The logic to process mod-extra.list finds modules that depend on the modules in that list and includes those in kernel-module-extra as well. This logic is broken, because it does not account for multi-layer dependency chains. Specifically, if a module depends on another module, and only the second one depends on something in mod-extra.list, then only the second would get moved. This leaves the first module in the kernel package, resulting in WARNING messages when the kernel rpm is installed: Preparing... ########################################### [100%] 1:kernel ########################################### [100%] WARNING: /lib/modules/3.2.0-0.rc3.git1.2.fc14.x86_64/kernel/drivers/scsi/bnx2fc/bnx2fc.ko needs unknown symbol cnic_register_driver WARNING: /lib/modules/3.2.0-0.rc3.git1.2.fc14.x86_64/kernel/drivers/scsi/bnx2fc/bnx2fc.ko needs unknown symbol cnic_unregister_driver WARNING: /lib/modules/3.2.0-0.rc3.git1.2.fc14.x86_64/kernel/drivers/scsi/bnx2i/bnx2i.ko needs unknown symbol cnic_register_driver WARNING: /lib/modules/3.2.0-0.rc3.git1.2.fc14.x86_64/kernel/drivers/scsi/bnx2i/bnx2i.ko needs unknown symbol cnic_unregister_driver WARNING: /lib/modules/3.2.0-0.rc3.git1.2.fc14.x86_64/kernel/drivers/usb/wusbcore/wusb-wa.ko needs unknown symbol wusbhc_reset_all WARNING: /lib/modules/3.2.0-0.rc3.git1.2.fc14.x86_64/kernel/drivers/usb/wusbcore/wusb-wa.ko needs unknown symbol wusbhc_handle_dn WARNING: /lib/modules/3.2.0-0.rc3.git1.2.fc14.x86_64/kernel/drivers/usb/wusbcore/wusb-wa.ko needs unknown symbol wusbd WARNING: /lib/modules/3.2.0-0.rc3.git1.2.fc14.x86_64/kernel/drivers/usb/wusbcore/wusb-wa.ko needs unknown symbol __wusb_dev_get_by_usb_dev WARNING: /lib/modules/3.2.0-0.rc3.git1.2.fc14.x86_64/kernel/drivers/usb/wusbcore/wusb-wa.ko needs unknown symbol wusbhc_giveback_urb WARNING: /lib/modules/3.2.0-0.rc3.git1.2.fc14.x86_64/kernel/drivers/usb/wusbcore/wusb-wa.ko needs unknown symbol wusb_dev_destroy WARNING: /lib/modules/3.2.0-0.rc3.git1.2.fc14.x86_64/kernel/drivers/usb/atm/xusbatm.ko needs unknown symbol usbatm_usb_disconnect WARNING: /lib/modules/3.2.0-0.rc3.git1.2.fc14.x86_64/kernel/drivers/usb/atm/xusbatm.ko needs unknown symbol usbatm_usb_probe WARNING: /lib/modules/3.2.0-0.rc3.git1.2.fc14.x86_64/kernel/net/bridge/netfilter/ebtable_broute.ko needs unknown symbol br_should_route_hook This patch reverses that logic, so that anything in kernel with a dependency on something in mod-extra.list will cause the module in mod-extra.list to be treated as if it was _not_ in that list. This prevents the WARNING lines above. Signed-off-by: John W. Linville <linville@tuxdriver.com> |
||
---|---|---|
scripts | ||
.gitignore | ||
acpi-ensure-thermal-limits-match-cpu-freq.patch | ||
acpi-sony-nonvs-blacklist.patch | ||
arm-omap-dt-compat.patch | ||
arm-smsc-support-reading-mac-address-from-device-tree.patch | ||
bcma-brcmsmac-compat.patch | ||
block-stray-block-put-after-teardown.patch | ||
config-arm-generic | ||
config-arm-omap-generic | ||
config-arm-tegra | ||
config-debug | ||
config-generic | ||
config-i686-PAE | ||
config-local | ||
config-nodebug | ||
config-powerpc32-generic | ||
config-powerpc32-smp | ||
config-powerpc64 | ||
config-powerpc-generic | ||
config-s390x | ||
config-sparc64-generic | ||
config-x86_64-generic | ||
config-x86-32-generic | ||
config-x86-generic | ||
cpupower.config | ||
cpupower.service | ||
die-floppy-die.patch | ||
disable-i8042-check-on-apple-mac.patch | ||
dmar-disable-when-ricoh-multifunction.patch | ||
drm-i915-sdvo-lvds-is-digital.patch | ||
drm-intel-next.patch | ||
efi-dont-map-boot-services-on-32bit.patch | ||
epoll-limit-paths.patch | ||
fix_xen_guest_on_old_EC2.patch | ||
floppy-Remove-_hlt-related-functions.patch | ||
kernel.spec | ||
linux-2.6-32bit-mmap-exec-randomization.patch | ||
linux-2.6-acpi-debug-infinite-loop.patch | ||
linux-2.6-acpi-video-dos.patch | ||
linux-2.6-compile-fixes.patch | ||
