| Summary: | /etc/init.d/pcmcia coding error (w/ rebuilt kernel) | ||
|---|---|---|---|
| Product: | [Retired] Red Hat Linux | Reporter: | Need Real Name <chagner> |
| Component: | kernel-pcmcia-cs | Assignee: | Arjan van de Ven <arjanv> |
| Status: | CLOSED CURRENTRELEASE | QA Contact: | Brian Brock <bbrock> |
| Severity: | medium | Docs Contact: | |
| Priority: | medium | ||
| Version: | 9 | ||
| Target Milestone: | --- | ||
| Target Release: | --- | ||
| Hardware: | i686 | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Fixed In Version: | Doc Type: | Bug Fix | |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2004-09-08 06:58:33 UTC | Type: | --- |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
This was fixed in an update to FC1, and in current trees. |
Description of problem: After rebuilding a kernel from the kernel-source rpm provided, I noticed that the pcmcia modules were not being loaded (reverting to the original install kernel still worked). After hunting down a number of things, here's what I came up with... 1. The kernel build/modules_install process creates a /lib/modules/<kernel_tag>/pcmcia directory. The original (installed) /lib/modules/2.4.20-8/ does not have this directory. 2. The /etc/init.d/pcmcia script keys of the existence of this directory and executes a different syntax for 'modprobe pcmcia_core.' I thought it wasn't possible (and I still think that), but calling 'modprobe pcmcia_core.o' will fail (the '.o' shouldn't be there... see code below). Here's the code block from the pcmcia script (with my comments added)... if [ -d /lib/modules/preferred ] ; then PC=/lib/modules/preferred/pcmcia else PC=/lib/modules/`uname -r`/pcmcia fi KD=/lib/modules/`uname -r`/kernel/drivers/pcmcia # MY COMMENT: THIS IS WHERE IT WAS MESSED UP. # MY COMMENT: THE NEW KERNEL's modules directory had a # MY COMMENT: top-level pcmcia dir which dropped into # MY COMMENT: this first if block. # The elif block is the right one. if [ -d $PC ] ; then echo -n " modules" # MY COMMENT: how does this work? I think the .o # MY COMMENT: shouldn't be there. # MY COMMENT: Looking at the same script for Redhat # MY COMMENT: 7.3, the lines have "$PC/" prefixing # MY COMMENT: the module name like so... # MY COMMENT: /sbin/modprobe $PC/pcmcia_core.o $CORE_OPTS # MY COMMENT: This makes _more_ sense, but I don't think # MY COMMENT: it'll work. /sbin/modprobe pcmcia_core.o $CORE_OPTS /sbin/modprobe $PCIC.o $PCIC_OPTS /sbin/modprobe ds.o elif [ -d $KD ] ; then /sbin/modprobe pcmcia_core /sbin/modprobe $PCIC /sbin/modprobe ds else echo $" module directory $PC not found." break fi I can fix it either by modifying the script or blowing away the /lib/modules/<kernel_tag>/pcmcia directory (since it just contains symlinks to the kernel directory versions. Version-Release number of selected component (if applicable): Standard RedHat 9 (Shrike) release. How reproducible: Recompiling everything now... but, assuming the error is in the pcmcia script, then it's not a kernel compilation error, it's a script logic error. Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info: