home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.arch
- Path: sparky!uunet!iWarp.intel.com|ichips!ichips!glew
- From: glew@pdx007.intel.com (Andy Glew)
- Subject: Re: trapping speculative ops
- In-Reply-To: schow@bqneh3.bnr.ca's message of Thu, 27 Aug 1992 00:53:40 GMT
- Message-ID: <GLEW.92Aug28174035@pdx007.intel.com>
- Sender: news@ichips.intel.com (News Account)
- Organization: Intel Corp., Hillsboro, Oregon
- References: <GLEW.92Aug25180333@pdx007.intel.com> <CLIFFC.92Aug26084159@medea.rice.edu>
- <1992Aug27.005340.6547@bcars64a.bnr.ca>
- Date: Sat, 29 Aug 1992 01:40:35 GMT
- Lines: 40
-
- >In article <CLIFFC.92Aug26084159@medea.rice.edu> cliffc@rice.edu (Cliff Click) writes:
- >>Let every register have some extra "trap" bits.
- >>A read of the register with it's trap bits set, causes the exception.
- >>A write to the register sets the trap bits according to the success of
- >>the operation.
- >>
- >>With this design, exceptions are triggered at the START of some operation,
- >>instead of in the middle of it.
- >
- >[schow@bqneh3.bnr.ca (Stanley T.H. Chow)]
- >Very nice model, but what to do about saving and restoring registers
- >across subroutine call, interupts, etc.?
-
- (1) It would be trivial for a compiler to arrange that all potential
- faults have been verified before a procedure call => no need to save
- the trap bits. Similarly, it would probably be not very necessary for
- a compiler to spill registers that contain potential faults.
-
- (2) If you have 64 registers, a single register can hold the trap bits
- for them all. I wouldn't feel too bad about adding that to the state
- to be saved/restored on an interrupt or context switch.
-
- (3) There are techniques that have been discussed for making the trap
- bits "hidden", not user visible. Myself, I think that goes against
- the RISC philosophy.
-
- (4) Adding "extra" bits to a register makes some other common
- operations, that do not fit into a standard 2input 1output format,
- much easier to do.
-
- --
-
- Andy Glew, glew@ichips.intel.com
- Intel Corp., M/S JF1-19, 5200 NE Elam Young Pkwy,
- Hillsboro, Oregon 97124-6497
-
- This is a private posting; it does not indicate opinions or positions
- of Intel Corp.
-
- Intel Inside (tm)
-