linux-2.6-crash-driver.patch | ||
linux-2.6-defaults-acpi-video.patch | ||
linux-2.6-defaults-aspm.patch | ||
linux-2.6-e1000-ich9-montevina.patch | ||
linux-2.6-i386-nx-emulation.patch | ||
linux-2.6-input-kill-stupid-messages.patch | ||
linux-2.6-intel-iommu-igfx.patch | ||
linux-2.6-makefile-after_link.patch | ||
linux-2.6-serial-460800.patch | ||
linux-2.6-silence-acpi-blacklist.patch | ||
linux-2.6-silence-fbcon-logo.patch | ||
linux-2.6-silence-noise.patch | ||
linux-2.6-upstream-reverts.patch | ||
linux-2.6-v4l-dvb-experimental.patch | ||
linux-2.6-v4l-dvb-fixes.patch | ||
linux-2.6-v4l-dvb-update.patch | ||
linux-2.6.30-no-pcspkr-modalias.patch | ||
Makefile | ||
Makefile.config | ||
Makefile.release | ||
merge.pl | ||
mod-extra.list | ||
nfsv4-include-bitmap-in-nfsv4_get_acl_data.patch | ||
oom-fix-integer-overflow-of-points.patch | ||
pci-Rework-ASPM-disable-code.patch | ||
quite-apm.patch | ||
README.txt | ||
revert-efi-rtclock.patch | ||
sources | ||
sysfs-msi-irq-per-device.patch | ||
taint-vbox.patch | ||
TODO | ||
utrace.patch |
Kernel package tips & tricks. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ The kernel is one of the more complicated packages in the distro, and for the newcomer, some of the voodoo in the spec file can be somewhat scary. This file attempts to document some of the magic. Speeding up make prep --------------------- The kernel is nearly 500MB of source code, and as such, 'make prep' takes a while. The spec file employs some trickery so that repeated invocations of make prep don't take as long. Ordinarily the %prep phase of a package will delete the tree it is about to untar/patch. The kernel %prep keeps around an unpatched version of the tree, and makes a symlink tree clone of that clean tree and than applies the patches listed in the spec to the symlink tree. This makes a huge difference if you're doing multiple make preps a day. As an added bonus, doing a diff between the clean tree and the symlink tree is slightly faster than it would be doing two proper copies of the tree. build logs. ----------- There's a convenience helper script in scripts/grab-logs.sh that will grab the build logs from koji for the kernel version reported by make verrel config heirarchy. ----------------- Instead of having to maintain a config file for every arch variant we build on, the kernel spec uses a nested system of configs. At the top level, is config-generic. Add options here that should be present in every possible config on all architectures. Beneath this are per-arch overrides. For example config-x86-generic add additional x86 specific options, and also _override_ any options that were set in config-generic. There exist two additional overrides, config-debug, and config-nodebug, which override -generic, and the per-arch overrides. It is documented further below. debug options. -------------- This is a little complicated, as the purpose & meaning of this changes depending on where we are in the release cycle. If we are building for a current stable release, 'make release' has typically been run already, which sets up the following.. - Two builds occur, a 'kernel' and a 'kernel-debug' flavor. - kernel-debug will get various heavyweight debugging options like lockdep etc turned on. If we are building for rawhide, 'make debug' has been run, which changes the status quo to: - We only build one kernel 'kernel' - The debug options from 'config-debug' are always turned on. This is done to increase coverage testing, as not many people actually run kernel-debug. To add new debug options, add an option to _both_ config-debug and config-nodebug, and also new stanzas to the Makefile 'debug' and 'release' targets. Sometimes debug options get added to config-generic, or per-arch overrides instead of config-[no]debug. In this instance, the options should have no discernable performance impact, otherwise they belong in the debug files.