Re: TI Graph Link Errors


[Prev][Next][Index][Thread]

Re: TI Graph Link Errors



At 07:44 AM 7/31/96 -0700, you wrote:
>Eric P. Anderson wrote:
>>
>> At 12:37 AM 7/31/96 -0700, you wrote:
>> >that run in them to simultaneously share IRQs.  OS/2 and LINUX allow
>> >sharing of IRQs, and the programs that run on them can both use the same
>> >IRQ.  They leave the scheduling up to the processor.  Windoze (either
>> >flavor) just does an "Oh crap" when confronted with shared IRQs, and
>> >solves the problem by shutting off one or the other
>>
>> >O.k., Dipweed:
>> >
>> >>Before you flame me for no reason whatsoever, you
>> >>should get your facts straight.  IRQs are HARDWARE related; they have
>> >>nothing to do with your operating system.
>> >
>> >Get your facts straight, for the second time;  IRQ HANDLING, the way the
>> >software in an OS works with the IRQ, which is different from the
>> >physical IRQ itself, is ENTIRELY OS DEPENDENT.
>>
>> Wrong.  Let me quote something from the Linux FAQ.
>>
>> Question 7.7. Can I have more than 3 serial ports by
>> sharing interrupts ?
>>
>> Yes, but you won't be able to use simultaneously two ordinary ports which
>> share an interrupt
>> (without some trickery). This is a limitation of the ISA bus architecture.
>
>Can be done with some trickery doesn't sound wrong to me.  Besides, how
>many non PCI bus pentiums are there?  Zero, so even though the ISA bus
>requires some work, its a moot point in a pentium.
>


Trickery doesn't means using IRQs simultaneously as you claim.  It means
reassigning devices so they all have unique IRQs.  It is simply not possible
to use an IRQ by more than one device at the same time.  It is a limitation
of the hardware on the PC.  Which operating system you are using has no
relevance in the matter.  In addition, just because a 486 or Pentium has a
local bus on it has no effect on this problem.  All PCs have this "problem".
The PCI or VL bus _never_ controls the IRQs.  That is still done by the ISA
bus, which is still there.  Just because the computer has a local bus
doesn't mean it is using IRQs through it.  Besides, doesn't it sound logical
for just one device to use an IRQ anyway?


>Apology accepted.  I'm glad to see you started to educate yourself,
>though.


Actually, it seems you have been educated.  This information supports MY
original argument.  Here, I'll even show you your incorrect statement.


>1)  Yes IRQs are hardware related.  How the software handles them is OS


>related.  That is called IRQ handling. Certain OS's can handle multiple
                                        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>devices on the same IRQ at the same time.  Others (DOS based) can't.
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^


I eagerly wait to see how you respond to this.


************************************************************
*  Eric P. Anderson                     "Who is General    *
*  crusader@mo.net                      Failure and why    *
*  http://walden.mo.net/~crusader/      is he reading      *
*  Finger for PGP public key.           drive C?"          *
*  Fprint:3E 3C 52 A2 C1 3F 61 B5 71 7F 10 F1 09 B4 D4 D7  *
*  This is a public notice that all unsolicited and/or     *
*  commercial emailers will be charged an archival and     *
*  downloading fee of U.S. $500.  Emailing me signifies    *
*  agreement to these terms.                               *
*  "The price of freedom is eternal vigilance." -Jefferson *
************************************************************


References: