asmoak@ncrcae.Columbia.NCR.COM (Andy Smoak) (09/14/90)
I have a question about the interrupt gates on the 376 processor.
The i486 MicroProcessor Programmer's Reference Manual (and the 386DX
manual) define the interrupt gates (starting at whatever location
+ 4) as having bits 0-4 as reserved, bits 5-7 to be 000, and bits
8-12 to be 01110 (bit 8 first). (This is on pg. 9-8 of my copy).
The 376 Embedded Processor Programmer's Reference Manual defines the
interrupt gates (starting at whatever location
+ 4) as having bits 0-4 as reserved, bits 5-7 to be 010, and bits
8-12 to be 00110 (bit 8 first). (This is on pg. 8-7 of my copy).
Two questions about this:
1) Could the 376 Manual be incorrect about these settings? I pointed this
out to our hardware guys, and they were suprised to see this difference
(esp. between the 386DX and the 376).
2) If the 376 Manual is correct, what types of problems would you see if
you set the interrupt gates up as if this were a 486 or 386 instead
of the 376 settings? I guess what I'm asking here is what are these
fixed bits used for?
Please e-mail your responses.
Thanks in advance.
Andy Smoak
Distributed Processing Center, NCR E&M Columbiapaulc@hp-lsd.COS.HP.COM (Paul Carroll) (09/15/90)
> The i486 MicroProcessor Programmer's Reference Manual (and the 386DX > manual) define the interrupt gates (starting at whatever location > + 4) as having bits 0-4 as reserved, bits 5-7 to be 000, and bits > 8-12 to be 01110 (bit 8 first). (This is on pg. 9-8 of my copy). > > The 376 Embedded Processor Programmer's Reference Manual defines the > interrupt gates (starting at whatever location > + 4) as having bits 0-4 as reserved, bits 5-7 to be 010, and bits > 8-12 to be 00110 (bit 8 first). (This is on pg. 8-7 of my copy). Just as a guess, since I haven't seen the 376 manual, but the 00110 bits should indicate an 80286-type interrupt gate while the 01110 bits are an 80386-type interrupt gate. This could be indicative of the software development environment used to create software for that processor. I don't know, but Intel may be showing a bias towards using the 80286 tool chain for producing executables for the 376. If this is not the case, then I would guess the 376 manual is hosed. As to the values in bits 5-7, those shouldn't matter according to the 80386/80486 scheme of doing things. ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ + Paul Carroll "I don't believe there is a single + + HP Logic Systems Division man, woman, or child alive in + + paulc%hp-lsd@hplabs.hp.com America today that doesn't enjoy a + + lovely beverage!" - David Letterman + ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
asmoak@ncrcae.Columbia.NCR.COM (Andy Smoak) (09/22/90)
As a followup to anyone that was interested about the interrupt gates on the 376 processor. I just got off of the phone with our intel technical rep. He informs me that the interrupt gates on the 376 are identical to those of the 386 and 486. So, the i376 Programmer's Reference Manual is screwed up. Thanks for the responses people. BTW, I tried to mail this to 386users@udel.edu and my stupid mailer could not find udel.edu (of course, it could be me who is stupid). If Bill Davidsen is out there, please forward this to this mailing list. Thanks. Andy Smoak NCR E&M Columbia