question: trap handler query....

question: trap handler query....

[Mukesh Rathor]

Hi all,
  I've a question. how would a cpu know from an IPI, if it's on a trap handler 
underneath. Looking at entry.S, I see that trap handlers push entry vector on 
stack, so I suppose one could walk down the frames (there's prob a better 
way). Moreover, what if the IPI comes right before the entry vector is pushed? 
Does x86 disable upon trap?

Thanks for any help,
Mukesh

Re: question: trap handler query....

[Daniel Stodden]

On Tue, 2008-01-22 at 12:40 -0800, Mukesh Rathor wrote:
> Hi all,
>   I've a question. how would a cpu know from an IPI, if it's on a trap handler 
> underneath. Looking at entry.S, I see that trap handlers push entry vector on 
> stack, so I suppose one could walk down the frames (there's prob a better 
> way). 

the better way is to check the interrupted privilege level in the pushed
CS register (bits 1-0). if it's zero, it's been in xen when it happened.

> Moreover, what if the IPI comes right before the entry vector is pushed? 

exception entry by itself is atomic. either you get a complete stack
frame (as pushed by the cpu, i.e. up to rip/error_code, i.e. not
including the SAVE_ALL macros) or remain at the instruction yet to cause
the trap. nowhere in between.

> Does x86 disable upon trap?

..interrupts you mean? typically no. only on interrupt gates.

regards,
daniel

-- 
Daniel Stodden
LRR     -      Lehrstuhl für Rechnertechnik und Rechnerorganisation
Institut für Informatik der TU München             D-85748 Garching
http://www.lrr.in.tum.de/~stodden         mailto:stodden@cs.tum.edu
PGP Fingerprint: F5A4 1575 4C56 E26A 0B33  3D80 457E 82AE B0D8 735B