delorie.com/archives/browse.cgi   search  
Mail Archives: djgpp/1992/05/18/09:06:41

Date: Mon, 18 May 1992 17:22 HKT
From: "Michael Shiu, HKUST CCST, (852)358-6205, ccmshiu AT usthk DOT ust DOT hk" <CCMSHIU AT usthk DOT bitnet>
Subject: Re: two kinds of exception handlers
To: djgpp AT sun DOT soe DOT clarkson DOT edu
Status: O

DJ,

>Some exceptions have been given their own TSS segments because I can't
>guarantee the availability of a stack for them to run in.  For
>example, if a page fault occurs while extending the stack, handling it
>with a regular interrupt would cause a double fault, so the page fault
>handler has it's own stack.  Same for timer, HD, and keyboard, since
>the page fault handler may need these interrupts to be in a valid
>state, I can't have these interrupts causing page faults.

>The remainder go through the common dispatch routines.

What if other H/W device (eg. serial port/network cards) interrupts while the
stack segment is not available? Should we give all H/W interrupts its own TSS?

thanks for your patience,
Michael.

- Raw text -


  webmaster     delorie software   privacy  
  Copyright © 2019   by DJ Delorie     Updated Jul 2019