Messages in this thread | | | From | alano@summanul ... | Subject | My 2 cent suggestion for 2.3.x | Date | Wed, 19 May 1999 10:58:01 -0700 |
| |
I am not certain if this has found its way into the tree yet. I have gotten bitten by it in the 2.0.x tree and have not seen anything about it being added in the 2.2.x or 2.3.x trees. If it has, or I am describing something that has been fixed already, then ignore me and just go on about your business.
Situation:
You have a stock install and/or rebuild the kernel without changing kernel versions. A driver was originally compiled as a module, but is now compiled directly into the kernel. The module is still in the modules directory for the kernel because the user did not clean it out or some sort of default is forcing it to use that directory. Opon boot, the kernel tries to use both the module and the internal code for the same device causing major chokeage. I have run into this on SCSI drivers more than once. (I am not certain if any other drivers die in quite the same manner.)
What I see as part of the module load process is some sort of check to say "Is this function already in the kernel? If not, don't load it."
I know what to look for when this happens, but I have seen a few other users greatly confused and concerned when it happens to them. (And I don't blame them from being frustrated. Unfortunatly frustration brings assignment of blame. "And It's Your Fault!!!")
And no, I don't have a patch for this. Yet.
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |