Re: enabling x87 interrupts

"Wendy E. McCaughrin" <spamtrap@xxxxxxxxxx> wrote in message
news:ecklk5$psb$1@xxxxxxxxxxxxxxxxxxxxxxxxxx
Is the 8259A (PIC) still used to enable interrupts? I unmasked

"Unmasked" meaning clearing the bit (i.e., zero or low). All bits not
triggering interrupts should be set (i.e., one or high).

bit #2
in the master PIC and bit #5 (a.k.a. #13) in the slave PIC it cascades
with, to enable FPU interrupts. I have unmasked all 6 exceptions in the
FPU's control-word and use STI. Yet, INT 75h is never accessed.


1) If there are any higher priority interrupts, IRQ0 timer or IRQ1 keyboard,
pending (uncleared), you may not get your desired interrupt until it is
cleared. This depends on how you setup the PICs EOI: Auto-EOI, non-specific
EOI, or specific EOI.
2) If you relocated the base addresses of the PICs (by setting ICW2) from
DOS values 0x08 and 0x70 for PIC1 and PIC2 respectively, you'll need to
check the relocated interrupt, not INT 75h., i.e., subtract 70h add your
base.


Rod Pemberton

.

Relevant Pages

  • Re: [tip:x86/apic] x86: Use EOI register in io-apic on intel platforms
    ... will be acked twice -- once via an EOI message sent from the local APIC ... EOI register) is when the level triggered interrupt gets registered at ... the cpu as an edge interrupt (in the local apic's trigger mode ... from accepting any I/O APIC interrupts. ...
    (Linux-Kernel)
  • Re: [RFC][PATCH] kdump: x86_64 timer interrupt lockup due to pending interrupt
    ... one extra EOI is being issued in check_timerto unlock ... Shouldn't we rather do this for all interrupts when the APIC is set up? ... Please find attached the patch. ... Basically ISR is a 256bit register. ...
    (Linux-Kernel)
  • Re: [RFC][PATCH] kdump: x86_64 timer interrupt lockup due to pending interrupt
    ... one extra EOI is being issued in check_timerto unlock ... Shouldn't we rather do this for all interrupts when the APIC is set up? ... I don't see how the timer is special here. ... Please find attached the patch. ...
    (Linux-Kernel)
  • Re: enabling x87 interrupts
    ... I'm not familiar with the FPU, so I'm not completely sure of everything you ... mask all unused interrupts both PICs. ... decide on EOI control: Auto-EOI, non-specific EOI, or specific EOI ... setup int 10h handler, using chosen EOI method (verify not a cpu ...
    (comp.lang.asm.x86)
  • Re: [RFC][PATCH] kdump: x86_64 timer interrupt lockup due to pending interrupt
    ... Problem happens because timer vector seems to be ... After a system crash, it is not safe to service interrupts any more, hence ... sends these interrupts to the CPU during early boot of second kernel. ... is serviced and CPU does not issue an LAPIC EOI because it think this ...
    (Linux-Kernel)