[lkml]   [1998]   [Apr]   [23]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: Fix for PCI detection in fdomain.c

> My apologies. I've obviously been working too hard on this problem.
> You're correct, I did only half apply your spelling fix to that line and
> it does indeed work properly when your spelling fix is completely
> applied.


> Question: Did you fix the other problem with the driver, the one that
> keeps it from compiling? I saw the patch you sent to Linus, and I
> didn't see if you added the declaration for "do_fdomain_16x0_intr". It
> was missing in 2.1.96.

No, it seems someone has been fixing all the SCSI drivers.

> Also, since you're maintaining the driver at the moment, you might want
> to do some of the following:

Well, I don't feel like being the right person to maintain the driver
as I don't have direct access any such controller.

> Fix the change history to note your driver re-write on
> 11 April 1998
> Up the driver version. It's been 5.45 since 1996, and
> it's not the same driver anymore.

OK, I'll do that.

> Remove James' e-mail address from the error message that
> appears in the 'dmesg' output. He'll probably appreciate
> that because he's not maintaining the code anymore.


| I agree that the old PCI bus scanning code in fdomain.c was
| not particularly elegant, it was basically just a hack to
| make it work. I tried using the BIOS32 routines ala Drew

It was in also all drivers -- when I was working on the new PCI code,
the most time-consuming part was modifying all drivers from the `single
architecture, poor PCI support' era to my new API. This usually involved
replacing 100 lines of code by 10 lines :-)

| Eckhard's method in his driver, but was unable to get them
| to work. Drew didn't seem very willing to help at the time
| so I just hacked at it until it worked, which it did for a
| few years.

Until I broke it :-)

| Portability was not an issue at the time since
| Linux had not really been successfully ported (I mean really
| usable) from the Intel platform. Please clean up the code
| and remove my email address from the detection failure print
| and replace it with your own if you wish.

I'll probably remove the address from the message at all.

Have a nice fortnight
Martin `MJ' Mares <>
Faculty of Math and Physics, Charles University, Prague, Czech Rep., Earth
"If at first you don't succeed, redefine success."

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

 \ /
  Last update: 2005-03-22 13:42    [W:0.040 / U:5.324 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site