Talk:Trap (computing)

This is not correct (or at least partially incorrect). A trap is NOT a fault, and it does NOT occur following an error. It is part of the proper procedure to run any system call (such as, open/close/read/write file, ask for the time of day, etc.) It isn't surprising, since I'm not really sure how reliable "the jargon file" is... I'd suggest using any Operating System book/ University Course site and correct this (not so small) glitch. Like here: http://www.edugrid.ac.in/webfolder/OpSystems/2_IntroductionII/Book-G_Nutt/trap.html Thoughts?


 * Why do you assume that the jargon file is my source? Try




 * Shmuel (Seymour J.) Metz Username:Chatul (talk) 18:46, 27 September 2010 (UTC)


 * The above references are woefully out of date. Try reading something that was published in, say, the last 40 years.  — Preceding unsigned comment added by 38.69.12.5 (talk) 03:42, 28 October 2013 (UTC)


 * The references may be old, but they are relevant to the historical meaning of the term. As to current usage, look at, e.g., z/Architecture Principles of Operation, SA22-7832-09. Shmuel (Seymour J.) Metz Username:Chatul (talk) 20:39, 28 October 2013 (UTC)

Horrible article
This article should be tossed and completely rewritten. Aside from inaccuracies, it's a lousy example of English prose. — Preceding unsigned comment added by 38.69.12.5 (talk) 03:40, 28 October 2013 (UTC)

Intel INT
I'm not sure how much belongs in the article, but Intel did not introduce the concept of an instruction intended for use by the software to cause a trap, although the detail of INT, of course, originated with Intel. Earlier examples are the Master Mode Entry (MME) instruction on the GE-600 series and the Supervisor Call instruction (SVC) on the IBM System/360.

Note that the call gate eliminates the need for INT. Shmuel (Seymour J.) Metz Username:Chatul (talk) 20:02, 7 September 2018 (UTC)

Replacing article with redirect to hardware exceptions section
I wrote a section Exception handling which has most of the content of this page. I don't see anything important on here missing from that section - any objections to my redirecting this page to that section? --Mathnerd314159 (talk) 05:44, 12 January 2022 (UTC)
 * There is relevant information not present in either Trap (computing) or Exception handling. Historically, trap has been used for several different types of interrupts
 * All interrupts, including I/O
 * All synchronous interrupts
 * Interrupt caused by a trap instruction; There is a discussion at Talk:Illegal opcode that discusses some of the trap instructions.
 * Conditionally
 * Unconditionally
 * There is also the question of whether OS interface instructions like Master Mode Entry (MME), SVC and UUO should be considered as trap instructions. --Shmuel (Seymour J.) Metz Username:Chatul (talk) 15:24, 12 January 2022 (UTC)
 * So as I read that list, a trap is an interrupt, an interrupt, or an interrupt. Then I guess redirecting to interrupt would be more appropriate. Regarding the history and various usages etc., that can go in the interrupt article as a section. I don't think "trap" has enough usage in a computing context to merit its own article. --Mathnerd314159 (talk) 02:53, 19 January 2022 (UTC)
 * I would recommend merging Interrupt and Trap (computing), with in Exception handling. --Shmuel (Seymour J.) Metz Username:Chatul (talk) 15:13, 19 January 2022 (UTC)
 * OK, I moved everything into Interrupt, so now this page is a redirect. Good? --Mathnerd314159 (talk) 05:29, 21 January 2022 (UTC)
 * Looks good. It probably needs clarification of hardware context versus software context, with examples. I made one small change to note that what the OS resumes after an interrupt need not be the interrupted process. --Shmuel (Seymour J.) Metz Username:Chatul (talk) 14:25, 21 January 2022 (UTC)