronous records the following is done:

  1. PACT is set TRUE

  2. Data is obtained for all input links

  3. Record processing is started

  4. The record processing routine returns

The asynchronous completion routine performs the following algorithm:

  1. Record processing continues

  2. Record specific alarm conditions are checked

  3. Monitors are raised

  4. Forward links are processed

  5. PACT is set FALSE.

A few attributes of the above rules are:

  1. Asynchronous record processing does not delay the scanners.

  2. Between the time that record processing begins and the asynchronous completion routine completes, no attempt will be made to again process the record. This is because PACT is TRUE. The routine dbProcess checks PACT and does not call the record processing routine if it is TRUE. Note, however, that if dbProcess finds the record active 10 times in succession, it raises a SCAN_ALARM.

  3. Forward and output links are triggered only when the asynchronous completion routine completes record processing.

With these rules the following works just fine:

\includegraphics{lockScanProcess_34}

When dbProcess is called for record ASYN, processing will be started but dbScanPassive will not be called. Until the asynchronous completion routine executes any additional attempts to process ASYN are ignored. When the asynchronous callback is invoked the dbScanPassive is performed.

Problems still remain. A few examples are:

5.9.1 Infinite Loop

Infinite processing loops are possible.

\includegraphics{lockScanProcess_1}

Assume both A and B are asynchronous passive records and a request is made to process A. The following sequence of events occur.

  1. A starts record processing and returns leaving PACT TRUE.

  2. Sometime later the record completion for A occurs. During record completion a request is made to process B. B starts processing and control returns to A which completes leaving its PACT field FALSE.

  3. Sometime later the record completion for B occurs. During record completion a request is made to process A. A starts processing and control returns to B which completes leaving its PACT field FALSE.

Thus an infinite loop of record processing has been set up. It is up to the application developer to prevent such loops.

5.9.2 Obtain Old Data

A dbGetLink to a passive asynchronous record can get old data.

\includegraphics{lockScanProcess_37}

If A is a passive asynchronous record then the dbGetLink request forces dbProcess to be called for A. dbProcess starts the processing and returns. dbGetLink then reads the field value which is still old because processing will only be completed at a later time.

5.9.3 Delays

Consider the following:

\includegraphics{lockScanProcess_40}

The second ASYN record will not begin processing until the first completes, etc. This is not really a problem except that the application developer must be aware of delays caused by asynchronous records. Again, note that scanners are not delayed, only records downstream of asynchronous records.

5.10 Cached Puts

The rules followed by dbPutLink and dbPutField provide for ``cached" puts. This is necessary because of asynchronous records. Two cases arise.

The first results from a dbPutField, which is a put coming from outside the database, i.e. Channel Access puts. If this is directed to a record that already has PACT TRUE because the record started processing but asynchronous completion has not yet occurred, then a value is written to the record but nothing will be done with the value until the record is again processed. In order to make this happen dbPutField arranges to have the record reprocessed when the record finally completes processing.

The second case results from dbPutLink finding a record already active because of a dbPutField directed to the record. In this case dbPutLink arranges to have the record reprocessed when the record finally completes processing. If the record is already active because it appears twice in a chain of record processing, it is not reprocessed because the chain of record processing would constitute an infinite loop.

Note that the term caching not queuing is used. If multiple requests are directed to a record while it is active, each new value is placed in the record but it will still only be processed once, i.e. last value wins.

5.11 putNotify

dbPutNotify (called when a Channel Access client calls ca_put_callback) is a request to notify the caller when all records processed as a result of the put are complete. Because of asynchronous records this can be complicated and the set of records that are processed because of a put may not be deterministic. The result of a dbPutNotify is the same as a dbPutField except for the following:

5.12 Channel Access Links

A channel access link is:

  1. A record link that references a record in a different IOC.

  2. A link that the application developer forces to be a channel access link.

A channel access client task (dbCa) handles all I/O for channel access links. It does the following:

Even if a link references a record in the same IOC it can be useful to force it to act like a channel access link. In particular the records will not be forced to be in the same lock set. As an example consider a scan record that links to a set of unrelated records, each of which can cause a lot of records to be processed. It is often NOT desirable to force all these records into the same lock set. Forcing the links to be handled as channel access links solves the problem.

CA links which connect between IOCs incur the extra overhead associated with message passing protocols, operating system calls, and network activity. In contrast, CA links which connect records in the same IOC are executed more efficiently by directly calling database access functions such as dbPutField and dbGetField, or by receiving callbacks directly from a database monitor subscription event queue.

Because channel access links interact with the database only via dbPutField, dbGetField, and a database monitor subscription event queue, their interaction with the database is fundamentally different from database links which are tightly integrated within the code that executes database records. For this reason and because channel access does not support the passing of a process passive flag, the semantics of channel access links are not the same as database links. Let's discuss the channel access semantics of INLINK, OUTLINK, and FWDLINK separately.

5.12.1 INLINK

The options for process passive are:

Maximize Severity is honored.

5.12.2 OUTLINK

The options for process passive are:

Maximize Severity is not honored.

5.12.3 FWDLINK

A channel access forward link is honored only if it references the PROC field of a record. In that case a ca_put with a value of 1 is written each time a forward link request is issued.

The available options are:

Maximize Severity is not honored.


next up previous contents index
Next: 6. Database Definition Up: AppDevGuide Previous: 4. Build Facility   Contents   Index
Andrew Johnson 2014-02-11