Matheus Pecoraro commented on a discussion on 
bsps/x86_64/amd64/interrupts/isr_handler.S: 
https://gitlab.rtems.org/rtems/rtos/rtems/-/merge_requests/95#note_108084

 >    /* Save the initial rsp */
 >    movq SAVED_RSP,   (0 * CPU_SIZEOF_POINTER)(rsp)
 >  
 > +  /* Save x87 FPU, MMX and SSE state */
 > +.set FXSAVE_SIZE, 512
 > +  /* Make space for FXSAVE */
 > +  subq $FXSAVE_SIZE, rsp
 > +  fwait
 > +  fxsave64 (rsp)

[This](https://www.agner.org/optimize/instruction_tables.pdf) was the only 
resource I could find and I don't think it's possible to determine a reliable 
number. 

I think "Latency" would be the best metric to determine this and in the ones 
it's reported for the 64 bit mode of the instruction it seems to average around 
~65 for FXSAVE and ~80 for FXRSTOR.

-- 
View it on GitLab: 
https://gitlab.rtems.org/rtems/rtos/rtems/-/merge_requests/95#note_108084
You're receiving this email because of your account on gitlab.rtems.org.


_______________________________________________
bugs mailing list
[email protected]
http://lists.rtems.org/mailman/listinfo/bugs

Reply via email to