let COMX depend on PROC_FS

<-- snip -->

...
#ifndef CONFIG_PROC_FS
#error For now, COMX really needs the /proc filesystem #endif
...

<-- snip -->

The following patch add a dependency to Kconfig to avoid compile errors with CONFIG_COMX and !CONFIG_PROC_FS:

@@ -62,7 +62,7 @@
 #
 config COMX
 	tristate "MultiGate (COMX) synchronous serial boards support"
-	depends on WAN && (ISA || PCI)
+	depends on WAN && (ISA || PCI) && PROC_FS
 	---help---
 	  Say Y if you want to use any board from the MultiGate (COMX) family.
 	  These boards are synchronous serial adapters for the PC,

cu
Adrian

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

Actually it still doesn't link with this because the procfs code in it is utter crap and relies on a symbol proc_get_inode that isn't exported since 2.3. As no one cared for this driver over years I'd suggest just removing it.

The proc_get_inode link problem only affects the modular build of comx.c .

The static build works fine in both 2.4 and 2.5.

cu
Adrian

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

But it's still broken :) This just shows no one actually tested it with actual hardware.

Christoph Hellwig:

I'm the current "maintainter" of the comx drivers (seriously lacked time up to now), so it's me to flame if you have some spare fuel. And forgive me for having forgot to update the maintainer line in comx.c (comx-* are fine). :)

The drivers are used by some hundreds of cards today but we tell users to get the small kernelpatch from www.itc.hu and the patch, among other things, exports proc_get_inode. There was a process to integrate the patch into the mainstream kernel last year but, due to lack of time on my part, it was suspended. I hope to be able to pick the line up again and clean things up.

s.

          ------------------------------------------------------------
          |  Programmers don't die, they just GOSUB without RETURN.  |
          ------------------------------------------------------------

So what about fixing it instead? The usage of proc_get_inode is broken and so is the whole profs mess in the comx driver. If you want to keep the API you need to add a ramfs-style filesystem instead of abusing procfs.

Christoph Hellwig:

Is the case the same with the SCSI drivers, IDE drivers, network core, filesystems and everything that creates directories and file entries in procfs?

              ---------------------------------------------------
              |  Widows '95 - The Micro$oft Solution Preventer  |
              ---------------------------------------------------

No. The problem with comx is that unlike other driver it doesn't not use the published procfs API but instead tries to implemented half of an own filesystem abusing procfs infrastructure.

Christoph Hellwig:

It'll get updated and fixed.

----------------------------------------------
| If you can't learn to do something well, |
| learn to enjoy doing it poorly. |
----------------------------------------------

"broken" is a very polite way to describe that driver. Starting with the idea of mkdir in virtual filesystem (procfs or otherwise) creating and populating a diretory (unmodifiable, BTW) and rmdir - removing it, even though it's non-empty (and can't be emptied, due to above).

Guys, that's _sick_. And that's aside of the shitload of races all over that code (no locking whatsoever). And kmalloc(..., GFP_KERNEL) with interrupts disabled. And shutting the hardware down before unregistering netdev (yes, you check that it's down; nothing guarantees that it will stay down while you do ->hw_exit() and friends). And so on, and so on...

IOW, driver needs a serious rewrite, starting with its API.