[lkml]   [1999]   [Jan]   [2]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: [linux-audio-dev] ISA Plug & Play support in kernel
George writes:
> On Sat, Jan 02, 1999 at 03:56:33AM -0500, Albert D. Cahalan wrote:

>> 4. If the hardware changes, such a config file may cause conflicts.
>> The system might not even boot after a hardware or Windows upgrade.
>> (Windows is known to change BIOS PnP settings!)
> If change hardware you have a good change of windows being unbootable as
> well, as such configuring the BIOS to just leave the PnP devices
> unconfigured works fine for %90 of the cases.

No, it goes like this:

BIOS loads kernel.
Kernel starts init, init runs scripts, etc.
PnP tool loads obsolete config file.
Modules load with obsolete parameters.

Even if the boot completes, you may have strange problems that are
hard to figure out.

>> 5. This is not PnP. It is software jumpers. PnP exists because it
>> is moronic to make the user tell the computer about the computer.
>> Computers ought to be able to determine their own innards.
> Use PCI then.

Linux makes you buy new hardware? That ought to sell...

Yes, Microsoft helped design this. Get over it. Every new PC
and ISA card supports this. Dynamic reconfiguration isn't as
nice as slot-based configuration, but is still quite usable.

>> 6. This won't let us reallocate devices. If I have n IRQs available
>> for the ISA bus and cards that need n+2 IRQs, I ought to be able
>> to use all but 2 devices at the same time without deciding that
>> 2 particular devices just won't ever function. The kernel needs
>> to move IRQs as needed so that I can freely use my hardware.
>> Userspace can't handle this at all.
> How can I best say this..
> Shoot, I'm not even going to /TRY/ to communicate this to someone who
> understands the internals so little as to suggest such a down right
> /INSANE/ and /EXTREMELY DANGEROUS/ thing go into the kernel..

It is perfectly reasonable. Would you rather share ISA IRQs?
That _would_ be extremely dangerous. How else to do this?

Kernel internals would change when such a patch is applied,
so that is no issue.

Cheap modern hardware sucks. If you don't like it, _you_ use PCI.

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
Please read the FAQ at

 \ /
  Last update: 2005-03-22 13:49    [W:0.047 / U:4.676 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site