A "joke" in the glibc manual

Posted Nov 8, 2018 23:31 UTC (Thu) by Paf (subscriber, #91811)
In reply to: A "joke" in the glibc manual by raegis
Parent article: A "joke" in the glibc manual

FWIW, as a minor detail, I believe in ATA the “master” device specifically controlled the bus and could decide what happened to the bus w/r/t the “slave” device. That doesn’t mean we have to keep using those terms, but it’s not exactly just “primary” and “secondary” either. (Though those would do fine, the level of information communicated by the different choice of terms is minimal, and it’s not like they communicated enough extra info to really aid in understanding. But there was a little bit of reason for the choice.)

A "joke" in the glibc manual

Posted Nov 8, 2018 23:39 UTC (Thu) by mpr22 (subscriber, #60784) [Link] (2 responses)

Wikipedia has the following to say on the subject:

----

Although they are in extremely common use, the terms "master" and "slave" do not actually appear in current versions of the ATA specifications. The two devices are simply referred to as "device 0" and "device 1", respectively, in ATA-2 and later.

It is a common myth that the controller on the master drive assumes control over the slave drive, or that the master drive may claim priority of communication over the other device on the same ATA interface. In fact, the drivers in the host operating system perform the necessary arbitration and serialization, and each drive's onboard controller operates independently of the other.

ATA master/slave

Posted Nov 11, 2018 3:45 UTC (Sun) by giraffedata (guest, #1954) [Link] (1 responses)

I always wondered what the relationship between the two devices on an ATA bus is.

While the spec may call them device 0 and device 1, that belies the actual relationship because I've seen the rules that you can't have just device 1 on a bus. If you have any device at all, you have to have device 0.

Does anyone know what the actual role of device 0/master is?

ATA master/slave

Posted Nov 11, 2018 12:34 UTC (Sun) by farnz (subscriber, #17727) [Link]

Look at section 9.16 Single device configurations in the ATA-5 draft. The only thing that's problematic is that device zero is supposed to identify the state of the PDIAG- and DASP- signals it sees, while device 1 asserts DASP- during reset; without device 0's presence, the host cannot determine the state of PDIAG-, and thus may have trouble identifying the cable type (as it cannot distinguish "faulty device 0" from "no device 0 present").